© 2016 University of South Africa

First edition © 2000

Second edition © 2002

All rights reserved

Printed and published by the

University of South Africa
Muckleneuk, Pretoria

LIN2601/1/2017–2023

70457409

InDesign

HSY_Style

ii
 1 Contents

Page

INTRODUCTION TO THE COURSE v

AP Hendrikse & PJ Sanderson

STUDY UNIT 1: Language and the brain 1

HA van Zweel

STUDY UNIT 2: Studying language acquisition 26

PJ Sanderson

STUDY UNIT 3: The transition to speech 53

PJ Sanderson

STUDY UNIT 4: Putting words together 72

PJ Sanderson

STUDY UNIT 5: Language input and individual differences 99

PJ Sanderson

STUDY UNIT 6: Language disability 119

EJ Pretorius

STUDY UNIT 7: The evolution of language 147

AP Hendrikse

GLOSSARY 176

LIN2601/1/2017-2023 iii
iv
 Introduction to the Course

What is this course about?

Language is a uniquely human ability. Although we all speak at least one language
and can probably understand more than one, particularly in the multilingual South
African situation, we seldom, if ever, ask any questions about this remarkable ability
to master a language or several languages. But, is it important to ask such questions?
For most speakers of a language such questions will probably only become crucial
if we lose this ability for some reason, or if we have a child with impaired linguistic
ability. Now, you are not an average speaker of a language in at least one significant
respect – you are a student of linguistics. We therefore presume that you do indeed
have certain questions about this uniquely human ability. Some of your questions
will be addressed in this course, but let us know if you have other questions that
you feel are important.
Since linguistics is a science, the questions we pose about the linguistic ability are
scientific questions. Are scientific questions any different from ordinary questions?
Yes and no! They have exactly the same form as everyday questions – Why?, How?,
Where?, When?, etc, but they differ from everyday questions in several respects. Not
every question based on curiosity qualifies as a scientific question. Scientific questions
arise from the intensive study of a particular phenomenon within a particular area
of reality and within a particular theoretical framework. Scientists begin to ask
questions only after they see certain patterns and regularities. These patterns and
regularities emerge from careful observation of phenomena over time, from careful
analysis of phenomena using scientific ‘tools’, from carefully designed experiments
in laboratories, and so on. Scientists are trained in methods and theories which help
them to formulate questions. Being trained to formulate questions in a specific way
is very important for the simple reason that the way we formulate our questions
very often determines the kind of answers we get.
Scientific questions are like spotlights; they focus on a very specific (sometimes very
small) area or domain. Everything inside their range is brightly lit, but whatever falls
outside their range remains in darkness. Using scientific questions like spotlights is
quite useful, but the problem is that spotlights may not focus on the most crucial areas
and are not very good at differentiating between what is relevant and what is irrelevant.
By using scientific questions like spotlights we run the risk of missing potentially
relevant things simply because these lie outside the area on which the light is focused.
The main question we will spotlight in this module is: How do normal children
acquire language? But in order to shed more light on the subject it is advisable to
use more than one spotlight. We believe that we can find out more about children’s
developing language by focusing on other questions as well, namely: How does
the language ability relate to other mental abilities? How do children with mental
disabilities acquire language? and Where does the language ability in human
beings come from? We think that if we didn’t ask these other questions we would
miss certain significant facts that help us answer the main question. In this module
we try to address some of these other questions by looking at the language ability
and the brain, the language ability of children with particular disabilities, and the
evolutionary origins of the language ability. Thus, we introduce a few additional
spotlights in an attempt to throw more light and to gain different perspectives on
the complex phenomenon of child language development.

LIN2601/1/2017-2023 v
 In order to become a competent speaker, a child is required to master a host of
linguistic skills, including the phonology, morphology, syntax, semantics and
pragmatics of their language. A brief overview of these subdomains of Linguistics
is provided for you here:

Phonology deals with the different kinds of speech sounds that occur in hu-
man languages and the kind of knowledge that speakers have about the sound
patterns of their particular language. This is the subdomain of linguistics
covered in LIN2603, entitled Sound and Sound Structure. Few of you have
probably ever given much thought to the sounds of which languages are made
up, except perhaps when you have learned a new language and encountered
sounds that are unfamiliar to your ear or tongue, or helped someone pronounce
sounds correctly in your own mother tongue. One of the fundamental aims of
phonology is to explain what it is that mother tongue speakers ‘know’ about
the sounds of their language. Much of our knowledge about our mother
tongue is unconscious knowledge. Unconscious knowledge is reflected in
the intuitive judgement or ‘gut feeling’ that mother-tongue speakers have
about what is ‘right’ and what is ‘wrong’ with the pronunciation of strings
of sounds in their language. This is evidenced by the fact that you use the
sounds of your language correctly – if you didn’t use these sounds correctly,
no one else who speaks your language would understand you when you spoke.
Other examples of unconscious phonological knowledge include the ability
to recognise a foreign accent, to make up new words and to know what is or
is not a sound in one’s language.

Morphology deals with the internal structure of words and the rules that
are used to build words. Morphology is covered in more detail in LIN1501
on Grammatical Patterns and Concepts and in LIN3701 on Approaches in
Linguistics. Most of you would also have done some morphology at school
when being taught a second language. For example, in English you would have
been taught that -ed is the ending that indicates past tense in jumped, walked,
played etc. In Zulu, for example, you would have been taught that um(u)- is
the class prefix for nouns like umuntu (‘person’) and umfazi (‘woman’). This
knowledge, being taught as a second language, would have been conscious
knowledge. The unconscious knowledge you have of your mother tongue is
even more interesting. For example, if you are English speaking you would
not have had to be told that plurals in English are formed by adding an -s to
the end of the noun as in tables, chairs, etc. You would have done this many
times without thinking about it, or even realising it. Morphemes are the
most basic units of meaning in a language. Stems like table, chair, walk etc
are morphemes and so are the meaningful prefixes like pre- and un- and word
endings like -ed and -s that we use to construct words.

Syntax deals with the internal structure of sentences and how the parts of a
sentence are related to each other. Every mother-tongue speaker knows which
words may occur where in a sentence. For instance, even a 5-year-old will
tell you that *Boy the ate his food is a ‘bad’ sentence. She may not be able to
explain why this sentence is ungrammatical, but she has unconscious syntac-
tic knowledge of her mother tongue that enables her to make that intuitive
judgement. A Zulu speaker knows that the word ukulala, which means ‘sleep’,

vi
 can be used as a verb or a noun. An English speaker will know that a verb
like break must be followed by an object, for example Susan breaks dishes.
The relationship between the transitive verb break and an obligatory object
is part of speakers’ unconscious syntactic knowledge and children have this
knowledge by the time they have acquired their mother tongue. Syntax is
studied in more detail in LIN1501 on Grammatical Patterns and Concepts,
and in LIN3701 on Approaches in Linguistics.

Semantics is concerned with the meaning of words and sentences. Speakers

of a language can produce and understand meaningful sentences seemingly
without any effort. This requires sophisticated semantic knowledge. Every
speaker knows the meaning of thousands of different words. The assumption
made in semantics is that this knowledge of word meanings is organised in
the mind in the form of a mental dictionary which represents words and their
meanings. For example, if told that somebody is an actress, we can immedi-
ately infer the person is human and female. Children also know the semantic
relationships between various words, for example whether they are synonyms
(words with similar meanings) like mother and mommy or antonyms (oppo-
sites) like happy and sad. They know how word meanings may be combined
to form sentence meanings and the restrictions that apply. Children recognise
the difference between semantically acceptable sentences and sentences which
are contradictory or meaningless. They will, for instance, recognise that the
sentence *My uncle is pregnant is not semantically acceptable, while the sen-
tence My aunt is pregnant is acceptable. Semantics is covered in LIN3701 on
approaches in Linguistics.

Pragmatics deals with the use of words, phrases and sentences in the actual
context of discourse. It is concerned with the appropriate use of language. You
would, for instance, not greet your professor with a Howzit bra, but rather
address him or her respectfully with a Good morning, professor. In the context
of a meeting between a student and the minister of education, an informal
manner is inappropriate – unless of course you know him personally! When
children acquire their mother tongue they need to learn the appropriate use
of language in different contexts.

What are the outcomes of this course?

As far as knowledge is concerned, in this course you will learn to:
 trace the course of first language development in children
 demonstrate an understanding of the relationship between language and the mind/
brain
 identify optimal conditions for language acquisition
 identify the consequences of non-optimal language acquisition
 identify universal patterns and individual differences in language acquisition
 explain the role of the caretaker in children’s language development
 identify similarities between language development in children and the evolution
of language in the human species.

LIN2601/1/2017-2023 vii
 You will also gain the following skills:
 to use appropriate tools to analyse actual language utterances and identify the
developmental stage of a child on the basis of a speech sample
 to apply your knowledge of first language development in order to interact ef-
fectively with babies and small children in a way that will facilitate and enrich
their linguistic development.

In addition to gaining knowledge and skills we hope that we can change some of
your existing attitudes and enable you to:
 use insights from language acquisition to develop an awareness of linguistic com-
plexity and human language competence
 develop an understanding of why and how language acquisition is studied and
an appreciation of the field, its concerns and its contribution
 develop understanding, sensitivity and tolerance for children with disabilities
and the way in which language acquisition differs in these children.

How do I go about studying this course?

Your study material for this module consists of this Study Guide (there is no
prescribed book). During the course you will be improving your ability to extract,
interpret and evaluate information, practising your reading and writing skills and
training yourself to think like a linguist!
We believe that the best way of learning is by doing. In other words, applying your
knowledge, developing skills and making connections between what you learn in
textbooks and what you observe around you in your everyday interactions with
others. By doing linguistics you will be able to master the theoretical concepts
presented in your study material. At the same time we will also try to get you to
think analytically and critically about language and to express your knowledge and
ideas about language in a scientific way.

How do I use this Study Guide?

The Study Guide is your primary source of information, so always start a study
unit by reading through the first page and the introduction. The first page of every
study unit is a plan of what the unit is about and provides important information
to guide your reading and learning.
The outcomes for each study unit are listed at the beginning of the study unit.
These are important ‘signposts’ as they point you to important sections of the work
and indicate what is expected of you. Come back to these outcomes after you have
worked through the study unit and make sure you have achieved each outcome.
You will also find a box that contains important terms and concepts. These terms
are listed in the order in which they are mentioned in the study unit and will give
you an idea of the kinds of topics and the central concepts that you will encounter
in the study unit. You need to make sure that you understand each term fully and
can give a definition of each term by the time you have finished working through
the study unit.

viii
 For example:

language acquisition CHILDES

developmental longitudinal study
psycholinguistics cross-sectional study
research question

The Preview box on the first page of each study unit gives you a short preview of
what the study unit is all about. Read through the preview so that you know what
to expect and can start thinking about some of the ideas that you will be reading
about. An example of a Preview box is given below:

Preview
In this study unit we introduce you to the process of language acquisition – how
children acquire their first language. We look at several reasons why we should
study language acquisition, including the fact that it teaches us what to expect
during a child’s crucial first few years and enables us to interact effectively with
young children and meet their language needs. We also take a look at several
different methods that researchers can use to study children’s language. We
find out why the common beliefs that language is acquired by imitation or
that it is taught by parents are misconceptions that seriously underestimate
the complexity of the language acquisition task. Finally, we include an overview
of some of the controversies and theoretical debates that exist between various
researchers and at how the focus of interest has changed over the last 60 years
in the field of language acquisition research.

After the preview you will find the main information you need to work through.
Look out for icons like the ones below:

The pencil icon pictured here is an indication that there is a task to complete. These
tasks give you an opportunity to test yourself and your understanding of what you
have read. Space is provided in the box for you to write your answers. We suggest
that you do these tasks as you come across them so that you familiarise yourself
with one concept before moving on to the next.
Feedback and/or model answers are given directly after each task, in the same box,
so remember to cover up the answers while you complete the task! Once you have
done the task, read through the feedback and correct your own work. If your answers
are correct or you have done what the feedback suggests, continue with the following
section. However, if some of your answers were incorrect or inappropriate, you should
go back to the relevant study material and reread it to see where you went wrong.
The left-hand margin has been used primarily to define difficult non-linguistic terms
that you may not be familiar with. We strongly suggest that you keep a dictionary

LIN2601/1/2017-2023 ix
 next to you while you study in case you come across any other unfamiliar words as
you read. Always look up unfamiliar words, and then use the left-hand margin to
write the definition as we have done. You will also find the light bulb icon below at
the end of each study unit, together with an activity that will help you to build your
academic vocabulary. Academic vocabulary refers to words that are not specifically
linguistic terms, but are more general words that are important at tertiary level and
that will come up again and again throughout your university studies. These tasks
are useful even though you will not be tested on them, and we strongly suggest
that you make an effort to work through them, especially if English is not your
mother tongue.

Academic vocabulary building

There are also opportunities for accessing the internet to watch

relevant videos or read articles. These are indicated by the
icon alongside.

One final icon is the pointing hand, which points you to optional
Further reading (a list of books or other source material for
students who want to find out more about a particular topic).

We hope that you enjoy the course and that it will enrich your understanding and
appreciation of the way children succeed in the incredible task of learning to use
language. Good luck with your studies!

x
 Study Unit 1 Study unit 1
Language and the brain
‘Biology gives you a brain. Life turns it into a mind.’
2

Jeffrey Eugenides (American novelist 1960–)

OUTCOMES
After you have studied this study unit you should be able to
 explain the relationship between the brain and language and why we study this
relationship
 briefly compare the disciplines of linguistics, psycholinguistics and neurolinguistics
 briefly describe the ways in which the relationship between the brain and lan-
guage is investigated
 identify the parts of the brain that are involved in specific language components
 identify the linguistic characteristics of two language disorders resulting from
brain damage
 provide definitions for the following important terms:

psycholinguistics lateralisation
neurolinguistics localisationists
neuron antilocalisationists
cerebral hemispheres aphasia
corpus callosum Broca’s aphasia (agrammatism)
contralateral control Wernicke’s aphasia (jargon aphasia)
localisation

Preview
We often take our ability to use language for granted, but when we realise
how complex language is and how intricate brain structures are, we are faced
with fascinating questions such as: What is the relationship between the brain
and language? Which disciplines study the relationship between language and
the brain? Where in the brain is language located? and How do we investigate
the language functions of the brain? In this study unit we will address these
questions and explore in some detail how damage to different parts of the brain
results in different language disorders.

LIN2601/1/2017-2023 1
 Introduction
This module is concerned with language acquisition – how young children learn to
speak their mother tongues. However, in order to understand what makes language
acquisition possible, we need to understand the relationship between language and
the brain.
In section 1 we will look at three interrelated disciplines that study language and
A discipline is a the brain: linguistics, neurolinguistics and psycholinguistics. A brief description
field of study. of each discipline will be given and then we will look at how these disciplines are
related to each other.
In section 2 we will briefly examine the structure of the brain in order to better
understand the relationship between language and the brain and to see how this
relationship affects language use and language acquisition.
In section 3 we will examine whether specific areas in the brain are involved in
language.
In section 4 we will discuss some of the methods used to investigate the brain and,
specifically, those used to investigate language areas in the brain.
In sections 5 and 6 we will take a closer look at Broca’s and Wernicke’s aphasia and
in section 7 we will summarise this study unit and see whether we can answer the
questions posed in the preview.

1 The disciplines that study language and the brain

Have you considered the fact that different disciplines can study the same phenomena?
Phenomena In this study unit we are interested in two phenomena, namely, the brain and
are things or
events that we language. You will discover that several disciplines study the same phenomena, but
observe and may from different perspectives.
want to describe
within a scientific For the purposes of this module, however, we will discuss only three disciplines:
discipline. linguistics, psycholinguistics and neurolinguistics. Note, there are many other
(Phenomena disciplines that study these two phenomena, but it would be impossible to discuss
is the plural of
phenomenon.) them all in one study unit. It is, however, important that you are aware that they
exist. These disciplines are psychology, neurology, cognitive psychology, education,
philosophy, computer science and artificial intelligence, to name but a few.

1.1 Linguistics
Linguistics studies the structure and function of human language. Every human
being knows at least one language. This knowledge of language is largely unconscious
and consists of knowledge about the sounds, words, sentences, meanings and the
appropriate use of language. These types of knowledge make up a person’s internal
grammar. Our internal grammar enables us to make intuitive judgements about our
mother tongue. For instance, a mother-tongue speaker of English will intuitively
know that Man the kicked ball the is a ‘bad’ or ungrammatical sentence, while The
man kicked the ball is a grammatical sentence. Linguistics attempts to describe this
intuitive, unconscious knowledge. As we saw in the introduction, Linguistics has
the following five subdomains: phonology, morphology, syntax, semantics and

2
 pragmatics (see the Introduction for brief descriptions of each subdomain). Each of
these subdomains studies a particular component of a person’s internal grammar.

1.2 Psycholinguistics
Cognitive science is the overarching discipline that investigates the mental processes
Cognition refers and operations that are involved in cognition. Psycholinguistics overlaps with
to our ability to
use our minds cognitive science, but has a narrower focus in that it studies only the mental processes
for attention, and skills involved in language, such as speech production and comprehension,
language, reading and writing. Within psycholinguistics there are several branches or
memory, subdisciplines, i.e. specialised fields that focus on a specific aspect of the mind–
perception,
problem solving, language relationship. The branch of psycholinguistics that is of interest to us here
creativity, is developmental psycholinguistics, where scientists study how children’s early
thinking and language develops through various stages as well as the later language developments
knowing.
that occur during the school years and in adult life. Developmental psycholinguistics,
in particular the study of children’s language acquisition, will be discussed in study
units 2 to 5. This branch of psycholinguistics overlaps with developmental
psychology, which examines the biological, social, emotional and cognitive changes
that occur as children develop into mature adults.

1.3 Neurolinguistics
Neurolinguistics is the scientific discipline that examines the relationship between
language and the brain. It looks at the biological basis of language, i.e. which brain
mechanisms underlie our ability to acquire and use language. Neurolinguistics is therefore
the study of how the structure and functioning of the brain enables us to possess and
use language. This discipline combines neurology (the study of the brain) and linguistics
(the study of language). One specific focus of this discipline is the relationship between
brain damage and speech and language deficits (see section 4.4 below).

Neurolinguists ask questions such as: How does the nervous system function to
produce and understand speech and language?; Where in the brain are speech and
language located?; What happens to language ability when particular areas of the
brain are damaged? and Are the components of language – phonology, morphology,
syntax, semantics – represented in distinct areas of the brain? We will be looking
briefly at some of these questions in this study unit.

You will have noticed that both psycholinguistics and neurolinguistics are
interdisciplinary undertakings. By interdisciplinary study we mean that several
disciplines are involved in studying the same phenomenon or phenomena. Knowledge
and methodology from all the disciplines involved are used to investigate the
relationship between language and the brain. In the case of neurolinguistics it is
neurology and linguistics that are involved, while in the case of psycholinguistics
psychology and linguistics are involved.

LIN2601/1/2017-2023 3
 Task 1.1
We have seen that different disciplines can study the same phenomena. Scientists from
other disciplines, such as psychology, neurology, speech therapy and cognitive science,
also study language and/or the brain, although their focus areas are somewhat differ-
ent. Try and match the disciplines on the left-hand side to their definitions on the right-
hand side by drawing lines from one column to the other.

DISCIPLINES DEFINITIONS

1. Developmental psychology A. is the scientific study of human language

2. Neurolinguistics B. studies children’s language as it develops

through various stages

3. Psycholinguistics C. investigates what mental processes are

involved in cognition

4. Developmental psycholinguistics D. is the study of how humans grow and

change across their lifespan

5. Linguistics E. examines the relationship between language

and the mind

6. Cognitive science F. studies the relationship between language

and the brain

Feedback
The correct matches are: 1 and D, 2 and F, 3 and E, 4 and B, 5 and A, 6 and C.
Have you noticed how similar the definitions for psycholinguistics and neurolinguistics
are? Psycholinguistics examines the relationship between language and the mind.
Neurolinguistics examines the relationship between language and the brain. Note that
the only difference between these two definitions is the last word: ‘mind’/‘brain’. Mind
and brain are two sides of the same coin: mind has to do with cognitive functions, whilst
brain refers to the biological structure that makes cognitive functions possible.

2 The structure of the brain

In this section you will come across some rather daunting terminology concerning
the structure of the brain. However, you need not memorise all the terms. You should
regard this section as a reference section that you can refer back to when reading
the other study units. All the important terms that you need to know and be able
to define are in the box on page 1.
The brain has a very complex structure and in the limited time available to us, we
cannot study the brain in a comprehensive or very detailed way. We can, however, gain

4
 some basic idea of the brain structures that affect language. It is important to realise
that we are dealing with different parts of the brain that interact and communicate
with each other, and with different parts of the body. Let’s start with the most basic
unit, the neuron.
Neurons are the brain cells that make up the building blocks of the nervous system.
They are specialised to transmit information. Every neuron consists of a cell body
and at least two extensions that receive and transmit impulses to other neurons. A
neuron can have connections with many other neurons and can pass on information
to other neurons using electrochemical processes.
Convoluted
means highly The brain is made up of white and grey matter, with the grey matter on the outside
folded. and the white matter on the inside. The grey matter is also called the cortex. The
cortex is a convoluted structure about 6 mm thick and is packed with approximately
ten billion neurons. The cortex is really ‘the brain of the brain’ – it integrates all
the functions of the nervous system. It is the decision-making organ of the body,
because it receives and transmits messages to and from all the sensory organs and
controls all voluntary actions.

Figure 1.1 Diagram of the brain1

The brain has three main parts: the brain stem, the cerebellum and the cerebrum.
The brain stem is the structure at the base of the brain that controls vital functions
such as respiration and heartbeat, as well as auditory input. In front of the brain
stem is the cerebellum, which regulates and maintains functions such as balance
and coordination of body movements, including those of the speech organs such as
the tongue, lips and vocal cords. The cerebrum is situated above the much smaller

1 Th is fi le is licensed under the Creative Commons Attribution-Share Alike 4.0 International licence. Th is image has been released as part of an
open knowledge project by Cancer Research UK. If reused, attribute to Cancer Research UK/Wikimedia Commons.

LIN2601/1/2017-2023 5
 cerebellum and consists of two asymmetrical halves known as cerebral hemispheres.
These two hemispheres are linked by a bundle of nerve fibres called the corpus
callosum, which acts as a bridge between the two hemispheres. This is a very
important part of the brain, because it allows the two hemispheres to communicate
with each other.

In each of the hemispheres the following four lobes can be identified:

 the temporal lobe
 the frontal lobe
 the occipital lobe
 the parietal lobe

In the following diagrams you will note the position of the different lobes. The deep
cleft separating the temporal and frontal lobes is known as the Sylvian fissure. It is
also important to note the position of the motor cortex, because it is this area that
controls our muscles. In other words, if this area of your brain is injured, you will
not be able to move your body or any body parts such as the lips, jaw and tongue
and will thus not be able to speak.

Figure 1.2 The four lobes in each hemisphere2

2 Henry Vandyke Carter [Public domain], via Wikimedia Commons. Th is image is in the public domain because it is a mere mechanical scan or
photocopy of a public domain original, or – from the available evidence – is so similar to such a scan or photocopy that no copyright protec-
tion can be expected to arise.

6
 Figure 1.3 The left hemisphere of the brain3

An important fact about our brains is that the two hemispheres control opposite
sides of our bodies. This is called contralateral control and it means that the left
hemisphere controls the right side of the body, while the right hemisphere controls
the left side of the body. In other words, if you move your right hand the message was
sent by your left hemisphere, and if you move your left hand the message was sent
by your right hemisphere. If someone suddenly loses the use of his right hand after
a brain injury, we know that a certain area of his left hemisphere has been damaged.

3 Th is work is licensed under the Creative Commons Attribution 3.0 Unported licence. To view a copy of this licence, visit http://creativecom-
mons.org/licenses/by/3.0/ or send a letter to Creative Commons, PO Box 1866, Mountain View, CA 94042, USA. By OpenStax College [CC
BY 3.0 (http://creativecommons.org/licenses/by/3.0)], via Wikimedia Commons.

LIN2601/1/2017-2023 7
 Figure
g 1.4 The two hemispheres
p as seen from above4

Task 1.2

Complete the following statements:

1. Neurons are the of the nervous system.
2. Every neuron consists of a and at least two
extensions.
3. The consists of two hemispheres.
4. The bridge between the two hemispheres is called
5. The grey matter of the brain is also called
6. The area of the brain that controls muscle movement of the face, tongue and
larynx is
7. Contralateral control means that and

Feedback
Neurons are the building blocks of the nervous system.
Every neuron consists of a cell body and at least two extensions.
The cerebrum consists of two hemispheres.

4 Th is fi le is licensed under the Creative Commons Attribution-Share Alike 3.0 Unported licence.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2
or any later version published by the Free Software Foundation; with no invariant sections, no front-cover texts, and no back-cover texts. A
copy of the licence is included in the section entitled GNU Free Documentation License.
.

8
 The bridge between the two hemispheres is called the corpus callosum.
The grey matter of the brain is also called the cortex.
The area of the brain that controls muscle movement of the face, tongue and larynx is
the motor cortex.
Contralateral control means that the left hemisphere controls the right side of the body
and the right hemisphere controls the left side of the body.

3 Language areas in the brain

In this section we want to find out which part of the brain (if any) is most involved in
language. Many scientists have tried to find the exact location of different cognitive
functions, including the language faculty, in the brain. This attempt to localise
specific cognitive functions is known as localisation.
Localisation and lateralisation are two related concepts. Lateralisation is a process
whereby one of the two hemispheres becomes dominant for certain cognitive
functions, and becomes responsible for handedness (left-handedness or right-
handedness). Lateralisation is an important process, because we want to know
whether one hemisphere becomes dominant for language and if so, whether we can
locate particular language areas in that hemisphere.

Task 1.3

Read the following extract from Akmajian, Demers, Farmer and Harnish (1990:440–441)
and then answer the questions that follow:
For over a century scholars have debated the question of speech and language
localization within the brain. In the 1860s, scientists known as localizationists
speculated that the functioning of specific regions in the brain was responsible
for language. Antilocalizationists argued that speech and language were the
consequence of the brain functioning as a whole.
In 1861, Paul Broca … described … a patient who in life had had extreme difficulty
in producing speech. Later, at autopsy, the patient was found to have damage in
the posterior inferior part of the frontal lobe in the left cerebral hemisphere, now
known as Broca’s area.
In 1874, Carl Wernicke published a monograph describing patients with speech com-
prehension deficits who had damage (lesions) outside Broca’s area, in the left
posterior temporal lobe.
Today scientists agree that specific neuroanatomic structures, generally of the left
hemisphere, are vital for speech and language, but debate continues as to which
structures are committed to the various linguistic capacities.

1. What do localisationists claim?

2. What do the antilocalisationists claim?

LIN2601/1/2017-2023 9
 3. What kind of deficit is caused by damage to Broca’s area?

4. What kind of deficit is caused by damage to Wernicke’s area?

5. Which hemisphere is vital for speech and language?

Feedback
You will have noticed that the localisationists and antilocalisationists hold opposing
views regarding the location of language. The localisationists claim that language
functions are located in specific areas of the brain, while the antilocalisationists claim
that the whole brain is responsible for speech and language. Both Broca and Wernicke
were localisationists. Damage to Broca’s area results in a speech production deficit.
Damage to Wernicke’s area results in a speech comprehension deficit. We will look at
these deficits more closely in later sections. From the above reading it is clear that the
left hemisphere is the dominant hemisphere for language. Berko Gleason (2005:16)
explains that about 85% of people are right-handed and almost all right-handed peo-
ple have their language functions represented in the left hemisphere. Studies such as
those of Knecht et al (2000) show that about half of left-handed people also have their
language functions lateralised for language in the left hemisphere. However, as you will
discover below, the left hemisphere cannot function on its own to produce language
and speech – the right hemisphere does play a part. Remember what links the two
hemispheres? The corpus callosum acts as a bridge between the hemispheres and
enables them to communicate with each other.

Since the 1990s, it has been recognised that both hemispheres contribute to many
Something is aspects of language and the interaction between hemispheres is now starting to
bilateral when receive more focus. Focus has therefore shifted to the brain’s integrating (rather than
it involves both
sides. lateralised) abilities. Research has shown that there are several language activities
(such as reading and speech comprehension) that involve both hemispheres (Cook
2002; Mundale 2002). Although speech production is strongly lateralised to the left
hemisphere, language understanding is more of a bilateral activity, with the right
hemisphere contributing by processing intonation, word connotations, metaphor
and humour (Cook 2002:171).
Berko Gleason (2005:16) also reports on gender differences related to language
processing: females have been shown to activate areas in both hemispheres while
processing speech sounds, whereas males use a much more restricted section of the
left hemisphere. It is also important to realise that there is considerable variability
in the way different individuals represent language in their brains. This is partly
due to the fact that our brains are plastic, which means that the brain can change
over the course of our lives. The brain can reorganise its neural pathways and make
new connections between neurons as we age, as we learn new things or in response
to a brain injury in which certain brain functions are lost.

10
 You may want to watch a short video entitled ‘Your brain on language’ at
https://www.youtube.com/watch?v=aUFdUiJXS4E

In the following section we will look at some of the techniques that can be used for
studying language and the brain.

4 Investigating the brain

There are various ways in which we can investigate the brain. One way would be to
‘open up’ someone’s brain while they are alive. (It would be impossible to see how
language functions in a corpse.) However, to open up someone’s brain when they’re
alive and healthy is risky and highly unethical! Therefore, scientists have devised
different methods of studying brain functions without harming the subjects. There are
at least four ways in which we can study brain function and hemispheric dominance.
These include sodium amytal tests, dichotic listening tests, brain-imaging techniques,
and aphasia studies. Each of these will be discussed in a subsection below.

4.1 Sodium amytal test

In order to determine which hemisphere is dominant for language, a test called the
sodium amytal test is used on some patients before brain surgery. Performing the
sodium amytal test before brain surgery allows doctors to mimic the effect that the
surgery might have (Troster & Parsons 2006:500). An anaesthetic, sodium amytal,
is injected into one of the two carotid arteries in the neck. Carotid arteries supply
the brain with blood. The left carotid artery leads to the left hemisphere, and the
right carotid artery leads to the right hemisphere.
Obler and Gjerlow (1999:28–29) explain that ‘if the drug is delivered to the language
side of the brain, a temporary paralysis of language function is experienced. The
patient cannot speak at all for several minutes and in the minutes after that, language
sounds aphasic’. They also claim that the results of this test show that 95% of
right-handers and 70% of left-handers experience temporary paralysis of language
function when the anaesthetic is injected into the left carotid artery leading to the
left hemisphere. This indicates that language is located in the left hemisphere for
all these people.

Listen to a short explanation of this test at

http://healthpad.net/publication/1585/what-is-a-wada-test/

4.2 Dichotic listening tests

A dichotic listening test is a way of determining which of the two hemispheres is

dominant for language. Hudson (2000:155) describes the test as follows:
In dichotic listening, subjects listen to a two-track stereo recording, one track
per ear, of a random list of syllables, numbers, or words, hearing different
signals on each track, so that different syllables, numbers, or words, are heard
simultaneously in the two ears. For most subjects, dichotic listening of such
pairs reveals a right ear advantage: the signal heard in the right ear tends to be

LIN2601/1/2017-2023 11
 recalled. Right-handers have shown about 80% left-dominance for language
in dichotic listening.

These results of the dichotic listening tests can be explained as follows: the fact that
the signals heard by the right ear are remembered more often means that the left
hemisphere is the one processing the signal. You will recall that contralateral control
means that the left hemisphere controls the right side of the body and the right
hemisphere controls the left side of the body. If the signals received in the right ear
are processed better, we can assume that the left hemisphere is doing the processing,
i.e. it is the language-dominant hemisphere.

Hudson (2000:157) claims that ‘dichotic listening in children shows that hemisphere
dominance for language is typically well established by about five years of age’.
This tells us something about lateralisation. You will recall from section 3 above
that lateralisation entails a process whereby one of the two hemispheres becomes
dominant for language and handedness. If Hudson’s claim is correct, we can assume
that lateralisation has taken place by the age of five years.

4.3 Brain-imaging techniques

Since the 1970s, technological innovations in medicine have enabled us to display
images of the live brain. These have provided important new and detailed information
about brain function. Brain-imaging techniques are an ideal way of investigating
structure–function relationships within the brain as studies can be performed on
normal subjects as well as brain-damaged patients and the techniques can explore
activated regions across the whole brain during the performance of language-related
tasks (Scott & Wise 2003:9).
PET scans (positron emission tomography) are a kind of X-ray of the brain, allowing
us to observe how areas of the brain operate. At the start of a scan, a harmless
radioactive molecule is injected into the blood. The subject lies on a table with their
head in the scanner. As Obler and Gjerlow (1999:9–10) explain:
In this sophisticated form of dynamic ‘X-ray’, a picture of the subject’s brain
can be seen on a computer screen. Different areas of the brain appear to light
up while a healthy person is undertaking certain language tasks whereas oth-
ers light up less or not at all, depending on the amount of activity in a given
brain area.

An fMRI (functional magnetic resonance imaging) is a non-invasive way of

obtaining images of the brain (or other tissues) by observing how magnetic fields
pass through it. In an fMRI experiment the subject’s whole body is inserted into a
scanner and changes in the blood flow in the brain are measured. Blood oxygen level
is assumed to be related to brain activity, and comparisons can therefore be made
between fluid patterns as the subject engages in various language-related tasks. This
technique is allowing gradually more detailed information to emerge about exactly
which areas of the brain relate to which aspects of language processing. To see how
fMRI shows hearing words, seeing words, speaking words and generating words go
to: http://cmudesignfiction.files.wordpress.com/2013/04/image011.gif

12
 Event-Related Potentials (ERPs) are another non-invasive way of measuring what
happens in the brain during language processing in babies, children or adults. These
use a cap fitted with electrodes to record voltage fluctuations on the surface of the
scalp in response to an external stimulus such as listening to words or sounds (Rispens
& Krikhaar 2010). ERPs reflect how long (in milliseconds) it takes us to process
a stimulus, and whether the stimulus is expected or unexpected. For example, the
technique can be used to measure the response to possible and implausible sentences,
with ERP measurements getting stronger as sentences become less and less plausible,
as in the three sentences below (Rispens & Krikhaar 2010:100):
1. The pizza was too hot to eat. (no negative deflection)
2. The pizza was too hot to drink. (weak negative deflection)
3. The pizza was too hot to read. (strong negative deflection)

4.4 Aphasia studies

Aphasia can be defined as a language impairment resulting from a brain lesion
caused by a stroke, a tumour, an infection or a head injury. There are many people
who suffer some form of damage to the left side of their brains and as a result of
this injury lose part of their ability to use or understand speech. In fact, aphasia
affects about 1 in 250 people (Mirman et al 2015). Most commonly this damage
is caused by a stroke – a blood clot prevents blood from flowing into a certain area
of the brain which then swells up and disrupts the functioning of the surrounding
areas of the brain. Stroke patients show fairly serious symptoms during the first few
days, but typically show some improvement due to the recovery of areas surrounding
the affected area.
By studying language in people with aphasia, we can try to improve our clinical
understanding of aphasia and get new insights into the various language centres in the
brain. The areas that have been identified so far are Broca’s area and Wernicke’s area.

Task 1.4

Read the following paragraph summarising the ways in which the brain can be investi-
gated. Comment critically on the structure of the paragraph and the writing style, and
suggest ways in which the paragraph could be improved:
1. The sodium amytal test is used on some patients before the brain surgery. This test
is done to determine which hemisphere is dominant for language in patients who
must undergo brain surgery. The left carotid artery leads to the left hemisphere,
and the right carotid artery leads to the right hemisphere.
2. Dicotic listening test is another way of determining which of the two hemispheres
is dominant for language in other words which hemisphere contains the language
centre. The fact that the signals heard by the right ear are remembered more often
means that the left hemisphere is the one processing the signal.
3. Aphasias is a language disorder that follows a brain lesion caused by a stroke,
a tumour, an infection or a head injury. This damage is caused by a blood clot
that prevents blood from flowing into a certain area of the brain
.

LIN2601/1/2017-2023 13
 Feedback
In this course you will not be expected to write essays, but you will be expected to
write a coherent paragraph (10–20 lines) in which you describe a process, present an
argument, discuss a theoretical viewpoint, etc. Writing paragraphs is an important skill
that you will need in assignments and the exam. Some of the problems you might have
picked up with the paragraph on investigating the brain are listed below:

 There is no introductory sentence that overviews the whole paragraph. Instead the
writer jumps straight into the content.
 There is repetition – the writer tells us twice that the sodium amytal test is used on
patients before brain surgery.
 Irrelevant information is included. The sentence about the carotid arteries does not
relate to the topic – ways of investigating the brain.
 Important information is omitted. There is no mention of brain-imaging techniques.
 There are three paragraphs instead of one coherent whole.
 Terms are spelled incorrectly – the word dichotic should have an h in it.
 There are some grammatical errors such as ‘Aphasias is ….’ (You will not be penalised
if your English is not perfect, as long as your language is clear and comprehensible.)
 The section on aphasia does not explain how the study of aphasia can tell us more
about the brain (i.e. it does not relate directly back to the question). It is important to
add that by studying the language of patients with damage to different areas of the
brain, we can identify the specific functions of these areas with respect to language.

Here is a better paragraph:

There are four ways in which we can investigate brain function and hemispheric
dominance. The first way is through a sodium amytal test which temporarily
paralyses one hemisphere of the brain and can tell us whether language is
located in the left or right hemisphere for a particular patient. Dichotic listen-
ing tests are another way of determining which of the two hemispheres is
dominant for language. In dichotic listening, subjects listen to a two-track
stereo recording, so that different syllables, numbers, or words, are heard in
each ear simultaneously. For most subjects, the signal heard in the right ear
tends to be recalled, showing that the left hemisphere is the one process-
ing the signal. Thirdly, sophisticated brain-imaging techniques like positron
emission tomography (PET), functional magnetic resonance imaging (fMRI)
and event-related potentials (ERPs) can provide visual images of the brain
and brain activity during language use. The fourth method is to study the
language of aphasic patients, i.e. patients who have a brain lesion resulting
from a stroke, a tumour, a gunshot wound, or an infection. By studying the
different kinds of language problems experienced by patients with different
damaged brain areas, we can find out the functions of various areas of the
brain such as Broca’s area and Wernicke’s area.

14
 The most important advice we can give you when writing a paragraph is to read the
question carefully. This means (a) identifying the topic (the content that you will have
to discuss) and (b) identifying the instructional key words (discuss, describe, explain,
etc.) Plan your paragraph for a few minutes before you start writing it, listing the main
ideas you will mention and the order in which these should come. Make sure these
ideas relate directly to the question. Start your paragraph with a topic sentence that
introduces the issues you will be discussing. Make sure you include definitions of all
key terms and that you spell them correctly. Make sure your ideas follow logically and
that you use linking words (and, however, in contrast, etc.) appropriately to indicate the
structure of your argument. Include examples to support the points you make. You will
get more marks for including original examples from your own language (with glosses)
than from simply repeating examples in the study material.

5 Broca’s aphasia
There are a few observations we can make about Broca’s area of the brain that seem
to indicate that it is associated with speech:
1. It is an area of the brain that is underdeveloped at birth and only reaches full
development when the child is about four years old.
2. It is an area of the brain that developed relatively late in the evolution of man.
3. It is an area of the brain that is underdeveloped in the brains of higher primates
such as chimpanzees.

5.1 Characteristics of Broca’s aphasia

Look at the following example of an utterance produced by a Broca’s aphasic (from
Goodglass 1976). What do you notice about this person’s use of language?
My uh mother died … uh … me … fi’teen. Uh, oh, I guess six month … my
mother pass away. An’ uh … an’en … uh … seventeen … seventeen … go … uh
high school. An uh … Christmas … well, uh, I uh … Pitt’burg.

If you have internet access, watch the YouTube video entitled Broca aphasia
on https://youtu.be/gocIUW3E-go in which a patient attempts to describe
what is happening in the ‘cookie-theft’ picture.

If you were able to watch the YouTube video, you would have noticed that Broca’s
aphasics talk slowly and with some difficulty as if they have trouble finding the
correct words. Severe cases of Broca’s aphasia show a great reduction in the flow of
their speech. Most Broca’s aphasics do not make use of intonation patterns or stress
patterns and their speech is therefore rather monotonous.

Note also that in this type of aphasia there are many words missing, including many of
the verbs. The speech of Broca’s aphasics also tends to lack function words, including
articles (e.g. this, that, the, a), prepositions, pronouns, auxiliaries, conjunctions, plural
markers and the possessive -s. These kinds of elements serve a grammatical function
and are characterised by their low semantic content (e.g. what does the mean?). Owing
to the loss of these grammatical words and morphemes this type of aphasic speech
is also referred to as agrammatism, a term meaning ‘without grammar’. The loss of
articles and other function words makes their speech seem like a telegram. However,
the basic structure of the sentence is usually preserved, and with a bit of practice a
listener can learn to add in the missing words when conversing with a Broca patient.

LIN2601/1/2017-2023 15
 Menn, Obler and Miceli (1989:168) quote the example of a patient who begins the
Red Riding Hood story as follows: Red Riding Hood. Pretty girl. Mother, grandmother
stick, uh wolf, um eat, um grandmother, uh wolves, wolf eat grandmothers. How would
you express the same message?

Mild forms of Broca’s aphasia consist only of the loss of the ability to embed
sentences. An embedded sentence is a sentence that is included in another, for example
in the sentence The man that kicked the dog came to my house, the main sentence is
The man came to my house and the embedded sentence is ((the man) kicked the dog).
A Broca’s aphasic might produce utterances such as The man is my boss and the man
brought me to the hospital instead of The man who brought me to the hospital is my
boss. Instead of embedding the sentence The man brought me to the hospital into the
first sentence (The man is my boss), the patient simply joins the two sentences with
the conjunction and. The lack of embedded sentences and the excessive use of the
word and to link sentences is typical of Broca aphasics’ speech.

Damage to Broca’s area seems to result mainly in problems with the production of
speech, including reading out loud. Patients typically read slowly and with many
hesitations and have particular difficulty reading grammatical elements such as
prepositions and articles. In some cases they are unable to read these words – i.e. they
will either pretend that the articles are not printed or, when forced to comment on
the articles, they will say something like One of those little words again! Menn et al
(1989:162–163) give the following example of an agrammatic Broca patient reading
aloud (the actual text is given first and the patient’s read-aloud version is below it):
I baked a basket of goodies. I want you to take them to your grandmother’s house
because she is sick in bed … At last she got to her grandmother’s house. She knocked
at the door, but there was no answer, so she walked into the house, and there she
found someone. (actual text)
I baked the basket of goodies. I want to y’ [long pause] to take them to her your
granma house because he, she is sick in bed … At last he, she got to her gramma,
granma house. She knocked at the door, but they was no uh answer, so uh she
walked, walked in, she walked in the door, and they here she found sometimes
someone. (patient’s version)

So far we’ve looked at English, but similar types of error occur in the speech of Broca
aphasics who speak other languages. For example, a Dutch patient describing the
cookie-theft picture gave the following explanation, lacking articles and pronouns
and joining sentences with and to avoid embedding:
En krukje bijna omgevallen. En binnen in pakt iets, koekjes of zo.
‘And [the] stool almost fell over. And inside in [the cupboard][he] takes some-
thing, cookies or something.’
(Menn et al 1989:271)

English and Dutch are not morphologically rich languages, so function words are
separate words that are often just omitted. In morphologically rich languages such
as the African languages we would expect that agrammatic patients will lose many
of the grammatical morphemes. In Swahili, Abuom and Bastiaanse (2012:15) found
that, as in English, agrammatism was associated with reduced spontaneous speech
output, slower speech rate, shorter utterances, limited use of embedded sentences
and large proportions of ungrammatical sentences. An unexpected finding was that

16
 grammatical morphemes were in fact seldom omitted, but they were frequently
substituted with incorrect morphemes. While agreement markers in agrammatic
Swahili patients tended to be correct, verb tense markers tended to be incorrect.
In milder cases of Broca’s aphasia where patients still have some speech left, they are
able to compensate for some of their problems in the following ways:
1. Using a series of simple sentences joined with and to compensate for the loss
of sentence embedding.
2. Using direct quotations since indirect speech requires the use of embedded
sentences. For example, in a story-completion test, patients were asked to com-
plete the following sentence: The children were being too noisy and mother was
annoyed, so she told … (them to be quiet). Because agrammatic patients cannot
use embedded sentences, they use direct quotations: The mother told them:
‘Little kids, be quiet man!’
3. Stringing nouns together in the same sequence as they would be in if there were
verbs and hoping that the listener is able to add the missing verbs.
4. Using gestures to compensate for their lack of language.

5.2 Summary
We have seen that when Broca’s area of the brain is damaged, the patient experiences
syntactic problems and produces a form of language in which most of the function
words are absent. This characteristic of this type of aphasic speech has earned it the
label agrammatism. Broca’s aphasics also tend to lose the ability to embed sentences.
Broca’s aphasics also experience phonological problems – they lose the intonation
and stress patterns in their speech.

Task 1.5

The following statements are incorrect. Use the spaces between statements to write
the correct versions of the statements.
1. Broca’s area lies in the temporal lobe of the right hemisphere.

2. All Broca’s aphasics have problems reading aloud.

3. Sentence embedding entails two sentences linked by the word and.

4. Broca’s aphasia is characterised only by the loss of articles.

5. Function words are words like nouns and verbs.

6. Broca’s aphasics use direct quotations because they can use embedded clauses.

LIN2601/1/2017-2023 17
 7. Broca’s aphasics do not use gestures.

8. Agrammatism means the overuse of function words that serve a grammatical function.

9. Broca’s aphasics compensate for the loss of nouns by stringing verbs together.

Feedback
Broca’s area lies in the frontal lobe of the left hemisphere. Only some Broca’s aphasics
have problems reading aloud. Sentence embedding does not entail two sentences
linked by the word ‘and’. Sentence embedding entails one sentence ‘embracing’ another
sentence, for example in ‘The girl who cries is her sister’, the sentence ‘The girl is her
sister’ embraces the sentence ‘who cries’. Broca’s aphasia is characterised by the loss
of grammatical elements, the loss of intonation and stress patterns and an inability to
use embedded sentences. Function words are not content words like nouns and verbs,
but are articles, prepositions, pronouns, auxiliaries and conjunctions. Broca’s aphasics
use direct quotations because they cannot use embedded clauses. Broca’s aphasics do
use gestures to compensate for their lack of language. Agrammatism means the loss
of the function words that serve a grammatical function. Broca’s aphasics compensate
for the loss of verbs by stringing nouns together.

6 Wernicke’s aphasia
In this section we will analyse the type of speech that is produced when the area of
the brain known as Wernicke’s area is damaged. Wernicke’s area lies at the upper
back part of the temporal lobe. There are a few observations we can make about
this part of the brain that seem to indicate that it must be associated with speech:
1. This is an area of the brain in which sensory data, particularly auditory data
(such as speech sounds) are processed.
2. It is also an area of the brain that developed relatively late in the evolution of
humans.
3. This area is also underdeveloped in the brains of higher primates.
Since Wernicke’s area lies so close to the sensory area of the brain (i.e. the part of the
brain that assimilates and interprets stimuli that the body receives from the outside
world) one would expect language comprehension problems to occur in these patients.

6.1 Characteristics of Wernicke’s aphasia

Look at the following utterance produced by a Wernicke’s aphasic and see if you
can identify some of the characteristics of this type of aphasia:
He wife saw the wonting to wofin to a house with the umbledor. Then he left the
wonding then he too to the womin and to the umbella upstairs …
(Goodglass & Kaplan 1972:59)

Or watch a YouTube clip of a patient with Wernicke’s aphasia on

https://youtu.be/dKTdMV6cOZw

18
 In severe cases of Wernicke’s aphasia like this it is impossible to understand what
the patient is saying because of the meaningless words that occur in the utterance.
Although Wernicke’s aphasics speak fluently and with well-preserved rhythm and
intonation, it is very difficult to understand what they are saying and they often
do not seem to understand what others are saying to them. Whereas Broca’s area
is associated with syntactic and phonological language abilities such as intonation,
embedding, function words and difficulty accessing verbs, Wernicke’s area seems
to be associated with the semantic and phonological components of language. Most
of these patients are not able to hear that they are making mistakes. The following
characteristics of Wernicke’s aphasia can be identified:
As you saw in the examples above, Wernicke’s aphasics tend to use formal,
technical terms and ‘made-up’ or nonsense words, also called neologisms. These
(new) combinations of speech sounds tend to follow the phonological rules of the
patient’s (original) language. Because of the high incidence of formal, scientific
jargon in their speech as well as the frequent occurrence of non-existent words or
neologisms (such as wofin) this type of aphasia is also known as jargon aphasia.
Weinstein, Lyerly, Cole and Ozer (1972:96) describe the jargon as consisting of
mispronounced words, neologisms, and standard English words put together in
meaningless sequence. As pointed out above, the patient seems unaware that he is
using these strange words. In fact, the more jargon a patient uses, the less it seems
to worry him. The jargon occurs in the writing of some patients while other patients
write the word correctly but then read it as a jargon word. It is as if words have lost
their meaning to these patients.
Weinstein et al (1972:99–100) give a very good example of jargon aphasia – the
patient, an American army sergeant, had a car accident in Germany (the neologisms
have been underlined):
Doctor: What is your main trouble?
Patient: Just my arms and my one leg. It has been ripped open and it’s getting
pretty good now.
Doctor: How did it come about?
Patient: They sent me back, sir, on my one arm. I had one arm here in
Koreejit and this one here on for a corgent. They took part of that on
the way back in (motioning to his cast) and my rove was renetted
up and we got a crane secker that’s where I got my injury in my head
and my base. Both of my eyes were blacked and this lady that
brought me back took it out. She was a captain, a nurse.
Note that the initial part of the patient’s sentences seem grammatical, while the
neologisms tend to appear later in the sentence. A similar pattern can be seen in
our first example: the noun phrases are always well structured and the first part of
the sentence (until the verb) seems correct. The neologisms tend to occur after the
verb. This seems to be a typical structure found in the language of many jargon
aphasics – the relatively well-formed sentence structure contrasts sharply with the
meaningless lexical material.
He wife saw the wonting to wofin to a house with the umbledor. Then he left the
wonding then he too to the womin and to the umbella upstairs …
(Goodglass & Kaplan 1972:59)

LIN2601/1/2017-2023 19
 Wernicke’s aphasics also tend to use indefinites such as thing, one, it, person, something,
anything, sometime, etc. The high incidence of these words (that have very little
meaning of their own) contributes to the difficulty that the listener has in trying to
understand the speech of these patients. Notice the use of indefinites such as here,
this one, that, it in the patient’s utterances in the example above. Interpreting the
patient’s utterances is extremely difficult as there is generally no preceding noun to
which these words refer. Since Wernicke’s aphasics are usually not aware of their
errors, they do not attempt to compensate for the shortcomings in their speech.
In a recent study that continues the localisationist tradition, Mirman et al (2015)
examined fMRI data from 99 people who had persistent language impairments
after a left-hemisphere stroke. They found further evidence that damage to different
areas of the brain resulted in different types of language problem: damage above
the Sylvian fissure was linked to phonological problems with speech production
(like saying ‘girappe’ for ‘giraffe’), while damage below the Sylvian fissure caused
phonological problems with speech recognition (like an inability to distinguish /b/
from /d). As expected, semantic production errors (like saying ‘zebra’ instead of
‘giraffe’) were linked to lesions around Wernicke’s area, while difficulty recognising
the semantic relationship between concepts (such as matching words to pictures) was
associated with damage to an area deep in the frontal lobe known as a ‘white matter
bottleneck’. The white matter bottleneck is a point of convergence among multiple
pathways in the brain, where a small amount of damage can have serious effects on
semantic processing, both visual and verbal. Mirman et al (2015:5) conclude that
production and comprehension, as well as semantic and phonological processing,
all occur in separate parts of the brain.

6.2 Summary
Wernicke’s area seems to be involved in the semantic integration of speech, and
damage to this area of the brain leads to speech that is fluent and well-structured but
often incomprehensible due to the number of neologisms and indefinites it contains.
Patients are typically unaware of the mistakes they are making. This suggests that
semantic information on the meaning of words and syntactic information on how
to combine them into sentences are stored in different places in the brain. We also
get the impression that the meaning of a word and the phonological information on
how to pronounce the word are not stored in the same place in the brain. It seems
that Broca’s area stores knowledge on how to structure simple sentences and that
Wernicke’s area links concepts to words. Thus in a Wernicke’s aphasic a concept is
incorrectly linked to various strings of sounds (resulting in neologisms) which are
then syntactically organised by the (intact) Broca’s area.

Task 1.6

Complete the following statements:

1. Wernicke’s area lies at the part of the
lobe.
2. Wernicke’s area is associated with the
component and the component of the grammar.

20
 3. Wernicke’s aphasia is characterised by fluent but
speech.
4. Wernicke’s aphasics tend to use such as
thing, anytime and something.
5. Jargon can be described as consisting of
and standard English words put together in
6. Neologisms tend to occur after the
7. Wernicke’s aphasics do not attempt to compensate for their mistakes, because
they
8. It is as if words have lost their to
Wernicke’s aphasics.
9. Wernicke’s aphasia indicates that information on the
of words and the information on how to
are stored in different places in the brain.
10. In Wernicke’s aphasia thoughts are
linked to strings of sounds.

Feedback
Wernicke’s area lies at the upper back part of the temporal lobe. Wernicke’s area is as-
sociated with the semantic component and the phonological component of the grammar.
Wernicke’s aphasia is characterised by fluent but incomprehensible speech. Wernicke’s
aphasics tend to use indefinites such as ‘thing’, ‘anytime’ and ‘something’. Jargon can
be described as consisting of mispronounced words, neologisms and standard English
words put together in meaningless sequence. Neologisms tend to occur after the verb.
Wernicke’s aphasics do not attempt to compensate for their mistakes, because they are
usually not aware of their errors. It is as if words have lost their meaning to Wernicke’s
aphasics. Wernicke’s aphasia indicates that information on the meaning of words and
the information on how to combine them into sentences are stored in different places
in the brain. In Wernicke’s aphasia thoughts are incorrectly linked to strings of sounds.

Task 1.7

Summarise your understanding of Broca’s and Wernicke’s aphasia by completing the

following table:

Type of aphasia Location in the Characteris- Linguis-

and alternative brain tics tic component(s)
name affected

Broca’s aphasia/ 1.
agrammatism
2.
3.
4.

Wernicke’s aphasia/ Wernicke’s 1. Fluent. Semantic and

area, temporal phonological
2
lobe of the left components
hemisphere 3

LIN2601/1/2017-2023 21
 Now use the information above to help you identify the type of aphasia in each of the
following examples, and justify your answer.
1. Pinker (1995:47) cites an example of a man who suffered a stroke at age 39:
‘Why, yes … Thursday, er, er, er, no, er, Friday … Bar-ba-ra … wife .. and, oh, car
… drive … purnpike … you know … rest and … TV’
2. ‘Boy, I’m sweating, I’m awful nervous, you know, once in a while I get caught up, I
can’t mention the tarripoi, a month ago, quite a little, I’ve done a lot well, I impose
a lot, while, on the other hand, you know what I mean, I have to run around, look
it over, trebbin and all that stuff’ (Pinker 1995:310).
3. Examiner: Tell me about how you got sick, what it was like.
Patient: It was, it was uh, we’ [well] I had a stroke an’ uh, it’s uh, um, when I
woke up it wa [was], I didn’ know my name, anything like that. An’ uh I can’t,
I can’t speak, my right side was blunt, lump [limp?] an’ I was crying uh any, any-
thing (Menn et al 1989:154).
4. Patient describing the ‘cookie-theft’ picture.
‘I can’t see well enough but I believe that either she or will have some food that’s
not good for you and she’s to get some for her, too … and that you get it there
because they shouldn’t go up there and get it unless you tell them that they could
have it’.
(Pinker 1995:312)

Feedback
The table you completed should have the following information:
Broca’s aphasia/agrammatism is characterised by:
1. Slow, hesitant speech,
2. the loss of the ability to embed sentences,
3. the loss of function words, and
4. the loss of intonation and stress patterns. Broca’s area is located in the frontal
lobe of the left hemisphere. The linguistic components associated with this
area are the syntactic and phonological components.
Wernicke’s aphasia/jargon aphasia is characterised by:
1. fluency,
2. the use of jargon, and
3. the use of indefinites. Wernicke’s area is located in the temporal lobe of the
left hemisphere. The linguistic components that are associated with this area
are the semantic and phonological components.

1. This is an example of Broca’s aphasia, because of the lack of articles, some miss-
ing verbs and the fact that it is not fluent.
2. This is an example of Wernicke’s aphasia, because of the fluency, the use of the
neologisms ‘tarripoi’ and ‘trebbin’ and the indefinite ‘all that stuff’.
3. This is an example of Broca’s aphasia, because of the lack of fluency, hesitations
and difficulty finding words. The patient uses ‘an’ (‘and’) frequently because he
cannot use embedded sentences.
4. This is an example of Wernicke’s aphasia, because of the fluency and the use of
indefinites like ‘she’, ‘her’, ‘some’, ‘it’ and ‘there’.

7 Summary
You will remember the questions that we posed in the preview were: What is the
relationship between the brain and language? Which disciplines study the relationship

22
 between the brain and language? Where in the brain is language located? and How
do we investigate language functions in the brain?
We now know that language functions are located primarily in the left hemisphere and
that disciplines such as linguistics, neurolinguistics and psycholinguistics are involved
in studying the relationship between language and the brain. Furthermore, we have
seen that there are various methods of studying language-related brain functions,
namely, by studying people who are aphasic or by making use of procedures such as
sodium amytal tests, dichotic listening tests and the various brain-imaging scans.
You will also be aware of some of the debates in neurolinguistics. You will recall that
early scientists claimed that language abilities were localised in the left hemisphere
while others viewed language as an activity that involved the whole brain. There is
still general agreement that Broca’s and Wernicke’s areas are the most important
centres for language, but as technology has become increasingly sophisticated and
precise, it is becoming clear that multiple different areas of both hemispheres of the
brain are involved in an interconnected way when we speak, listen, read and write.
In the following units you will need your knowledge of brain structures in order
to better understand language acquisition, when things go wrong in the brain and
how the human brain evolved.

Academic vocabulary building

Learning a few Latin and Greek prefixes will help you work out the meaning of many
technical terms. Here are some useful ones that we came across in this study unit:
uni- or mono- means ‘one’, as in words like monolingual, universal
di- or bi- mean ‘two’ as in words like bicycle, bilingual, dichotic listening
tri- means ‘three’ as in words like tricycle, triangle, triad, triplet etc
hemi- means ‘half’ as in words like hemisphere (half a sphere)
a- means ‘without’ or ‘lacking’ as in words like agrammatism (lacking gram-
mar) and asymmetry (lacking symmetry)

Use your knowledge of prefixes to work out the answers to these questions:
1. How many legs does a tripod have?
2. How many languages does a bilingual speak?
3. In a dichotic listening test sounds are presented to
ears simultaneously.
4. If -phasia is the Greek word for speech, then an aphasic person is a person
speech, i.e. who has suffered a loss of language
abilities.
5. When a voice is described as monotonous it has only
tone or pitch.

You should also know the following position words:

posterior – the back part, behind

LIN2601/1/2017-2023 23
 anterior – the front part, in front
inferior – the bottom or lower part
superior – the top or higher part

Further reading

Aphasia is a short, award-winning movie produced in 2010 by Carl McIntyre, a

former actor who developed aphasia and wanted to document his story. See www.
aphasiathemovie.com or watch the trailer on YouTube.
http://www.theaphasiacenter.com/aphasia-case-studies/
Akmajian, A, Demers, RA, Farmer, AK & Harnish, RM. 2010. Linguistics: an in-
troduction to language and communication. 6th edition. Cambridge, MA: MIT
Press. (Chapter 10 is on speech production and comprehension and Chapter
12 is on language and the brain.)
Crystal, D. 2010. The Cambridge encyclopedia of language. 3rd edition. Cambridge:
Cambridge University Press. (Chapter 46 is on language disability, including
aphasia.)
Fromkin, V, Rodman, R & Hyams, N. 2014. An introduction to language. 10th edi-
tion. Boston, MA: Wadsworth/Cengage Learning. (Chapter 10 deals with the
brain and language.)
Gardner, H. 1974. The shattered mind. New York: Vintage. (Contains case studies
on aphasics.)
Hale, S. 2002. The man who lost his language: a case of aphasia. United Kingdom:
Allen Lane. (This is a fascinating and readable personal account of a couple’s
experiences.)

Bibliography

Abuom, TO & Bastiaanse, R. 2012. Characteristics of Swahili–English bilingual

agrammatic spontaneous speech and the consequences for understanding
agrammatic aphasia. Journal of Neurolinguistics 25(4):1–18.
Akmajian, A, Demers, RA, Farmer, AK & Harnish, RM. 2010. Linguistics: an
introduction to language and communication. 6th edition. Cambridge, MA:
MIT Press.
Crystal, D. 1987. The Cambridge encyclopedia of language. Cambridge: Cambridge
University Press.

24
 Cook, ND. 2002. “Bihemispheric Language: How the two hemispheres collaborate
in the processing of language”. In Crow, TJ. 2002. The speciation of modern
Homo sapiens. Oxford: Oxford University Press.
Fromkin, V & Rodman, R. 2014. An introduction to language. 10th edition. Boston,
MA: Wadsworth/Cengage Learning.
Goodglass, H. 1976. Agrammatism, in Studies in neurolinguistics, Volume 1, edited
by H Whitaker & HA Whitaker. New York: Academic Press.
Goodglass, H & Kaplan, E. 1972. The assessment of aphasia and related disorders.
Philadelphia: Lea & Febiger.
Goodglass, H & Kaplan, E. 1983. The Boston Diagnostic Aphasia Examination.
Boston, MA: Lea & Febiger.
Hudson, G. 2000. Essential introductory linguistics. Oxford: Blackwell.
Knecht, S, Dräger, B, Deppe, M, Bobe, L, Lohmann, H, Flöel, A, et al. 2000.
Handedness and hemispheric language dominance in healthy humans. Brain:
A Journal of Neurology 123(12):2512–2518.
Menn, L, Obler, LK & Miceli, G (eds). 1989. Agrammatic aphasia: a cross-language
narrative sourcebook. Amsterdam: John Benjamins.
Mirman, D, Chen, Q, Zhang, Y, Wang, Z, Faseyitan, OK, Branch Coslett, H, et
al. 2015. Neural organization of spoken language revealed by lesion–symptom
mapping. Nature Communications 6. Art no. 6762 doi:10.1038/ncomms7762
Mundale, J. 2002. Concepts of localization: balkanization in the brain, in
Brain and mind. 3:313. doi:10.1023/A:1022912227833. Dordrecht: Kluwer
Academic Publishers.
Obler, LK & Gjerlow, K. 1999. Language and the brain. Cambridge: Cambridge
University Press.
Pinker, S. 1995. The language instinct: the new science of language and mind. New
York: Penguin.
Rispens, J & Krikhaar, E. 2010. Using event-related potentials in language acquisi-
tion research, in Experimental methods in language acquisition research, edited
by E Blom & S Unsworth. Amsterdam/Philadephia: John Benjamins, 95–114.
Scott, SK & Wise, RJS. 2003. Functional imaging and language: a critical guide to
methodology and analysis. Speech Communication 41:7–21.
Slobin, D. 1979. Psycholinguistics. Glenview: Scott Foresman.
Troster, AI & Parsons TD. 2006. Sodium amytal testing and language. In Encyclo-
pedia of Language and Linguistics 11:500–502.
Weinstein, EA, Lyerly, OG, Cole, M & Ozer, MN. 1972. Meaning in jargon aphasia,
in Aphasia, edited by MT Sarno. New York: Appleton-Century-Crofts, 92–104.

LIN2601/1/2017-2023 25
 Study Unit 2 Study unit 2
Studying language acquisition
‘Children have to be educated, but they have also to be left to educate themselves.’
2

Abbé Dimnet (French priest and writer of Art of Thinking, 1866–1954)

OUTCOMES
After you have studied this study unit you should be able to
 explain why we study language acquisition in children
 describe various tools and methods used to carry out research in language
acquisition
 describe some basic observations that have been made about language acquisition
 clarify the relationship between observations, explanations and theories
 provide arguments to show why it is not possible to acquire all aspects of lan-
guage by imitation
 explain why the notion that parents teach their children language is empirically invalid
 state the basic views of behaviourist, nativist, cognitive, social interactionist,
connectionist and emergentist theories and identify the main differences
 recognise controversial areas of debate in the field of language acquisition research
 provide definitions for the following important terms:

language acquisition behaviourism

language learning nativism
transcription competence
CHILDES performance
longitudinal study language acquisition device
cross-sectional study Universal Grammar
naturalistic observation linguistic universals
semi-structured elicitation parameter setting
experimental research cognitive theory
nature–nurture debate social interactionism
rationalism connectionist models
empiricism emergentism

Preview
In this study unit we introduce you to the process of language acquisition – how
children acquire their first language. We look at several reasons why we should
study language acquisition and take a look at several different methods that
researchers can use to study children’s language. We find out why the common
beliefs that language is acquired by imitation or that it is taught by parents are
misconceptions that seriously underestimate the complexity of the language
acquisition task. Finally, we include an overview of some of the controversies
and theoretical debates that exist between various researchers and we look at
the way the focus of interest has changed over the last 70 years in the field of
language acquisition research.

26
 Introduction
Learning to speak our mother tongues is an amazing intellectual achievement.
It is one of the most difficult tasks that humans are faced with, and yet children
learn to speak a language during the first few years of life. The question of how this
happens is a problem that has always fascinated not only linguists, but also parents,
psychologists, philosophers and educators. Try answering these questions to see how
much you know about children’s language learning.
Do children need to be taught language?

At approximately what age do children start to use recognisable words?

Do adults and children learn language in the same way?

Here are the answers to the test-yourself questions: Children do not require any
explicit teaching, but simply seem to pick up language quickly and effortlessly from
hearing it spoken around them and participating in conversational interaction with
others. On average, first words emerge when children are about a year old, although
this differs from child to child. Children seem to be better equipped for picking up
languages than adults, as you will know if you have experienced the hard work, slow
progress and frustrations of learning a new language in later life. But adults have
an advantage over children in that they have more advanced academic abilities and
better concentration spans. Because the process of language learning differs markedly
between adults and children, linguists use two different terms. The unconscious and
informal process by which a child learns language in the preschool years is known
as language acquisition. Linguists use the term language learning to refer to
the more conscious or formal learning process that takes place when people learn
a second (or additional) language after puberty. At primary school level, children
learn through a combination of conscious and unconscious processes, in other
words, partly through language acquisition and partly through language learning.
In this module we will look only at language acquisition – the way in which young
children learn their mother tongues. The focus in this study unit will be on why the
subject is important, what it can tell us about the human mind, and how we study
this topic. If you are interested in second language learning you should consider
taking the third-year module on ‘Learning and teaching an additional language’
(LIN3703).
The scientific study of how children learn language in the preschool years falls into
the area of study known as psycholinguistics. Although people have always been
interested in the miraculous way in which children’s language unfolds, it was only
during the 1950s that psycholinguistics became a recognised field of study. As you
learnt in study unit 1, psycholinguistics uses theories and research methods from
both linguistics (the study of language) and psychology (the study of the mind) in
order to explore the mental processes that are involved in using language and in
learning to speak. The particular area of psycholinguistics that we focus on in this

LIN2601/1/2017-2023 27
 course is known as developmental psycholinguistics as it deals with the ways that
children’s knowledge of language develops over time.
Many of you will have some idea of the various linguistic stages through which
children progress from general knowledge or from your own experience with younger
brothers and sisters. Some of you may have observed the development of your
own children, or the pupils that you teach, while others may have taken previous
psychology or linguistics modules such as LIN1502. This knowledge will assist you
in making sense of the course and getting the most out of it. However, since many
of you may not have spent lengthy periods of time with young children, this course
doesn’t assume any prior knowledge about language acquisition. All the various
terms and concepts you need to know will be explained as you come across them.

1 Why do we study language acquisition?

Why is it important to study children’s language development? Firstly, knowledge
of the way in which children acquire language is important for anyone involved in
the care and education of children. It provides greater insight into the way the child
develops and enables parents, caretakers and teachers to provide the rich speech
environment children need to develop to their full potential. Secondly, as students and
academics, we take an interest in the world around us and are continually questioning
why and how things happen. Instead of simply taking things for granted, we consider
it a challenge to gain a greater understanding of a phenomenon that forms such an
important part of human life. Thirdly, the more we know about language development,
the better we will be at recognising and addressing a variety of educational problems
such as delayed language development. Finally, researchers interested in artificial
intelligence (programming computers to ‘understand’ and generate language) can
learn a lot from studying children’s language acquisition. As Bloom (1993:6) puts it:
‘Computer scientists interested in language might study children for the same reason
that Leonardo da Vinci, who was interested in building a flying machine, chose to
study birds.’ Once we know enough about children’s language to make valid statements
and come to valid conclusions, these conclusions can be related to real-world issues.
Our knowledge of the process of child language acquisition can therefore help us come
up with meaningful solutions to language problems.

2 How do we study language acquisition?

How exactly does one go about investigating how children acquire language? For
An utterance the same reasons we saw earlier, one cannot open up their heads to see what is
is any sound, happening inside, nor can we simply ask young children to answer questions about
word or sentence
that someone their language use. It is important to notice that adult language can be investigated
produces. in several ways that are not open to child language researchers. For example, I could
find out about your language competence by asking you whether you consider
various sentences to be grammatical or not, or by asking you to provide sentences
in your language which demonstrate a particular grammatical construction like the
past tense rule. Children do not have the specialised vocabulary or sophisticated
knowledge of how language works (known as metalinguistic awareness) necessary
to understand and answer these questions. Instead of simply asking children what
they know, we have to be more indirect in our investigations, and work out what
they know by looking closely at what they say. Most of what we know about language

28
 acquisition comes from studying what children actually say, i.e. from collecting and
analysing the utterances children produce.
According to Foster-Cohen (1999:1–2), one of the most fascinating topics entails
how children are able to learn languages:
There are those who believe that children ‘know’ a great deal about language
– much more than might at first appear from what they say (or are able to say).
There are others, however, who believe that children know very little about
language, and must work it all out from hearing (or seeing, in the case of sign
languages) the language of others and from their own attempts to use language.
Ethics are moral The reason why we cannot decide what infants know or don’t know is that we cannot
principles or rules
about what is observe their language directly. We cannot get inside children’s heads, but have to
right and wrong. use more or less subtle methods of observation and experimentation that we hope
will give us the clues we need.

2.1 Ethical aspects of research

An important point to bear in mind when undertaking research is the question of
ethics. Crystal (1997:230) tells the story of the Mogul Emperor of India, Akbar the
Great (1542–1605), who undertook an experiment to find out whether children who
were isolated from human contact would learn language. He built a mansion where
newborn babies were kept in isolated captivity, with nurses who were instructed
not to utter a word in the children’s presence. After some years he visited the house
to find out the results of his experiment. No speech or even cries were heard in the
house – despite the fact that the children were already 4 years old.
The ethical problem with this kind of research is obvious – no one has the right to
deliberately deprive others of their human rights, such as the right to liberty and the
right to grow up in a family or society that encourages communication. Although
this kind of experiment would never be allowed to happen nowadays, researchers
still have to consider ethical questions. To give one example, making recordings of
people’s speech using a hidden microphone that they know nothing about is also
viewed by many as ethically questionable, as people have the right to know when their
conversations are being recorded and for what purpose the recordings will be used.

2.2 Research methods and research design

One of the first tasks a language acquisition researcher faces is to decide on a research
question. For example, a researcher might decide that she wants to try and find
out the answer to the following question: At approximately what age do children
start to ask questions? When scholars undertake research they do not do so with a
blank mind. Instead, they will usually formulate a hypothesis, which is an educated
guess at what the answer to the question might be or how/when something happens.
The hypothesis is then tested to find out whether it was correct or incorrect. In our
example, the researcher might hypothesise that children’s first questions emerge at
about 18 months.
In order to answer the research question and find out whether the hypothesis was
right or wrong, he or she has to collect data (examples of children’s language). Any
researcher must decide on the methodology he or she will use in the study. To a large

LIN2601/1/2017-2023 29
 extent, the aims of the study will determine the methodology used. Methodology
basically refers to the way in which the study is carried out; for example how many
children are included in the study, how the children are selected, how long the
study will last, how data will be collected, what kind of recording equipment will
be used, etc. While researchers formerly used notebooks to write down children’s
utterances, a more accurate, objective and up-to-date recording method is to use a
smartphone or video camera to record children’s utterances. Of course anyone who
has spent time with young children will know that they seldom sit still for long,
and that recording sustained sections of speech is very difficult when the subjects
keep disappearing out of range of the recording equipment!
The next stage involves making a transcription of the data. A transcription is a
written version of the spoken data, and often contains detailed additional information
such as the context within which the utterance occurred, the pronunciation used, the
pauses within and between utterances etc, whether words were whispered/shouted/
inaudible on the tape, etc. Take a look at the following sample transcription (based
on Berko Gleason & Bernstein Ratner 1998:352):
Participants: CHI Adami Child, MOT Nan Mother
Age of child: 4;3
Date: 10-APR-1991
Coder: Rachel Brown
CHI: toys
CHI: af’er (2) my book.
CHI: dis is not my book
MOT: this is not your book?
CHI: you’re just saying it
MOT: Well, these are the pictures of our trip to Disney World.
CHI: i want two Aprils in my ( )

You will see that the participants in the conversation are listed at the top of the
page, with abbreviations so that the transcription doesn’t have to include the full
name for each utterance. In this case the participants are the child and the mother.
Information is provided on the age of the child, date on which the recording was
made and who transcribed the information from the tape. Notice that there is
very little punctuation as punctuation is associated with written language rather
than spoken language. Punctuation is used occasionally where it is clearly audible,
for example the question mark indicates the rising intonation associated with a
question. The English spelling is modified so as to provide information about the
pronunciation used by the child. If a researcher wants to be even more precise about
transcribing the exact sounds used by the child, he or she can use the International
Phonetic Alphabet, which assigns a separate symbol to all the speech sounds in
human language. The (2) in the child’s second utterance indicates a two-second
pause at that point. At the end of the transcription you will notice an empty set of
brackets. These indicate that the child’s utterance was unclear. Transcriptions also
often include information about which words are stressed (e.g. by writing these in
capital letters), and about non-verbal responses such as laughter.
As Wardhaugh (1992:152) points out: ‘The conclusions you draw about the behaviour
of any group are only as good as the sample on which you base your conclusions.’
In this case, the sample is the number of children that a researcher investigates.

30
 The larger the sample, i.e. the more children used in the study, the more valid the
results are likely to be. The children who are studied are known as subjects. In order
to come up with valid generalisations about language acquisition, researchers also
need to ensure that their sample includes a wide range of different subjects – girls,
boys, urban and rural children, children from differing language groups, cultures
and social classes, etc.
An important development in the study of language acquisition was the setting up
A database of a large computerised database known as CHILDES (Child Language Data
contains Exchange System) by psycholinguists Brian MacWhinney and Catherine Snow in
information
stored in an the 1980s. This database is growing all the time and contains transcriptions from
organised form thousands of research projects in many different languages, including Afrikaans
on computer. and Sesotho. This resource allows researchers to study large amounts of data without
having to collect it themselves.

The CHILDES database can be found on the internet at

http://childes.psy.cmu.edu

2.2.1 Longitudinal and cross-sectional studies

Research into children’s language can be undertaken in different ways, depending
on the aims and the time available. The earliest studies of children’s language were
diary records of the language development of an individual child, usually kept by
the parents. Charles Darwin, for example, kept detailed notes on his son’s language
development. From about the 1960s researchers began to collect tape-recorded data
on children’s language. One of the first major studies was that of Roger Brown at
Harvard University, who recorded the language development of three children
(Adam, Eve and Sarah) every month for several years and transcribed the recordings
for researchers to study.
Longitudinal studies involve a detailed record of the utterances of a particular child
or children. By continuing to collect data from the same children over a period
of months or even years, researchers develop a good idea of how their language
changes and develops over time. Longitudinal studies thus come back to the same
group of children at a later period or periods. Because longitudinal studies are so
time-consuming, they are also expensive and can only study a limited number of
individual children.
Longitudinal studies are necessary to answer certain kinds of questions, such
as those that deal with the effects of children’s early linguistic environment
on their later acquisition of language. For instance, a longitudinal design was
used to show that infants whose mothers spoke to them frequently and in short
sentences at 9 months (time 1) performed better on tests of language compre-
hension at 18 months (time 2) than did infants of less talkative mothers (Mur-
ray, Johnson & Peters 1990 in Berko Gleason & Bernstein Ratner 1998:353)

Cross-sectional studies, on the other hand, look at the language of at least two
groups of subjects at one particular point in time. An example of a cross-sectional
study is one which aims to find out how the questions produced by 2-year-olds and
3-year-olds differ. By comparing two groups of children (2-year-olds and 3-year-olds)

LIN2601/1/2017-2023 31
 we would be able to see whether we could make any generalisations about the kinds
of questions they produce at each age. Cross-sectional studies enable researchers to
obtain a great deal of data in a short time, rather than over months or years. Studying
many subjects also makes it more likely that study results can be generalised.

Task 2.1

Match each statement below with a type of research design (longitudinal or

cross-sectional).
1. This kind of study is often completed in a short period of time.

2. Brown’s study of Adam, Eve, and Sarah.

3. This kind of study compares children of different age groups on a particular

language task at a particular point in time.

4. This kind of study looks at the language of the same children at different points
in time.

5. Parent diaries are an example of this kind of study.

Feedback
1. A cross-sectional study is often completed in a short period of time.
2. Brown’s study of Adam, Eve, and Sarah is a longitudinal study as it traces the
language ability of a small group of children over a long period of time.
3. A cross-sectional study compares children of different age groups on a particular
language task at a particular point in time.
4. A longitudinal study looks at the language of the same children at different points
in time.
5. Parent diaries are an example of a longitudinal study.

Task 2.2

If you wanted to carry out the following research, would you choose a longitudinal or
cross-sectional research design?
1. Aim: To find out how many new words a particular child learns to use between
age 2 and 3.

32
 2. Aim: To find out whether bilingual children have bigger/smaller vocabularies
than monolinguals of the same age.

3. Aim: To find out whether children who talk early also learn to read early.

Feedback
1. To find out how many new words a particular child learns to use in a year, we
would have to use a longitudinal study.
2. To find out whether bilingual children have bigger/smaller vocabularies than
monolinguals of the same age you would need a cross-sectional study. A group of
monolingual children would have to be observed and their vocabularies compared
with a group of bilingual children at the same developmental stage.
3. To find out whether children who talk early also learn to read early, we would
need a longitudinal study as we are studying the same group of children at two
different points in time, namely, at age 1 to 2 when they learn to talk, and at ages
5 to 8 when they learn to read

Interestingly, one of the oldest methods of studying child language is making a come-
back, namely, the parent report. While parents are certainly not as objective as
researchers should ideally be, they are ideally placed to engage in detailed long-term
data collection. Dale and Goodman (2005:42) believe that ‘one of the most striking
developments in the study of child language, and indeed developmental psychology
more generally, over the past 20 years is the revival of the parent report as a trustworthy,
and trusted, research technique’.
The type of parent report most often used in recent research follows a set format
and is known as the MacArthur-Bates Communicative Development Inventories
or MCDI (Fenson, Marchman, Thal, Dale, Reznick & Bates 2007). There are two
forms of the MCDI, one for itemising the words and gestures of children aged 8
to 16 months and the other for the words and sentences of children aged 16 to 30
months. In the 8 to 16 month MCDI, parents complete a checklist of 396 words
and 63 gestures, indicating for each item if the child understands the word/gesture
or if the child understands and uses that word/gesture. In the case of the 16 to 30
month MCDI, there are 680 words in the checklist, and a grammatical component
in which parents look at sentence pairs like kitty sleep and kitty sleeping and indicate
which one is closest to their children’s own language use.
The MCDI can be used cross-sectionally to compare large samples in order to
identify commonalities and individual differences. It can also be used longitudinally,
for example as a monthly report used to trace developing vocabulary and syntax.
The MCDI has also proved to be a good predictor of children’s later development.
For example, children with language delay (MCDI scores in the lowest 10%) at
24 months generally continue to experience difficulties at age four, although a
few children overcome their slow starts and catch up to normal children (Dale &
Goodman 2005:50).

LIN2601/1/2017-2023 33
 2.2.2 Observational and experimental studies
Naturalistic observational studies focus on collecting children’s spontaneous
An observation speech. Researchers can learn a lot from recording children’s utterances in natural
is a piece of everyday situations. However, people tend to talk in a different way when they are
information
obtained by being recorded, and researchers are thus faced with the ‘observer’s paradox’ – they
studying or have to try and observe people’s normal behaviour, but simply by being there as an
watching observer they are altering the ‘normal’ situation. Even though the speech behaviour
something. collected in these circumstances can never be truly natural, the researcher has to
make every effort to be unobtrusive and not to interfere with the child’s natural
everyday use of language. A child that feels uncomfortable, shy or threatened will
probably talk very little, while the microphone can also be distracting. One solution
is to help the children get used to the recording equipment for a few sessions before
recording actually begins. Other techniques may be to keep the video recorder in
the corner of the room or out of sight, or to provide a topic or toy that the child is
very interested in, as people tend to be less self-conscious and talk more when they
are excited.
Let’s see how Katherine Demuth tried to ensure that she did not affect the Sesotho
children’s natural use of language during her research (Demuth 1984):
After having visited the children’s homes a number of times, becoming fa-
miliar with the household routine and giving the children an opportunity to
get to know me, audio recording was commenced. I would check with the
mother about her next day’s activities and then visit with tape recorder, pen
and notebook. I generally sat off to the side while children interacted with
members of the family or with playmates … Questions directed to me by a
third party or by the child were answered, while initiation on my part was
practically nil.

Task 2.3

Read the extract above and identify three ways in which Demuth tried to minimise
her influence on the children’s language:

34
 Feedback
Demuth made sure the children knew her and were familiar with her, she sat off to the
side so she did not intrude too much on the conversation, and she did not initiate (start)
any conversation. Demuth (1984:10–11) states ‘While my influence on the children’s
verbal and social interaction must be taken into account, it was a least minimized by
being familiar to the family and the children and by playing a background rather than a
major participant role in the interaction which took place.’

Another problem with naturalistic observation is that often there are certain types
of sentences or grammatical structures that don’t appear in the data a researcher has
collected. For example, a child may have uttered many questions starting with the
words What and Why, but no questions starting with the word How. In cases like
this it is difficult to know whether the child has not yet learnt this particular form, or
whether it is simply accidental that the data didn’t include an example of this form.
In order to address this problem, semi-structured elicitation can be a useful way of
getting children to say things and use grammatical forms that would be infrequent
in naturalistic conversation. Elicitation means ‘to draw out’ language data by asking
specific questions or providing specific tasks. The researcher plays a game, watches
a silent movie or looks at an object together with the child in an effort to encourage
speech production, while still keeping the situation as natural as possible. For example,
researchers in many countries have used the wordless ‘frog story’ picture book entitled
‘Frog, where are you?’ as a tool for eliciting narrative descriptions (Berman & Slobin
1994). The first three pictures in this story can be seen at:
http://torontopubliclibrary.typepad.com/kids-books/2015/12/who-needs-words-
anyway.html
Another way of supplementing naturalistic data is to undertake experimental
research. Experimental research is basically a controlled way of collecting information
about specific aspects of language, and usually involves researchers asking children
to perform specific activities in a laboratory setting. The technique of elicitation
is frequently used in language acquisition experiments. For example, children
may be asked to complete a word or phrase to see whether or not they can use a
particular grammatical structure. A famous example of this kind of study is Jean
Berko Gleason’s ‘wug test’, designed to test how children used English plural forms
(Berko 1958). Researchers taught children several nonsense words by showing
them pictures of non-existent creatures and repeating the name, for example This
is a wug. Children’s knowledge of the plural form was then tested by asking them
to complete a phrase like ‘This picture shows two ____’. http://www.onbeing.org/
blog/sunday-morning-exercise-take-wug-test/2510

If you have a minute, you can watch Jean Berko Gleason conducting a
w test at https://www.youtube.com/watch?v=MgB2iMuEZAA
wug

Another way of testing children’s knowledge of grammatical forms is to ask them

to repeat various sentences. It has been found that children cannot repeat complex
sentences that are above their level of language development, i.e. sentences that they

LIN2601/1/2017-2023 35
 could not have produced themselves at that stage. Thus, a child who has not yet
learnt the plural form would hear the phrase two potatoes but would say two potato.
(You would experience a similar thing yourself if you tried to repeat a sentence in
a language you did not know!) Children’s understanding of various sentences, for
example passive constructions such as The boy was hit by the girl, can be tested by
asking them to act out sentences using puppets or other toys, or to point out the
correct one from a series of pictures.

Children’s errors are also an extremely important source of information on the

language acquisition process. Linguistic errors can tell us a lot about the child’s
grammar and the hypotheses he or she is making about language structure. When
faced with children’s errors, psycholinguists typically ask themselves the following
questions: Do children learning language X typically make this kind of an error?
Why do children make this error? How long do errors of this kind persist? When
do they disappear from the child’s speech? Note that children do not make errors
in all areas of language – some parts of grammar seem to be more susceptible to
errors than others. For example, many children have trouble acquiring the past-
tense form, but hardly ever make errors in acquiring the word order rules of their
language (Bloom 1993:32).

Experimental testing procedures can even be used on babies as young as six months.
For example, the headturn preference procedure, developed in 1985, is used to test
differentiation of grammatical as opposed to ungrammatical forms or familiar as
opposed to unfamiliar forms (Johnson & Zamuner 2010). Two different sound files
are played to the baby while it sits on the caregiver’s lap and the researcher measures
how long the baby focuses on each sound before looking away. Babies typically listen
for longer periods to stimuli that are grammatical or familiar than stimuli which
are ungrammatical or unfamiliar. Since babies have short attention spans, the entire
test seldom lasts more than five minutes.

3 Issues in language acquisition research

As with any subject which is not yet fully understood, many different theories relating
to language acquisition have emerged over the last 50 years or so. The ultimate aim
of any theory of language acquisition is to account for children’s language at any
point in their development (i.e. to explain why they say certain things and make
certain errors) and to explain the processes responsible for the gradual increase in
language competence (i.e. to explain how children eventually end up speaking like
adults). If the average person in the street were asked how children acquire language,
they would probably reply that their parents teach them the language or that they
learn it by imitation. Common knowledge about language acquisition is thus based
on two viewpoints:
 Parents teach their children language.
 Children acquire language by imitating the language of their parents.
Do you agree with the statements above? What are your own ideas on the subject?

36
 We will look at possible answers to these questions in a minute, but let’s reflect
first on the differences between opinions, theories and facts. You should be aware
that theories are not factual; they are basically different people’s informed opinions
about how language acquisition works. The opinions you gave above wouldn’t be
considered theories as you probably don’t know enough about the subject to give
an informed opinion based on extensive observation of the facts. Theories use
facts (data) to support their arguments, but different theorists undertake different
kinds of research and interpret the research findings in different ways. Theories
may differ quite widely and give rise to lively and sometimes angry debate between
psycholinguists. In your reading, you need to be able to identify when the author
is giving you facts and when the author is describing how various theories inter-
pret these facts. Look out for phrases like According to X, X claims that …, In X’s
view …, as these indicate that we are talking about a particular theoretical point
of view rather than a fact.
Most of what we know about language acquisition comes from studying what
children actually say, i.e. from their output, and to a lesser extent from looking at
the speech addressed to children during their interactions with other people, i.e.
the input that children hear. After decades of intensive research psycholinguists
now have a fairly comprehensive picture of the course of language development,
especially of English, since this is the most intensively studied language in the world.
This means that we have a fairly clear picture of the various stages that children go
through, the types of words they learn first and the types of mistakes they make.
However, despite the accumulation of a fairly detailed corpus of data on language
development, we still don’t really know how it happens. The mechanisms that produce
these accomplishments are still being debated and are not yet well understood.
It is important to bear in mind that observations or data relating to children’s
language are not self-explanatory, i.e. they do not ‘speak for themselves’. They have
to be interpreted, and as you well know, the same set of data can be interpreted in
different ways. Take for example, the following developmental observation:
When children enter the one-word stage shortly after their first birthday,
they initially produce a few words only. At about the age of 16 months, they
suddenly seem to undergo a vocabulary spurt, after which they start learning
and producing new words at quite a remarkable rate.

This observation is informative because it tells us something about the way in which
You postulate the child’s language is progressing, but it doesn’t tell us why progress suddenly speeds
that something up, i.e. it is a description not an explanation. This observation must be interpreted
is true when you
assume it. (i.e. explained), and interpretations may involve speculation about factors that are
not observable or testable. The vocabulary spurt itself is testable – we can observe
other children who are about the same age and we will find a similar marked increase
in the children’s output. However, we do not know what is happening inside the
child’s head in terms of cognitive development or language development, we still
don’t really know how it happens. The mechanisms that produce these accomplishments
are still being debated and are not yet well understood.

LIN2601/1/2017-2023 37
 One explanation for this vocabulary spurt is that the child reaches a crucial stage
Empirical means of development at which she realises (unconsciously) that sounds can be used to
based on practical
experience rather represent concepts or things in the world. Once she makes this discovery, she
than theory. purposefully starts naming or labelling objects and events she encounters in her
environment. Now this is not an observation (i.e. a ‘fact’), it is an explanation. This
explanation postulates that there are unobservable forces underlying the vocabulary
spurt (such as the discovery that words can stand for concepts). This explanation is
a hypothesis or informed scientific guess rather than a description of observable
factors. A hypothesis (being a guess) can be wrong. Not all psycholinguists accept
the explanation given above and other explanations have been put forward to account
for the same observation, i.e. different interpretations of this observation or different
theoretical viewpoints have been proposed. A theory is therefore not part of the
phenomenon being studied, it is an explanation of (aspects of) that phenomenon.
A theory is thus not a truth, nor is it an observation or empirical ‘fact’; it is simply
one explanation (among many others) of a particular phenomenon. Observations
and theories are obviously intimately interrelated: data are continuously used to
support or falsify theories, to test their validity, while theories are continuously being
formulated and modified to interpret and explain data.
Contrasting viewpoints in child language research are often formulated as questions,
The word for example Do parents teach their children language, or does language simply
nurture means unfold according to a genetic programme? Many of these questions are still
to look after a
growing child. unanswered, and different theories emphasise different answers to each question.
In many cases the most realistic answer lies somewhere in between the two options.
In the following section we look at some of these questions or debates in more detail.
One of the major tasks in psycholinguistics is to find out how much of our language
ability is innate or inborn and how much is shaped by environmental factors such
as the conditions or surroundings in which we grow up. This question is known
as the nature–nurture debate. In order to understand what language is, how it
works, and how young children can acquire such a complex system, we need to try
and sort out how language relates to factors internal to us (our minds) and external
to us (our environment). Most researchers nowadays agree that both innate and
environmental factors play a role in language development, but argue as to which is
the more important of the two. Some scholars believe that internal or innate factors –
the nature factors – are the most dominant factors in this relationship, while others
believe that external or environmental factors – the nurture factors – are dominant.

4 Dominant theories in developmental psycholinguistics

The nature–nurture debate is part of a larger, longstanding philosophical debate
about how knowledge is acquired. The argument basically hinges on the part played
in this process by the mind and human reasoning ability on one hand and human
senses and experiences on the other. Those who emphasise the role of reason are
known as rationalists, while those who emphasise the role of experiences and
information obtained through our senses are known as empiricists (remember that
something is empirical when it is based on practical experience rather than theory).
Early empiricists believed that the mind of a newborn baby is like a blank slate
(tabula rasa is the Latin term) until his or her experiences of the world leave their
imprint upon it. Rationalists disagree with this view, believing that the mind of a
newborn already has a particular structure that imposes constraints on the way in
which knowledge, including language, is acquired.
38
 Another fundamental difference is that empiricists undertake active research to find
out more about children’s language and use only observable and measurable data
in their explanations. Rationalists, on the other hand, do not mind using evidence
that is not directly observable, for example making claims about the structure of
the mind and mental processes. The rationalist approach relies more on theoretical
deductions and assumptions about what children know, rather than on actual
empirical research. The rationalist approach believes that although young children
produce very limited utterances, they have a complex system of abstract language
knowledge. This is known as the competence–performance distinction. Competence
is basically the abstract linguistic knowledge stored in our brains. Performance
involves putting that knowledge into practice by actually speaking or using language,
and may be less than perfect due to factors like lack of concentration, slips of the
tongue, a bad cold, etc. Researchers in the empiricist tradition are interested only
in performance, and refuse to deal with such non-observable and non-measurable
concepts such as ‘competence’.

4.1 Behaviourism
This debate between rationalism and empiricism has played a very important part in
the history of psycholinguistics. One version of empiricism that dominated psychology
and linguistics from the 1930s to the 1960s is known as behaviourism. Behaviourists
believed that internal mechanisms (like the language acquisition process) cannot
be studied except through their observable outputs. The most famous behaviourist
in linguistics was BF Skinner. Skinner published a book in 1957 entitled Verbal
behavior, which claimed that speech was a learnt behaviour that develops through
imitating and interacting with others. He claimed that children acquire language as
a result of parents reinforcing and shaping children’s utterances – in the same way as
laboratory rats can be trained to run through a maze. Behaviourists believed that parents
and caretakers teach children language through negative and positive reinforcement,
demonstrating language use, correcting errors and offering them opportunities to
practise their new skills. In summary, behaviourists believe that:
 Observable behaviours are the only way to investigate mental processes.
 Language learning is manifested by a change in observable behaviour brought
about by stimuli coming from the environment.
 Language is learned in the way in which other habits are learned.
 Experimental methods should be used to investigate language acquisition.
 The environment shapes behaviour.
 Children learn words by associating sounds with objects, actions and events.
 Children learn words and syntax by imitating others.
 Adults enable children to learn words and syntax by teaching and reinforcing
correct speech.

4.2 Nativism
At the other end of the spectrum are the researchers who take a rationalist view and
emphasise the nature factors in the acquisition of language knowledge. This view
is known as nativism. Some of the evidence that has been put forward to support
the nativist viewpoint includes the following:

LIN2601/1/2017-2023 39
  All children acquire the language to which they are exposed, which means that
a child has the potential to acquire any language.
 Children acquire language at a time when they are intellectually immature and
most other intellectual tasks are beyond them.
 Children acquire language fairly effortlessly and in a short period of time. This
contrasts strikingly with adult learners, who need to make a conscious effort over
several years in order to become proficient in the new language.
 All children go through the same stages of development within the first 12 to 18
months, irrespective of what language they are learning.
 Children acquire language without being taught explicitly.
 All children achieve basic linguistic competence at approximately the same age
(6 or 7).
 Children continually come up with new words and sentences that they have
never heard before.

From the observations above it is evident that nature (internal factors) plays an
important role in language acquisition. Nativists believe that the ability to learn
language is an innate or inborn human ability, allowing children all over the world
to learn successfully that behind the directly observable stream of speech sounds is an
underlying system or set of patterns that determines how these speech sounds combine
into words and how words combine into phrases and sentences.
The most important researcher in the nativist tradition is the American linguist Noam
Chomsky. Chomsky was born in 1928 and is one of the most important men in
the field of linguistics. His radical and often controversial ideas have revolutionised
linguistics and have had an important influence on other fields including philosophy,
psychology and artificial intelligence. In 1959 Chomsky wrote an extremely critical
review of Skinner’s book Verbal behavior, suggesting that behaviourist theory focused
too much on the role of parents and caretakers, rather than on the child’s innate
language knowledge. This review marked a turning point in language acquisition
research, and a shift away from behaviourism as a convincing theory of language
acquisition. Since 1960 almost all theories of language acquisition have assumed at
least some degree of innate linguistic knowledge. As we have mentioned, extreme
versions of nativism or empiricism are seldom found in current theories of language
acquisition. Researchers now agree that language acquisition is determined both
by a child’s innate capacities and his or her linguistic experiences, although they
typically stress one factor at the expense of the other.
Some of Chomsky’s arguments against behaviourism are summarised below:
Chomsky argued convincingly against imitation as a language acquisition strategy,
showing that if language were learned in this way, it would place a large burden on
memory, since every utterance would have to be stored separately in the memory
(rather than simply storing one rule for constructing all similar sentences). Secondly,
if language were learned by imitation, children would not be able to use language
creatively to say and understand new things not heard before. As we know, children
often produce utterances that never occur in adult speech. Consider the child who
looked out of the window at night and said Allgone outside. Bye bye outside. He must
obviously have heard the words all, gone, outside and bye bye, but he was unlikely to
ever have heard an adult utter these words in the order in which he uttered them.

40
 These utterances cannot be considered to be imitated language, nor is it likely that
his parents taught him sentences like these.
A third piece of evidence against imitation is the fact that children seem unable to
imitate adult utterances that are too complex for their current level of development.
Crystal (1997:236) cites the following example:
Child: Nobody don’t like me.
Mother: No, say ‘Nobody likes me’.
Child: Nobody don’t like me.
(Eight repetitions of this dialogue)
Mother: No, now listen carefully: say ‘Nobody likes me’.
Child: Oh! Nobody don’t likes me.

These observations suggest that imitation cannot explain all the facts of child
language development. In fact, imitation operates only on a very simplistic level
(such as learning words and certain basic fixed phrases).
Chomsky also criticised the behaviourist view that parents teach their children
language. It is true that parents spend a lot of time teaching children the names of
things (Look, here’s a cat … this is a kitty-cat, and here’s a doggie is not an uncommon
form of conversation with a 2-year-old!) and they provide the child with formulaic
politeness phrases like Say ‘Thank you for the present, Granny’. In some cultures
caretakers provide the child with many prompts, i.e. getting the child to repeat verbal
routines used in certain social situations. Consider the following Sesotho examples
of a mother (M) interacting with her daughter (D), aged 30 months, prompting her
to ask questions in a routine format (Demuth 1984:32–33). The daughter repeats
her version of the mother’s utterances.
M: Ere: ke ngoan’a mang?
Say ‘whose child is he?’
D: ana mang?
M: Ere: o moholo?
Say ‘Is he big?’
D: o holo?

The routine nature of such formats obviously provides the child with practice in
asking questions, but the actual structural components of the sentence and their
functions are never explicitly taught (as they would be to an adult learning Sesotho
as a second language, for example). Although the child is repeating her version of
what the adult says, she still needs to convert this input into a form that can be used
to construct new questions not heard before.
Behaviourists claimed that children learnt language partly through having their
mistakes corrected by their parents. Research has subsequently shown, however, that
parents seldom correct their children’s grammar although they may correct errors of
fact. For example, if the child points to a picture of a lion and says Dere kitty cat the
caretaker may say No, that’s not a kitty cat, that’s a lion. Chomsky refers to this as
the ‘absence of negative evidence’, i.e. although children receive positive evidence of
grammatical utterances by hearing well-formed sentences around them, they seldom
receive an overt correction (negative evidence) alerting them to the fact that they
have made a grammatical error. Research such as that of Strapp (1999:377) suggests

LIN2601/1/2017-2023 41
 that correcting children is not very effective – children are unlikely to adopt an adult
form after overt teaching or being corrected. Although practically all the utterances
that young children produce up to the age of about 3½ years of age contain errors
of some kind, parents do not correct every utterance. The child would probably be
so demotivated that he or she would stop talking altogether if this were to happen!
Because children are corrected so infrequently, it seems unlikely that corrections
and indirect teaching can account for language acquisition.

Task 2.4

Paragraph questions may require you to agree or disagree with a statement and then
justify your answer. This means that you must explain why you agree/disagree with the
statement and give further information on the topic. Look back at Task 1.4 for some hints
on writing paragraphs and then write a paragraph on whether you agree or disagree
with the statement below:
Children learn language by imitating their parents’ utterances.

Feedback
Here is one possible answer.
I disagree. While behaviourists believed that language was acquired mainly by
imitation and reinforcement, the nativist view is that imitation is only of limited
use in the language acquisition process, for example for learning words and set
phrases. Firstly, children often make errors and make up words that they would
not have heard, and use sentences that they are unlikely to have heard. This sug-
gests that children have an innate capacity to learn language. Another piece of
evidence against imitation is that it would place a large burden on memory, since

42
 every utterance would have to be memorised separately (rather than simply stor-
ing one rule for constructing all similar sentences). Thirdly, children do not seem
to be able to imitate utterances that are far more complex than their current level
of development. These observations suggest that imitation cannot explain all the
facts of language acquisition.

You should be aware that Chomsky’s version of nativist theory is both idiosyncratic
Something is and controversial. While his contribution to the field of psycholinguistics cannot be
idiosyncratic denied, certain aspects of his theory are not generally accepted. One of his most central
when it is unusual
and associated (and most controversial) beliefs is that knowledge of grammatical principles is to a
with a particular large extent innate. Chomsky claims that children use many grammatical features
individual. that they could not possibly have learnt on the basis of language exposure alone, and
that this knowledge must therefore be innate. Chomsky claims that children are born
with a ‘language acquisition device’ or abstract mental blueprint of universal linguistic
rules and principles known (also referred to as Universal Grammar or UG). This
language acquisition device, which is assumed to be a physiological part of the brain,
allows children to detect underlying regularities in the language around them and
construct a set of basic rules that allows them to be infinitely creative in the way they
use their language. Foster-Cohen (1999:108–109) says:
Clearly, whatever UG is, it must be very flexible. It cannot contain informa-
tion about any particular language, or we would be led to the obviously silly
position of saying that children are born ‘knowing’ all the world’s languages.
Clearly, the form of linguistic knowledge children possess must be abstract.
It must be able to recognise the specifics of particular languages once they
are presented, and to cope with what appear to be pretty extreme differences
between languages.

Since the aspects that are common to all languages (the linguistic universals) are
already present in the mind, Chomsky views language acquisition as basically a matter
of filling in the language-specific details in order to become competent speakers.
Even the language-specific details he believes are not learnt from scratch by listening
to the language of the environment, but are selected from a set of options present
in the language acquisition device. These options are known as parameters, and
the process of selecting the appropriate option for a particular language is known
as parameter setting. When all the parameters have been set, the child has a core
grammar of his or her particular language.
The establishment of core grammar completes the work of UG in language
acquisition. It does not, however, complete the work of language acquisition
itself. There’s a lot more to be done – lots more to really learn from the input:
all the vocabulary of the language, the bound morphemes, the idioms etc.
The list may be very long. Language acquisition is thus a co-operative effort
between UG and learning from the input, both in the sense of the input trig-
gering the parameter settings and in the sense of the input providing language
forms from which rules can be deduced by general learning mechanisms not
special to language. (Foster-Cohen 1999:109–110)

Chomsky has devoted much of his life to spelling out the nature of the language
acquisition device and to detailed investigations into the structure and rules of
language and what languages have in common. He is first and foremost a theoretical

LIN2601/1/2017-2023 43
 linguist rather than a psycholinguist, and is often criticised by other psycholinguists
for making sweeping claims about the language acquisition process without consulting
empirical research on children’s language to back up his assumptions.
Chomsky believes strongly that the development of language is a completely different
process from other kinds of cognitive development such as the ability to think and
Something is reason, including categorising, problem solving and increasing the number of items
autonomous that can be stored in memory. He believes that language development and cognitive
when it is separate
from other parts development are autonomous processes (i.e. they develop separately without
and functions influencing each other). This question of whether language is a separate, autonomous
independently. faculty of the human mind, or is dependent on other kinds of cognitive development
is still a matter of considerable debate.
Another controversial element of Chomsky’s theory is his claim that the language
input that children hear around them provides insufficient evidence for language
acquisition to take place:
Consider first the nature of primary linguistic data. This consists of a finite
amount of information about sentences, which, furthermore, must be rather
Something is
degenerate when restricted in scope, considering the time limitations that are in effect, and
it is atypical or of fairly degenerate in quality. For example, certain signals might be accepted
poor quality. as properly formed sentences, while others are classed as nonsentences …
(Chomsky 1965:31)

This argument is termed the poverty of the stimulus argument, as it stresses

the limitations of input in the language acquisition process. The claim is that the
language children hear is degenerate – full of errors and short sentence fragments
rather than perfect grammatical sentences – and so does not form a good enough
model for children to use as the basis for deducing linguistic rules. Chomsky thus
concludes that most of these rules are unlearnable and must be innate.

Task 2.5

As you read through any textbook or academic journal you need to remember that the
authors may not be neutral and may be adopting a particular theoretical perspective
on certain issues. In other words they may have certain assumptions that influence the
way they talk about language acquisition. Read the following extract and see if you can
identify whether the author is an empiricist or a rationalist:
I am persuaded that some aspects of the linguistic system are innate and specific
to language … I think there is reasonable evidence that certain parts of syntax
(the way sentences are structured out of words), morphology (the way words
are structured out of smaller units known as morphemes), phonology (the way
morphemes are structured out of sound units) and semantics (the meanings that
words and sentences encode) are so ‘weird’ that only accidents of genetics could
account for them (Foster-Cohen 1999:11).

44
 Feedback
You should have picked up that Foster-Cohen is not writing from a neutral position but
from a rationalist perspective. Words like ‘innate’, ‘specific to language’ and ‘accidents
of genetics’ indicate that she has adopted a Chomskyan nativist perspective on certain
issues.

Let’s turn now to some of the other major theories and developments that have
shaped the field of language acquisition over the last 70 years.

5 Overview of 70 years of language acquisition research

5.1 The 1950s – behaviourism, nativism and cognitive theory

As we have mentioned, it was in the 1950s that researchers first began to interest
themselves in language acquisition as an area of study. During the early years the
focus was primarily on the nature–nurture debate, with behaviourists arguing with
nativists such as Chomsky about whether language was a learned behaviour or an
innate ability.
Another argument was going on in the 1950s between Chomsky and the Swiss
psychologist, Jean Piaget. As we have said, Chomsky believed that language was
an autonomous process, separate from other aspects of cognitive development.
Piaget and his followers, on the other hand, believed that language was dependent
on thought, i.e. that language could not develop unless certain cognitive skills had
developed first. This theory of language acquisition is known as cognitive theory.
Cognitive theory aims to identify the sequence of cognitive development and illustrate
how these developments are reflected in language acquisition. For example, Piaget
(1926) refers to the period from birth to 2 years as the sensorimotor period because
this is the time when children interact with and learn about their environment
through their senses and through physical (‘motor’) activity. Some of the important
cognitive developments during this period include the child’s gradual realisation
that when something disappears from view, it doesn’t cease to exist. This concept is
known as object permanence. Very young children cannot conceive of things that
are not physically present in their environments, but by the end of the sensorimotor
period, children realise that concepts and objects can exist without being physically
present. The development of object permanence is a prerequisite for the emergence
of words like gone, and in fact, for the emergence of words in general, since children
cannot learn words unless they can conceptualise objects and activities that are not
physically present and use words to symbolise concepts.

5.2 The 1960s

In the 1960s, researchers began to put these theories to the test, and undertook
research into children’s acquisition of syntax (grammar). Longitudinal studies of
children provided detailed information about the order in which various grammatical
constructions were acquired. Roger Brown’s (1973) study of the language of Adam,
Eve and Sarah was based on research carried out monthly for several years from
1962, and provided a great deal of what we now know about the language acquisition

LIN2601/1/2017-2023 45
 process. Cross-linguistic studies of language acquisition in various parts of the world
were undertaken, and revealed that many aspects of the process were universal.

5.3 The 1970s and the emergence of social interactionism

In the 1970s research began to focus on issues other than grammar, including the
meanings and functions of children’s utterances. Research such as that of Rosch
explored Piaget’s idea that language wasn’t a separate module of the mind, but was
rooted in overall cognitive mechanisms that were not specific to language. These
included principles of human categorisation. For example, Rosch et al (1976) showed
that 3-year-olds typically learnt the words like cat and lion first, before more specific
words like Siamese or lioness or mane, and before broader-category words like feline
or mammal. These words are learnt first, according to cognitive theorists, because
distinctions drawn at this basic level are the most useful distinctions, since it is at
this level that we most readily perceive different phenomena and objects around us.
Cognitive linguistics later became a fully-fledged discipline, as you will see if you
continue with LIN3701 Approaches in Linguistics.
Researchers also began to take an interest in input, in other words the language
addressed to children by their parents and caretakers. The aim of this research was
to find out what kind of language children heard around them in their everyday
environment and whether or not the language addressed to children was ‘degenerate’
as Chomsky claimed.
Psychologists such as Vygotsky and psycholinguists such as Jerome Bruner and
Catherine Snow carried out research that suggested that language acquisition can only
occur through interaction with others, and a theory known as social interactionism
emerged which is still current today. Social interactionists believe that mere exposure
to language on the radio or television or from hearing adults talking to each other
is not enough – it is the special language used by caretakers to children that enables
children to process and produce language. While social interactionism is similar
to behaviourism in the stress laid on the role of the environment, it differs from
behaviourism in two ways. Firstly, it does accept a certain innate language ability
as a given and, secondly, it sees the child as interacting actively with his or her
environment rather than as a passive participant whose language is ‘shaped’ by parents.
Through their communicative attempts and their responses to parents’ utterances,
children cue their parents to adapt their language and supply the appropriate level
of language input the child requires at that point in his or her development. Social
interactionists study the communicative function of language, the ways that adults
tailor their input to meet children’s needs and the effects of different kinds of input
on children’s developing language (Berko Gleason & Bernstein Ratner 1998:386).

For a quick review of the nativist, behaviourist and interactionist theories

you could watch https://www.youtube.com/watch?v=8FhKgzm4Za4

5.4 The 1980 and 1990s

A feature of the decades since 1980 has been the increase in cross-linguistic research,
including a shift of focus from identifying cross-linguistic similarities to identifying
cross-linguistic differences in the language acquisition process. Dan Slobin has led

46
 this initiative, with a research project that compared the acquisition of 28 languages
from a wide range of language families (e.g. Slobin 1997). The main question that
cross-linguistic research addresses is whether and how the actual course of language
acquisition is influenced by the syntactic and other differences that exist between
languages. Stoll (2009) points out that we only have acquisition data for about 2%
of the world’s 6000 to 7000 languages, with the vast majority of studies focusing
on English, so there is plenty of scope for additional cross-linguistic research.
Another important development in the 1980s was the emergence of connectionist
models such as those of Bates and McWhinney (1987). These models see language
development as a matter of developing connections between the neurons in the brain
and are thus more empiricist than nativist. However, like Piaget, they argue that
general cognitive processes can account for language learning. Using language is
seen as a matter of activating certain neurons in the brain that then activate a whole
series of related concepts. Here is Berko-Gleason and Bernstein Ratner’s (1998:386)
explanation of how connectionist models work:
A child develops such connections over time through exposure to the forms
of language associated with external events. For example, a child may hear
the word bottle under varying circumstances and thereby establish neural
associations to the word, to the initial sound /b/, to the word milk and so
forth. Ultimately those interconnected associations become the ‘meaning’ of
the word. Information in such a neural network is conveyed through many
interconnected units or nodes. The nodes have activation levels or thresholds
that can turn them on or off, and learning consists essentially of adjusting the
strengths of the connections in a direction that produces the desired outputs.

Language learning is thus seen as a process of detecting regularities and patterns in

the input. As the child is exposed to more and more language input, correct patterns
are strengthened and incorrect patterns are weakened so that errors eventually
disappear. Since this view of learning is similar to the way in which a computer
‘learns’, connectionists use computers to simulate human learning in order to test
their theories.

5.5 2000 to the present day

A widely adopted approach to language acquisition in the last two decades is
emergentism, an approach which has its roots in social interactionism and
connectionist models (see 5.3 and 5.4 above). Emergentism explains the development
of particular new aspects of language as resulting from the interaction of other
more general learning mechanisms and social experiences. In this view, language
is neither innate nor triggered, but emerges gradually as the child constructs a new
system using strategies they have already learnt. Learning language is thus not seen
as an autonomous task, but as a task in which multiple sources of information are
integrated and coordinated for communicative purposes while children are engaged
in doing everyday things. For example, Hollich et al (2000) offer an emergentist
explanation for word learning, based on children’s existing general abilities such as
paying attention to novel forms, following an adult’s eye gaze, and recognising the
link between pointing and labelling. Children also exploit linguistic clues such as
the presence of articles and other morphemes – children as young as 17 months will
pay attention to an object when asked to look at ‘the blick’ and will pay attention to

LIN2601/1/2017-2023 47
 an action when ‘blicking’ is requested (Hollich et al 2000:20). As children’s word-
learning principles emerge and develop, word learning becomes more and more
effective over the course of the second year of life.
If language emerges or is constructed in individual children and does not simply
uncover an innate set of knowledge, then it follows that there will be many different
routes in the course of language acquisition. The last twenty years has therefore
also seen an increased interest in individual differences rather than generalisations
about language learning. For example, while, on average, vocabulary development
does indeed involve a vocabulary spurt at around 16 months, when word learning
accelerates dramatically, this pattern does not hold for all children. Some children
accumulate words in a steady fashion with no spurt, or display two vocabulary
spurts, or start their spurts much earlier or later. The study of individual differences
has confirmed many of the early findings and generalisations about child language,
while allowing a much more fine-grained analysis of early language to take place
(Dale & Goodman 2005:67).  

6 Summary
In this study unit we have focused on some of the theories that dominate the field
of developmental psycholinguistics and seen that each of these theories stresses
one side of the nature–nurture equation at the expense of the other. However, it is
generally acknowledged that both nature and nurture influence the process, and that
none of the theories can account for the whole story of child language development
and how it works. It should be clear to you by now that theories are not fixed, but
change over time as new questions are asked, new research evidence emerges or
new techniques for investigating the human mind are invented. Psycholinguists
still have a lot of work to do before this fascinating process is fully understood! As
Crystal (1997:237) reminds us:
It is not possible in the present state of knowledge, to choose between these
various approaches … Doubtless imitative skills, a general language-learning-
mechanism, cognitive awareness and structured input all play their part in
guiding the course of language acquisition. Unravelling the interdependence
of these factors constitutes the main goal of future child language research.

Task 2.6

Test your understanding of the study unit by choosing the option that fits best with the
given statement in the following multiple-choice questions. Circle the correct answer
([1], [2], [3] or [4]) and then check your answers against those given below.
1. The central problem of developmental psycholinguistics is
a how to explain the process of language acquisition
b how to explain language
c how to develop models for language teaching
d how to describe the stages of language learning.

2. Rationalists believe that

a the mind of a newborn baby is a tabula rasa
b the mind of a newborn baby already has a particular structure that influences
the way language is acquired

48
 c all evidence must be directly observable
d all evidence must be obtained through our senses rather than theoretical
assumptions.

3. The nature–nurture controversy in psycholinguistics concerns

a the relationship between innate and inborn factors
b the role the environment plays in language acquisition
c the role that caretakers play in motivating their children to learn language
d the way that factors internal to the child interact with the child’s social and
linguistic environment

4. Nativists believe that

a children acquire language by storing in memory all the words and sentences
they encounter
b the linguistic system is acquired slowly because of the immaturity and limited
experience of the acquirers
c children are genetically equipped to acquire any language
d imitation plays a crucial role in ‘cracking the linguistic code’

5. Emergentists argue that

a children use their general learning mechanisms and social experiences to
help them learn language
b children from different cultures acquire language in different ways
c language emerges instantaneously
d there are universal stages of language development

6. Because language acquisition is at least in part genetically determined, we know

that if a child born from English parents is reared in a monolingual Zulu-speaking
environment:
a the child will acquire Zulu
b the child will acquire English
c it is uncertain which of the two languages the child will acquire
d the child will be bilingual

7. Nativists believe that

a language is learnt by trial and error
b word meanings are programmed into the mind
c children are born knowing specific languages
d the ability to acquire language is genetically determined

8. Cognitive theory claims that

a complex problem solving cannot take place unless a certain level of lan-
guage competence has been reached
b children can only start to speak if certain cognitive skills develop first
c language development and cognitive development do not influence each other
d cognitive development and language development proceed simultaneously
and are dependent on each other

9. The theory that focuses on parental input and its influence on children’s language
development is known as
a nativism
b cognitive theory
c social interactionism
d connectionism

10. Which of the following statements does not reflect Chomsky’s theory of lan-
guage acquisition?
a In their first few years of life children develop a ‘language acquisition device’.
b Children have the ability to detect underlying regularities in the language

LIN2601/1/2017-2023 49
 around them and construct a set of basic rules that allows them to be infinitely
creative in the way they use their language.
c Children are born with a mental blueprint of universal linguistic rules and
principles known as Universal Grammar.
d The ability to learn language is an innate or inborn human ability in the same
way as learning to walk.

Feedback
1. [1] 2. [2] 3. [4] 4. [3] 5. [1] 6. [1] 7. [4] 8. [2] 9. [3] 10. [1]

Academic vocabulary building

Match words 1 to 8 with the appropriate definition on the right-hand side. Look
up the remaining four words in a dictionary and add their definitions in the space
provided. Look back at the definitions in the left-hand column of the study unit if
you need to.

1 empirical a) You something if you add extra

information to it.

2 distinction b) evidence is based on practical

experience and hard data rather than theory.

3 postulate c) A is a practical skill or method.

4 valid d) An argument is if it is based on the

truth or on sound reasoning.

5 supplement e) You reach a about something when

you form an opinion about it after considering it.

6 conclusion f) You make a between similar things if

you identify the ways in which they differ.

7 criteria g) are the principles you use

to help you make a judgement or decision. The singular
of this word is ‘criterion’.

8 technique h) You that something is true when

you assume it.

50
 9
controversies

10 indirect

11
phenomenon

12 observations

Further reading

Those of you who are interested in this topic might like to read the following:
Berko Gleason, J. 2005. The development of language. 6th edition. Boston, MA:
Pearson. Chapter 7 is on theoretical approaches to language acquisition.

Some of the main journals that address issues in language acquisition are the follow-
ing: Journal of Child Language, First Language, Language Acquisition, Applied
Psycholinguistics and Child Development.

There are some interesting FAQs about language acquisition on the Linguist List
website at https://linguistlist.org/ask-ling/lang-acq.cfm

Or watch a collection of video clips about child language acquisition at http://www.theen-

glishzone.org.uk/cla.html

Or search for articles and video talks by experts such as Chomsky and Pinker at
www.bigthink.com

Bibliography

Bates, E & McWhinney, B. 1987. Competition, variation, and language learning,

in Mechanisms of language acquisition, edited by B McWhinney. Hillsdale, NJ:
Erlbaum, 157–194.
Berko, J. 1958. The child’s learning of English morphology. Word 14:150–77.
Berko Gleason, J. 2005. The development of language. 6th edition. Boston, MA:
Pearson.
Berko Gleason, J, Johnson, J & Peters, J. 1990. Fine-tuning of utterance length
to preverbal infants: effects on later language development. Journal of Child
Language 17:511–526.
Berko Gleason, J & Bernstein Ratner, N. 1998. Psycholinguistics. 2nd edition. New
York: Holt, Rinehart & Winston.

LIN2601/1/2017-2023 51
 Bloom, P. 1993. Overview: controversies in language acquisition, in Language acquisi-
tion: core readings, edited by P Bloom. Hemel Hempstead: Harvester Wheatsheaf.
Brown, R. 1973. A first language: the early stages. Cambridge, MA: Harvard Uni-
versity Press.
Chomsky, N. 1965. Aspects of the theory of syntax. Cambridge, MA. MIT Press.
Crystal, D. 1997. The Cambridge encyclopedia of language. 2nd edition. Cambridge:
Cambridge University Press.
Dale, PS & Goodman, JC. 2005. Commonality and individual differences in vocabu-
lary growth, in Beyond nature-nurture: essays in honor of Elizabeth Bates, edited
by M Tomasello & DI Slobin (eds). Mahwah, NJ: Lawrence Erlbaum, 41–78.
Demuth, KA. 1984. Aspects of Sesotho language acquisition. Bloomington, IA: Indiana
University Linguistics Club.
Department of Linguistics. 1990. Study Guide 2 for LNG100S (Language in use).
Pretoria: Unisa.
Department of Linguistics. 1992. Only Study Guide for LNG2043. Pretoria: Unisa.
Department of Linguistics. 1996. Only Study Guide for LNG100S. Pretoria: Unisa.
Fenson, L, Marchman, V, Thal, D, Dale, P, Reznick, JS & Bates, E. 2007. MacArthur-
Bates Communicative Development Inventories: users’ guide and manual. 2nd
edition. Baltimore: Paul Brookes.
Foster-Cohen, SH. 1999. An introduction to child language development. New York:
Addison Wesley Longman.
Hollich, GJ, Hirsh-Pasek, K, Michnick Golinkoff, R, Brand, RJ, Brown, E, Chung,
HL, Hennon, E, Rocroi, C & Bloom, L. 2000. Breaking the language barrier:
an Emergentist Coalition Model for the origins of word learning. Monographs
of the Society for Research in Child Development 65(3):1–135.
Johnson, E & Zamuner, T. 2010. Using infant and toddler testing methods in lan-
guage acquisition research, Experimental methods in language acquisition research,
edited by E Blom & S Unsworth. Amsterdam: John Benjamins.
Piaget, J. 1926. The language and thought of the child. New York: Harcourt Brace
Jovanovich.
Rosch, EH, Mervis, CB, Gray, WD, Johnson, DM & Boyes-Braem P. 1976.
“Basic objects in natural categories”. Cognitive Psychology. 8(3):382–439.
doi:10.1016/0010-0285(76)90013-X
Slobin, DI. 1997. The Cross-Linguistic Study of Language Acquisition Vol 5: expanding
the contexts. Mahwah, NJ: Lawrence Erlbaum.
Stoll, S. 2009. Crosslinguistic approaches to language acquisition, The Cambridge
handbook of child language, edited by EL Bavin. Cambridge: Cambridge Uni-
versity Press.
Wardhaugh, R. 1992. An introduction to sociolinguistics. 2nd edition. Oxford: Basil
Blackwell.

52
 Study Unit 3 Study unit 3
The transition to speech
‘The mind is not a vessel to be filled, but a fire to be kindled.’
2

Plutarch (Greek historian, 45–120 AD)

3

OUTCOMES
After you have studied this study unit you should be able to
 distinguish between the prelinguistic and linguistic stages of language develop-
ment and describe their characteristic features
 distinguish between children’s perception and production of speech sounds and
explain why children’s production is less advanced than their perception
 describe the expected order in which words emerge, based on their meanings
 identify different kinds of errors in children’s early use of words
 provide definitions for the following important terms:

prelinguistic stage holophrase

linguistic stage substitution
joint attention assimilation
babbling reduction
reduplication underextension
protoword overextension
one-word stage

Preview
In this study unit we look at a child’s early years and discuss the question of
when children’s utterances move from meaningless cooing and babbling to the
meaningful use of language as a system for conveying messages. We look at
children’s early ability to perceive and distinguish different speech sounds and
at the development of the ability to produce speech sounds in the babbling stage
and throughout the preschool years. The child’s early words are also investigated
in terms of both their meanings and pronunciation.

LIN2601/1/2017-2023 53
 Introduction
The earliest sounds that babies make include crying, cooing, gurgling, coughing,
sneezing, etc. Would you consider these sounds as examples of language or not?
Take a minute or two to think about this question and fill in your answers in the
space provided before you continue reading.

These early sounds are not viewed as language because they are usually involuntary
A precursor responses to various sensations such as hunger, discomfort or wellbeing. Language,
is something in contrast, is voluntary and non-automatic – we say what we choose to say and are
that comes before not controlled by our physical state or environment. Psycholinguists refer to the
and is a sign of
a coming event. earliest period of language development, which covers approximately the first 12 to
18 months of the child’s life, as the prelinguistic stage. Although early vocalisations
like cooing and crying can be seen as the precursors of linguistic communication,
they differ from language in that the sounds do not yet have the formal structure
and symbolic characteristics of language.
What do we mean when we say language has a symbolic function? Symbolism means
using one thing (a symbol) to represent another thing or idea, which is exactly what
language does. Language uses strings of sounds (words) to represent various objects
and actions in the world. The task that children face is to learn to attach a label (a
string of sounds, i.e. a word) to a referent (i.e. object/event) and then generalise this
label to all other objects that belong to the same class or category. For example, the
child first associates the sound string d-o-g with a specific dog and later learns to call
all animals that belong to the class ‘dog’ by this label, irrespective of their shape, size
or colour, and irrespective of whether they are animate (i.e. alive) or inanimate (e.g.
a toy dog, a picture of a dog, etc.). The emergence of the child’s first words marks
the beginning of the linguistic stage – the stage at which children have worked
out the symbolic function of language and start using language to communicate
intentionally and purposefully.
In this study unit we focus on the prelinguistic stage of language development and
the transition from the prelinguistic to the linguistic stage. The child’s language
development during this period is placed in the broader context of the child’s overall
development and maturation.

Task 3.1

Instead of writing a child’s age in months, for example 18 months, psycholinguists

frequently use the notation 1;6. The first number (before the semicolon) is the child’s
age in years, and the second number (the number after the semicolon) is the additional
number of months, so in this case the child is 1 year and 6 months old, i.e. 18 months.
Try writing the following ages using the semicolon notation explained above.
1. 2 months

54
 2. 12 months

3. 18 months

4. 2 years

5. 4 years and 3 months

Feedback
Make sure that you use a semicolon and not just a comma when you use this notation.

1 The prelinguistic stage

Before a baby learns to match words and the concepts they refer to, she has to
pass through several important preliminary stages, and familiarise herself with
nonlinguistic forms of communication such as facial expressions, gaze and gestures.
In their first year, children are often heavily dependent on non-linguistic methods
Gaze refers to of communication such as actions, gestures, facial expressions and tone of voice. Some
the direction in of the earliest sounds include involuntary grunts and sighs. Smiling begins within
which someone is the first couple of months and is initially an involuntary muscle spasm. Gradually
looking.
babies realise that smiling can elicit a response from other people and begin to use it
as a conscious behaviour. Crying is also initially a response to an internal physical
state – to hunger, discomfort etc – but soon becomes used as a way to elicit responses
from others. From the second month of life babies also start to use voluntary contented
cooing sounds. Months 4 to 7 are characterised by vocal play, with the baby
experimenting with the vocal organs, producing shrieks, murmurs, growls and shouts.
Throughout these early stages of vocalisation, children show quite a lot of
responsiveness to other people. They prefer to vocalise when the other person
is not, rather than at the same time (i.e. they engage in rudimentary turn-
taking), to imitate sounds produced by others, and generally behave more and
more as if they are in genuine conversation (Foster-Cohen 1999:20).

Pointing and reaching are two of the earliest gestures used by children. Again they
are initially used merely to grasp or to explore an object with a fingertip – without
any communicative intention. An important milestone is reached sometime after 9
months when children realise that smiles, cries, coos and gestures such as reaching
and pointing can be used intentionally to communicate with others. These gestures
then become associated with a meaning – ‘give me that’ in the case of reaching and
‘look at that’ in the case of pointing. This allows children to share interest in an
object with another person, a phenomenon known as joint attention.
Children who don’t realise that they can get adults to do their bidding, i.e.
who don’t realise that gestures communicate, will simply gaze at the object
reached for or pointed at. Children who realise that these gestures commu-

LIN2601/1/2017-2023 55
 nicate will glance to check that the person they are with has seen the gesture
and is either complying with the request or responding to the point. Children
who realise that gestures communicate will also tend to continue pointing
or reaching until a response is forthcoming, and may also show distress at an
object or person until there is a response (Foster-Cohen 1999:27–28).

The realisation that sounds and gestures can be used in a symbolic way to convey a
particular message is an important step on the way to realising that sound strings
(words) can be used symbolically.

Task 3.2

Classify each of the following behaviours as either unintentional or intention-

al communication.
1. The child shakes her head to signal ‘no’.

2. The child smiles to get the caretaker to interact with her.

3. The child uses a reaching gesture to grasp objects.

4. The child uses pointing to direct the caretaker’s attention to something.

5. The child ‘smiles’ because of indigestion.

6. The child produces cooing sounds in response to the caretaker’s utterances.

7. The child cries as a result of the discomfort of a wet nappy.

Feedback
Behaviours 3, 5 and 7 are examples of unintentional communication. These actions are
responses to environmental sensations and are not intended to communicate a mes-
sage to anyone else. Behaviours 1, 2, 4 and 6 are examples of intentional communica-
tion. The child uses these behaviours in a voluntary way to communicate something
or to elicit particular behaviours from others. Note that there are three elements in the
intentional behaviours – the child, the message and the caretaker, indicating the child’s
ability to engage in joint attention.

1.1 Early perception and production of speech sounds

There is strong biological evidence that indicates that the ability to perceive human
speech sounds is operative from birth. Ingenious techniques have been devised to

56
 measure speech perception abilities in newborn babies. One technique is to use a
dummy connected to a sound system. Every time the baby sucks on the dummy, a
sound is produced. Babies tend to suck frequently at first, an action which researchers
interpret as an indication that the baby is attending to the sound. As the baby loses
interest or get bored with the sound, the sucking becomes less frequent. As soon
as the researcher changes the sound that is produced, the baby starts to suck more
vigorously again (Berko Gleason 2005:71). Experiments that use these techniques
show that newborn babies can discriminate between human speech sounds and all
other sounds that occur in their environment.
The ability to hear and distinguish between various sounds is an important prerequisite
for acquiring language. It seems that from birth children are genetically capable of
‘tuning into’ human speech and ‘tuning out’ nonspeech. (In study unit 6 you will
read about autistic children whose ability to ‘tune into’ human speech is seriously
impaired.) It also appears that babies’ speech perception is fairly refined. For example,
babies as young as three days old can distinguish their own mother’s voice from other
female voices. This not only helps in the fight for survival, it also ensures that babies
are particularly receptive to the speech of the person who is most likely to interact
and talk to them the most, i.e. the mother. Babies can perceive contrasts between
sounds just as adults can, for example, one-month-old infants can perceive differences
between various consonants (e.g. pa : ba), and vowels (e.g. pa : pi). Patricia Kuhl (e.g.
Kuhl, Stevens, Hayashi, Deguchi, Kiritani & Iverson 2006, Moon, Lagercrantz &
Kuhl 2012) has shown that infants can also categorise speech sounds on the basis
of common or shared features, regardless of whether the sounds are from their own
or another language. In her research she found that babies who were 5 to 10 months
old could recognise the same vowel even when the speakers’ voices differed with
regard to pitch and sex of speaker.
This natural ability to distinguish certain contrasts between sounds seems to diminish
as the child grows older and gains more exposure to the sounds of the particular
languages spoken in his or her linguistic community. It was found, for example, that
6-month-old English babies could discriminate between different kinds of r-sounds
used in Hindi but not in English. In other words, these babies had the potential to
distinguish subtle contrasts between sounds even though these sounds did not occur
in their own language environment. However, when the English babies were tested
at 8 months, they failed to distinguish the differences between these same sounds.
Researchers concluded that between 6 and 12 months, children lose the ability to
discriminate between sounds that are non-functional (not used to distinguish words)
in the language they are acquiring (Kuhl et al, 2006, Moon et al 2012).

For interest, you may want to watch this long but fascinating video in
which psycholinguist Professor Patricia Kuhl explains some of the latest
w
experimental techniques for studying babies’ language at
https://www.youtube.com/watch?v=M-ymanHajN8

From the above evidence it is clear that children are born with fairly sophisticated
Auditory means auditory mechanisms and brain functions that enable them to focus on human
related to our speech sounds and to be able to categorise these sounds. However, children’s ability
sense of hearing. to produce speech sounds takes longer to develop. Even though the ability to perceive
speech sounds is present from birth, the ability to produce sounds that resemble
human speech sounds only emerges gradually as a result of maturational changes

LIN2601/1/2017-2023 57
 in the first months. A newborn baby’s articulatory system is not the same as an
adult’s. In particular, a baby’s tongue is relatively large in proportion to the oral
cavity and fills almost the entire mouth, allowing little room for movement. This is
to prevent the baby from choking when drinking milk. The tongue size and fairly
inflexible position means that a baby’s earliest crying consists mostly of the [æ] sound
as in the word cat. In adults the tongue can assume a much larger variety of
configurations. The larynx or voice box is also much higher in babies, which makes
it impossible to produce all the vowels that occur in adult speech. Compare the size
and position of the infant’s vocal organs https://www.researchgate.net/
figure/12780226_fig1_Fig 1 Major anatomic differences between the infant and
the adult vocal tracts Drawing.
During the second and third month of life, the head and neck areas grow fairly
rapidly, enabling the baby to raise his or her tongue and move it around with
greater flexibility and to change the shape of the lips, allowing more varied sound
production. Between the third and fourth month consonants start to appear in
children’s cooing sounds.
Another reason why children’s ability to produce speech sounds takes time to develop
is that the brain weighs only about 300 g at birth because the brain cells (neurons) are
not all connected to each other. Babies are born with all their brain cells and do not
develop any additional ones after birth. However, from about the fifth month until
about 2 years, brain development occurs as more and more neuronal connections
are made. The brain gradually increases in weight to about a kilogram. This allows
greater communication between the brain and the articulatory organs and therefore
greater control over the speech sounds produced.
The adjective
cortical is derived Figure 3.1 Cortical development over time
from the noun
cortex. (http://www.urbanchildinstitute.org/sites/all/files/databooks/2011/ch1-fg3-synapse-
density-over-time.jpg)

2 Babbling
The babbling stage begins when children are about six months old (0;6). Children
start to explore the abilities of their vocal organs, and soon discover that they can
make loud and soft noises, and can alter the pitch of their utterances from high
shrieks to deep grumbles. The early period of babbling does not seem to be intentional
or communicative. The first stage of babbling often involves reduplication, i.e.
producing vocalisations containing repeated sequences of consonants and vowels
such as baba or dadada. Most babbling consists of open syllables, i.e. syllables that
consist of a consonant followed by a vowel (CV). Babbled syllables ending in a
consonant (closed syllables) or containing consonant clusters (two consonants next
to each other) are less frequent. Consonants like /b/, /p/, /m/, /w/, /n/, /t/ and /d/
are amongst the first to be used, and are an important milestone, indicating that
the baby has gained considerable control over the vocal organs – tongue, lips and
vocal cords.

58
 A note on phonetic transcription
The forward slashes around the speech sounds above are used for transcrip-
tions which accurately reflect the child’s pronunciation. Square brackets are
sometimes used instead of forward slashes, but the principle is the same. Some
of you may have been introduced to phonetic transcription in other language
courses, or in the LIN2603 course on sound and sound structure. Phonetic
transcription is basically a way of writing utterances using symbols to represent
sounds. Each symbol represents one sound, so these transcriptions are clear
and unambiguous, unlike normal spelling. Phonetic transcriptions provide a
useful way of indicating the actual pronunciation of the child. You do not need
to know how to write sounds using phonetic symbols, but we will gradually
start introducing you to phonetic transcriptions together with explanations
of the pronunciation these transcriptions represent. If you want to download
some free phonetic fonts for your device go to http://www.sil.org/resources/
software_fonts/search?f[0]=field_sf_category%3A36519.

Children often produce sounds that do not occur in the languages spoken around
them. Babbling occurs spontaneously and universally and the sounds that are
babbled tend to be similar in all children, irrespective of the language they hear
around them, i.e. the babbling of babies from speech communities as diverse as
Zulu, English, Chinese and Arabic will sound the same. Deaf babies also coo and
babble in the same way as hearing babies even though they cannot perceive auditory
stimuli. Some studies have shown that parents’ attention and responsiveness to
children’s babbling can increase the quantity but not the quality of babbling, i.e.
the children may babble more but they do not increase the range of sounds that are
produced. This evidence suggests that the emergence of babbling relies largely on
internal biological factors, rather than on external factors such as linguistic input
and encouragement from parents and caretakers.
From about the ninth month children’s babbling becomes more varied. For example,
instead of the reduplicative baba one may hear babbled utterances like maba, titu,
anadi, etc. New consonants like /f/, /s/ and /r/ also start emerging. During this stage
children also learn to imitate the stress or intonation patterns of the sentences they
hear around them. Babbled strings of sounds are produced with a variety of stress
and intonation patterns that resemble those of simple sentences, for example rising
intonation at the end of a sequence to signal a question. Children often babble
like this in play situations, for example when they talk on a toy phone or when
they page through books – from their vocalisations they appear to be talking to
someone, commenting on the pictures in the books or telling a story.
There is no clear break between the prelinguistic and the linguistic stages, between
babbling and the emergence of the child’s first words. Even when children start to
use words, they do not suddenly cease babbling. There is a period of overlap between
the two during which babbling is interspersed with real words. Babbling gradually
decreases as the child acquires more words and speech increases in frequency. The
following table gives data on the percentage of babble in two children’s early utterances.
(The two columns do not add up to 100% as some of the children’s vocalisations
were grunts of various kinds):

LIN2601/1/2017-2023 59
 Table 3.1 Percentage of babble in two children’s utterances from 11 to 16 months
(based on Vihman 1996:132)

Age in Babble (%) Words (%)

months
Lerato
11 73 18
12 61 8
13 67 12
14 42 49
15 28 52
16 17 51
Nkateko
11 92 0
12 85 0
13 69 4
14 73 10
15 35 12
16 52 24

Task 3.3

Look at the table above and then answer the following questions.

1. At what age did Lerato start including words in her utterances? Give your answers
using the semicolon notation explained in Task 3.1.

2. At what age was there a sudden surge in the percentage of words used by Lerato?

3. At what age did Lerato’s utterances contain more words than babble?

4. At what age did Nkateko start including words in his utterances?

60
 5. How do the two children compare in terms of their rate of development?

Feedback
1. Lerato started including words in her utterances before 11 months (0;11). We are
not given data before this so we cannot be sure exactly when.
2. There was a sudden surge in the percentage of words used by Lerato at 14 months
(1;2). At 13 months only 12% of her utterances were words and at 14 months 49%
of her utterances were words.
3. Lerato’s utterances contained more words (49%) than babble (42%) from 14
months (1;2).
4. Nkateko started including words in his utterances at 13 months (1;1).
5. Lerato’s development is slightly faster than Nkateko’s. Lerato used words earlier
and has a smaller proportion of babble in her speech at each age. Note that the
difference is only a few months though – Lerato was 13 months old and Nkateko
was 15 months old when 12% of their utterances were words. The basic pattern
of development is the same although the rate may differ slightly from child to child.

During the later stage of babbling, certain forms start being used fairly consistently
Behaviour is in certain contexts, accompanied by eye contact with the caretaker and gestures like
idiosyncratic reaching, pointing, grasping and rejection movements. These word-like forms are
if it is unusual known as protowords and are basically idiosyncratic words invented by the child
and associated
with a particular rather than modelled on actual adult words. Protowords are used fairly consistently
individual. to refer to a particular concept in an attempt to communicate. Reich (1986:37) gives
the example of his son, Quentin, who at age 1;1 used the sound ‘uhuh’ together
with a pointing gesture to mean something like What is this called? or Give me the
name of this object.

Task 3.4

The data below (Reich 1986:56) gives information about the forms and meanings of
Joan’s first few words. Which of Joan’s words (in the first column) are protowords, and
which are attempts to produce the adult word? Place a tick next to the words that you
consider to be protowords. (The symbol /a/ represents the sound u as in cup.)

Child’s word Meaning

/ap/ up
/ba/ bottle
/bas/ bus
/baza/ put on
/da/ down
/bat/ pocket
/ba/ pie
/bap/ lamb
/nas/ nice

LIN2601/1/2017-2023 61
 Feedback
You should have ticked the protowords /baza/ ‘put on’ and /bap/ ‘lamb’ as these appear
to be made-up words that are unique to the child and are not based on adult words.
The other words all appear to be genuine attempts to produce the adult forms, although
many of these are simplified in some way as a result of the child’s limited control over
her vocal organs.

3 The one-word stage

The period from the child’s first word to the time when the child starts to put two
words together is known as the one-word stage. This period lasts roughly from age
1 to 2. Why does language emerge in all children at about the same time? Those
who subscribe to a nativist view (see section 4.2 in study unit 2 above) answer
this question by suggesting that language is set in motion by an internal ‘alarm
clock’, preprogrammed to go off at a certain time, in the same way as our bodies
are preprogrammed to mature sexually at about age 13. Like sexual development,
language development cannot take place before this ‘alarm clock’ has gone off,
i.e. before the necessary maturation of body and brain has taken place. Imagine
language as a seed that is planted in our minds before birth. The seed contains the
potential to be a mature tree but it only starts to grow after a period of time has
passed. In the same way, the emergence of language is not the result of a conscious
decision made by the baby, but of the inner biological clock that preprogrammes
the pace of development.
One of the major problems in the study of the one-word stage is determining what
the child actually means when she uses a word. Because context, gestures and
intonation patterns always accompany early word use, they provide caretakers (and
researchers) with clues as to what the child might possibly be talking about. Even so,
we cannot always be sure about the child’s meaning. For example, if a child points
to the bath, on the edge of which her yellow rubber duck is sitting, and says barf, she
could be referring to (i.e. naming or labelling) the concrete object ‘bath’, the activity
of bathing, or the rubber duck which she associates with the activity of bathing. The
child’s early words are referred to as holophrases (holo means ‘complete’) because
the child seems to compress so much meaning into a single word. A holophrase is
basically a one-word sentence – using a single word to express a complex idea.

3.1 Children’s acquisition of speech sounds

Children’s first words follow the same articulatory patterns as babbling, i.e. a
preference for reduplication and open syllables, with the crucial difference being that
words have a fairly stable form and are used to communicate a particular meaning.
Consonants made with the lips and the tip of the tongue, such as /p/, /b/, /m/, /w/,
/t/, /d/ and /n/ are the most common. As we have seen with babbling, the vowel
sounds used by children in their early words usually include /a/ (as in mama), /i/
(pronounced ee as in see) and /u/ (pronounced oo as in spoon).

IIf you want to hear an example of a baby in the one-word stage, watch
https://www.youtube.com/watch?v=xwglVbiIvtc&ebc=ANyPxKrbQjkR
doFl8jsoGXcgGLEgQQOJaQnUZ8-A8roOYP9rIfZ2I3GUrMKx5quM
BqHvgf4trE7evvywbfIzwyvgwQzGMEYHbg

62
 The syllable structure of children’s words is also fairly simple. If we use C as the
abbreviation for a consonant, and V as the abbreviation for a vowel, the most likely
patterns are V (a vowel on its own), CV (consonant followed by a vowel) and VC (a
vowel followed by a consonant). Reduplicated CVCV also occurs frequently, where
the second syllable is the same as the first, for example baba, dada etc.

Task 3.5

The following data lists the sounds and syllable structures used by my son Sibusiso
in his first 25 words. Look at the data and then answer the questions that follow. The
phonetic symbols you need to know are:
/a/ = cup
/i/ = see
/o/ = door
/æ/ = cat
/ʋ/ = look
/u/ = zoo
/j/ = yes
// = think

FORM MEANING FORM MEANING

/aju/ hello /na/ no
/ap/ up /nana/ mommy
/api/ up /mama/ mommy
/baba/ byebye /po/ pour
/bədi/ bird // fish
/bʋs/ push /si/ see
/bit/ bed /siæt/ see that
/da/ that /ta/ give it to me
/de/ there /ta/ take this
/dada/ daddy /jat/ light
/hat/ hot /u/ me (refers to himself)
/kɑ/ car /wəti/ water

1. Give an example of a word with a CV syllable structure in Sibusiso’s vocabulary.

2. Give an example of a word with a VC syllable structure in Sibusiso’s vocabulary.

3. Give an example of a word with a CVCV syllable structure in Sibusiso’s vocabulary.

4. Which of these three syllable structures is the most common in Sibusiso’s early
words?

5. Give an example of a reduplicated word in Sibusiso’s vocabulary.

6. What are the most common consonant sounds in Sibusiso’s early words?

7. What are the most common vowel sounds in Sibusiso’s early words?

LIN2601/1/2017-2023 63
 Feedback
1. In Sibusiso’s first 25 words, /da/, /ka/, /na/, /po/, etc. have a CV syllable structure.
2. /ap/ has a VC syllable structure.
3. /bədi/, /baba/, /wəti/ etc. have a CVCV syllable structure. (You only had to give
one example in each case.)
4. Sibusiso produces 7 words with a CV syllable structure, so this is the most com-
mon syllable structure for this child.
5. Reduplication is the repetition of syllables, as in Sibusiso’s words /baba/, /dada/
and /mama/.
6. As expected, the most common consonant sounds in Sibusiso’s early words are
/b/, /p/, /t/ and /d/, all produced with the lips or tongue tip.
7. The most common vowel sounds in Sibusiso’s early words are /a/ and /i/.

Usually, children’s pronunciation improves gradually and their words come to sound
more and more like the adult target words. However, children’s pronunciation
sometimes gets worse before it gets better (Foster-Cohen 1999:35). The following
example (Reich 1986:55) shows that a child who could pronounce the word pretty
correctly at 10 months, later goes through almost a year of various incorrect attempts
to pronounce this word.
0;10 pretty (adult-like pronunciation)
1;0 prrty (with no vowel)
1;1 pretty, preety
1;3 pshitty
1;4 pwitty, pity, yitty
1;9 biddy

The thing to notice here is that although children can often produce sounds from
an early age, it takes several months to master a sound, i.e. to gain sufficient control
over their vocal organs to produce the same sound every time it is required. Some
sounds take longer to master than others. We have already mentioned that sounds
produced with the lips and the tip of the tongue are the earliest to be used and are
mastered by the age of two in the average child. Sounds like /r/, /l/ and /s/ are only
mastered at about 3 years, while the sound /ʒ/ (the consonant in the middle of the
word Asia) is only mastered by about 6 years. Of course not all languages use all
of these sounds, but the sounds that occur very frequently in the world’s languages
tend to be the ones that children learn first. In contrast, rare sounds like /ʒ/ and the
click sounds of Zulu, Xhosa and other Nguni languages are often mastered fairly
late by children learning these languages. The table below (adapted from Reich
1986:57) illustrates the approximate sequence and age at which various consonants
are mastered. Examples of words containing each sound are given if it is non-obvious.
Consonant Average age mastered
/p/ /m/ /n/ /h/ /w/ /b/ 1;6
/k/ /ɸ/ /d/ /t/ // as in sing 2
/f/ /j/ as in you 2;6
/r/ /l/ /s/ 3
/t∫/ as in chew, /∫/ as in shoe 3;6
/z/ /v/ /dʒ/ as in jump 4

64
 /θ/ as in thing 4;6
/ð/ as in then 5
/ʒ/ as in Asia 6

Task 3.6

Which of the following words would you expect a 3-year-old child to have difficulty
producing? Consult the table above to help you answer this question.
Pig zip cat bath

Feedback
The p and g sounds of ‘pig’ should not pose a problem to a 3-year old as these sounds
are acquired at approximately age 1;6 and 2 respectively. A 3-year-old may have prob-
lems saying ‘zip’ as the [z] sound is only mastered round about 4 years. The /k/ and /t/
sounds in ‘cat’ are acquired at about age 2. However, the /θ/ in ‘bath’ is one of the last
sounds to be mastered, at about 4;6 years. ‘Bath’ and ‘zip’ may thus pose pronunciation
difficulties for a 3-year-old.

Since we have shown that children’s perception abilities are well developed from
birth, it appears to be children’s control over their vocal organs that prevents them
mastering certain sounds, rather than an inability to discriminate between sounds.
Children often understand words and sound contrasts that they are not yet able
to pronounce themselves. For example a child may be able to point correctly to a
picture of a coat or a goat when each is mentioned, while pronouncing both as coat.
What strategies do children use when they cannot yet produce a certain speech sound?
One option is simply to avoid using words that contain the problematic sound. Here
are some other strategies, with examples from Sibusiso’s first 25 words listed earlier.
 substitution occurs when one sound is replaced by another one, e.g. that becomes
/dat/ (d is substituted for the more difficult /ð/-sound.)
 assimilation occurs when a sound becomes more like the one next to it, e.g. when
mommy becomes /mama/ the child makes the second vowel the same as the first
one. Assimilation can also affect consonants, e.g. when rabbit is pronounced as
/bæbit/ the child makes the first consonant the same as the second one.
 reduction occurs when children omit sounds, e.g. see that becomes /siæt/. This
often happens in unstressed syllables or at the end of a word.

Task 3.7

Which of the three strategies mentioned above are used in the following child pronuncia-
tions? Hint: Some of the words may make use of more than one simplification strategy.

LIN2601/1/2017-2023 65
 /da/ down

/ba/ bottle

/bʋs/ push

/za/ that

/bat/ pocket

/dodi/ doggy

Feedback
/da/ involves reduction (deletion) of sounds at the end of a word, a very typical strategy
in children’s early speech. /ba/ involves reduction of the unstressed second syllable of
‘bottle’. /bʋs/ involves two instances of substitution – /p/ is pronounced /b/ and the sh-
sound /ʃ/ is pronounced /s/. /za/ uses sound substitution as the initial sound of ‘that’ is a
late sound to be acquired, and also uses reduction of the final consonant. /bat/ involves
reduction of the unstressed second syllable and substitution as the initial /p/ of ‘pocket’
becomes a /b/. Assimilation takes place in the word /dodi/ as the /g/ is changed so that
the two consonants are pronounced the same.

3.2 Which words come first?

Children’s first words refer to people and objects in their immediate environment,
sounds, games, action verbs and words that facilitate social interaction. Reich
(1986:44) classifies children’s words into slightly more detailed categories. The
categories at the top of the list are the most common, and the categories towards
the bottom of the list occur less frequently:
 objects and living things that move, e.g. mama, dog
more common
 objects that the child can manipulate, e.g. ball, juice
 action verbs, e.g. give, push
 adjectives and adverbs that describe characteristics of something, e.g. dirty, outside
 personal-social words (words that express emotional states and perform social
functions), e.g. no, please
 function words (short words like to, the, can, is, etc. that serve a grammatical
function). Although function words occur very frequently in a language, children
less common
tend to ignore them. Two possible reasons include the fact that function words
are generally unstressed and therefore more difficult for children to perceive than
other words, and the fact that they add little to the meaning and may be omitted
without affecting communication.
 names for clothing the child cannot easily manipulate, e.g. nappy, trousers
 names for places in the environment, e.g. bedroom

66
 Task 3.8

Look again at Sibusiso’s first 24 words. Do the word categories at the top of Reich’s list
occur more frequently than those further down the list?

FORM MEANING FORM MEANING

/aju/ hello /na/ no
/ap/ up /nana/ mommy
/api/ up /mama/ mommy
/baba/ byebye /po/ pour
/bədi/ bird // fish
/bʋs/ push /si/ see
/bit/ bed /siæt/ see that
/da/ that /ta/ give it to me
/dea/ there /ta/ take this
/dada/ daddy /jat/ light
/hat/ hot /u/ me (refers to himself)
/kɑ/ car /wəti/ water

Feedback
The first thing we notice about Sibusiso’s vocabulary is that he has different forms that
have the same meaning, for example /nana/ and /mama/ both mean ‘mommy’. This
indicates that he has not yet mastered the pronunciation of these words. Sibusiso has
6 words for objects and living things that move (2 words for ‘mommy’ and one each
for ‘daddy’, ‘fish’, ‘bird’ and ‘me’). He has 4 words for objects he can manipulate, in this
case ‘car’, ‘water’, ‘bed’ and ‘light’ (I spent a lot of time holding him up to switch on light
switches and sitting in the car while he manoeuvred the steering wheel …). He has 8
action words (the words for ‘push’, ‘pour’, ‘see’, ‘see that’, ‘take this’, ‘give it to me’ and
2 words for ‘up’). He uses only one adjective (/hat/ = ‘hot’), 3 personal-social words (the
words for ‘no’, ‘hello’ and ‘byebye’) and 2 function words (‘that’ and ‘there’). It seems
to be true that most of Sibusiso’s early words refer to people/objects (10 out of 24) and
action verbs (8 out of 24). As Reich suggests, there are relatively few adjectives and
function words in his vocabulary.

Evidence from children’s early words show that children do not merely imitate the
language of the environment. If they did, we would expect them to use the most
frequently used words that occur in adult language first. In a language such as
English, articles and pronouns like the, a, which and that would then be amongst
the first words. Yet in most languages these are the words that are usually acquired
late. Everywhere in the world children begin by acquiring content words such as
nouns and verbs. Chomsky would argue that children are born with an advanced

LIN2601/1/2017-2023 67
 idea of what language looks like and that they therefore ‘know’ what types of
structures they are looking for when they begin to acquire language. On the other
hand, social interactionists would argue that children learn content words like
nouns and verbs first because these are particularly salient or observable in the
environment, are frequently pointed out or labelled by parents and are particularly
useful for communicative purposes.
When children acquire their first words, they may use them in a rather restricted way,
for example the word apple may initially be used to refer only to reddish apples and
not also to green or yellow apples. A word like doggie may initially refer only to the
family dog and not to dogs in general. The technical term for this is underextension,
as the child’s range of referents for a word is narrower than in adult language. The
opposite can also happen. When the child overextends words, the range of referents
is greater than that of adult language, for example the word apple may be used to
refer to all kinds of roundish looking fruit like apples, pears, peaches and apricots.
To give another example, one 22-month old child used the word bunny to refer to
rabbits, pictures of kangaroos in a book, and long haired cats, due most probably
to the common features of hopping (rabbits and kangaroos) and long fur (rabbits
and cats). According to Foster-Cohen (1999:136), children overextend over 40%
of their first words.
The number of words in a child’s vocabulary increases rapidly, and a vocabulary
spurt (i.e. a sudden increase in the rate at which new words are learnt) has been
observed to occur at around 50 words. The vocabulary spurt takes place at around
16 months on average (Tomasello & Slobin 2005:67). Clark (2009:283) indicates
that between the ages of one and two, most children learn 200–600 words, and this
grows to around 14 000 words (nine new words a day) by age 6. By early adulthood,
vocabularies range from 50 000–100 000 words.

4 Summary
In this study unit we have considered the transition from prelinguistic communication
to children’s first meaningful words. As children engage in interaction and joint
attention with their caretakers their utterances develop from cooing to babbling,
through to protowords and holophrases. We considered the developing phonological
abilities that occur during the first year and the types of words and lexical errors
that typically appear first. We also looked at some statistics relating to the number
of words children use and at how this relates to the input they receive at home. In
the next study unit we look beyond the one-word stage to the earliest beginnings
of grammar.

Task 3.9

Test your understanding of the study unit so far by choosing the option that fits best
with the given statement in the following multiple-choice questions. Circle the correct
answer ([1], [2], [3] or [4]) and then check your answers against those given below.

1. The child’s vocalisations during the prelinguistic stage are not considered to be
real language because
a they are not made up of the same speech sounds as language

68
 b they are responses to physical sensations and are therefore not voluntary
c they do not have the symbolic characteristics of language
d both [2] and [3].

2. 16 months can also be expressed in the following way:

a 1;6
b 1:6
c 1;4
d 1:4
3. Which of the following words would you expect to find in a child’s earliest vocabulary?
a please, yes, byebye
b cold, hungry, red
c in, there, the
d daddy, bottle, doggy
4. The words please, yes, byebye in Question 3 above are known as
a function words
b personal-social words
c protowords
d adjectives
5. The words in, there, the in Question 3 above are known as
a function words
b personal-social words
c protowords
d modifiers
6. A child who produces contentless utterances such as wawawa and bobo is prob-
ably in the stage
a reduplicative babbling
b variegated babbling
c protoword
d one-word
7. A word invented by a child and used in a fairly consistent form to refer to a par-
ticular thing is known as
a a function word
b a holophrase
c a protoword
d an action word.
8. Sounds that are often mastered fairly late in children’s speech include
a /p/, /b/, /m/ and /n/
b /k/, /g/ and /t/
c /f/ and /s/
d /tʃ/ (‘ch’), /ʃ/ (‘sh’) and /θ/ (‘th’).
9. The Northern Sotho baby-talk word nana (‘baby’) is formed from the adult term
ngwana through the strategy of
a substitution
b assimilation
c reduction
d All of the above.
10. A Northern Sotho child who produces the word fiša (‘hot’) as šaša is using the
strategy of
a substitution
b assimilation
c reduction
d All of the above.

LIN2601/1/2017-2023 69
 11. An example of underextension is
a when a child uses the word juice to refer to all drinks
b when the child uses the word juice to refer to milk but not juice
c when the child uses the word juice to refer to orange juice but not other kinds
of juice
d when the child uses the word juice in the same way as an adult.

Feedback
1. [4] 2. [3] 3. [4] 4. [2] 5. [1] 6. [1] 7. [3] 8. [4] 9. [3] 10. [2] 11. [3]

Academic vocabulary building

Choose the correct word from the left-hand column to complete the definitions on
the right-hand side. Look up the remaining words in a dictionary if you are unsure
of their meanings, and write down their definitions in the margin.
references/preferences Your are the things you like
or prefer.
strategies/strategists are plans about how to do
something.
perceive/receive You something
when you notice it or become aware of it.
unique/universal Something is when it is very
widespread or found all over the world.
evidence/evident is anything that suggests or
proves that something is true or has happened in a certain
way.
approximate/estimate You something when you
try to judge it without measuring exactly.
proportion/proposition The of something is how
big that part is in relation to the whole thing.

Further reading

Those of you who are interested in this topic might like to read the following:
Berko Gleason, J. 2005. The development of language. 6th edition. Boston: Pearson.
Chapters 3 and 4.

70
 Bibliography
Berko Gleason, J. 2005. The development of language. 6th edition. Boston, MA:
Pearson.
Clark, EV. 2009. First language acquisition. Cambridge: Cambridge University Press.
Department of Linguistics 1990. Study guide 2 for LNG100S (Language in use).
Pretoria: Unisa.
Department of Linguistics 1992. Only study guide for LNG2043. Pretoria: Unisa.
Department of Linguistics 1996. Only study guide for LNG100S. Pretoria: Unisa.
Foster-Cohen, SH. 1999. An introduction to child language development. New York:
Addison Wesley Longman.
Hart, B & Risley, T. 1995. Meaningful differences in the everyday experience of young
American children. Baltimore, ML: Paul H. Brookes.
Hoff, E. 2013. Interpreting the early language trajectories of children from low SES
and language minority homes: Implications for closing achievement gaps.
Developmental Psychology 49(1):4–14.
Kuhl, PK, Stevens, E, Hayashi, A, Deguchi, T, Kiritani, S & Iverson, P. 2006. Infants
show a facilitation effect for native language phonetic perception between 6
and 12 months. Developmental Science 9:13–21.
Moon, C, Lagercrantz, H & Kuhl, PK. 2012. Language experienced in utero affects
vowel perception after birth: a two-country study. Acta Pediatrica 102:156–160.
Nation, ISP. 1995. A brief critique of Hart B & T Risley. Meaningful differences in
the everyday experience of young American children. Baltimore: Paul H. Brookes.
Reich, PA. 1986. Language development. New Jersey: Prentice-Hall.
Tardif, T, Fletcher, P, Liang, W, Zhang, Z, Kaciroti, N & Marchman, VA. 2008.
Baby’s first 10 words. Developmental Psychology 44(4):929–938.
Tomasello, M & Slobin, DI. 2005. Beyond nature-nurture: essays in honor of Eliza-
beth Bates. Mahwah, NJ, London: Lawrence Erlbaum.
Vihman, MM. 1996. Phonological development: the origins of language in the child.
Clinical Linguistics and Phonetics 12(2):149–152. 

LIN2601/1/2017-2023 71
 Study Unit 4 Study unit 4
Putting words together
‘When putting words together it is good to do it with nicety and caution,
2

your elegance and talent will be evident if by putting

3

ordinary words together you create a new voice.’

4

Horace (Roman poet, 65–8 BC)

5

OUTCOMES
After you have studied this study unit you should be able to
 describe children’s linguistic developments from age 2 to 5
 describe children’s linguistic developments during the primary school years
 calculate the MLU of a speech sample
 provide definitions for the following important terms:

two-word stage prereading stage

telegraphic speech initial reading
MLU fluency stage
multiple-word stage communica- reading to learn
tive competence metalinguistic development
‘top-down’ vs ‘bottom-up’ word
recognition

Preview
In this study unit we investigate some of the ways that children’s speech changes
and develops after age 2. Language development during this period is still
fairly uniform among children from different language environments. Focus-
ing on what is universal in language acquisition, we look at children’s growing
language competence as reflected in the increasing length and complexity of
their sentences. We will focus on describing children’s speech at all linguistic
levels – phonological, morphological, syntactic, semantic and pragmatic, and
will start to put our skills and knowledge into practice by analysing some actual
samples of child language. Finally we focus on language development in the
school years, including the development of reading and storytelling skills and
the ability to think and talk about language and how it works.

72
 Introduction
We mentioned in study unit 3 that the linguistic phase signals the emergence of
intentional, communicative use of language. Children start by producing one-word
utterances, then two-word utterances. As they get older, more words are added and
their utterances get longer and more complex. In this study unit we look at some of
the major linguistic developments from age 2 to the early school years.

1 The two-word stage

An important stage of linguistic development is reached when children start putting
two words together to form simple sentences. This usually occurs between the ages
of approximately 15 months and 2 years and marks the beginning of the two-word
stage. Children now tune into the unique structural properties of the language to
which they are exposed. As we now know, children don’t simply imitate the language
they hear around them but often come up with creative utterances like byebye car
and allgone biscuit. This kind of evidence suggests that children are unconsciously
constructing utterances according to some pattern, although this pattern may differ
from the rules and patterns of adult grammar. Even the simplest two-word utterances
show evidence of syntax, i.e. that children are creating sentences by following rules,
rather than just randomly stringing words together.
Just before children move into the two-word stage, they begin to use various strategies
to extend their one-word utterances. For example, some children tag on a meaningless
‘filler’ syllable like /a/ or /i/ to the beginning or end of their one-word utterances
to produce utterances like a write or i ball (Foster-Cohen 1999:25–26). Another
strategy is simply to repeat a word, producing utterances like car car. Although these
are not genuine two-word utterances, they provide evidence of a transition period
between the one-word and the two-word stage.
Look at the following two-word utterances, produced by a 2-year-old girl called
Kendall (Bowerman 1973). The context and the possible meaning are given in
brackets after each utterance.
Bed pretty (pointing to the bed she slept in)
Want juice (a statement requesting juice)
Pillow here (pointing to the pillow on her bed)
Kendall swim (comment made while she was swimming)
Mummy read (a statement requesting her mother to read to her)
Kendall book (showing someone her new book)
There cow (pointing to cow in picture book).

Like one-word utterances, two-word utterances consist mainly of combinations of

content words like nouns and verbs. These are the kinds of words that carry the main
semantic load in sentences. Function words such as the, is, in and that and words
that are not absolutely essential for conveying the semantic message are omitted, as
they would be in a telegram. Children’s speech from the two-word stage onwards
is thus sometimes referred to as telegraphic speech, as it includes only the words
that carry the main message. Note that we can understand children’s telegraphic
speech because the missing function words are usually predictable from the context.

LIN2601/1/2017-2023 73
 Some of the few function words that do appear fairly early in children’s speech are pronouns
(e.g. I, me, you, my, mine, yours), demonstratives (e.g. this, that, there) and prepositions (e.g.
up, off, on).

Task 4.1

Assuming that the meanings given in brackets are correct interpretations of Kendall’s
utterances, fill in the function words and grammatical morphemes missing from each
utterance. What kinds of function words does she leave out? The first one has been
done for you as an example.
Bed pretty (pointing to the bed she slept in) = My bed is pretty
Kendall doesn’t use the possessive pronoun my or the verb is.
Want juice (a statement requesting juice)

Pillow here (pointing to the pillow on her bed)

Kendall swim (comment made while she was swimming)

Mummy read (a statement requesting her mother to read to her)

Kendall book (showing someone her new book)

There cow (pointing to cow in picture book)

Feedback
Kendall’s utterances don’t include function words like pronouns (I, my, me), the verb ‘to
be’ (is, am), prepositions (to), articles (a) and grammatical morphemes like -ing. Possible
absent function words are suggested below.
Bed pretty (pointing to the bed she slept in) = My bed is pretty
Want juice (asking for juice) = I want some juice
Pillow here (pointing to the pillow on her bed) = My pillow is here
Kendall swim (comment made while she was swimming) = I am swimming
Mummy read (asking her mother to read to her) = Mummy, read to me
Kendall book (showing someone her new book) = This is my book
There cow (pointing to cow in picture book) = There is a cow

74
 What is interesting about these early two-word utterances is that children all seem
to express the same kinds of semantic or meaning relationships. In order to be able
to analyse the meanings and functions of children’s utterances we need to remind
ourselves of the semantic functions of the words in an adult sentence. Look at the
following example:
Ntombi bought her new jacket at the flea market.
The verb here is bought. It is the action word of the sentence. All the nouns in the
sentence also have different functions or roles. Ntombi is the agent, as she is the
person who performs the action, in this case, buying. The jacket is the object, as
it is the thing which is affected by the action of the agent (the term patient is also
sometimes used for the object of the sentence). The flea market fulfils the role of
location, as it is the place where the action happens. And we also have a possessor
or owner of something, in this case her (referring back to Ntombi). The word new
is an attribute, as it describes a characteristic of the jacket.
The psycholinguist Roger Brown classified children’s two-word utterances into the
following ten categories, using many of the functional roles explained above. His
classification looks at each of the roles of the words in a two-word utterance and the
semantic relationship that exists between these two words. The categories can also be
applied to one-word utterances if the context is used to add in the missing element.
 Nomination (naming), e.g. This doll, Here baby
 Recurrence (expressing that something is happening again or that the child wants
it to happen again), e.g. More noise, Another raisin
 Non-existence, e.g. Allgone egg, Nomore noise
 Agent-action, e.g. Mommy write, Daddy go
 Action-object, e.g. Throw ball, Eat raisin
 Agent-object, e.g. Mommy book (Mommy is putting the book away)
 Action-location, e.g. Go movie, Put there
 Agent-location, e.g. Mommy kitchen (Mommy is in the kitchen)
 Possessor-possessed, e.g. Mommy dress (Mommy’s dress)
 Object attribute, e.g. Stove hot

Now try the task below.

Task 4.2

1. What is the semantic relationship between the two words in each of Kendall’s
utterances below?
Bed pretty (pointing to the bed she slept in)

Want juice (a statement requesting juice)

Pillow here (pointing to the pillow on her bed)

Kendall swim (comment made while she was swimming)

LIN2601/1/2017-2023 75
 Mummy read (a statement requesting her mother to read to her)

Kendall book (showing someone her new book)

There cow (pointing to cow in picture book)

1. What are the semantic relationships between the two words in the following
Afrikaans utterances (Vorster 1983:234–237)? Possible meanings are given
for you in brackets after each utterance. Remember that the word order differs
in Afrikaans so the order of the elements may differ from the English examples
above, although the semantic relationships will be the same.
Utterance Wynand water
Gloss Wynand water
Meaning Wynand is falling in the water (Wynand val in die water)

Utterance: bang kat

Gloss: scared cat
Meaning: He is scared of the cat (Hy is bang vir die kat)

Utterance: groot hond

Gloss: big dog
Meaning: It’s a big dog. (Dis ’n groot hond)

Feedback
1. Bed pretty (pointing to the bed she slept in) = object + attribute
Want juice (asking for juice) = action + object
Pillow here (pointing to the pillow on her bed) = nomination
Kendall swim (comment made while she was
swimming) = agent + action
Mummy read (asking her mother to read to her) = agent + action
Kendall book (showing someone her new book) = possessor + possessed
There cow (pointing to cow in picture book) = nomination

2. Wynand water = agent + object

Bang kat = action + object
Groot hond = attribute + object

Just like adult utterances, children’s utterances can be analysed on various linguistic
levels. In addition to a semantic analysis, like Brown’s above, we need to look at the
pragmatic level (the communicative function of the utterances), the syntax (the word
classes and word order, i.e. the grammar of the utterance), the morphology (how
words are made up) and, if possible, the phonology (pronunciation) of the child. To
take one simple example, the two-word utterance Mommy bwead (produced when
mother was busy baking bread) can be analysed as follows:

76
 Phonologically the child seems to have mastered sounds like /m/, /b/ and /d/ but
has substituted a /w/ sound for the /r/ in bread. According to the table on page
73 the /r/ sound is usually mastered at about age 3, so this kind of phonological
substitution would be expected in the two-word stage.
Morphologically the utterance consists of two words, each of a single morpheme.
Syntactically there are two nouns, i.e. a subject (mommy) and an object (bread),
but no verb so the utterance is syntactically incorrect but quite usual for the two-
word stage.
Semantically this falls into Brown’s category agent-object as the mother is the doer
of the action and bread is the object that is affected by the agent’s action.
Pragmatically the utterance functions as a statement describing a state of affairs.
It is a comment on what is going on in the child’s environment.

Task 4.3

Analyse the utterance Toby shoes, said by a small child named Toby when showing
his new shoes to his older sister. The task requires you to say something about each
of the linguistic levels. Use the following questions to help you:
Phonological level: This level can only be addressed if we are given a tran
scription that reflects pronunciation. The kinds of ques
tions you would ask yourself here would be those we
looked at in study unit 3:
How does the child’s pronunciation differ from an adult’s?
What sounds does the child have difficulty pronouncing?
What simplification strategies is the child using, for
example assimilation, reduction, substitution?

Morphological level: Can the words be broken down into smaller meaningful
elements? e.g. eats consists of the verb eat + the sin
gular verb ending -s.
Which word endings are missing or incorrectly used?

Syntactic level: What parts of speech are used, e.g. nouns, verbs,
adjectives, etc?
Which words are missing?
Are the nouns used as agents or objects of the
sentence?

LIN2601/1/2017-2023 77
 Is the word order the same as or different from the
equivalent utterance in adult speech?

Semantic level: Which of Brown’s 10 semantic categories does the utter

ance fall into?
What does the sentence mean in the given context?

Pragmatic level: What is the function of the utterance in the context in

which it is used, e.g. is it a statement, question, re
quest, warning, command etc?

Feedback
Phonologically the utterance ‘Toby shoes’ seems to be adult-like as no evidence to
the contrary is given in the form of a more detailed transcription.
Morphologically the utterance consists of three morphemes (‘Toby shoe-s’) indicating
that the child can form regular plurals. The possessive morpheme -’s is omitted.
Syntactically there are two nouns, i.e. ‘Toby’ and ‘shoes’, but no verb. The utterance
is syntactically incorrect but quite usual for the two-word stage.
Semantically this falls into Brown’s category possessor-possessed as Toby owns the
shoes.
Pragmatically the utterance functions as a statement describing a state of affairs.

2 Measuring MLU (mean length of utterance)

How do researchers measure children’s linguistic progress? In some cases, researchers
count the average number of words in a child’s sentences, but more often they count
the average number of morphemes. (A morpheme is the smallest meaningful unit of
language, so a word like cats would consist of two morphemes: cat + -s. The ending
-s is considered a morpheme since it has the meaning ‘more than one’.)
The figure that researchers come up with is known as the mean length of utterance or
MLU, and is an objective way of measuring a child’s linguistic development. MLU
is calculated by counting the total number of words/morphemes and dividing by the

78
 total number of utterances to obtain the mean (average) number of words/morphemes
per utterance. For the purposes of this course, we will use morphemes as the basis of
our MLU calculations. For example, if the child produces 100 utterances, containing
a total of 170 morphemes, the MLU is 170 ÷ 100 = 1.7. An MLU of 1.7 means that
the average length of the child’s utterances is between 1 and 2 morphemes. MLU
figures are useful because they are independent of age or of how much children
talk and are still seen as a valuable assessment method, although not after MLU 4
(Huttenlocher, Vasilyeva, Cymerman & Levine 2002:339). These days, MLU can
be calculated by computer, making it an even more efficient method for assessing
children’s stage of development.
Here is a summary of Brown’s rules for calculating MLU:
 An utterance is a sentence or a portion of a sentence.
 Only fully transcribed utterances are used. Do not include sentences with unclear
portions. (Unclear portions are often marked by ( ) or xxx in a transcription.)
 Count each morpheme, including possessive -’s, plural -s, third person singular
-s, past tense -ed, progressive -ing etc.
 Count repeated words and phrases.

Notice that there are one or two exceptions to the basic principle of counting each
morpheme:
 Don’t count words like um and oh, but do count words like yes, no and their
equivalents – mmhm, uhuh etc.
 Compound words like birthday, choo-choo which refer to single concepts count
as 1 morpheme, not 2.
 Irregular past tense forms like woke, went, saw count as 1 morpheme, not 2.
 Diminutives like mommy, kitty count as 1 morpheme, not 2.

Let’s work through an example together.

In this example the child, aged 4;8, is talking to her uncle (Foster-Cohen 1999:63).
Calculate the MLU of the child.
Adult: Do you go to school?
Child: Yep, I go to pre-school.
Adult: Oh really? Is that the same school that S goes to?
Child: No, S goes in kindergarten.
Adult: Kindergarten, OK. What d’you do in pre-school?
Child: We-e-e-ell, I made a snowman (laughs) a army snowman today.
Adult: An army snowman! You had snow at your preschool?
Child: No, it didn’t – we – te – made it out of paper. We teared it with our hands.
Adult: Oooooh, OK
Child: Yeah, I made a uuuh (…) my uh snowman’s body (…) shaped like triangle.

Let’s start by counting the number of utterances. The child has five speaking turns,
but the fourth turn contains two sentences, making a total of six utterances. Now
on to the morphemes:
Utterance 1 (Yep, I go to pre-school) contains 5 morphemes (pre-school is one concept).

LIN2601/1/2017-2023 79
 Utterance 2 (No, S goes in kindergarten) is 6 morphemes (goes is 2 morphemes: go +
third person singular -s).
Utterance 3 (We-e-e-ell, I made a snowman (laughs) a army snowman today) is 9
morphemes. (made is an irregular past tense so counts as 1 morpheme).
Utterance 4 (No, it didn’t – we – te – made it out of paper) is 10 morphemes. (Didn’t
is 2 morphemes: did + n’t, the te is not counted as it is just a hesitation, and made
is 1 morpheme as it is an irregular past tense).
Utterance 5 (We teared it with our hands) is 8 morphemes (teared is two morphemes:
tear + -ed and hands is 2 morphemes: hand + -s)
Utterance 6 is not counted as portions of the utterance were unclear and could not
be transcribed.
We therefore have 5 + 6 + 9 + 10 + 8 = 38 morphemes divided by 5 utterances =
an MLU of 7.6. This means that the child uses an average of 7.6 morphemes per
utterance.
There are often debatable or problematic aspects in calculating MLU, and not everyone
will come up with exactly the same answer when asked to calculate the MLU of the
same data, so we will mark this kind of question leniently in exams and assignments.
However, you need to make sure you know the basic calculation method (number
of morphemes divided by number of utterances) and that you can make an attempt
to count morphemes accurately following the guidelines above. Although MLU is
a useful way of classifying the early stages of language development, it has been
used predominantly with English and may be less successful or yield very different
results when applied to structurally different languages such as African languages.

Task 4.4

Use the following data (based on Foster-Cohen 1999:59) as a basis for calculating the
MLU of Avi, aged 1;9.
1. back-pack 16. here mama get that
2. lap 17. see you later alligator
3. right now 18. I pooted
4. Hi 19. (…) a clown
5. peekaboo 20. a hooks
6. bottle 21. I reach that
7. here mama 22. get that
8. by lap Aja 23. kick it
9. I sit down here 24. powder
10. I read that 25. I try it
11. I get it 26. Oh cinnamon roll
12. corner 27. peanut butter
13. by corner 28. thank you
14. throw it 29. have a look
15. I throw it 30. here Papa, bite that

80
 Feedback
There are 30 utterances here but utterance 19 should be discarded since it contains an
untranscribable portion. The Oh in utterance 26 is not counted. Words like back-pack,
peekaboo, cinnamon roll, peanut butter and thank you are compound words referring
to a single concept so they each count as 1 morpheme. The words pooted and hooks
each have 2 morphemes. This makes a total of 65 morphemes divided by 29 utterances
= a MLU of 2.24. Don’t worry if your answer differs slightly, as long as it is close to 2.2.
You may also want to exclude utterance 17 since it seems to be a learned phrase rather
than a genuine four-word-sentence.

3 The multiple-word stage

Once children start to put words together, their utterances quickly begin to include
two-word sentences, three-word sentences and even longer sentences. This period of
language acquisition is known as the multiple-word stage. The length of children’s
sentences increases gradually throughout the preschool and early school years as
they learn the syntactic rules necessary for using function words and creating
more complex sentences.

You can listen to an example of a child at the multiple-word stage

https://www.youtube.com/watch?v=UrRKLHq25UA

Children from different language environments all achieve linguistic competence

at roughly age 6 in spite of structural differences between languages. By the time
children start school, the majority of the rules of the language have been internalised
– an incredible intellectual task to have accomplished before school-going age.
The order of acquisition of complex constructions seems fairly consistent across
English children although it may differ in other languages. The typical order of
English acquisition is as follows:
present progressive -ing, e.g. walking
learnt first. prepositions in/on
plural, e.g. cats
irregular past tense, e.g. ate, made
possessive ’s, e.g. the cat’s tail
copula (the verb ‘to be’), e.g. The dog is here
articles the, a
regular past tense, e.g. walked, played
third person singular -s, e.g. She walks
auxiliaries, e.g. can, do, might etc.
As we have seen with word meanings, children sometimes get these syntactic rules
learnt last. wrong in various ways. One common error is overgeneralisation, which occurs when
children apply rules too widely. For example, English-speaking children soon learn
the most common rule for plural formation: add -s to a singular noun to form the
plural:
dog + plural  dogs

LIN2601/1/2017-2023 81
 This rule is then consistently applied, even to exceptions. Thus, the child produces
forms like songs, books, tables as well as *foots/feets, *oxes, *sheeps, *childs and so on.
(An asterisk (*) before a word or sentence indicates that it is ungrammatical.)
One interesting phenomenon relating to overextension is that children frequently
pass through a stage of using exceptions correctly before they start overgeneralising.
For example, an English child may use correct plural forms like feet and sheep and
only later start using overgeneralised plurals like *foots and *sheeps. The explanation
usually given for children’s correct use of exceptions at an early age is that they have
not yet learnt the rule, and have merely memorised unanalysed forms. (So imitation
does play a limited role in children’s early acquisition!) Once children start analysing
these forms and work out the rule, they frequently overgeneralise the rule for a period
of time. Eventually the child learns the exceptions to the general rules and starts to
use correct plurals again.
Overgeneralisation is a common strategy used by all children, irrespective of the
language they are acquiring. For example, Zulu children have a formidable task
facing them when acquiring their language, as Zulu nouns are classified into 15
classes and each noun must have the appropriate noun class prefix (see below).
Noun class prefi x
Class 1 (singular) u- e.g. utata ‘father’
Class 2 (plural) aba- e.g. abatata ‘fathers’
Class 3 (singular) umu-
Class 4 (plural) imi-
Class 5 (singular) i- e.g. ijezi ‘jersey’
Class 6 (plural) ama- e.g. amajezi ‘jerseys’
Class 7 (singular) si-
Class 8 (plural) zi-
Class 9 (singular) i-
Class 10 (plural) zi-
Class 11 (singular) lu-
Class 12 (plural) zi-
Class 14 bu-
Class 15 ku-
Studies of the acquisition of noun class prefixes in several African languages (Demuth
2003:5) report the following three partially overlapping stages of development
during ages 2–3:
1. Full or partial noun stems with no prefixes, e.g. tata, jezi
2. Overgeneralised ‘placeholder’ vowel or nasal prefixes, e.g. utata, itata or ntata
This strategy involves putting in a ‘placeholder’ or ‘filler’ syllable for the prefixes
which have not yet been mastered.
3. Full and phonologically appropriate noun class prefixes, e.g. utata, amajezi, etc
In addition, verbs and adjectives in Zulu and other African languages must have an
agreement marker that shows agreement with the class of the noun. For example, in
the sentence below the noun class marker u- on utata is associated with the obligatory
agreement marker u- on the verb -hamba ‘go’.
Utata uhamba
‘The father goes’

82
 Gxilishe, De Villiers and De Villiers (2007) present data from Xhosa children aged
1;0–3;3 which shows that they use correct agreement markers 20% of the time
by age 2, 50% of the time by age 2;6 and 80% of the time by age 3. Between the
ages of 18 and 36 months, the correct agreement markers are therefore used with
increasing frequency and on more and more different verbs. Interestingly, however,
while children omit agreement markers in the initial stages of acquisition, and then
use a vague overgeneralised agreement marker for a while, they do not seem to use
incorrect subject agreement markers after age 3, for example ujezi or itata. Demuth
(2003) suggests that the relatively early and ‘error free’ acquisition of Bantu noun
class and agreement systems suggests that learning complex morphological paradigms
is easy when there are clear phonological patterns for children to follow.

3.1 The development of questions

Asking and answering questions is obviously a very important interactional skill
that children must learn. They soon learn to recognise the rising intonation typically
associated with questions and to give a response, although the response is not always
appropriate at first. Before we look at children’s questions, let’s look at some of the
syntactic and semantic features of questions that need to be mastered.
Semantically, a question can function either as a request for action (i.e. for someone to
do something) or a request for information (i.e. for someone to tell you something).
For example, Can you shut the door? is a request for action, while What time is the
party? is a request for information.
Syntactically, there are three basic types of question in English:
Tag questions are formed by adding a short tag onto a statement, e.g. We are
going to the shop now, aren’t we?
Yes-no questions require the answer yes or no and are formed by taking a
statement like I can swim, and switching the position of the auxiliary verb and
the subject, e.g. Can I swim?
Wh-questions contain a wh-word like who, what, where and require an answer
other than yes or no. The auxiliary also comes before the subject noun, e.g.
Where can I buy some?

At first, 2-year-olds typically form questions by placing a wh-word like who, what or
where at the beginning of the utterance, for example What dolly doing? Where baby
Sarah rattle? The auxiliary verb is simply omitted. The wh-words when, how and why
are only used slightly later, for example Why he can’t go? One reason for this is that
the concepts of time, manner and causality are more abstract and require longer and
more complex answers than the concepts expressed by who, what and where, which can
be answered with a single word or phrase (Tager-Flusberg 2005:174). At the same time,
children are developing competence in forming yes-no questions, but tend to place the
subject before the auxiliary, for example I can have some? After age 3, children learn
that questions require a switch in the position of the auxiliary verb and the subject,
for example Why can’t he go?, Did you fix it? or Can I have some? Tag questions are the
last question type to emerge, usually some time between 3;6 and 4 years.
The frequency of children’s questions increases gradually, and reaches a peak at
about age 5, when questions make up nearly 20% of the child’s utterances (Reich

LIN2601/1/2017-2023 83
 1986:118). When parents fail to give answers to these questions, children persist in
asking for the information, suggesting that the goal of this behaviour is to gather
the information they need in order to learn about the world.

3.2 The development of negatives

Negatives are used for several different purposes, including saying that something is
absent or nonexistent, rejecting a suggestion, and denying the truth of a statement.
An utterance like no cake can have any one of the three purposes described above,
for example
absence (‘There is no cake’)
No cake rejection (‘I don’t want any cake’)
denial (‘That isn’t cake’)

The earliest negative forms are usually produced by simply adding the word no or
not, usually at the beginning of an utterance, for example No Mommy do it (Tager-
Flusberg 2005:171). The next stage in the development of negative forms is the use
of contracted negatives like aren’t, mustn’t and don’t. Uncontracted negatives like
are not and do not appear slightly later, at about age 3, once the child realises that
the contracted negatives are made up of an auxiliary verb + not. However, as in
adult speech, contracted negatives remain more common than uncontracted ones.

3.3 Joining sentences together

In the early years, children’s utterances contain only one clause (i.e. a phrase
containing one main verb). In about the third year they start to join clauses together
to produce complex sentences. At first, the word and is the most common way
of joining clauses together, while children later learn to use more sophisticated
connectives like because, then, but and so (Foster-Cohen 1999:72, 76).
Another way of joining sentences together is to use a relative clause. Relative clauses
are embedded clauses that give further information about something, using a relative
pronoun (that, who, which). For example, in the sentence, Let’s eat the cake that you
made, the relative clause that you made is embedded in another sentence – Let’s eat the
cake. In their early attempts at relative clauses, children frequently omit the relative
pronoun, for example Let’s eat the cake you made, or use an incorrect pronoun, for
example Let’s eat the cake what you made. Mastery of this construction usually only
occurs at about age 5 or 6.

3.4 Developing communicative competence

Developing communicative skills is a lifelong process, starting off with a baby’s
initial ability to express pleasure or discomfort and ending up with adult speakers’
ability to use language to adapt their language in subtle ways to make it appropriate
to different situations. For example, adults need to be able to use formal language
in written business reports, and in-group slang when chatting to their friends. This
pragmatic knowledge and ability – the ability to use language to communicate in
ways that are appropriate to the context – is known as communicative competence.
We focus here on some of the more important skills that develop in the preschool
years. (In study unit 6 you will see what happens when children have problems with
the pragmatic aspects of language.)

84
 Conversational skills develop throughout childhood. We have seen that even in the
first weeks of life, caretakers interact with babies by responding to the baby’s coughs,
cries and other vocalisations as if they were conversing with the baby. Through this
early interaction, babies acquire an understanding of some of the rules of conversation,
such as the fact that conversation involves communication between two or more
people and is organised through turn-taking, so that each participant takes turns
to talk and to listen.
In the one-word stage, when their conversation is still obviously very limited, children
start to initiate conversation by using a word to request things or comment on objects
and events around them. Children also start to recognise the differences between
questions and statements and learn to respond appropriately. Another important
aspect of conversation which children need to learn is that utterances should relate
to the preceding utterance and should stick to one main topic or related topics rather
than jumping from one unrelated topic to another.
One important development is the fine-tuning of the turn-taking skills that emerged
earlier. Children learn some of the subtleties involved in conversing with others, for
example how to leave appropriate gaps between the end of one person’s speaking turn
and the beginning of the next speaker’s turn. By about age 3;6 children typically
leave one- or two-second gaps between turns, while the adult norm is to leave only
about half a second between turns (Foster 1990:112). They learn to follow the
politeness rules appropriate to their culture, for example to say Please and Could I
have … instead of just Want biscuit or Gimme another one. Children also develop
their ability to make their conversation relevant and appropriate to what has gone
before, for example to respond to a question by providing information of some kind,
or to initiate conversation by commenting on something.
One final important development is learning what to say to whom, i.e. adjusting
one’s speech to the person one is addressing. An understanding of different social
groups based on age, sex, status and familiarity is a prerequisite for this kind of
language adjustment. By age 4, children have enough background knowledge to
adapt their speech when acting out different roles such as doctor, patient, teacher
and pupil (Foster 1990:127).
Now that we’ve read the theory, let’s look at some actual samples of children’s
developing language. There are samples of 1-year-old, 2-year-old, 3-year-old, 4-year-
old and 5-year old speech. The transcriptions all come from the CHILDES database
and are of American children.

SAMPLE 1
In this transcription Geraldine’s mother is talking to Geraldine (aged 1;6). (The
database can be found on the internet at www.childes.psy.cmu.edu/cornell/haas1917).
MOT: What’s this?
CHI: Hat.
MOT: No.
MOT: What’s this?
CHI: Hat.
MOT: Hair.
MOT: What’s this?
CHI: Nose.

LIN2601/1/2017-2023 85
 MOT: Nose.
MOT: Yeah.
MOT: What’s this?
CHI: Tongue.
MOT: Tongue.
MOT: Gerry, what are these?
CHI: Um.
MOT: What are these?
MOT: What are those?
CHI: Hands.
MOT: Hands, yeah.
MOT: You’re a good girl.
CHI: All gone.
MOT: All gone.

SAMPLE 2
The next transcription is of Geraldine (aged 2;5) and her mother during Gerry’s bath
time (www.childes.psy.cmu.edu/cornell/haas1929). The line marked %act indicates
an action rather than a verbal utterance.
CHI: I took off my socks.
CHI: Mommy.
CHI: Mommy, see see.
MOT: Good Girl.
MOT: Take off your coat now.
CHI: No, not my coat.
MOT: Well, you can’t go in the tub with your coat on.
MOT: I’ll be right back, hon.
CHI: It’s not a coat.
MOT: Take it off now.
%act: Gerry is crying
MOT: What’s the matter.
CHI: I hit my, my xxx.
MOT: You did all right.
MOT: You did fine.
CHI: I ripped it.
MOT: No, it’s okay.

SAMPLE 3
Mary (aged 3;0) is chatting to her aunt: (www.childes.psy.cmu.edu/cornell/horn1603)
CHI: Where’s my coffee?
AUN: Where’s your coffee
CHI: Uhhuh.
CHI: I want to drink it.
AUN: Children don’t have coffee.
CHI: My coffee is right here, right here, see?
AUN: Yeah.
CHI: It’s cold.
AUN: You a big girl now, huh?
AUN: You drink coffee.

86
 CHI: Uhhuh.
CHI: I want to drink it.

SAMPLE 4
In the following conversation Matthew (aged 4;0) and his mother are talking about
the weather (www.childes.psy.cmu.edu/cornell/mom06).
CHI: Cold.
CHI: I think outside it’s a little warmer.
CHI: Put it on my chair too.
MOT: Want me to put it on the chair too?
MOT: It’s not summer any more is it sweetheart?
CHI: No.
MOT: Okay.
CHI: It’s Fall.
MOT: Is it Fall?
CHI: Yes, it’s cold out in the morning.
MOT: Oh, is that what happens when it’s Fall?
CHI: Yeah.
MOT: What else happens when it’s Fall?
CHI: Well, gets colder than the summer.
MOT: Yeah, it does.
MOT: Now, what happens after Fall?
CHI: I don’t know.
CHI: Winter.
MOT: Winter.

SAMPLE 5
In the last sample Peter (aged 5;0) is in hospital, about to have an operation. His
father and mother are with him (www.childes.psy.cmu.edu/cornell/Isno11).
CHI: When are we going to take an operation?
MOT: Right here.
CHI: No, I’m not.
FAT: They’re going to let me do it, Peter.
MOT: They’ve got a special room where they’ll be, the anesthetist who puts you to
sleep, some nurses maybe and Dr Kelly.
CHI: Dr Kelly makes me go to sleep.
MOT: No, I think, I think the anesthetist gives you the things to go to sleep.
MOT: Dr Kelly does the operation.
CHI: Dr Kelly?
CHI: I don’t like this Momma.
CHI: Why don’t we got one of these things in ( …)
MOT: Well yeah, what do you think they’re going to have you sit up in there while
they’re doing it, huh?
CHI: No, I think the doctor is bringing me.
MOT: Oh, they’ll probably have you walk.
CHI: No.

LIN2601/1/2017-2023 87
 Read the transcriptions and then try the task that follows.

Task 4.5

Read samples 1 to 5 of children’s language, paying careful attention to the children’s

ages in each sample. Try and describe in broad terms (1–2 paragraphs) the ways in
which children’s language differs at age 2, 3, 4 and 5, paying particular attention to the
length and complexity of utterances at each age.

Feedback
The sentence length and complexity of sentence structure clearly increase as children
mature. As far as sentence length is concerned, 1-year-old speech contains single-
word utterances. The 1-year-old speech seems to contain 1 to 5 words per utterance,
indicating that this child is in the telegraphic stage. The sample of 3-year-old speech
contains sentences up to 8 words long. The 4-year-old and 5-year-old speech also

88
 contains up to 8 words per utterance. We shouldn’t, however, read too much into these
similar results as we are only given a small amount of data in the task and the contexts
of the conversations may differ. The important point is that from about age 3 children
can use fairly long sentences and express complex ideas.
In terms of syntactic complexity, the fact that the data is American makes it difficult for
us to analyse the grammaticality of some of the sentences. However, in broad terms,
2-year-old speech includes simple vocabulary and also some function words. By age
three, the vocabulary and syntax is still simple but most sentences are grammatical. By
age four and five, more sophisticated vocabulary and complex constructions like com-
paratives are being used. A few ungrammatical utterances, like ‘Why don’t we got …’
still remain. In terms of communicative competence, even the child in Sample 1 shows
an ability to stick to the topic and answer questions. In later samples the children show
an increasing ability to answer appropriately and also to initiate conversation by intro-
ducing topics and shifting the conversational topic.

4 Later linguistic developments

We do not know as much about language development after about the age of 5 as
we do about language development up to this time. Here are some reasons for this
state of affairs.
While the research literature in education has always focused on later language
development, it has only been fairly recently that linguists have turned their
attention to language development after the age of about five or six. There are
a number of reasons for this relative neglect. One is the sense that the hard
part is done by the age of five, and that change thereafter is slow, and mostly
schooled rather than developing naturally. Another is that older children are
generally shyer and more silent around researchers than younger children
and thus exhibit less ‘data’ for capture (Stephens 1988). Moreover, since older
children are aware that their language can be deliberately modified for the
audience, their performances are often stilted and less natural than those of
younger children.

And finally, older children are much more mobile. The most interesting lan-
guage may be happening at twenty miles an hour on a skateboard. Clearly,
these are not ideal recording conditions! (Foster-Cohen 1999:171).

Children aged 5 to 7 understand approximately 20 000 words, including synonyms

(two words with the same meaning), homonyms (one word with different meanings)
and more abstract descriptive words like carefully, worried or clever. They can use the
past and future tense and join sentences using conjunctions. They start to use language
for different purposes, such as persuading others or discussing more complex ideas.
From age 8 onwards, learners ‘fine-tune’ their existing linguistic competence, learning
finer details such as irregular constructions and exceptions to rules. Although
negation errors like I didn’t did it and overgeneralisation of past tense forms like
runned, drawed and bited continue well into the school years, children gradually
learn the exceptions to the past tense and various other rules. At school children
start to make use of a wider range of clause types than before. For example, they
learn to use passive sentences (e.g. The ball was thrown by the boy), which are rare
in English, even in adult speech. Children’s ability to interpret complex sentences
improves, and their descriptions become better as they start to use adjectives and

LIN2601/1/2017-2023 89
 other modifiers to produce phrases like, for example, The girl with the red shirt. They
develop language skills such as understanding other points of view, giving reasons
and explaining choices, predicting outcomes and drawing conclusions.
One obvious area of language development in the school years is in the number of
vocabulary items a child can produce and comprehend. According to estimations,
a Grade 3 pupil (aged 8 or 9) has a vocabulary of 4 000 to 10 000 words while
a matric pupil has a vocabulary of about 40 000 words (excluding names, place
names and idioms). This huge growth in vocabulary from age 8 to age 18 implies
that children learn 3 000 words a year or 8 new words a day, not only from direct
teaching at school, but more importantly, from reading (Foster-Cohen 1999:176).
In the next three sections we explore some of the major language developments that
occur after children start school, namely, the skills of telling stories, reading, and
learning to think and talk about language and how it works.

4.1 Learning to tell stories

During the school years children learn to produce longer stretches of text, including
stories. Children’s pragmatic development includes an increasing ability to tell
stories that hang together and that are logical in terms of the order or sequence of
events described and the notion of causality (what causes something to happen).
Children gradually become familiar with the various elements of a story, namely,
characters, plot, setting and theme. Let’s look at some actual examples of children’s
stories to find out how their storytelling skills improve over time. The three samples
were all based on the ‘frog sequence’ of pictures (see Figure 2.2 in Study Unit 2):
see: http://torontopubliclibrary.typepad.com/kids-books/2015/12/who-needs-
words-anyway.html.
Age 3;9 (www.childes.psy.cmu.edu/frogs/slobin/03f)
And, um, he, this dog is looking into the bowl. And then the frog is still in there.
And now he got away. The dog got away, and then he tried to go in but he didn’t,
couldn’t go in!
Age 5;11 (www.childes.psy.cmu.edu/frogs/slobin/05k)
When the boy and the dog were asleep the frog jumped out of the jar and then the
boy and the dog woke up, the frog was gone. Then the boy got dressed and the dog
stuck his head in the jar. And then the boy opened up his window and called out
for his frog and the dog still had the jar on his head.
Age 9;6 (www.childes.psy.cmu.edu/frogs/slobin/09c)
Well there was a boy and he had a frog and a dog. He loved his frog very much,
and maybe his dog did too. One night when he was asleep the frog climbed out
of his jar. The boy woke up the next morning, the frog was gone. The dog looked
concerned. The boy quickly got dressed and the dog put his head in a jar. He got
stuck there, but the boy did not pull it off.

Task 4.6

Read the three stories above, paying careful attention to the children’s ages in each
sample. Try to describe in broad terms (1–2 paragraphs) the ways in which children’s
storytelling ability improve in these three samples. Focus specifically on the length

90
 and complexity of the story, the vocabulary and specifically the connectives used, and
the logical sequencing of events.

Feedback
The samples clearly show that children’s stories get longer and more detailed as they
get older. In the first sample (produced by a child aged 3;9) the story consists of short
sentences and relies heavily on the connective ‘and’ to join sentences together. The
child’s vocabulary is rather limited, for example he uses the term ‘bowl’ rather than the

LIN2601/1/2017-2023 91
 more appropriate ‘jar’ or ‘bottle’. Also notice that the use of the pronouns ‘he’ and ‘his’
is confusing as it isn’t always clear who the pronoun refers to.
In sample 2 (from a child aged 5;11) the story starts rather abruptly without introducing
the characters, the boy, the dog and the frog. The child makes use of connectives such
as ‘and’ and ‘then’ but joins sentences that have no logical connection, for example
‘Then the boy got dressed and the dog stuck his head in the jar.’
At age 9;6 the child begins the story as an adult would, by introducing the three main
characters. The story describes the events in a logical order, and tells us when each
event occurred, for example ‘One night when he was asleep the frog climbed out of
his jar.’ Connectives like ‘and’ and ‘but’ are used appropriately, as is sophisticated vo-
cabulary like ‘concerned’. He even describes the characters’ feelings and motivation,
telling us that ‘the boy loved his frog very much, and maybe his dog did too’. Pronouns
such as ‘he’ and ‘his’ are used in appropriate ways, showing that the child is able to
omit non-essential information, such as repeated names. When necessary, the child
reintroduces the term ‘the boy’ to prevent ambiguity.

4.2 Learning to read

There are several skills that make up our ability to read. These include phonological
awareness (i.e. awareness of the sounds of language), awareness of the link between
letters (spelling) and sound, the ability to recognise familiar words and the ability
to decode unfamiliar words using both letter–sound relationships and contextual
knowledge. According to Ehri (1991:63):
Understanding the processes involved in learning to read turns out to be
a complex task because reading entails a host of processes operating together in
concert. For example, as readers’ eyes move across a line of text, they are pick-
ing up visual information about letters, words and sentences at the same time
as they are mentally processing the information by computing the structure
and meaning of the words and sentences; integrating this with information
they have already read and their world knowledge to understand the text, and
forming expectations about forthcoming text.

Reading requires us to recognise printed words so that we can access their meaning
and pronunciation, both stored in our memories. There are two ways to recognise
printed words: the first is simply to recognise a word because we have read it so many
times before. This recognition of whole words without breaking them down into
parts is known as top-down word recognition. The second way is to break a word
up and map the letters onto the sounds they represent. This process of building up
words from smaller units such as letters and syllables is known as bottom-up word
recognition.
Why do we need two different ways of recognising printed words?
Typically, common words are recognised as unanalysed wholes, using a top-down
strategy. Top-down word recognition is also essential for reading irregular words like
two and answer as these do not follow the conventional letter-sound rules. Bottom-
up word recognition is necessary because we need to be able to read and pronounce
words that we have never seen in print before. Children typically learn to read their
first few words as unanalysed wholes, using a top-down strategy. At some point,
however (usually age 6–7), children must acquire bottom-up word recognition skills,

92
 learning to divide words into sounds and match sounds with letters. Research has
shown that good readers make more use of top-down reading than unskilled readers.
Learning to read occurs in four stages: prereading, initial reading, fluent reading
and reading to learn. These are explained in more detail below:

The prereading stage begins during the preschool years when children start to
participate in literacy events, such as listening to adults reading stories aloud,
playing with books, learning nursery rhymes, learning the alphabet etc. During
the prereading stage children also increase their phonological awareness, i.e. the
ability to focus on the sounds of words rather than their meaning. Examples of
this growing phonological awareness include the ability to recognise rhyme and
the ability to identify words that begin or end with the same sound. Research has
shown that phonological awareness is a key ingredient in successful reading and
that ‘improving children’s ability to manipulate sound segments of spoken words,
and to link sound segments with letters, caused them to become better readers and
spellers’ (Stuart 1995:33).
The initial reading stage is when children begin to understand how to map sounds
onto letters and to understand the principles of the alphabetic spelling system. The
‘phonics method’ of teaching reading focuses in detail on teaching children the
sound-letter relationships in language, for example that the letter combinations
ee and ea ‘say’ /i/, that -tion represents the sound ‘shin’, that gh is usually silent
in English and that the ‘silent e’ at the ends of English words changes the vowel
sound, etc. Children learn to recognise common words and read simple text. They
learn to figure out unfamiliar words and follow story lines in text. Exposure to a
wide range of books and learning to associate reading with a pleasant environment
also play an important part in getting children to make the connection between
meaning and print.
The fluency stage is reached in about the second year of reading instruction when
children become faster and more skilled at the activities learnt in the initial reading
stage. As their vocabularies increase they learn to recognise more and more words
and to effectively figure out unfamiliar words. Children who read a lot acquire larger
vocabularies and more effective comprehension skills and begin to outperform those
who do less reading.
The final stage of learning to read is known as reading to learn, and begins
somewhere around the fourth year of reading instruction. From this point onwards,
children are sufficiently skilled readers to be able to use reading to find out new
information and comprehend unfamiliar ideas in all their school subjects. Even
later, children and teenagers learn to analyse and react to what they read, and to
recognise multiple viewpoints in text.

If you are interested in literacy rates and reading programmes in South

A
Africa you could look at http://www.unesco.org/uil/
litbase/?menu=4&programme=154 or http://www.read.org.za

LIN2601/1/2017-2023 93
 4.3 Metalinguistic development
Metalinguistic development is the development of the ability to think and talk
about language. It involves reflecting on language and how it works (as we do
when we study linguistics!) There are many different skills that make up a child’s
growing metalinguistic awareness, including the ability to judge the grammaticality
of sentences, to define words and appreciate linguistic humour – jokes, riddles,
tongue twisters, etc. Before age 5, children’s metalinguistic knowledge of language
is largely implicit, but by age 7 they can consciously access this knowledge and
describe it verbally.
One of the first metalinguistic skills to emerge is the ability to correct one’s own
speech errors and slips of the tongue. Children as young as 18 months can correct
their own utterances, for example if they realise they have made a syntactic or
pronunciation error. For example in the first frog story above, the child corrects his
syntax as follows: ‘… and then he tried to go in but he didn’t, couldn’t go in!’ This
shows that children are actively monitoring their own speech. By age 7 children
can recognise whether sentences are grammatical or not and correct most errors.
Some of the metalinguistic abilities that emerge in the primary school years include
the ability to use grammatical terms like noun and verb, and the ability to give
appropriate definitions of words. Preschoolers’ definitions of words tend to be
immature in that they are concrete and are based on personal experiences, while at
school children gradually learn to define words in more abstract terms, for example
by assigning words to categories and mentioning characteristics and functions of
objects (Taylor & Taylor 1990:268–269). For example, an adult definition of a word
like knife might be something like the following: A knife is a tool (category name)
with a sharp metal blade (appearance) that is used to cut (function). A 4-year-old
child’s definition might involve pointing to an actual knife, saying This is a knife.
Between age 5 and 10 they might give definitions like A knife is for cutting (function),
or You cut with a knife and you cut yourself (personal experience) or A saw is like a
knife (giving the name of a similar object).
Now check your understanding by completing the task below:

Task 4.7

Write a paragraph in which you summarise some of the most important language de-
velopments in school-aged children.

94
 Feedback
Syntactic and lexical (vocabulary) development continues during the school years.
According to estimations, from age 8 to age 18 children learn 3 000 words a year.
Children’s utterances become gradually more adult-like as they learn the exceptions
to the past tense and various other rules and learn to use a wider range of sentence
structures than before. These include the passive, adjectives and other modifiers and
descriptive prepositional phrases. During the school years children learn to produce
longer stretches of texts, including stories that hang together in a logical manner. At first
the main connective they use is ‘and’. Later they use more varied connectives as well
as more sophisticated vocabulary, sentence construction, etc. School-aged children
also learn how to do more things with language, for example persuading, justifying one’s
views and drawing conclusions. A major milestone is reached in the early school years
when children learn to read by matching sounds with letters (usually age 6–7). Reading
assists language development by familiarising children with new words and sentence
structures and is an important source of new information. Finally, the school years are
an important time for metalinguistic development – the development of the ability to re-
flect on language and talk about how it works. There are many different skills that make
up a child’s growing metalinguistic awareness, including the ability to correct errors in
speech, judge the grammaticality of sentences, define words and appreciate linguistic
humour. Before age 5, children’s metalinguistic knowledge of language is largely implicit,
but by age seven they can consciously access this knowledge and describe it verbally.

5 Summary
In this study unit we have looked at the development of syntactic, semantic and
pragmatic skills from the two-word stage through to the primary school years. We
analysed real language data and learnt how to calculate the MLU (mean length of
utterance) of a speech sample as an objective way of measuring a child’s linguistic
development. We looked at some of the errors children make as they learn to produce
more complex morphology and grammar, including Zulu noun class prefixes and
English questions and negatives. We explored the notion of communicative competence
– the ability to use language to communicate in ways that are appropriate to the
context. We saw how children’s storytelling skills develop over time and the four
stages involved in learning to read. In the next study unit we turn our attention to
the input that children hear and to individual differences that can affect the speed
or route of language learning.

Task 4.8

Test your understanding of the study unit so far by choosing the option that fits best
with the given statement in the following multiple-choice questions. Circle the correct
answer ([1], [2], [3] or [4]) and then check your answers against those given below.

1. Telegraphic speech includes

a one-word utterances
b one-word utterances with ‘filler’ syllables attached
c two-word utterances without function words
d three-word utterances with function words

2. Some of the earliest function words to appear are

LIN2601/1/2017-2023 95
 a prepositions and auxiliaries
b demonstratives and auxiliaries
c pronouns and relative pronouns
d prepositions and pronouns

3. The semantic relation between the words in the utterance kick ball is
a agent–action
b action–object
c action–location
d object–attribute

4. The linguistic level that looks at the communicative function of an utterance in the
context in which it is used is the level.
a phonological
b morphological
c syntactic
d semantic
e pragmatic

5. The MLU of a child who had just entered the two-word stage would probably be
a close to 1
b between 1 and 2
c exactly 2
d more than 2.

6. The question What he did buy me? is

a a yes–no question
b a wh-question where the auxiliary and subject have been switched
c a wh-question where the auxiliary and subject have not been switched
d a tag question

7. Put the following negative statements in the likeliest order in which a child would
start to use them, from earliest to latest.
a Don’t touch
b No touch
c Do not touch
i ABC
ii B C A
iii C B A
iv B A C

8. In MLU calculations, diminutives like doggy and horsie count as

a 1 morpheme
b 2 morphemes
c 3 morphemes
d They are not counted.

9. A child who is overgeneralising the English past tense rule will produce forms like
a rode, bit, ate
b ride, bite, eat
c rided, bited, eated
d rides, bites, eats

10. During the school years children learn a total of about new words
per year.
a 10 000
b 40 000
c 800–1 200
d 3 000

96
 Feedback
1.[3] 2. [4] 3. [2] 4. [5] 5. [2] 6. [3] 7. [4] 8. [1] 9. [3] 10. [4]

Academic vocabulary building

Many of the verbs used in this study unit can be turned into nouns. The verb
describes an action, and the noun usually describes the result of this action. Look
at the verb used in each of the following sentences, and fill in the corresponding
noun in the space provided. The first two have been done for you as an example.
Hint: all except one of these nouns end in -tion.
If you define a word you produce a definition.
If you calculate MLU you do a calculation.
If you investigate a topic you do an
If you classify something you come up with a
If you overgeneralise a rule you make an
The process of acquiring a language is known as
If you converse with someone you engage in a
If you distinguish between two things you make a
If you utter words you produce an
If you communicate with someone you engage in
If you simplify something, the result is a

Further reading

Those of you who are interested in this topic might like to read one of the following:
Gopnik, A, Meltzoff, AN & Kuhl, PK. 1999. The scientist in the crib: minds, brains,
and how children learn. New York: Harper Collins.
Berko Gleason, J (ed.). 2005. The development of language. 6th edition. Boston, MA:
Pearson. Chapters 5, 6 and 10.

Bibliography
Berko Gleason, J (ed.). 2005. The development of language. 6th edition. Boston: Pearson.

LIN2601/1/2017-2023 97
 Bowerman, M. 1973. Early syntactic development: a cross-linguistic study with special
reference to Finnish. Cambridge: Cambridge University Press.
Demuth, K. 2003. The acquisition of Bantu languages, in The Bantu languages,
edited by D Nurse & G Philippson. Surrey, England: Curzon Press, 209–222.
Department of Linguistics 1990. Study guide 2 for LNG100S (Language in use).
Pretoria: Unisa.
Department of Linguistics 1992. Only study guide for LNG2043. Pretoria: Unisa.
Department of Linguistics 1996. Only study guide for LNG100S. Pretoria: Unisa.
Ehri, LC. 1991. The development of reading and spelling in children: an overview, in
Dyslexia: integrating theory and practice, edited by M Snowling & M Thomson.
London: Whurr.
Foster, SH. 1990. The communicative competence of young children: a modular ap-
proach. London: Longman.
Foster-Cohen, SH. 1999. An introduction to child language development. New York:
Addison Wesley Longman.
Funnell, E & Stuart, M (eds). 1995. Learning to read: psychology in the classroom.
Oxford: Blackwell.
Gxilishe, S, De Villiers, P & De Villiers, J. 2007. The acquisition of subject agree-
ment in Xhosa, in Proceedings of the 2nd Conference on Generative Approaches
to Language Acquisition North America (GALANA), edited by A Belikova.
Somerville, MA: Cascadilla Proceedings Project, 114–123.
Huttenlocher, J, Vasilyeva, M, Cymerman, E & Levine, S. 2002. Language input
and child syntax. Cognitive Psychology 45:337–374.
MacWhinney, B. 2000. The CHILDES Project: Tools for analyzing talk. 3rd edition.
Mahwah, NJ: Lawrence Erlbaum.
Reich, PA. 1986. Language development. Englewood Cliffs, NJ: Prentice-Hall.
Stuart, M. 1995. Recognizing printed words unlocks the door to reading: how do
children find the key? in Learning to read: psychology in the classroom, edited
by E Funnell & M Stuart. Oxford: Blackwell.
Tager-Flusberg, H. 2005. Putting words together: morphology and syntax in the
preschool years, in The development of language, edited by J Berko Gleason. 6th
edition. Boston: Pearson, 148–190.
Taylor, I & Taylor, MM. 1990. Psycholinguistics: learning and using language. Engle-
wood Cliffs, NJ: Prentice Hall.
Vorster, J. 1983. Aspects of the acquisition of Afrikaans syntax. DLit et Phil thesis.
University of South Africa.

98
 Study Unit 5 Study unit 5
Language input and individual
differences
‘If you don’t acknowledge differences, it’s as bad as stereotyping or
2

reducing someone’
Aasif Mandvi (Indian actor, 1966–)
3

OUTCOMES
After you have studied this study unit you should be able to
 identify the characteristics of caretaker speech and discuss its role in language
acquisition
 identify and account for individual differences that may occur in the endpoint,
strategy and rate of early language acquisition
 identify and account for cross-linguistic differences in early language acquisition
 provide definitions for the following important terms:

critical period expansion

input analytic vs gestalt learning
caretaker speech referential vs expressive learning
baby-talk words cross-linguistic research

Preview
In this study unit we focus on the language directed at young children, and
investigate the problem of how children use the language they hear around
them as a basis for working out the rules of the language in a way that will
enable them to use language creatively. To balance our earlier focus on univer-
sal aspects of language acquisition, this study unit also focuses on aspects of
language acquisition that differ from child to child, such as the rate and style
of acquisition. We conclude by looking at some differences in the course of
language acquisition across differing language groups.

LIN2601/1/2017-2023 99
 Introduction
Imagine a situation in which a child grows up hearing no spoken language – a baby
abandoned in a forest, for example, and brought up by wolves. What language
or communication system would you expect the child to use? Write your answer in
the space provided before you continue reading.

Children who have been isolated from human contact do not acquire language,
although they may use grunts and gestures of various kinds as a primitive form
of communication. A real-life example of this situation occurred in France in 1799,
when an 11-year-old boy was found in the countryside around Aveyron. He had
grown up without human contact, feeding off forest plants and fruits, and he spoke
no language. He was nicknamed Victor, the wild boy of Aveyron, and was later
extensively trained to try and get him to use spoken language:
A young French doctor named Itard took over the task of attempting to train
Victor for normal life, and subjected him to a strict training regimen. It is
particularly noteworthy, therefore, that despite this regimen, Victor did not
progress very far linguistically. With great effort, he apparently mastered the
concept of word reference, initially treating the words he learned as names
which could only be applied to the objects on which he had been trained, and
then gradually coming to extend them to any of a class of similar objects. He
even over-generalised creatively in the way normal children are known to do.
Having mastered some nouns, Victor was then taught a few adjectives (e.g.
‘petit’ (‘small’), ‘grand’ (‘big’)) and some action verbs (e.g. ‘jeter’ (‘throw’),
‘toucher’ (‘touch’)). This highly skeletal communication system appeared to
have been the height of Victor’s linguistic achievement. He did manage to
write the words in his system, but the system itself does not appear to have
become any more complicated: no pronouns, no auxiliaries, no embedded
clauses (Foster-Cohen 1999:123–124).

A similar case is that of deaf children who grow up in families with hearing parents.
Often these children do not acquire sign language as a mother tongue because they
lack a signing environment at home. Instead they learn sign language after age 7,
when they encounter it for the first time from their peers in schools for the deaf.
Research has shown that these children never become fluent signers, regardless of
how long they are exposed to sign language (see Unisci 2002 http://www.unisci.com/
stories/20021/0104026.htm). This evidence suggests that there is a critical period
during which a first language must be acquired, some time before age 7, in order
for effective language acquisition to take place.
If you are interested, you could read about the case of the ‘languageless’ 13-year-old child
Genie and its implications for the existence of a critical period for language acquisition at
http://psychology.about.com/od/historyofpsychology/a/genie.htm

100
 1 Caretaker speech
The examples we have discussed above are exceptions to the normal environment in
which children acquire their first language. Most children are exposed continually
to language addressed to them by caretakers. In this context the term ‘caretaker’
refers to any person involved in the bringing up of children, for example mothers,
fathers, older siblings, grandparents, domestic helpers, preprimary school teachers,
etc. According to the social interactionists, if we want to understand how children
acquire a complex system like language, then we need to examine more closely the
nature of this input – the language children hear – as well as the environment in
which the child grows up. By environment we refer to both the social (i.e. the culture,
class and ethnic group in which the child grows up) and the linguistic environment
(i.e. the number of languages and the nature of the input that the child is exposed
to). This problem can be broken down into the following kinds of questions:
1. Do adults speak in a different way to young children than they do to older
children and adults?
2. If adult speech to young children is different from ordinary speech, how does
it differ, i.e. what are its characteristics?
3. Do all members of a speech community and do all cultures speak to children
in the same way? In other words, are the characteristics of caretaker speech
universal? (Here we are interested in the way in which different societies as
well as different communities within the same society speak to children.)
4. Is caretaker speech necessary, or is it simply a useful facilitator in the acquisition
process?

Research has shown that adults do indeed speak differently when addressing young
children than when addressing older children and adults. Psycholinguists use the
terms caretaker speech, motherese or child-directed speech to refer to speech
directed at young children. You should be aware that the social organisation and
the status and role of children in a particular society affect the way that adults
interact with them and the amount and kind of speech addressed to them. In some
cultures adults interact very little with children, believing they should be seen and
not heard. For example, the Kaluli people in Papua New Guinea do not engage in
much sustained interaction with their children until they are at least 18 months
old. Other societies believe that it is important to interact with their children from
an early age, and encourage children to express themselves and contribute to the
conversation.

A short video on caretaker speech in different cultures is available at

https://www.youtube.com/watch?v=AtYojRu7ajY

Let’s look at a South African example to get an idea of some of the characteristics of
caretaker speech. In the following conversation a grandmother is interacting with
her granddaughter of 2;6 in Sesotho (Demuth 1984:34–35).
Grandmother: ’Neuoe uena le etsa joang?
What are you doing?
Neuoe (2;6) Re isa (Adult form = rea itlhatsoa)
We’re washing.
Grandmother: E?
What?

LIN2601/1/2017-2023 101
 Neuoe: Rea isatsoa (Adult form = rea itlhatsoa)
We’re washing ourselves.
Grandmother: Lea itlhatsoa?
Are you washing yourselves?
Neuoe: E.
Yes.
Grandmother: Le itlhatsoa hokae?
Where are you washing?
Mother: Ere nokaneng.
Say ‘at the creek’.
Grandmother: Le itlhatsoa hokae?
Where are you washing?
Neuoe: Linokanengeng (Adult form = linokaneng)
At the creeks.

Research has identified the following characteristics of caretaker speech (adult-to-

Diminutives are child language) as compared to ordinary adult-to-adult language:
words indicating
a small or  Sing-song, high-pitched intonation.
young version of  Use of special vocabulary. The simplified words and diminutives like bunny
something, e.g.
in English, dog (‘rabbit’), dindins (‘dinner’) and doggy (‘dog’) used by adults when speaking to
becomes doggy, very young children are known as baby-talk words. This phenomenon is found
in Sepedi mpša in many different languages, for example in Sepedi the adult word dijo (‘food’)
becomes mpšana becomes jaja and the adult word ngwana (‘baby’) becomes nana.
etc.
 Shorter utterance length.
 It is generally grammatical.
 It focuses on the ‘here and now’, i.e. on things that are observable in the child’s
immediate environment at that particular time. In the example above, the grand-
mother is asking questions about the game the girls are currently engaged in and
the actions they are miming.
 Slower speech rate than in adult-to-adult language, with frequent pauses between
utterances.
 Clear enunciation.
 Frequent questions and commands. For example, in the Sesotho example above,
notice the frequent use of questions to encourage the child to speak.
 Frequent repetition and expansion. Expansion is the term used when a caretaker
repeats a child’s utterance in a grammatically correct form, often adding elements
omitted by the child. For example, when a child says bean, the mother might
respond Yes, eat your beans. Expansions are very common in caretaker speech,
making up about 20% of all utterances addressed to children (Strapp 1999:386).
 Caretakers often fine-tune their speech so that it is slightly above the current
level of their children’s language output.

Task 5.1

Look at the transcript of caretaker speech below, in which a father and his daughter
Ella aged 2;4 are having breakfast.

102
 You can also watch a video of the interaction at
http://childes.psy.cmu.edu/browser/index.php?url=Eng-UK/Forrester/biggirl.cha

12 *F: do you want some banana or some toast → ▶

13 %gpx: child touching camera near high-chair
14 *E: no I like (nu)tella → ▶
15 *F: nutella alright → ▶
16 *E: I want(nu)tella → ▶
17 *E: I don’t know (wh)at (th)at → ▶
18 *E: a camera → ▶
19 *E: ohhh waa gaaa → ▶
20 *F: what’re you doing with my camera ↗ ▶
21 *E: he he
22 (0.5) hmm holding it holding it → ▶
23 *F: be careful with it → ▶
24 *E: oh, why ↗ ▶
25 *F: you know why → ▶
26 *E: why ↗ ▶
27 *F: because it’s ⌈not → ▶
28 *E: ⌊wha wha wha → ▶
29 (0.5)
30 *F: its not a to:::y → ▶
31 *E: he (.) hey hey &=laugh → ▶
32 *E: &=sings → ▶
33 %gpx: continuing to touch camera .
34 *F: what pictures do we take with the camera ↗ ▶
35 (1.)
36 *E: I can’t remember → ▶
37 *F: you can’t remember ↗ ▶
38 *F: did we take pictures when Ella was very small ↗ ▶
39 (2.)
40 *E: no::o → ▶
41 *F: she was tiny tiny baby → ▶
42 *E: no::o → ▶
43 *F: +, no → ▶
44 *E: I’m nott tiny baby, dada da → ▶
45 *F: I didn’t say you were (.) → ▶

1. Give the child’s age in months and the stage of linguistic development she is at.

2. What is the MLU of the father’s utterances (in morphemes)?

3. Would you expect the child’s MLU to be lower or higher than the father’s?

4. Would you agree that the father’s conversation focuses on the ‘here and now’?
If so, why?

LIN2601/1/2017-2023 103
 5. Give an example of repetition in the father’s speech.

6. How many questions does the father ask Ella?

7. Give an example of exaggerated intonation from the father’s speech.

Feedback
1. The child is 28 months old and in the multiple word stage.
2. A striking feature of the father’s speech is the use of very short utterances. In the
example above the sentences are all between 1 and 10 morphemes long. The
MLU of his utterances is 5.4 (65 morphemes divided by 12 sentences). Remember
to discard the utterance in line 27 as it has an unclear portion and don’t forget to
count what’re, doing, it’s, can’t, didn’t as two morphemes each.
3. We would expect the child’s MLU to be slightly lower than the father’s, and in fact
her utterances are all between 2 and 6 morphemes long, averaging around 4.
4. Yes, the conversation focuses on the ‘here and now’ because it focuses on the
breakfast activity happening at the time and the camera in the kitchen. The father
does, however, mention baby photographs taken in the past.
5. Repetition of words and phrases by caretakers is common, for example ‘she was
tiny tiny baby’. The father repeats some of Ella’s utterances, for example in lines
36 and 37 and lines 42 and 43.
6. Caretaker speech is characterised by frequent questions and commands. In the
example above, the father uses 4 questions in 12 utterances. Notice that although
there are no question marks in the transcript, the little arrows at the end of the
lines indicate intonation. There is rising intonation for most of the questions and
level intonation for the statements and commands.
7. The video footage is striking in its use of sing-song intonation and lengthened
vowels. You can see this in the transcript, for example, where the father lengthens
the vowel in ‘toy’.

2 Enriching a child’s linguistic environment

The general consensus seems to be that children who are brought up in a rich
verbal environment (i.e. one that provides plenty of opportunities for interaction,
practice and vocabulary growth) tend to fare better linguistically than those whose
environment is not verbally enriching. In other words, although all children will
acquire basic linguistic and communicative competence, children from a verbally
rich and stimulating environment are likely to have a wider vocabulary and greater
language awareness and communicative skills.
According to Leffel and Suskind (2013:267), enriching a child’s linguistic environment
at an early age has significant positive effects on the child’s mental and social
development. Children from an enriched environment acquire language at a faster
rate than children from an average environment, for example using two-word
sentences at 12 to 13 months instead of 20 to 22 months. The child’s problem-

104
 solving abilities, school performance, independence and motivation all benefit from
an enriched environment during infancy, and improve even more when enrichment
is continued throughout the child’s first years.
Research by Tamis-Lemonda, Bornstein and Baumwell (1998, 2001) indicates that
children with responsive caretakers reach the important milestones of 50 words and
the two-word stage earlier than other children. Responsive caretakers in this study
were defined as those who are sensitive to their children’s interests, talk about what
the child is doing and pay attention and react to children’s own utterances. In a
similar study, Kloth, Janssen, Kraaimaat and Brutten (1998) found that mothers’
interactional styles fell into one of three categories (the study focused on mothers
but applies equally to all caretakers):
 ‘Non-intervening’ mothers don’t ask their children a lot of questions and don’t
pressurise the child to respond.
 ‘Explaining’ mothers are talkative and dominate the conversation, providing
information, labelling objects and interrupting the child.
 ‘Directing’ mothers try and control the child’s behaviour by issuing commands,
warnings, criticising, saying ‘no’, etc.

One of their findings was that children with explaining mothers tend to have a
higher level of understanding than children with the other maternal interaction
styles. A possible reason for this is that these children simply hear more input. Leffel
and Suskind (2013) indicate that children who receive a lot of feedback that relates
back to their own utterances, in the form of imitation, expansion etc, tend to make
more rapid progress. Rapid progress is also associated with children whose mothers
initiate and facilitate conversation rather than using language mainly to control the
child’s behaviour.
Hart and Risley (1995) suggest that the amount of language children hear from their
parents has a direct influence on the size of their vocabulary. In a controversial but
influential study, they argue that because parents in families with low socio-economic
status (SES) speak less with their children, these children have smaller vocabularies
than children from families with high SES. In this study, parents with higher SES
talked much more to their children and thereby enriched both the quantity and
quality of language that their children heard. By age 4, they suggest that there is a
‘thirty million word gap’ between the largest parental input and the lowest; that is,
a child from a high-income family will be exposed to 30 million more words within
the first four years of life than a child from a low-income family. Hart and Risley
(1995) also claim that these differences in vocabulary size at age 3 have a direct
effect on children’s performance at school at age 9. Hart and Risley’s research has
resulted in the Thirty Million Words Initiative (teaching parents how to teach their
babies and young children effectively in order to develop their brains). See http://
thirtymillionwords.org/tmw-initiative/.

Figure 5.1 below gives an indication of vocabulary growth and the number of words
in a child’s productive vocabulary until age 3 in high, middle and low SES families.

LIN2601/1/2017-2023 105
 Figure 5.1 Disparities in early vocabulary growth (Hart & Risley 1995)
http://www.learningandthebrain.com/blog/math-for-babies/

Task 5.2

Figure 5.1 (Hart & Risley 1995) plots the number of words in children’s productive vo-
cabulary (on the vertical axis or side of the graph) against the age of the child (on the
horizontal axis or bottom of the graph).
At what age does the vocabulary spurt occur according to this data?

Use the data above to estimate how many words a child with a high SES knows at 24
months and compare it to a child with a middle and low SES at the same age. (Use a
ruler to draw an imaginary line upwards from the point marked 24 to the graph itself.
Make sure the ruler is straight.)
According to the data, approximately how many words do children from high, middle
and low SES learn between age 2 and 3?

106
 Feedback
The graphs are fairly flat at first but suddenly start to go up steeply from 16 months for
children from all socioeconomic groups. This indicates the beginning of the vocabulary
spurt.
The data indicates that at 24 months the child with a high SES has a vocabulary of about
300 words. At the same age a child with the middle SES has around 140 words and a
child with a low SES has just over 100 words. Since this is just an estimate read off the
graph, don’t worry if your answer differs by 25 words or so, but read the instructions
and check the graph again if your answers differ by more than this.
You can work out approximately how many new words these children learnt in the third
year of life by subtracting the total number of words used at 24 months from the number
of words used at 36 months. (This figure won’t be 100% accurate as some words drop
out of the child’s vocabulary at various stages.) Since the children with a high SES know
an average of 1 100 words at age 3, this means they learn about 700 words (1 100 –
300) in the third year. The middle income group learns around 450 words (590 – 140)
and the low income group only about 250 words (350 – 100).

There is considerable controversy about these findings, with critics arguing that
the methodology was faulty and that children from lower income groups and their
parents simply feel less comfortable around researchers and therefore talked less
than they might otherwise. Critics such as Michaels (2013) and Nation (n.d.) also
say this model stigmatises poor families and blames them for their children’s lack
of academic success:
The Hart and Risley study may be viewed as an attempt to explain why some
children do not succeed at school. It does this by using a deficit model. In
its crudest terms, this model says that low socio-economic status children do
not get enough language-based interaction with their parents and as a result
have a limited vocabulary. Because of this limited vocabulary, their cognitive
functioning suffers. Because an adequate level of cognitive functioning is the
basis of learning to read and succeeding in school subjects, low socio-economic
status children do not do well at school. This vocabulary deficit limits them
for the rest of their lives. Because of the severe methodological flaws in the
study, these conclusions are unwarranted (Nation 1995).

Despite these criticisms, there is subsequent research that supports the findings of
Hart and Risley (see for example the discussion in Hoff [2013] and Leffel & Suskind
[2013]). As with all controversies, you will have to consider the arguments of both
sides very carefully and make up your own mind about the best ways of interacting
with young children in order to support optimal learning.

If you are interested you could watch an interview with Professor Risley on
https://www.youtube.com/watch?v=fOLzEUmXV5s&feature=youtu.be&
list=PLLxDwKxHx1yIdFbP6nWJKqjczZWmwlcXx or listen to a mom
explaining how she is trying to enrich the way she speaks to her child on
https://www.youtube.com/watch?v=7BrrPOiPISw

Now that we have looked at some of the academic findings about caretaker speech,
let’s try and use those findings to formulate some practical advice about the best way

LIN2601/1/2017-2023 107
 to interact with small children. If you are a parent or caregiver, you probably want
to know how to go about enriching a child’s language environment, and the good
news is that it doesn’t require a huge investment in effort, money or time. One can
start stimulating children with enriching activities at any age, but the important
thing to realise is that children can benefit from hearing language and face-to-face
conversation from the time they are born. From birth, caretakers need to talk with
the baby face to face, imitate the baby’s vocalisations and make new sounds for him
or her to imitate. From about three months caretakers should join in as the baby
plays, name the things the infant looks at or touches, draw the infant’s attention to
things and name them, and begin to use picture books. From nine months caretakers
should expand on the child’s utterances and introduce the child to rhymes, songs
and story books. From 14 months adults should read stories and information books,
involve the child in conversations and encourage the child to talk about his or her
experiences. Children should also be encouraged to role-play with their peers.
Here are some ideas for improving your own interactions with young children
(Fowler 1990, Bornstein & Putnick 2013):
 Talk to them a lot. Engage the child’s attention and be sensitive to when the
child is interested or is losing interest.
 It’s OK to use caretaker speech.
 Speak clearly and not too quickly.
 Avoid correcting children’s errors.
 Don’t use incorrect grammar like ‘Baby wan’ go home?
 Don’t bombard the child with complex language too early. Start with simple
sounds and words and increase the degree of difficulty gradually.
 Use appropriate movements and actions during songs and stories.
 Read books together every day. Joint book reading exposes children to vocabu-
lary and concepts that are not commonly used in everyday conversations. Start
with simple picture books with one picture per page, then gradually work up to
more complex scenes. You can make your own picture books by cutting pictures
out of magazines or find amazing books online at www.bookdash.org or the In-
ternational Children’s Digital Library http://en.childrenslibrary.org/index.shtml.
Continue the book-reading sessions until the child can read alone.
 Tell oral stories. Like reading, storytelling promotes a range of language and
literacy skills from new words and grammatical constructions to imagination
and originality.
 Take the child to different areas of the house/neighbourhood and talk about
the various places you go to and activities you do. Talk about these places and
activities afterwards as well to encourage reflection.

3 Individual differences
For many years, as a result of the dominance of Chomsky’s theory of Universal
Grammar, there was a lack of interest in individual differences, but this changed in
the 1980s. Although universal patterns are very striking, individual differences also
characterise the language acquisition process. Not all children fit into the typical
pattern we have learnt about so far, and in this section we show you that the language
acquisition process can differ in both rate (speed of development) and route (pattern

108
 or sequence of development). Like adult speakers, children use language differently
depending on their gender, social class and culture. Individual personality traits
can also have an effect on the rate of development and the acquisition strategies
children prefer.
Wells (1992:112) identifies four groups of factors which can influence children’s
language acquisition and cause it to vary in rate or route. Variation is a result of an
interaction between all these factors:
 inherited factors like sex, intelligence and personality
 the social and cultural background of the child and the family structure
 the style of interaction, e.g. interpersonal relations and child-rearing methods
 situational factors like the setting, activity and the status of the participants

3.1 Differences in the route of language development

Acquisition styles and strategies may also differ from child to child. For example,
a distinction has been made between analytic and gestalt learning strategies.
Children who follow an analytic style tend to analyse the input into its constituent
parts and then gradually build up utterances in the usual way, starting with one-word
utterances and moving gradually to two- and three-word sentences. Children who
prefer a gestalt style focus on whole utterances, and often produce larger chunks of
speech at an early age, even if the meaning of each of the parts is not known. For
example, phrases like What’s it doing?, Lemme have it or I don’t want it might be
used at a stage where analytic children are using only one-word utterances (Foster-
Cohen 1999:130–131). Most children fall somewhere between these two extremes
and use a combination of both strategies in their early utterances.
One of the earliest studies to highlight differences between children was Nelson’s
(1973) study of the early words used by children about a year old. The children in
the study all had words for the people and objects in their everyday environments,
but the children seemed to fall into two groups depending on how many nouns
were present in their first 50 words. The referential children had a high proportion
of nouns in their vocabularies (more than 50% of their vocabularies consisted of
nouns). They tended to use an analytic strategy (acquiring single words rather than
phrases) and acquired words faster than the other group – the expressive children.
The expressive children had fewer words for objects in their vocabularies (i.e.
nouns made up less than 50% of the total vocabulary), but used more pronouns
and multiword personal-social expressions like Go away, Oh dear, Thank you and I
want it. These children acquired words at a slower rate than the referential children.
Because of the number of multiword gestalt expressions used during the one-word
stage, the transition between the one-word stage and the two-word stage was less
clear in the expressive children.
Nelson concluded that referential and expressive children use language to fulfil
different functions: referential children were learning language to categorise the
objects around them, while expressive children were more socially orientated, using
language to talk about themselves and others, to express feelings, needs and social
forms. You should note that there were only 18 children used in Nelson’s research,
so it shouldn’t be accepted without reservation. Additional research has suggested
that most children fall somewhere between these two extremes, and only a few
children exhibit a distinctly referential or expressive strategy.

LIN2601/1/2017-2023 109
 Task 5.3

1. Look at the following data from Nelson (1973:22) and classify each child as either
referential or expressive. Remember to justify your answer.

Elizabeth’s first 50 words

object names 24
action words 6
adjectives 6
personal-social words 12
function words 2

Rachel’s first 50 words

object names 41
action words 4
adjectives 2
personal-social words 1
function words 1

2. Look back at Task 3.5 and decide whether Sibusiso is a referential or expressive child.

Feedback
1. Elizabeth is an expressive child as less than half of her vocabulary (24 out of 50
words) consists of object names. Twelve of her first words are personal-social
words, compared to Rachel who uses only one personal-social word. Rachel is
a referential child as 41 of her first 50 words (82%) are object names.
2. The child whose vocabulary we looked at earlier, Sibusiso, is an expressive child. Of
his first 24 words, 9 (less than half) refer to objects. However, Sibusiso has a pronoun
(/u/ ‘me’) in his early vocabulary and many personal-social words like /ta/ ‘give it to me’ and
/baba/ ‘byebye’, and therefore seems to prefer an expressive learning strategy.

The nature–nurture controversy raises its head again as researchers have attempted
to account for the referential-expressive distinction. Are these different styles a result
of innate cognitive differences between children (nature), or do they result from
different ways in which caretakers interact with their children (nurture)? Again the
answer seems to be both. Referential children spend a lot of time playing with toys
and objects, while expressive children tend to spend less time playing and more time
interacting with adults. Mothers’ interaction with their children also has an influence.
Mothers of referential children tend to spend a lot of time reading books to their
children or playing with toys together. These activities provide opportunities for
naming objects and learning new nouns. Mothers of expressive children may prefer
to engage in conversational routines with their children and other forms of play that
do not involve toys. The kinds of expressions learnt from this kind of mother–child
interaction tend to be social words like Thank you and Hi and phrases like Let me

110
 see. Additional evidence that it is the caretaker’s input that determines the child’s
early acquisition style comes from Bates, Bretherton and Snyder (1988), who show
that a bilingual child was highly referential in her one-word vocabulary in English,
but expressive in her later learning of Italian, for which she had a different caretaker.

3.2 Differences in rate of acquisition

As mentioned above, children also differ in their rate of acquisition. Some children
start developing later than others and some move through the developmental stages
faster than others. For example, some children are reported to produce little or
no meaningful speech before 16 months of age, while others have vocabularies of
more than 100 words at or before 16 months (Caselli et al 1995). Some children
accumulate words in a steady fashion with no obvious vocabulary spurt, or display
two vocabulary spurts, or start their spurts much earlier or later than the 16-month
average (Tomasello & Slobin 2005:67). Normal children can differ by up to three
years in their rate of language development, though the stages they pass through
are generally the same regardless of how stretched out or compressed the stages are.
As we saw earlier, the amount of talk the child hears and participates in is one variable
that affects the rate of acquisition. For this reason, firstborn children are often faster
developers because they receive more of their parents’ undivided attention and
engage in shared activities more often than younger siblings. Blind children seem
to be slightly slower to talk, as the nonverbal cues and context they can draw on are
limited, although they usually catch up to sighted children by age 3 (Hoff 2013:338).
Research suggests that learners whose language develops faster than average use
more nouns and complex noun phrases and make more grammatical errors. They
appear to focus on the task of communicating, rather than on the fine details of
grammar. This kind of ‘risk-taking’ approach results in rapid progress. Slower
learners, in contrast, use fewer noun phrases but more verb phrases, and perform
better than fast learners in comprehension tests. Their approach to language is
more cautious and conservative, and focuses on the details of language structure,
resulting in a slower rate of acquisition (Goldfield & Snow 2005:306). Here is Reich’s
account of the differences between his own children. (Also see if you can identify
the phonological simplification strategy used by Quentin in the given example. Is
it reduction, substitution or assimilation?):
The fact that some children are more reticent than others in using words that
they know they can’t pronounce correctly may be one cause of variation in the
time a child first starts to speak. It may also explain why sometimes previously
acquired words drop out of the vocabulary of a child for a period. Quentin
was more conservative than Athena. He was slow at beginning to speak, but
when he did start, his speech was easy to understand. This desire to be cor-
rect was evident even after he began to speak. Once Quentin, a Star Trek
fan, tried to say Mr. Spok. All he could produce was Mr. Spop. He broke into
tears when he found he couldn’t say it properly. On the other hand, Athena
was more concerned with doing things that she saw others doing than she was
in doing them right. She began to speak early, but her phonology was much
worse than that of her brother (Reich 1986:58–59).

LIN2601/1/2017-2023 111
 4 Cross-linguistic differences
We mentioned earlier that in the last thirty years there has been a marked increase
in cross-linguistic research, i.e. research that compares language acquisition in
different languages and cultures. The focus was initially on identifying cross-linguistic
similarities, but these days the focus is more often on identifying cross-linguistic
differences in the language acquisition process. Dan Slobin (1997) outlines the main
aims of this research as, firstly, charting cross-linguistic diversity by engaging in
detailed studies on the acquisition of different languages and, secondly, determining
the limitations on this diversity, if any.
Thanks to a generation of cross-linguistic studies, it has become clear that patterns
of language learning do vary considerably across different language communities.
For example, the culture in which a child grows up has been shown to affect the
acquisition of early words. The following table lists the 20 most common first
words in English-speaking and Cantonese-speaking children aged 8 to 16 months
at the very earliest stage when the child uses less than 10 words. Words in italics
are common to both languages.

Table 5.1 The 20 most common first words in English and Cantonese-children
(Tardif, Fletcher, Liang, Zhang, Kaciroti & Marchman 2008:932)

United States Beijing

English-speaking (n 264) Cantonese-speaking (n 336)
Daddy (54) Mommy (87)
Mommy (50) Daddy (85)
BaaBaa (33) Grandma – Paternal (40)
Bye (25) Grandpa –Paternal (17)
Hi (24) Hello?/Wei? (14)
UhOh (20) Hit (12)
Grr (16) Uncle – Paternal (11)
Bottle (13) Grab/Grasp (9)
YumYum (13) Auntie – Maternal (8)
Dog (12) Bye (8)
No (12) UhOh (Aiyou) (7)
WoofWoof (11) Ya/Wow (7)
Vroom (11) Sister – Older (7)
Kitty (10) WoofWoof (7)
Ball (10) Brother – Older (6)
Baby (7) Hug/Hold (6)
Duck (6) Light (4)
Cat (5) Grandma – Maternal (3)
Ouch (5) Egg (3)
Banana (3) Vroom (3)

Tardif et al (2008) note that, as we would expect, both lists are dominated by
people words and objects that can be manipulated. However, there are significant
cross-cultural differences between the two lists, notably with regard to the number

112
 of words for family members in Cantonese and the number of animal names in
English. Kinship is an important cultural feature of Cantonese and the words of
various kinship relations are therefore among the first to be acquired. In societies
like the USA where material possessions such as toys and books form an important
part of child culture, children quickly acquire the names of various toy objects and
animals in their environment.
Grammatical differences between languages also affect the course of language
acquisition. According to Caselli et al (1995), children’s ‘first sentences are always
tailored to and shaped by the structural properties of their native language’. This
means that the telegraphic utterances of children from different language groups
will differ in terms of word order and frequency of different constructions, and
the order in which grammatical structures appear will differ as well, depending on
how common and how complex these are in the adult language input that children
hear. For example, Katherine Demuth (2003) observes that Sesotho children start
using passives at an average age of 2;8, years ahead of children learning English
(who typically start using passives after age 5). She attributes this to the frequency
with which passives are used in adult Sesotho.
Additional evidence of cross-linguistic differences is provided by Slobin (2002:8–9).
Using the CHILDES database, he compared the verbs relating to ‘manner of motion’
produced by English, Spanish and French children aged 2 to 4:
English: bump, chase, climb, crawl, creep, dance, float, flop, fly, hike, hop, jog,
jump, march, paddle, pounce, race, roll, run, rush, scoot, skip, slide, slip, sneak,
step, swim, tread, trip, trot, walk, wiggle
French: courir, faire du ski, glisser, nager, sauter, voler [= run, ski, slip, swim,
jump, fly]
Spanish: bailar, caer(se), correr, chocar(se), escaparse, nadar, saltar, volar [= dance,
fall, run, crash, escape, swim, jump, fly]

It is striking that the English children use 34 different ‘manner of motion’ verbs,
many more than the French and Spanish children. While English children learn
very specific vocabulary to differentiate many different sorts of movement – crawl vs
creep, run vs jog, etc, French and Spanish children have a much smaller and simpler
vocabulary in this area. This difference mirrors the structural difference between
English, which uses verbs of manner, and Spanish and French, which tend to use
a simple verb, adding an adverbial expression to describe the manner if necessary.
Manner of motion seems to be a matter of everyday attention for speakers of English,
while speakers of Spanish and French seem to attend to manner of motion only
when it is especially significant. Slobin (2002:14) concludes that children are trained
by their language – from very early on – to adopt particular perspectives on events,
based on dimensions of experience and activity that are seen as important in their
language. The implication is that acquiring a first language gives us a particular
perspective on the world, and acquiring additional languages provides us with new
perspectives.
There is still a lot of work to be done in the area of cross-linguistic language acquisition
research. For example, of the world’s 6 000 or 7 000 languages, only 10% have
been analysed grammatically and only 30 are represented in the online CHILDES
transcripts. Detailed language acquisition studies require a large investment in time
and money, with native-speaker researchers and research assistants, translators,

LIN2601/1/2017-2023 113
 recording equipment, etc. As these studies are undertaken in more and more
languages, the effect of language diversity on the language acquisition process will
become clearer.

5 Summary
In this study unit we have moved away from considering universal commonalities
and turned our attention to some of the differences in the rate and route of language
acquisition. We looked at how the input children hear affects their language acquisition,
as do internal traits such as personality and other individual differences. We also
identified some cross-linguistic differences in language acquisition relating to the
cultural environment and the structure of different languages. In the following study
unit we will continue to focus on differences in the path of language acquisition as
we look at language development in children with various disabilities.

Task 5.4

Test your understanding of the study unit by choosing the option that fits best with the
given statement in the following multiple-choice questions. Circle the correct answer
([1], [2], [3] or [4]).

1. Caretaker speech differs from adult-to-adult language in the following way:

a Caretaker speech is faster than adult-to-adult language in words per minute.
b Caretaker speech contains many vocabulary items not found in adult-to-adult
language.
c Caretaker speech contains fewer questions and commands than adult-to-
adult language.
d Caretaker speech contains less repetition than adult-to-adult language.

2. Which of the following statements is true?

a Caregivers often underextend words.
b Caretaker speech utterances are usually slightly shorter than the child’s ut-
terances at that stage.
c Caretaker speech is characterised by disfluencies and grammatical errors.
d Caretaker speech is characterised by pauses between utterances.

3. Which of the following are not baby-talk words in English?

a blanket, bottle
b kitty, doggy
c tummy, bunny
d mommy, daddy

4. Which of the following statements describes referential children?

a They use many multiword personal-social expressions like Go away and
Thank you.
b Less than 50% of their vocabularies consists of nouns.
c They use language mainly to express feelings and needs.
d They acquire words at a faster rate than expressive children.

5. The ‘30 million word gap’ research suggest that children from disadvantaged homes
a will never have a productive vocabulary of 30 million words
b have vocabularies that are 30 million words below those of children from
wealthier homes at age 4 due to poverty and poor nutrition

114
 c hear an average of 30 million fewer words than children from high income
groups by the time they are 4
d achieve academically as well as advantaged children by the time they are 9

6. Cross-linguistic research shows that

a patterns of language learning are constant across different language
b communities
c the culture in which a child grows up affects the acquisition of early words
d the telegraphic utterances of children from different language groups follow
a similar structure
e the order in which grammatical structures appear is universal

Feedback

1. [2] 2. [4] 3. [1] 4. [4] 5. [3] 6. [2]

Academic vocabulary building

Here are some academic terms with tricky plurals. Make sure you know which
endings are singular and which are plural. Test yourself by filling in the correct
word ending in the sentences below:
Singular Plural
phenomenon phenomena
criterion criteria
hypothesis hypotheses
thesis theses
appendix appendices
stimulus stimuli
data data
Many interesting linguistic phenomen have been observed
by listening to children’s utterances.
Append 5 is on page 275.
He spends a lot of his time reading thes on the subject of language
acquisition.
MLU is the most important criteri for deciding on a child’s stage of
development.
Several hypothes have been put forward to account for the language
acquisition process.
The need to communicate acts as a stimul for children to learn language.
The dat shows that children adopt different language acquisition
strategies.
(Data is usually used as an uncount noun with a singular verb like shows. Some
people use it as a plural form, and would use the plural verb show.)

LIN2601/1/2017-2023 115
 Further reading

Those of you who are interested in this topic might like to read one of the following
books:
Berko Gleason, J (ed.). 2005. The development of language. 6th edition. Boston, MA:
Pearson. Chapter 8 is on individual differences.
Deutscher, G. 2010. Through the language glass: why the world looks different in other
languages. London: Random House.
Whitehead, MR. 2007. Developing language and literacy with young children 0–8.
3rd edition. London: Paul Chapman.

Bibliography

Bates, E, Bretherton, I & Snyder, L. 1988. From first words to grammar: individual
differences and dissociable mechanisms. New York: Cambridge University Press.
Bloom, P. 2001. Roots of word learning, in Language acquisition and conceptual de-
velopment, edited by M Bowerman & SC Levinson. Cambridge: Cambridge
University Press.
Bornstein, MH & Putnick, D. 2012. Cognitive and socioemotional caregiving
in developing countries. Child Development 83(1):46–61 doi:10.1111/j.1467-
8624.2011.01673.x (accessed 30 March 2016).
Caselli, MC, Bates, E, Casadio, P, Fenson, J, Fenson, L, Sanderl, L & Weir, J. 1995.
A cross-linguistic study of early lexical development. Cognitive Development
10(2):159–199.
Demuth, KA. 1984. Aspects of Sesotho language acquisition. Bloomington, IA: Indiana
University Linguistics Club.
Demuth, KA. 1992. Acquisition of Sesotho, in The cross-linguistic study of language ac-
quisition Vol 3, edited by D Slobin. Hillsdale, NJ: Lawrence Erlbaum, 557–638.
Demuth, KA. 2003. The acquisition of Bantu languages, in The Bantu languages,
edited by D Nurse & G Philippson. Surrey, England: Curzon Press, 209–222.
Department of Linguistics 1990. Study guide 2 for LNG100S (Language in use).
Pretoria: Unisa.
Department of Linguistics. 1992. Only study guide for LNG2043. Pretoria: Unisa.
Department of Linguistics. 1996. Only study guide for LNG100S. Pretoria: Unisa.
Foster-Cohen, SH. 1999. An introduction to child language development. New York:
Addison Wesley Longman.
Fowler, W. 1990. Talking from infancy: how to nurture and cultivate early language
development. Cambridge, MA: Brookline Books.

116
 Goldfield, BA & Snow, CE. 2005. Individual differences, in The development of lan-
guage, edited by J Berko Gleason. 6th edition. Boston, MA: Pearson, 292–323.
Hoff, E. 2013. Language development. 5th edition. Belmont, CA: Wadsworth Cen-
gage Learning.
Huttenlocher, J, Vasilyeva, M, Cymerman, E & Levine, S. 2002. Language input
and child syntax. Cognitive Psychology 45:337–374.
Kloth, S, Janssen, P, Kraaimaat, F & Brutten, GJ. 1998. Communicative styles of
mothers interacting with their preschool-age children: a factor-analytic study.
Journal of Child Language 25:149–168.
Leffel, K & Suskind, D. 2013. Parent-directed approaches to enrich the early language
of children living in poverty. Seminars in Speech and Language 34(4):267–277.
Lenneberg, E. 1967. Biological foundations of language. New York: Wiley.
Michaels, S. 2013. Déjà vu all over again: what’s wrong with Hart & Risley and a
‘linguistic deficit’ framework in early childhood education? Learning Landscapes
7(1):23–41 Available at: http://www.learninglandscapes.ca/images/documents/
ll-no13/michaels.pdf (accessed 29 March 2016).
Nation ISP. n.d. A brief critique of Hart, B & Risley, T. (1995). Available at:
http://www.victoria.ac.nz/lals/about/staff/publications/paul-nation/Hart_
and_Risley_critique.pdf (accessed 29 March 2016)
Nelson, K. 1973. Structure and strategy in learning to talk. Monographs of the Society
for Research in Child Development 38(1–2).
Pinker, S. 1995. The language instinct: the new science of language and mind. New
York: Penguin.
Reich, PA. 1986. Language development. Englewood Cliffs, NJ: Prentice-Hall.
Rice, M, Smolik, F, Perpich, D, Thompson, T, Rytting, N & Blossom, M. 2010.
Mean length of utterance levels in 6-month intervals for children 3 to 9 years
with and without language impairments. Journal of Speech, Language and Hear-
ing Research 53(2):333–349.
Slobin, DI. 1997. The cross-linguistic study of language acquisition Vol 5: expanding the
contexts. Mahwah, NJ: Lawrence Erlbaum.
Slobin, DI. 2002. Cognitive and communicative consequences of linguistic diver-
sity, in The diversity of languages and language learning, edited by S Strömqvist.
Lund, Sweden: Lund University, Centre for Languages and Literature, 7–23.
Stoll, S. 2009. Crosslinguistic approaches to language acquisition, in The Cambridge
handbook of child language, edited by EL Bavin. Cambridge: Cambridge Uni-
versity Press, 89–104.
Strapp, CM. 1999. Mothers’, fathers’, and siblings’ responses to children’s language
errors: comparing sources of negative evidence. Journal of Child Language
26:373–91.

LIN2601/1/2017-2023 117
 Tamis-Lemonda, CS et al. 1998. Predicting variation in the timing of language
milestones in the second year: An events history approach. Journal of Child
Language 25:675–700.
Tamis-Lemonda, CS, Bornstein, M & Baumwell, L. 2001. Maternal responsive-
ness and children’s achievement of language milestones. Child Development
72(3):748–767.
Tardif, T, Fletcher, P, Liang, W, Zhang, Z, Kaciroti, N & Marchman VA et
al. 2008. Baby’s first 10 words. Developmental Psychology 44.4:929–938.
DOI: 10.1037/0012-1649.44.4.929.
Tomasello, M & Slobin, DI. 2005. Beyond nature-nurture: essays in honor of Eliza-
beth Bates. Mahwah, NJ, London: Lawrence Erlbaum,
Unisci. 2002. Critical period for acquiring nonverbal language. Available at:
http://www.unisci.com/stories/20021/0104026.htm (accessed 29 March
2013).

118
 Study Unit 6 Study unit 6
Language disability
‘To measure the success of our societies, we should examine how well those
2

with different abilities, including persons with autism, are integrated

3

as full and valued members’

4

Ban Ki-moon (South Korean politician and 8th secretary-general

5

6 of the United Nations, 1944)

OUTCOMES
After you have studied this study unit you should be able to
 explain what Specific Language Impairment, dyslexia, Down syndrome, Wil-
liams syndrome and autism spectrum disorder are
 identify the causes or possible causes of these disabilities and the parts of the
brain that are affected in each case
 describe how language development is affected as a result of the specific disabilities
 explain why the study of language development in children with disabilities is
important for psycholinguistics
 outline what recent research on children with disabilities has revealed about the
relationship between language and the brain, and language and cognition
 provide definitions for the following important terms:

Specific Language Impairment Autism Spectrum Disorder

(SLI) delayed vs deviant language
dyslexia development
association/dissociation between Asperger’s syndrome
language and cognition mutism
intellectual disability
echolalia
Down syndrome
Williams syndrome

Preview
In this study unit you will be able to make use of the knowledge that you have
built up so far (in study units 1–5) to examine an important topic, that of
atypical language development in children who are born with, or acquire at an
early age, some kind of disability that impairs their ability to receive, process
or produce language. We will look at language development in children with
Specific Language Impairment, dyslexia, Down syndrome, Williams syndrome
and Autism Spectrum Disorder. In each case we shall see how the particular
impairment affects language development and what steps are typically taken
to aid the child in developing effective linguistic and communicative strategies.

LIN2601/1/2017-2023 119
 Introduction
In the previous study units you learnt something about where specific language
functions are situated in the brain and you became acquainted with the different
stages of language development that ordinary children go through. You also made
acquaintance with some of the main controversies in developmental psycholinguistics,
namely, the nature–nurture debate and the question of whether language and
cognition are separate abilities or whether language is dependent on cognition. The
kind of knowledge you have acquired so far will stand you in good stead for this
unit, which deals with language development in children with cognitive disabilities.
The domain of children with language problems is a vast and interesting one and
An impairment covers language development in children who have sensory impairments (e.g. deaf
is a disability or blind children), physical impairments (e.g. cerebral palsy), cognitive impairments
which negatively (e.g. Down syndrome), neurological impairments (e.g. autism and stuttering) and
affects normal
functioning. children with delayed language development in the absence of any apparent cognitive
or neurological disability. This domain is too vast to be dealt with in a single unit
and we have thus necessarily had to narrow our focus. We will look specifically at
the various ways in which language is affected in children with an unexpected
A syndrome is Specific Language Impairment or reading disorder such as dyslexia, and in children
a collection of
clinical signs and with Down syndrome, Williams syndrome and Autism Spectrum Disorder. In each
symptoms. case we shall briefly discuss the possible causes of the impairment, identify the
symptoms that characterise the disorder, describe ways in which the impairment
affects language and the nature of the communicative difficulty, and briefly consider
possible remedial therapy or intervention programmes that can help to minimise their
linguistic problems and aid them to develop more effective communicative skills.
Why is it important to study the language of children with disorders? We believe
it is important for three reasons. First of all, our knowledge about how language
develops in ordinary children can help us understand how the process of language
development in children with disorders is similar to or different from normal language
development. Secondly, data about the language of children with disorders can provide
us with valuable insight into the relationship between language and other factors
such as biological, physical, cognitive and neurological factors. Thirdly, by examining
the relationship between specific disorders and language development, we hope to
understand these disorders better and find ways to treat them more effectively and
maybe even prevent them in the future. As Bernstein Ratner (2005:325) points out,
it has been estimated that ‘8 to 12% of school-age children demonstrate patterns
of communicative development that may be termed “delayed” or “disordered” ’.
This is a relatively large proportion of children that is affected. It is therefore crucial
that we advance our understanding of their problems and find ways to help these
children improve their linguistic skills. After all, it is through language that we
express ourselves and interact with other people and our world in meaningful ways.

1 Specific language impairment (SLI) and dyslexia

The incidence The last twenty years have seen an increasing focus on a disorder known as specific
of a disability or language impairment or SLI. This refers to children who have normal health,
disease is the rate hearing, and cognitive abilities, but who nevertheless display language skills below
or frequency with
which it occurs. those expected for their age (in the bottom 10%) and below their nonverbal IQs.
These children therefore have specific problems with language even though they do

120
 not have problems with other cognitive abilities like maths or problem-solving. SLI
is difficult to diagnose until about age 4, but the incidence of SLI in 5-year-old
children is approximately 7% (Tomblin 1996 cited in Bernstein Ratner 2005:352).
Children with SLI experience delayed language acquisition and most are late talkers
(though not all late talkers have SLI) (Bernstein Ratner 2005:353). They often
experience difficulties with word retrieval, for example referring to an English muffin
as ‘something round and English’ or forgetting the word ‘spoon’ and saying ‘you eat
breakfast with it’. Their sentences are shorter than those of children their age, and
grammatical morphemes are often omitted. They have difficulties with phonological
discrimination, morphology and syntax, and later with reading. The poor language
abilities associated with SLI typically persist throughout the early school years, and
sometimes on into adulthood (Tomblin 2009:425).
Here’s a short transcript of Mark, a child with SLI, aged 4;3 (http://childes.psy.
cmu.edu/browser/index.php?url=Clinical-MOR/Conti2/Mark/mark06.cha). Notice
the short MLU and the missing grammatical morphemes in Mark’s utterances:
591 %com: tower of bricks collapses
592 *CHI: hey that one not fell [?] .
594 %gpx: points to standing tower
595 *MOT: no it hasn’t has it ?
597 *CHI: 0 .
598 %par: squeaks
599 %act: touches tower so that it wobbles
600 *MOT: that’s mine though isn’t it ?
603 *CHI: yeah .
605 *MOT: what you gonna do ?
607 *CHI: 0 .
608 %act: pushes head gently against tower
609 %com: tower collapses
610 *MOT: oh !
613 *CHI: 0 .
614 %par: laughs
615 *MOT: fallen down now hasn’t it ?
617 *CHI: that [?] do again .
619 %act: starts to build
620 *MOT: you what ?
622 *CHI: do again .
The causes of SLI are unclear but it is known to run in families, and may therefore have
a genetic component. It has also been suggested that children with SLI have a slower
processing ability and/or problems with working memory (Montgomery 2003 cited
in Bernstein Ratner 2005:361). Recent research using MRIs on children with SLI has
shown that Broca’s area is sometimes smaller than expected and there is sometimes a
greater volume and greater connectivity in the cortex (Tomblin 2009:428). ‘Although
connectivity is generally considered important for neural networks, excessive
connectivity could result in inefficient and noisy processing’ (Tomblin 2009:428).
Language intervention for children with SLI is important, and involves modelling
appropriate responses, expansion of the child’s utterances and providing structure
for the children’s own attempts. Here are some examples from Bernstein Ratner
(2005:363):

LIN2601/1/2017-2023 121
 Parent: I am rolling the clay. You are too. Say ‘I am rolling’.
Child: I am rolling
Child: This clay no good.
Parent: This clay isn’t any good. It is too dry.
Parent: Look at these snakes I made. This one is very big. This one is very small. And this
one is …
Child: Skinny!

SLI often co-occurs with dyslexia, a ‘disorder manifested by difficulty in learning

to read despite conventional instruction, adequate intelligence, and socio-cultural
opportunity’. While SLI relates to unexpected difficulties in acquiring spoken
language, dyslexic children have an unexpected disability when it comes to reading.
For example, a 12-year-old dyslexic child may have normal intelligence but a reading
age of 8–10 years and a poor ability to match regular spelling and sounds and therefore
to write and comprehend written text. An adult dyslexic may have a spelling age of
only 12 years, a very low reading speed and a tendency to lose concentration while
reading. Similar-looking words like snip and ship, beg and bog, dig and big pose
particular problems for dyslexics.
Forty-three per cent of children with SLI are later diagnosed with dyslexia, and 55%
of dyslexic children meet the diagnostic criteria for SLI (Newbury et al 2011). The
exact causes are still uncertain although, like SLI, dyslexia tends to cluster in certain
families, indicating that there is a genetic component to the disability. There is a
higher prevalence of dyslexia in males than females, which also points to a genetic
or chromosomal cause.
Some researchers believe the symptoms of dyslexia result from a visual impairment
in which eye movements are not well controlled and 3-D perception is poor, while
others see it as a result of a poor phonological ability – the ability to segment speech
into sounds. Quite often the poor ability to segment words into strings of sounds
goes hand in hand with earlier speech problems such as lisping, stuttering or
delayed language development (Stein 1991:38). In fact most dyslexics seem to have
both phonological impairments and visuo-spatial impairments. One theory is that
dyslexia is associated with inadequate lateralisation of the two brain hemispheres.
You will recall from study unit 1 that lateralisation involves the brain hemispheres
specialising in different skills. While our linguistic ability, including the ability to
segment speech into sounds, is controlled by the left hemisphere, our visuo-spatial
ability is controlled by the right hemisphere. Reading thus differs from speaking and
listening in that it requires simultaneous processing in both hemispheres. Evidence
from dichotic listening tests and PET and MRI scans suggests that dyslexia is
characterised by an over-reliance on one brain hemisphere for processing written
text. In some dyslexics it is the right hemisphere that is dominant for processing
text and in others it is the left hemisphere.
Proposed therapies for dyslexia include drug therapy, high vitamin doses and
chiropracty, but none of these have been proved conclusively to help. Reading
intervention such as training in letter-sound correspondences can improve the spelling
and reading comprehension of dyslexics to some extent. One theory suggests that
‘multisensory’ approaches are beneficial in helping dyslexics make the link between
reading a word, writing a word and hearing a word (Thomson 1991:244). For

122
 example, children are taught to hear a word, say it aloud, spell it out aloud, write it
down while saying each letter as it is written and then read the word.
Research on children with SLI and dyslexia is important because it relates to the
question we raised in study unit 2, namely, whether language and cognition are
separate faculties or dependent on each other in some way. SLI and dyslexia, by
definition, result in impaired language without any associated cognitive deficiencies
and therefore provide some support for Chomsky’s view that language development
and cognitive development are indeed independent.

2 Intellectual disability and language development

In section 1 above, we looked at specific language difficulties in children with
normal intelligence. In this section we look at how language development is affected
in children with an intellectual disability. An intellectual disability is a deficit in
cognitive functions, such as reasoning, problem-solving, planning, abstract thinking,
judgement, academic learning, learning from experience, or practical understanding
(American Psychiatric Association 2013). Children with an intellectual disability
develop slower than others, and their mental age is always lower than their actual
age (e.g. an intellectually disabled child of 15 years may have the mental age of a
5-year-old). They experience difficulty functioning in one or more activities of daily
life, for example communicating, interacting with others or achieving personal
independence. There are many different causes of intellectual disability, with different
cognitive and linguistic patterns of ability and behaviour.
It is estimated that between 1 and 4% of school-aged children have some form of
intellectual disability. For example, Christiansen et al (2002) report an incidence of
3,56% of intellectual disability in their study of 6 692 children aged 2 to 9 around
the rural area of Bushbuckridge, South Africa. This brings us to the controversial
question concerning the measurement of intellectual ability. Although not everyone
agrees on the best way to assess intellectual functioning and ability, intelligent
quotient (IQ) tests continue to be used as a standard measure of intellectual ability.
An IQ score of 100 is regarded as average, and children with intellectual disabilities
tend to have IQ scores of around 70 or below.
Table 6.1 below presents the four categories of intellectual disability based on the
ability of the individual to cope with the requirements of everyday life:

Table 6.1 Categories of intellectual disability

Mild Some developmental delays, some language disabilities, no

unusual physical characteristics, can be educated successfully.

Moderate Noticeable developmental delays, physical signs of impair-

ment, basic communication skills, can be taught basic self-
care routines, e.g. toilet training, dressing themselves, etc.

Severe Considerable developmental delays, understand speech but

have little ability to communicate, can be taught some basic
self-care routines.

LIN2601/1/2017-2023 123
 Profound Significant developmental delays in all areas, extremely
limited communication ability, require constant care in a
protected environment.

It is difficult to separate the brain processes that control language from those that
control general cognitive functioning because, in the general population, ability in
both go hand in hand. However, by studying children with disorders researchers
can tease apart these processes. As we saw earlier in the introduction to this unit,
the relationship between cognition and language is still imperfectly understood,
and different opinions are expressed on this matter by researchers and practitioners
alike. Some (like Piaget) believe that some aspects of cognitive development are
necessary precursors to linguistic development and provide a scaffold for language
development, while others (like Chomsky) believe that cognition and language
are two separate systems that develop in parallel ways. The study of language
development in children with cognitive impairments provides a unique window
onto this relationship and allows us to examine it in ways that are not possible for
normal language development.
Sections 3 and 4 below focus on children with Down syndrome and William
syndrome, both of which are associated with a degree of intellectual disability.

3 Down syndrome
Down syndrome is a genetic condition in which a person is born with an abnormal
number of chromosomes. It is named after an English doctor, Dr Langdon Down,
who first described the symptoms of this condition in 1866. The condition was first
referred to as ‘Down’s syndrome’ but the preferred term for this condition since
the 1990s is ‘Down syndrome’ or idawuni sinidromu in Zulu, sindromo ya Down
in Setswana and Sepedi, and vhulwadze ha u sa fhelela vhuluvhini in Tshivenda.

3.1 What causes Down syndrome?

Down syndrome is often defined as a ‘chromosomal abnormality’. Chromosomes
are small structures that occur in pairs in all cells and carry the genes. Most people
have two pairs of 23 chromosomes per cell (i.e. 46 chromosomes per cell), which
originate from 23 chromosomes in the mother’s egg, and 23 in the father’s sperm.
People with Down syndrome have an extra chromosome per cell, i.e. they have 47
chromosomes per cell. The extra chromosome comes specifically from chromosome
#21.
Babies with Down syndrome can be born to parents of every age, social class and
ethnic group and it is important to realise that parents are not responsible for having
‘caused’ their child to be born with Down syndrome. However, the risk of having
a child with Down syndrome increases the older the mother. Despite advances in
recent years in genetic research, the factors that trigger the chromosomal abnormality
in the foetus are not yet fully understood. The incidence of Down syndrome is
estimated to be one in every 1 000 live births in developed countries and one in
every 650 live births in developing countries. In South Africa it is roughly one in
every 500 (www.downsyndrome.org.za, Christianson 1996).

124
 Down syndrome is an irreversible condition, i.e. it cannot be cured. However,
children can be stimulated and developed to their full potential, and many children
with Down syndrome can reach quite high levels of functioning. As you will
remember from study unit 5, a stimulating environment can make a big difference
to a child’s cognitive and language development, irrespective of whether the child
has a disability or not.

3.2 Symptoms and characteristic features

There are two areas of the brain that are affected by the extra chromosome in #21,
namely, Broca’s area and the cerebellum. Refresh your memory by referring back to
study unit 1 to locate these areas in the brain and then do Task 6.1.

Task 6.1

Use the knowledge that you acquired in study unit 1 to make predictions about how
language may possibly be affected in children with Down syndrome:
1. If the cerebellum is affected in Down syndrome, what effects can we predict that
this will have on a child’s language?

2. If Broca’s area of the brain is affected in Down syndrome, what effects can we
predict that this will have on a child’s language?

3. Do you think there is possibly a link between these affected areas?

Feedback
1. The cerebellum controls motor coordination. If this area is affected, then the motor
coordination of the articulatory organs will probably be affected.
2. Broca’s area controls speech production. If Broca’s area is affected, then children
with Down syndrome will probably have problems with speech production.
3. Yes, there is a link. The cerebellum and Broca’s area are both areas of the brain
that control motor coordination. If these areas are impaired, then speech produc-
tion is likely to be affected.

LIN2601/1/2017-2023 125
 Here are some of the typical features that occur in children with Down syndrome
(visuals from Down Syndrome South Africa 1998). Not all these features will be
present in every child:
 A rounded face, with slanted eyes (the shape of the eyes led to the earlier use of
the term ‘mongolism’, a term which has since fallen into disuse).
 Slightly open mouth with protruding tongue. The tongue is large in relation to
the size of the oral cavity.
 Low muscle tone and looseness of joints.
 Eye defects in about 60% of cases.
 Hearing defects and problems with short-term memory.
As you can predict, three of the characteristics of Down syndrome will have a direct
effect on language acquisition and development:
 The large tongue restricts movement in the mouth and inhibits clear articulation
of sounds.
 Low muscle tone affects speech production and causes difficulty in coordinating
articulatory movements to combine sounds in the production of syllables, words,
phrases and sentences.
 Hearing defects can adversely affect language development because the child does
not hear linguistic input clearly. Problems with auditory short-term memory have
important implications for language acquisition because most of early language
is acquired through listening to speech. In order to process sentences for mean-
ing, syntactic forms and functions, children need to be able to hold a number of
sentences in short-term memory.
Having identified some of the features of Down syndrome and the way in which they
can affect language, let us now build up a broader picture of language development
in children with Down syndrome.

3.3 Language development

The path of language development in children with Down syndrome is similar to

A ceiling effect that of normal children except that it occurs more slowly and there appears to be a
is a limitation ceiling effect that is probably imposed by the genetic condition, but could also be
that occurs that caused by other factors. Language development is delayed but not deviant (i.e. not
hampers further
development. qualitatively different). It must be stressed however that, like ordinary children,
children with Down syndrome also display variation in their rate of language
development. Some acquire a fairly high degree of linguistic proficiency while others
may reach a ceiling quite early.
Babies with Down syndrome go through the usual babbling stages that normal
babies do, and they also engage in social interaction with their parents or caregivers
during the first year, such as listening, establishing eye contact and joint attention,
and taking turns in early ‘conversations’ via babbling. However, their development
in these early communicative acts is slower and less well developed than in normal
children (Martin, Klusek, Estigarribia & Roberts 2009).
Vocabulary is also learnt more slowly by children with Down syndrome. For example,
only 12% of 1-year-old Swedish children with Down syndrome could produce at
least one word (Berglund, Eriksson & Johansson 2001). As you learnt in study unit

126
 3, in ordinary children a vocabulary spurt seems to occur around 1;6 years when
children start acquiring the names of many new things or objects. A similar spurt
occurs in children with Down syndrome, only it occurs later, at about the mental
age of 2;6 years.
The two-word stage is also delayed in children with Down syndrome, and they
experience greater difficulty in picking up grammatical rules. Their speech, even
when they’re older, often tends to be telegraphic (e.g. No want to swim instead of I
don’t want to go swimming). Some studies have shown that older children with Down
syndrome show continuing syntactic development up to 20 years of age, but they
tend to omit function words (e.g. articles, prepositions and auxiliaries in English)
(Buckley 1993:6).
One must also not overlook the relationship between language development and
linguistic input that the children receive from the environment. Studies show that
adults who interact with teenagers with Down syndrome tend to ask more closed
questions than open-ended questions (e.g. Did you have a nice day at school? as opposed
to What did you do at school today?), they use many repetitions and clarifications,
and tend to complete children’s sentences for them. Buckley (1993:64) suggests that
although such conversational strategies may be helpful in the short term, in the long
term they may curb the development of longer sentences in the child.
Owing to the impairments in the cerebellum and Broca’s areas, children with Down
syndrome have problems with speech production and this affects their intelligibility.
Because their tongues are large, they often have difficulty with the production of
consonants and consonant clusters. Research indicates that pronunciation is often
better when short utterances are used, and this increases intelligibility. Buckley
(1993:6) suggests that since the main aim of communication is to be understood, it
could be that children with Down syndrome use shorter utterances because shorter
MLU improves the intelligibility of their utterances, even though they may be capable
of generating more complex sentences. The poor intelligibility of many children with
Down syndrome continues into adulthood and remains a lifelong challenge (Martin
et al 2009; Oliver 2010). In her study of 87 teenagers with Down syndrome, Buckley
(1993) found that while more than 70% of the children were regularly producing
utterances of five words or more, about 20% of them used language with a MLU
of three words or less. Furthermore, although these utterances were intelligible to
those who knew the children, intelligibility decreased when the teenagers attempted
to communicate with strangers.

If you want to hear an example, listen to ‘The language of Down

syndrome’, where Zachary interviews his 14-year-old brother Seth who
has Down syndrome
https://www.youtube.com/watch?v=HF1Ur_9VYVs

3.4 Therapy and intervention

Having a better understanding of where the problems lie and how physical problems
affect social interaction and language development, researchers, practitioners and
caregivers can use communication and language techniques that aid in improving
problem areas. Some of the specific language problems that children with Down
syndrome have are remediable to some degree. Intelligibility can be improved to
some extent with speech therapy and motor-speech programming techniques.
LIN2601/1/2017-2023 127
 Our assumptions about what people with cognitive impairments can or cannot do
affect our beliefs about what treatment will be effective and appropriate for them.
In the past many children with Down syndrome were not taught to read because it
was thought that they were not up to it. However, research into language disorders
has shown that children can learn language through different channels or mediums.
For normal children, language acquisition typically occurs via the auditory channel
(i.e. through hearing spoken language). For children with disabilities relating to
the auditory channel, the visual channel provides a stronger medium for language
acquisition. This has led researchers and practitioners to use intervention methods
that rely on visual processing to help support language development in children
with Down syndrome. These methods include signing and reading.
Children with Down syndrome can be successfully taught to read (even from as early
as 3 years). Research during the past three decades shows that reading actually helps
them develop their language skills and their speech production to a higher degree
than is usually achieved via the auditory medium (Martin et al 2009). Most children
with Down syndrome find reading useful, enjoyable and stimulating. Reading seems
to improve their auditory memory span, language development, general memory
development and spoken language. Furthermore, children with Down syndrome
who read have been shown to improve their scores on standardised IQ tests.

4 Williams syndrome
Williams syndrome is a rare and, until recently, a relatively unknown genetic disorder
Remediation that was first recognised as a distinct syndrome in 1961 by a New Zealand heart
is a therapy specialist, Dr Williams. The reason why we have included it for discussion here is
or teaching that developments in genetic research since the 1980s have led to some important
intervention that
can help. discoveries about this condition that throw more light on the complex relationship
between the mind and language. There are also some interesting differences in the
language of children with Down syndrome and children with Williams syndrome.
By comparing differences in the language development of such children we can try
to identify the cognitive and linguistic features that characterise a specific disorder
and so gain greater insight into the way in which the brain is organised. Greater
insight into a disorder naturally enables us to make better informed decisions about
treatment and remediation programmes for the individuals concerned.

4.1 What causes Williams syndrome?

Findings from genetic research have revealed that children with Williams syndrome
have a chromosomal abnormality – a small part of chromosome #7 is missing. This
missing part of the chromosome affects several genes that relate to brain function
and development, including visuo-spatial abilities. Williams syndrome is quite rare
and is estimated to occur in about one in 7 500 live births (Mervis & Velleman
2011). Like Down syndrome, it is a congenital condition and neither of the parents
can be ‘blamed’ for its occurrence.

4.2 Symptoms and characteristic features

Children with Williams syndrome have characteristic facial and physical features
and they often look more like other children with Williams syndrome than like
their relatives. Here are some of the features associated with the syndrome:

128
  Small stature, upturned nose, long upper lip and wide mouth, puffiness around
the eyes, and a small chin, resulting in a pixie-like appearance.
 Mild to moderate intellectual disability with developmental delays in walking
and talking, and overall slow growth and development.
 Sensitive hearing or hearing loss.
 A friendly, sociable and empathetic nature; the children enjoy social contact with
adults but are often anxious or insecure.
 Symptoms of attention deficit hyperactivity disorder (ADHD) in about 65% of
children (Mervis & Velleman 2011).

4.3 Language abilities

Although there is a delay in the onset of language development amongst children
with Williams syndrome, many of them go on to develop quite impressive language
skills. They are socially interactive children who readily participate in conversations
and as they become older, they can develop quite vivid descriptive skills. Because
they are so ‘chatty’ and have good language skills, their cognitive handicap is often
masked or overlooked. However, because of visuo-spatial impairments, children
with Williams syndrome may have problems learning to read and write.
Lexical development almost always starts later than in normal children. For example,
Capirci, Sabbadini and Volterra (1996) report that Elissa, an Italian girl with Williams
syndrome, only produced ten words at age 2;3 (in normal Italian children, 10 words
are produced at around 1;6 years, while an average of 334 words are produced at
age 2;3). At age 4;10 Elissa still had problems naming colours, and would use a
strategy of associating colours with the names of things that bore these colours.
For example, if something was brown she said it was like chocolate, and the colour
green she called grass.
Unlike normal children and children with Down syndrome, first words appear several
months before children with Williams syndrome start pointing at objects to draw
caretakers’ attention (Mervis & Velleman 2011). The absence or late development
of joint attention is consistently raised as a concern by parents of toddlers with
Williams syndrome. Despite this, children with Williams syndrome tend to have
good pronunciation and expressive and fluent spoken language skills. Some children
use varying intonation, pitch, volume and dramatic flair while they talk, and are
considered by many to have the ‘gift of the gab’. However, half of the 33 children with
Williams syndrome studied by Huffman, Velleman and Mervis (2011) experienced
low muscle tone resulting in an unusually harsh vocal quality.
A remarkable characteristic of many older children with Williams syndrome is their
relatively rich and wide vocabulary knowledge, especially of low frequency words.
Naming objects, actions, and descriptors has consistently been identified as their
area of greatest strength (Mervis & Velleman 2011). They have larger vocabularies
than would be expected for their mental age and can usually use the words correctly
in context. For example, Van, 11 years old and with an IQ of 50, could use the
words peninsula, slumbering, husk, solemn, dissecting and hoisting appropriately in
his conversations. They can also generally give accurate explanations of a word’s
meaning. For example, when asked to explain what the word poor meant, Crystal,
15 years old and with an IQ of 49, replied: ‘Means they are hungry, they don’t have
any money, they are dying from exposure’. In contrast, a particular weakness in the

LIN2601/1/2017-2023 129
 language of children with Williams syndrome appears to be their understanding of
conjunctions like unless and although and of words relating to time, quantity and
dimensions (Mervis & Velleman 2011).

In terms of syntax, children with Williams syndrome were observed to follow

the same sequence of development as normal children, but they also made more
morphosyntactic errors than normal children, and they also made errors that
normal children seldom make. The grammatical abilities of children with Willams
syndrome are at about the level expected for their overall (usually mildly impaired)
intellectual abilities. Many individuals have considerable difficulty with complex
morphological or grammatical constructions, for example grammatical gender (i.e.
the masculine/feminine/neuter status of nouns and adjectives in some languages),
agreement between nouns and adjectives, tense and verb conjugations, the use of
articles, prepositions and auxiliaries (Karmiloff-Smith & Grant 1992 in Capirci et
al 1996:1035). Typical errors include using intonation rather than grammar to mark
a question, overgeneralising the past tense (e.g. comed, eated), or double-marking
(e.g. Is the bug is hungry?) (Mervis & Velleman 2011).

Although they occasionally make morphological or pronoun errors, the expressive

language use of children with Williams syndrome shows grammatical complexity
and is well formed. Here is a sample of speech produced by Ben, a teenager with an
IQ of 54, describing what happened when he hurt his foot.
After it stopped hurting I was told I could go to school again and do whatever
I feel like doing. If it got really infected they would have taken my toe off.
They had to give me ether so I wouldn’t feel the pain (Bellugi, Marks, Bihrle
& Sabo 1988).

This next example is from Lenhoff, Wang, Greenberg and Bellugi (1997):
What an elephant is, it is one of the animals. And what an elephant does, it
lives in the jungle. It can also live in the zoo. And what it has, it has long,
grey ears, fan ears, ears that can blow in the wind. It has a long trunk that
can pick up grass or pick up hay. If they’re in a bad mood, it can be terrible. If
the elephant gets mad, it could stomp, it could charge. Sometimes elephants
can charge. They have big long tusks. They can damage a car. It could be
dangerous. When they’re in a pinch, when they’re in a bad mood, it can be
terrible. You don’t want an elephant as a pet. You want a cat or a dog or a bird.

The ability to use language appropriately and meaningfully in different contexts is

part of a person’s pragmatic ability. As referred to earlier, children with Williams
syndrome are often quite chatty and have good conversational skills. However, they
may rely too heavily on stereotyped phrases, or have difficulty producing appropriate
responses to questions. According to Mervis and Velleman (2011), the most common
difficulties were providing too little information, providing irrelevant information,
misinterpreting what the researcher had said, or not producing a response that
continued the conversation.
Some researchers have also tested the metalinguistic ability of children with Williams
syndrome. Metalinguistic ability (i.e. the ability to reflect on language use and to
detect and correct incorrect sentences) requires greater cognitive-linguistic resources
because it requires an awareness of grammatical and semantic rules. The teenagers

130
 that Bellugi et al (1988) tested on metalinguistic awareness showed an ability to
handle metalinguistic tasks. For example:
Examiner: The log swam lazily away.
Crystal: Logs cannot do that. Logs float lazily in water (Bellugi et al 1988:189)
Bellugi et al (1988:189) conclude that the results on the metalinguistic tasks indicate
that children with Williams syndrome have linguistic abilities that ‘involve awareness
and manipulation of grammatical structure’.

4.4 Conclusion
The research into the cognitive and linguistic abilities of children with Williams
syndrome shows a complex yet interesting profile of development in both language and
cognition – one of selective impairment in development. These individuals show peaks
of ability in some aspects of language and cognition, and disabilities in other aspects.

Task 6.2

Below are examples of speech from two teenagers, one of whom has Down syndrome
and the other Williams syndrome. Both teenagers were asked to invent a story for the
‘frog’ picture sequence. Using your knowledge of how language is affected differently in
people with these two syndromes, compare the linguistic features of the two teenagers
and then identify which teenager has which syndrome.
A. Teenager, age 18, IQ 55: syndrome
The frog is in the jar. The jar is on the floor. That’s it. The stool is broke. The clothes
is laying there.
B. Teenager, age 17, IQ 50: syndrome
Once upon a time, when it was dark at night, the boy had a frog. The boy was look-
ing at the frog, sitting on the chair, on the table, and the dog was looking through,
looking to the frog in a jar. That night he sleeped, and slept for a long time. The
dog did too. But the frog was not gonna go to sleep. The frog went out from the jar.
And when the frog went out, the boy and the dog were still sleeping. Next morning
it was beautiful in the morning. It was bright and the sun was nice and warm. Then
suddenly when he opened his eyes, he looked at the jar and then suddenly the frog
was not there. The jar was empty. There was no frog to be found.
(Taken from the data of Ursula Bellugi, in Lenhoff et al, Scientific American,
December 1997:45)

Feedback
Teenager A has a very limited linguistic and storytelling skills and uses only short, simple
sentences. The very low MLU at age 18 (and of course the low IQ) suggests intellectual
disability. There are grammatical errors in some of the sentences, for example ‘The
stool is broke. The clothes is laying there’. Teenager A has Down syndrome.
Teenager B has a good command of language and knows many of the conventions
for telling a story, for example starting with ‘Once upon a time’. The language is flu-

LIN2601/1/2017-2023 131
 ent, with only one error that is later corrected, ‘That night he sleeped, and slept for a
long time’. The language is not significantly different from that of a normal 17-year-old,
although Teenager B has an IQ of only 50. This evidence suggests that Teenager B
has Williams syndrome.

5 Autism Spectrum Disorder

We come now to the third syndrome to be examined in this unit, namely autism.
Autism is more correctly referred to as Autism Spectrum Disorder because autistic
individuals vary greatly with regard to the severity of the disorder, occupying varying
positions along a spectrum of symptoms and abilities. Unlike Down and Williams
syndromes, which have a genetic basis and represent fairly homogeneous groups of
individuals, autism spectrum disorder is a more complex syndrome and one that is
not yet fully understood. Autism is a lifelong developmental disorder that generally
(but not always) emerges within the first three years of life. The syndrome was first
diagnosed and described by the child psychiatrist, Leo Kanner, in 1943. He identified
the characteristics of a group of children who showed profound withdrawal from
social contact, were more attached to objects than humans, engaged in repetitive
behaviour, and were either mute (i.e. they failed to acquire language) or else if they
did acquire language, they used it in inappropriate ways. Some individuals acquire
high levels of independent functioning, while others never acquire language (they
are called nonverbal autistics) and require constant care and supervision.

If you have internet access, you may want to watch this Youtube clip
explaining briefly about autism spectrum disorder.
http://www.monkeysee.com/play/
5041-what-is-autism-and-how-prevalent-is-it

Autism occurs in all race, ethnic and class groups, and its incidence is higher amongst
boys than girls – about 80% of autistic people are males (Bateman 2013). In the 1980s,
estimations of the occurrence of autism suggested that 1 in every 1000 people had
the disorder. Over the last thirty years, however, the incidence has been increasing
rapidly, partly but not only because the diagnostic criteria for autism have changed.
In 2014, the Centers for Disease Control and Prevention’s Autism and Developmental
Disabilities Monitoring Network reported that approximately one in 68 children in
the United States has an Autism Spectrum Disorder (1 in 42 for boys, and 1 in 189
for girls) (see http://autismsciencefoundation.org/what-is-autism/how-common-is-
autism/). No reliable figures are available for South Africa, but in the Western Cape
alone, ten children a week are diagnosed with Autism Spectrum Disorder between
the Red Cross Children’s Hospital, Lentegeur and Tygerberg hospitals. With just
nine specialised schools in the entire country, an estimated 135 000 autistic children
are not getting the education and support they need (Bateman 2013).

If you have time, watch this report on a new school for autistic learners
in Soweto, Johannesburg at
https://www.youtube.com/watch?v=uNQ1stae_5k

Before looking at language development in autism, let us look more closely at the
causes of autism and the behavioural symptoms of autistic individuals. This will
provide a broader context within which to examine the linguistic features of autism.

132
 5.1 Causes of autism
The question as to what causes autism is a very controversial one that can stir up much
heated debate. Unlike disorders such as Down syndrome and Williams syndrome,
which are caused by specific chromosomal abnormalities, the causes of autism are
far more complex and not yet properly understood.
It was initially believed that autism had a psychological cause, but it is now generally
accepted that the cause is biological and not psychological. Some possible causes
are ascribed to environmental toxins such as food allergies, heavy metals, viruses,
adverse reactions to vaccinations and brain injury. Whatever the causes that trigger
this disorder, the result is an interference in normal brain development. The main
problem seems to stem from a dysfunctional brain stem, corpus callosum and
central nervous system. In fact, the central nervous system in autistic individuals
seems to scramble sensory input (e.g. input relating to sight, sounds, touch, taste).
Consequently, these individuals have severe problems interpreting and understanding
stimuli – in effect, they have difficulty making sense of the world into which they
are born and the people who inhabit it.

5.2 Characteristics of autism

In order to understand the characteristics of autistic language, one must first
understand the characteristics of autistic behaviour in general. The problems with
describing autistic spectrum disorder are that not all autistic people show the same
symptoms and there are varying degrees of severity in the symptoms. For example,
some autistic people are intellectually disabled while some are not; some autistic
people may eventually acquire high levels of linguistic competence while others
remain mute all their lives; some autistic people may reach high levels of functioning
and independence in adult life, while others require constant care. Despite different
Triad means a manifestations of autism and despite differences in the severity of the disorder, there
group of three. are specific social, cognitive and linguistic symptoms that characterise autism. This
is referred to as the triad of impairments.
Among the ‘classic’ symptoms of autism (sometimes diagnosed as Kanner’s autism)
are an intense dislike of being touched, avoidance of eye contact, a failure to express
overt affection or to play with other children. As young children, they do not engage
in symbolic play such as pretending to be driving a toy car around. Instead, they
engage with toys as objects in repetitive activities. For example, they will hold a
toy car up and spin its wheels for hours on end. They also engage in other forms
of repetitive behaviour, such as flicking their fingers or flicking a light switch on
and off, flapping their arms, banging their heads or rocking themselves for hours
on end. They do not share these activities with anyone and seem to be lost in their
own world, oblivious to the presence of others.
Autistic children prefer to do the same things at the same time and in the same order
every day and they get very upset if their routines are disturbed or interrupted, and
may scream and throw tantrums. As they grow older their repetitive behaviours may
be replaced by more complex repetitive behaviours, for example they may insist on
going to the shops by the exact same route every time.
Severely autistic children do not seem to ‘tune into’ human speech and in fact their
lack of response to speech often leads their parents or caregivers to suspect that they
are deaf. Usually they are not deaf and in fact they often have an oversensitive sense

LIN2601/1/2017-2023 133
 of hearing. For example, in her book Donna Williams describes how stroking the
bristles of a brush and listening to their swishing sound – a sound that children
would not normally hear or listen to – soothed her fear and anxiety. Many children
in this group are mute and, as we shall see later in section 5.3, those who are
verbal show delayed language acquisition. For severely autistic children, speech is
not a means of social interaction (e.g. for talking to people, finding out about the
world and how it works). These children also have problems copying other people’s
gestures (e.g. smiles, frowns, hand movements, etc) and behaviour. Many of the
problems that autistic people experience stem from the fact that they basically lack
an understanding of the rules of social behaviour – they do not understand why
people behave the way they do.
This is how one autistic person describes the problems she experiences:
Verbatim means
‘word for word’. As it was a long time before I realized that people might actually be speak-
ing to me, so it was a long time before I realized that I too was a person … I
was not able to pick people out as being different from objects. When I did
realize that people were supposed to be more important than objects, things
began to take on a new and more difficult light. … I cannot remember ever
thinking about where my mother, father, sister, brother were, they did not
seem to concern me. I think this was because for a long time I did not realize
that they were people and that people are supposed to be more important
than objects (Joliffe et al 1992 in Peeters 1997:87).

They have unusual reactions to sensory stimuli – sound, light, vibrations, touch.
They may either ignore a particular stimulus (e.g. human speech), be attracted to a
stimulus (e.g. be fascinated with spinning objects or with colours) or dislike a stimulus
intensely (e.g. having their hair brushed). They have no fear of real dangers (e.g. they
may have no fear of heights or they will walk into moving traffic) but paradoxically
they may be frightened by everyday things that seem perfectly harmless to us (e.g.
a certain colour, a row of trees in a street).

One of the milder forms of autism spectrum disorder is known as Asperger’s

syndrome (named after the Swedish doctor who identified the symptoms in
the early 1940s). Individuals with this autistic syndrome typically develop good
language competencies but, like all autistic people, they engage in repetitive behaviour,
show little flexibility in thought patterns and they have problems with socialisation
and communication. Their speech patterns may have strange inflections, and they
tend to have poor volume control – for instance, they may speak very loudly in
public. Poor motor coordination is frequent, and they tend to be clumsy. Behaviour
problems are common in this group, for example aggression or taking things that
don’t belong to them. Such behavioural problems arise from a fundamental lack of
understanding of the rules of social behaviour.

Despite their impairments in specific areas of cognitive activity, some autistic

individuals show ‘islands of skill’, for example some have an amazing ability to
draw accurate and minutely detailed pictures, some have amazing calendar memory
(they can work out past and future dates and times) and are good at arithmetical
calculations, some can successfully imitate music tones or rote memorise long chunks
of information (e.g. they can repeat, verbatim, entries in a telephone directory or
weather reports they heard months previously).

134
 5.3 Language development

The following are some of the characteristic features of language among autistic
The term people:
echolalia is
derived from the  mutism (i.e. not developing expressive language ability)
word echo.  echolalia (i.e. repeating verbatim sentences spoken by other people)
 strange speech patterns, such as monotonous intonation, pronoun reversal (e.g.
exchanging I for you), and a tendency to understand only the literal use of language
 severe impairment in acquiring the pragmatic aspects of language (i.e. knowing
what language is for and how to use it appropriately in different contexts)

You may want to watch the following video summarising the characteristic
language problems experienced by autistic children:
http://www.monkeysee.com/play/5044-what-are-some-examples-of-
language-difficulties-with-autism

Many autistic children do not develop language and remain mute, while some
high functioning autistic people develop language in all its syntactic complexity
and learn to read and write. When autistic children do develop language, the onset
of language development is delayed, and once it emerges, development tends to be
slow and irregular. Although verbal autistic children do not have major problems
acquiring the morphology and syntax of language, they have profound problems
using the language appropriately in different social contexts. Let us now look at
these linguistic characteristics more closely.

5.3.1 Language and communication in early development

As you will recall from study unit 3, normal infants engage in reciprocal social
behaviour from the first few months of life. For example, they show greater interest
in people than objects, they are genetically ‘tuned into’ the human voice above other
sounds, they establish mutual gaze and reciprocal smiling, and they unconsciously
coordinate their movements and vocalisations with a partner (e.g. a baby lying on
her back on a bed may smile, kick her legs, wave her arms in glee and make cooing
sounds when she sees her mother). Even intellectually disabled children engage in
and respond to such social behaviour.
Because of the social unresponsiveness of autistic children, they have difficulties
establishing and maintaining joint attention, initiating a topic, handling different
topics in a conversation, and dealing with different interlocutors. They have problems
understanding gestures, gaze, vocalisation and facial expressions. All these different
aspects of lack of socialisation have negative effects on the language development
of autistic children. In their longitudinal study of the language development of six
autistic children, Rollins and Snow (1998) found that the autistic children who
seldom established or maintained joint focus of attention made very little progress
in their language development – their MLU was shorter and there was less syntactic
variety in their utterances compared to autistic children who managed to maintain
more joint attention.

LIN2601/1/2017-2023 135
 Task 6.3

Look at the table below which describes the development of language and communica-
tion in autistic children in the first four years of life (taken from Peeters 1997:50). Now
in the column before it, you should compare the development of normal children at the
different age milestones listed below.
Normal children Age in Autistic children
months
6 Crying is difficult to interpret
8 Limited or unusual babbling (e.g.
squeals or screeches)
No imitation of sounds, gestures,
expressions
12 First words may appear, but often not
used meaningfully
Frequent, loud crying; remains difficult to
interpret
24 Usually fewer than 15 words
Words appear, then drop out
Gestures do not develop; few points to
objects
36 Word combinations rare
May echo phrases, but no creative lan-
guage use
Odd rhythm, tone or stress
Poor articulation in about half of speak
ing children
Half or more are without meaningful
speech
Takes parent by hand and leads to
object
Goes to customary location and waits to
be given object
48 A few combine two or three words
creatively
Echolalia persists; may be used
communicatively
Makes requests, Mimics
TV commercials

Feedback
As you can see from this table, language emerges much later in autistic children than
in normal children, and takes longer to unfold. Babbling appears at 8 months (6 months
in normal children) and is limited in the range of sounds. Within the first year normal
children start understanding some words and engaging in games such as peek-a-boo
and naming rituals. At around 12 months normal children begin to realise that words
refer to things, after which there is a rapid increase in receptive and expressive vo-
cabulary. In autistic children the first words also appear at about 12 months but are not
used meaningfully. By 24 months, when normal children are at the two-word stage,
autistic children still use fewer than 15 words. The vocabulary spurt doesn’t occur and
the multiple-word stage is only reached at age 4 (soon after age 2 in normal children)

136
 and is limited to two or three-word sentences. Your comparative table should look
something like the following:

Normal children Age in Autistic children

months
Reduplicative babbling 6 Crying is difficult to interpret
Imitation of sounds, gestures, 8 Limited or unusual babbling (e.g. expre-
sions squeals or screeches)
No imitation of sounds, gestures,
expressions
First words appear, and are used 12 First words may appear, but often not
used meaningfully
Frequent, loud crying; remains difficult to
interpret
Two-word stage begins 24 Usually fewer than 15 words
Creative use of telegraphic Words appear, then drop out
language Gestures do not develop; few points to
objects
Multiple word utterances 36 Word combinations rare
Increasing vocabulary and May echo phrases, but no creative lan-
guage grammatical competence use
Odd rhythm, tone or stress
Poor articulation in about half of speak-
ing children
Half or more are without meaningful
speech
Takes parent by hand and leads to
object
Goes to customary location and waits to
be given object
Fluent, with good grammatical 48 A few combine two or three words
and communicative competence creatively
Echolalia persists; may be used
communicatively
Makes requests,
Mimics TV commercials

5.3.2 Mutism
One of the major problems in studying language in autistic children is that
Influx is the approximately 20% of autistic children remain mute all their lives (they are called
arrival or inflow nonverbal autistics) (Luyster & Lord 2009:448). The reason for their mutism is not
of a large volume always easy to establish. As we saw earlier in section 5.2, autistic people have problems
of something.
controlling and understanding the stimuli they receive. Obviously, if information
is not registered in the first place, it cannot be interpreted. The apparent deaf
behaviour of many autistic children is related to their attempts to shut down the
influx of auditory stimuli that they receive because they have problems filtering
auditory input – such as shutting out background noise or controlling auditory
volume.
Research in the past twenty years has shown that mute autistics have smaller brain
stems than normal and show abnormalities in brain stem functioning – an area of
the brain that controls auditory input (Grandin 1995:72) (see fig.1.1 in study unit 1).

LIN2601/1/2017-2023 137
 Because their visuo-spatial processing is less impaired than their auditory processing,
autistic children are often better at acquiring language through the visual medium
(e.g. signs, reading) than through the auditory (i.e. spoken) medium. In this they
are similar to children with Down syndrome. Speech therapists have found that
some nonverbal autistic children start acquiring language only after they first learn
to sing – apparently the brain circuits used for singing are more normal in autistic
children than the brain circuits for spoken language.
Sometimes they may ‘know’ language but they remain mute because they do not
know how to use language as a social tool to convey meaning. They use gestures to
indicate their needs, for example an autistic child may indicate his need for food by
taking his mother by the hand to the fridge. Here is one autistic boy’s description
of his silence during childhood.
Because I didn’t use speech to communicate until I was 12, there was consid-
erable doubt as to whether I would ever be able to learn to function indepen-
dently. No one guessed how much I understood because I didn’t say what I
knew. And no one guessed the critical thing I didn’t know, the one missing
connection that so much else depended on: I didn’t communicate by talking,
not because I was incapable of learning to use language, but because I simply
didn’t know that that was what talking was for. Learning how to talk follows
knowing why to talk – and until I learned that words have meaning, there
was no reason to go to the trouble of learning to pronounce them as sounds
… I had no idea that [using words] could be a way to exchange meaning with
other minds (Jim Sinclair 1992 in Peeters 1997:64–65).

You may also want to watch this incredible story about Carly, a
non-verbal autistic who ‘speaks’ through her computer at
https://www.youtube.com/watch?v=xMBzJleeOno

5.3.3 Echolalia
Some autistic children have a tendency to repeat the speech of others, a phenomenon
known as echolalia. There are two types of echolalia – immediate (i.e. the child
repeats a sentence immediately after it has been spoken by another person) and
delayed (i.e. the child repeats a sentence or sometimes even an entire dialogue a
few hours, weeks or even months after it was originally spoken). It is sometimes
difficult to establish whether the repeated language forms are properly internalised
or whether they are just rote memorised pieces of speech. For example, Jeremy, an
autistic boy, often repeated the utterance Hands off the radio, you’ ll break it! This
echolalic utterance may seem meaningless, but it was noted that Jeremy used this
phrase whenever he wanted to listen to music. He first heard this warning when
he went near a radio. He thus associated this phrase with the radio, which in turn
he associated with music, so whenever he wanted to hear music, he reproduced
this sentence (Peeters 1997:54). He was unaware that there were alternative ways
of expressing needs, such as I want to listen to music or Please play me some music.
It must be noted that echolalia is not restricted to autistic children. It also occurs in
normal children during a specific period in their language development, and it also
occurs among intellectually disabled children who are not autistic. It is considered
to be a strategy that some children adopt to help them acquire language. What

138
 distinguishes echolalia in autistic children from other children is the constant
repetitive nature of the echolalia, at times the sheer length of the delayed echolalic
utterance, and the fact that the repeated sentence is often said in the same intonation
pattern and tone of voice as the original speaker.
The faithful reproduction of language as reflected in echolalia indicates that autistic
children do not have problems with low-level perceptual processes. Instead, the
problems seem to lie at a higher level of processing, that of interpreting or making
sense of incoming stimuli. This echolalic behaviour suggests that the linguistic
stimuli the autistic child hears are unprocessed or unanalysed. The child has difficulty
breaking language down into component parts and attaching meaning to the parts
and then to the whole, so repeating whole chunks of language is a coping strategy.
In her autobiography, Donna Williams (1992) reports that if she did not repeat
utterances addressed to her, she only understood 5 to 10% of what was said.
Although some echolalic utterances are self-stimulatory and seem to be an extension
of repetitive behaviour, Peeters argues that echolalia is not the meaningless repetition
of language as was once thought, but rather ‘an attempt to take control of a situation
with the limited means available’ (1997:56).

5.3.4 Phonological, morphological and syntactic development

Autistic children do not seem to have problems producing speech sounds, and their
phonological development on the whole does not seem to be negatively affected
by the syndrome. However, their intonation is often described as ‘monotonous’,
‘mechanical’ or ‘wooden’ (Fay 1988:199).
The general consensus seems to be that morphosyntactic development in autistic
children is similar to that of normal children but occurs at a much slower rate.
However, even though autistic children acquire fairly complex syntactic structures (i.e.
the form of language), their use of such structures (i.e. the function) is inappropriate.
This has led several researchers to conclude that there is a ‘dissociation’ between form
and function in the language of autistic children. De Villiers and De Villiers (1987)
suggest that the semantic, syntactic and pragmatic knowledge is woven together in
normal children, but in autistic children these strands become unravelled.
In study unit 3 you saw that a common feature of early language learning involves the
overextension of words and the overgeneralisation of grammatical rules. For instance,
a child may call any feline looking creature a ‘cat’ (e.g. lions, tigers, cats may be called
cat). Autistic children are unlikely to use this strategy. Instead, they may associate a
label with a specific thing in a specific context, or use several labels to describe one
thing. Here is how a mother described her autistic son’s labelling activity:
What I noticed about Thomas’s early language development was that he made
up his own names for different objects which were really the same, but didn’t
look exactly alike. For various things we call bicycle he had other names. He
had a bicycle, a tractor, a wheels in the mud, a wheels in the grass and also a
feet on the pedals. Everyone called him a ‘creative young man’ but I knew
that it had more to do with helplessness. He didn’t understand what I meant
when I said ‘Get on your bicycle’ if his feet on the pedals was in front of him.
He couldn’t understand that we called his feet on the pedals a bicycle as well.
(Hilde de Clercq in Peeters 1997:51)

LIN2601/1/2017-2023 139
 The ability to generalise is a fundamental aspect of normal human cognition. Being able
to perceive patterns underlying the things we perceive in the external world helps us
organise our world and ascribe meaning to it. Tests of verbal memory show differences in
the way that perception and meaning are linked in autistic children and other children
at the same mental age. In such tests, autistic children usually show superior recall for
random, unstructured lists of words (e.g. chair, apple, truck, into, spade, tickle, horse).
However, they have problems perceiving the underlying categories when structured
lists of words are presented to them, such as apple, banana, orange, peach.
Another interesting aspect of autistic language is the deviant use of personal pronouns.
Autistic children often tend to switch the pronouns you and I, using you to refer to
themselves and I to refer to their conversational partner, for example:
Stan helped you (instead of Stan helped me)
Help you please (instead of Help me please)
I’m wearing glasses (instead of You’re wearing glasses) (data from Tager-Flusberg
1994:185)
This unusual reversal of pronouns reflects difficulty in understanding shifting
reference between speaker and listener. As we shall see in the next section, autistic
children have problems understanding that people have different perspectives, i.e.
that another person may perceive and respond to the world in a way that is different
from our own perspective.

5.3.5 Pragmatic disabilities

Although the formal aspects of language (e.g. morphology and syntax) may be
present in the language of autistic children, they have severe problems with regard
to the pragmatic aspects of language (i.e. the appropriate use of language). They
are not socially competent, they do not pay attention to social cues and they have
problems understanding how and why normal people think and behave as they do.
Autistic children have problems understanding and expressing emotions. They have
great difficulty reading or interpreting gestures and faces and associating emotions
with gestures, for example associating a hug with friendship or interpreting a
finger to the lips as a request to keep quiet. On the basis of findings such as these,
researchers have concluded that autistic people do not have problems with perception
itself; instead, the problem seems to occur at a higher level of cognitive functioning,
namely, interpreting what the stimuli mean, adding meaning to what is perceived.
Thomas can’t read facial expressions at all. He didn’t use to be able to un-
derstand when an adult was cross with him, he often started to laugh when
someone’s face changed expression suddenly. Since then I’ve taught him the
difference between ‘happy’ and ‘frightened’ and ‘angry’. We’ve practised with
dozens of pictures and Thomas can now name them all perfectly. But un-
fortunately for him we don’t all have the same face and it is still difficult for
him. When his sister Elisabeth looked at him crossly the other day, he asked
quite seriously, ‘Elisabeth, what are all those lines on your face for?’ (Hilde
De Clercq in Peeters 1997:122).
Because their ability to interpret stimuli is disrupted, autistic individuals have a
very literal understanding of language, so the figurative, metaphorical aspects of
language are puzzling to them. For example, Peeters (1997) relates how one autistic

140
 child, on being told to Wipe your feet after coming back from a walk, took off his
shoes and socks and literally wiped his feet. Another was surprised on hearing that
his mother had cried her eyes out and looked at her face to see what she looked like
without eyes. They have difficulty understanding jokes and understanding implied
messages in, for example, requests. For instance, if asked Can you tell me where you
live? an autistic child might reply Yes but not give the address.

5.4 Autism and intervention

The question of what kind of intervention is best for autism is a controversial one
Adherents are that can cause much heated debate, and it is not the objective of this unit to debate
the followers of a these issues here. Treatment for autism is usually a very intensive, comprehensive
particular theory undertaking that involves the child’s entire family and a team of professionals.
or practice.
Multiple therapies maybe required, including music therapy, speech and language
therapy, occupational therapy, social skills training and other specialised programmes.
Parents should be aware that each therapy has a different approach to treating autism,
that each therapy has its own group of adherents, and that their child might respond
in different ways to each therapy. Given the different degrees of autism spectrum
disorder, it may well be that different approaches will prove to be effective for specific
kinds of autism stemming from different underlying causes. However, parents are
advised to avoid counsellors who believe that the causes of autism are psychological.
Whichever therapy is chosen, the general consensus is that the social abilities of
autistic individuals tend to improve with age, and also with educational intervention.

5.5 Summary
The different types of autism should be perceived as part of a spectrum, ranging
from mild to severe. Although language development in autistic children is delayed,
they seem to go through the same stages of language acquisition with regard to
morphosyntactic development as normal children, albeit at a slower pace. Some aspects
of language development are unusual, such as pronoun reversal, and the semantic
categorisation and generalisation of words. Even though the symptoms of autism may
abate in some children as they get older, the pragmatic aspects of language continue
to pose problems. Educational intervention can help them to achieve higher levels
of functioning, or at least to reduce some of the behavioural problems.
Studies on the language development of autistic children have led to some interesting
speculation about the relationship between language and the mind. The fact that
morphology and syntax seem to be relatively unaffected (at least in the acquisition
of English) by the disorder has led some researchers (e.g. Tager-Flusberg) to the
conclusion that the language faculty is separate from the mind. However, this
position is not supported by everyone. Rollins and Snow (1998), for example, argue
that the pragmatic disabilities of autism have an impact on the acquisition of the
formal aspects of language. The formal grammatical components of language are
meaningless without the simultaneous acquisition of pragmatic competence. The
failure to integrate incoming information and to interpret it at a higher level of
processing and give it meaning can result in profound and pervasive social isolation.
As we saw above, the social disabilities of autistic people are regarded as one of the
defining characteristics of the syndrome. These social disabilities have profound
effects on autistic individuals’ pragmatic competence.

LIN2601/1/2017-2023 141
 6 Summary
The discussions of language disabilities associated with SLI, Down syndrome,
Williams syndrome and Autism Spectrum Disorder in this study unit have provided
us with valuable insight into the relationship between language and other factors
such as biological, physical, cognitive and neurological factors. However, there are
many questions about the causes and treatments of these disabilities that remain
unanswered. What the existing research does indicate is that although the human
brain is massively interconnected, an impairment in one aspect of cognitive processing
does not affect all aspects of processing, and islands of skill can exist in the face
of severe disability. Children with disabilities are born with many qualities and
abilities and parents and caregivers should develop these to the full. With effective
intervention programmes, many of these children can lead more fulfilling and
independent lives. A slogan that parents are encouraged to remember is: Never lower
your expectations of your child!

Academic vocabulary building

1. There are several phrasal verbs in English (i.e. verbs followed by a preposi-
tion). These phrasal verbs must be learnt as a phrase. Here are some useful
ones to know for academic discourse, so make sure that you know what
they mean.
to account for something
to throw light on/upon
to be prone to
to engage in
2. Make sure you know the difference in meaning between the following simi-
lar sounding or similar looking pairs of words:
affect – effect (when something is affected (V), there is some kind of notice-
able effect (N))
adolescents – adolescence (adolescents (N) are teenagers and adolescence (N) is
the teenage phase)
relative – relevant (something is relative (Adj) when it is compared to
something else, something is relevant (Adj) when it is closely connected or
appropriate).
3. When we learn words, we unconsciously organise them into word fami-
lies, i.e. words that mean the same but have different forms. For example,
the words happy, happiness, happily, unhappy belong to the same word fam-
ily. Researchers in vocabulary development suggest that when we learn a
word, we can work out the meaning of similar words in the word fam-
ily. For example, if we know the meaning of happy, we will probably be
able to work out the meaning of happiness and happily when we encounter
them for the first time. Work out word families for the following words:
remediation controversial

142
 4. Match the words in column A with their appropriate meanings in column B.
A B
1. abate a. escapes from
2. eludes b. the start, beginning
3. congenital c. not to notice something
4. remediation d. openly done, can be seen
5. onset e. to get less, to diminish
6. overlook f. disputed; not everyone agrees
7. controversial g. means of treating or counteracting a disability or
behavioural problem
8. overt h. born with; existing from birth

Further reading

Bavin, EL (ed.). 2009. The Cambridge handbook of child language. Cambridge: Cam-
bridge University Press. (Chapter 25 is on language development and autism
and chapter 26 is on language development in children with Down and Wil-
liams syndrome.)
Crystal, D. 2010. The Cambridge encyclopedia of language. 3rd edition. Cambridge:
Cambridge University Press. (Chapter 46 is on language disability.)
http://www.williams-syndrome.co.za Williams Syndrome Association of South Africa
www.dyslexicsouthafrica.com Dyslexic Association of South Africa

Down syndrome
Movies about Down syndrome include Duo (1996), The Memory Keeper’s Daughter
(2008) and Where Hope Grows (2014).
Hanson, MJ. 1996. Teaching the infant with Down syndrome: a guide for parents and
professionals. 2nd edition. Austin, TX: Pro-Ed.
http://downsyndrome.nacd.org/language_acquisition.php links to a research article by
Robert Doman about language acquisition in children with Down syndrome.
Down Syndrome South Africa provide information and resources at www.down-
syndrome.org.za
Hunt, N. 1966. The world of Nigel Hunt. London: Darwen Findlayson. (An autobi-
ography of a boy with Down syndrome. It has an introduction written by his
father, describing how Nigel’s mother taught him to read and of the ignorance
and prejudice they endured from lay and professional people alike in trying to
secure an appropriate and stimulating education for Nigel.)
Oelwein, P. 1995. Teaching reading to children with Down syndrome. Bethesda, MD:
Woodbine House. (This helpful book provides detailed advice and innovative
ideas on how to teach reading to children of all ages with Down syndrome,
using a language teaching approach.)

Autism
Three well-known movies about autism include Rain Man (1988), Adam (2009) and
Temple Grandin (2010).

LIN2601/1/2017-2023 143
 Enjoyable novels with autistic characters include The curious incident of the dog in
the night-time (Mark Haddon 2003) and House rules (Jodi Picoult 2010)
www.afa.org.za Association for Autism
www.aut2know.co.za Autism South Africa
Gerland, G. 2003. A real person: life on the outside. London: Souvenir Press. (An
autobiography that gives unique insight into what it means to be autistic.)
Grandin, T. 1986. Emergence: labelled autistic. Navato, CA: Arena.
Grandin, T. 1995. Thinking in pictures. New York: Doubleday.
Hart, C. 1987. Without reason. New York: Harper & Row.
Williams, D. 1991. Nobody nowhere. London: Corgi Books.
Williams, D. 1995. Somebody somewhere. New York: Time Books

Bibliography

American Psychiatric Association. 2013. Diagnostic and statistical manual of men-

tal disorders. 5th edition. Washington, D.C: APA.
Bateman, C. 2013. Autism: mitigating a global epidemic. South African Medical
Journal 103(5):276–277.
Bellugi, U, Marks, S, Bihrle, A & Sabo, H. 1988. Dissociation between language
and cognitive functions in Williams syndrome, in Language development in
exceptional circumstances, edited by D Bishop & K Mogford. Edinburgh:
Churchill Livingstone: 177–189.
Berglund, E, Eriksson, M & Johansson, I. 2001. Parental reports of spoken lan-
guage skills in children with Down syndrome. Journal of Speech, Language,
and Hearing Research 44:179–191.
Bernstein Ratner, N. 2005. Evidence-based practice in stuttering: some questions
to consider. Journal of Fluency Disorders 30:163–188.
Buckley, S. 1993. Language development in children with Down syndrome: rea-
sons for optimism. Down Syndrome Research and Practice 1(1):3–9.
Buckley, S. 1993. Developing the speech and language skills of teenagers with
Down’s syndrome. Down Syndrome Research and Practice 1(2):63–71.
Buckley, S & Bird, G. 1992. Mainstream or special school: how do we decide?
Portsmouth Down Syndrome Trust Newsletter 2(2).
Capirci, O, Sabbadini, L & Volterra, V. 1996. Language development in Williams
syndrome: a case study. Cognitive Neuropsychology 13:1017–1039.
Christianson, AL. 1996. Down syndrome in sub-Saharan Africa. Journal of Medi-
cal Genetics 33:89–92.

144
 Christianson, AL, Zwane, ME, Manga, P, Rosen, E, Venter, A, Downs, D &
Kromberg, JGR. 2002, Children with intellectual disability in rural South
Africa: prevalence and associated disability. Journal of Intellectual Disability
Research 46:179–186.
Cohen, DJ, Paul, R & Volkmar, FR. 1987. Issues in the classification of pervasive
developmental disorders and associated conditions, in Handbook of autism
and pervasive developmental disorders, edited by DJ Cohen, AM Donnellan
& R Paul. New York: John Wiley & Sons.
De Villiers, PA & De Villiers, JG. 1987. Commentary on language and autism,
in Handbook of autism and pervasive developmental disorders, edited by DJ
Cohen, AM Donnellan & R Paul. New York: John Wiley & Sons.
Down Syndrome South Africa. 1998. We are very much like you: a basic introduc-
tion to Down syndrome. Pretoria: National Department of Health.
Fay, WH. 1988. Infantile autism, in Language development in exceptional circum-
stances, edited by D Bishop & K Mogford. London: Churchill Livingstone.
Foster-Cohen, SH. 1999. An introduction to child language development. New York:
Addison Wesley Longman.
Frith, U. 1991. Autism: explaining the enigma. Oxford: Basil Blackwell.
Frith, U & Baron-Cohen, S. 1987. Perception in autistic children, in Handbook of
autism and pervasive developmental disorders, edited by DJ Cohen, AM Don-
nellan & R Paul. New York: John Wiley & Sons.
Frith, U. 2013. Autism and dyslexia: a glance over 25 years of research. Perspectives on
Psychological Science. Available at pps.sagepub.com (accessed 25 January 2016).
Huffman, MJ, Velleman, SL & Mervis, CB. 2011. Motor speech characteristics of
children with Williams syndrome. Symposium on Research in Child Language
Disorders. Madison, WI. Published online 2011 Oct 1. doi:  10.1044/lle18.3.98.
Laws, G, Buckley, S, Bird, G, MacDonald, J & Broadly, I. 1995. The influence of
reading instruction on language development and memory in children with
Down Syndrome. Down Syndrome Research and Practice 3(2):59–65.
Lenhoff, HM, Wang, PP, Greenberg, F & Bellugi, U. 1997. Williams syndrome
and the brain. Scientific American December:42–47.
Luyster, RJ & Lord, C. 2009. The language of children with autism, in The Cam-
bridge handbook of child language, edited by EL Bavin. Cambridge: Cam-
bridge University Press: 447–458.
MacKenzie, J. 1997. Language and communication in children with Down syn-
drome. Down Syndrome South Africa 1:4–5.
Martin, GE, Klusek, J, Estigarribia, B & Roberts, JE. 2009. Language charac-
teristics of individuals with Down syndrome. Topics in Language Disorders
29(2):112–132. Available at http://www.ncbi.nlm.nih.gov/pmc/articles/
PMC2860304/ (accessed 26 January 2016).

LIN2601/1/2017-2023 145
 McTear, MF & Conti-Ramsden, G. 1992. Pragmatic disability in children. Lon-
don: Whurr.
Mervis, CB & Velleman, SL. 2011. Children with Williams syndrome: language,
cognitive, and behavioral characteristics and their implications for inter-
vention. Perspectives on Language Learning and Education 18(3):98–107.
doi:10.1044/lle18.3.98.
Newbury, DF, Paracchini, S, Scerri, TS, et al. 2011. Investigation of dyslexia
and SLI risk variants in reading- and language-impaired subjects. Behavior
Genetics 41(1):90–104.
Oliver, C. 2010. Down syndrome and language development. MA dissertation,
Indiana University. Available at http://opensiuc.lib.siu.edu/cgi/viewcontent.
cgi?article=1238&context=gs_rp (accessed 26 January 2016).
Paul, R. 1987. Communication, in Handbook of autism and pervasive developmen-
tal disorders, edited by DJ Cohen, AM Donnellan & R Paul, R. New York:
John Wiley & Sons.
Peeters, T. 1997. Autism. London: Whurr.
Ratner, N. 2005. Atypical language development, in The development of language,
6th edition, edited by J Berko Gleason. Columbus, OH: Merrill.
Rollins, PR & Snow, CA. 1998. Shared attention and grammatical development
in typical children and children with autism. Journal of Child Language
25:653–673.
Snowling, M & Thomson, M (eds). 1991. Dyslexia: Integrating theory and practice.
London: Whurr.
Stein, JF. 1991. Vision and visual dyslexia. Boca Raton: CRC Press.
Tager-Flusberg, H. 1994. Dissociations in form and function in the acquisition of
language by autistic children, in Constraints on language acquisition: stud-
ies of atypical children, edited by H Tager-Flusberg. Hillsdale, NJ: Lawrence
Erlbaum.
Thomson, M. 1991. The teaching of spelling using techniques of simultaneous
oral spelling and visual inspection, in Dyslexia: integrating theory and prac-
tice, edited by M Snowling & M Thomson. London: Whurr.
Tomblin, JB. 2009. Children with specific language impairment, in The Cam-
bridge handbook of child language, edited by EL Bavin. Cambridge: Cam-
bridge University Press.
Wing, L & Attwood, A. 1987. Syndromes of autism and atypical development,
in Handbook of autism and pervasive developmental disorders, edited by DJ
Cohen, AM Donnellan & R Paul. New York: John Wiley & Sons.

146
 Study Unit 7 Study unit 7
The evolution of language
‘It is not the strongest of the species that survive,
2

nor the most intelligent, but the one most responsive to change’
3

Charles Darwin (1809–1882 English naturalist famous for his theory of evolution)
4

OUTCOMES
After you have studied this study unit you should be able to
 contrast the development of the linguistic ability in the human species and in
human infants and show why these two domains of study are of relevance for
each other
 describe the differences between the two main theories of the evolution of lan-
guage with reference to the kinds of problems they try to account for and the
nature of their explanations
 present evidence in support of the hypothesis that language universals are innate
 describe some of the evolutionary developments in the ancestors of the human
species that predisposed humans for the evolution of language
 present evidence in support of the hypothesis that language universals develop
from more general principles of cognition and information processing
 provide definitions for the following important terms:

evolution hominid
phylogenesis primate
ontogenesis continuity controversy
genetic mutation/saltation precursors/preadaptations
gradual evolution chunking

Preview
This study unit takes a different approach to the acquisition of language from
all the previous chapters, focusing on the development of the language abil-
ity in the human species around 100 000 or 200 000 years ago. We look at
two competing theories about how language might have evolved – one that
assumes a sudden genetic change, and one that sees language emerging based
on the gradual evolution of other aspects of human biology, including our eyes,
our upright posture and our vocal tracts.

LIN2601/1/2017-2023 147
 Introduction
Up to this point the study units of this module have dealt with questions such as
How do children acquire language?, How is language organised in the brain?, and
What can go wrong with language ability because of language disorders and brain
damage? All of these questions deal with the language ability of individuals. We
now turn to questions about the language ability in the human species as a whole.
In this study unit we try to answer the question ‘Where does the uniquely human
ability to communicate by means of language come from?’.
This study unit is all about evolution, more specifically, the evolution of the human
language ability. Evolution refers to the natural way that plants and animals have
developed over time to become the modern plants and animals we know today.
Modern humans have a highly developed intellectual capacity that enables us to
organise sophisticated information and perform the complex tasks of everyday life.
It comes as no surprise then that human beings are the only animal species with
elaborate communication systems such as oral language, writing systems and sign
language. Yet, despite the significance of the linguistic ability in human life, we
seldom, if ever, ask questions about its origins. Have you ever thought of where
the languages of the world come from, whether they have a common ancestor, and
whether human language and animal communication systems are related? Think
for a moment about these questions.
Speculations about the origins of language make rather entertaining reading. For
example, according to one theory, language originated as imitations of natural sounds,
such as animal sounds. Another theory suggested that the involuntary grunting we
make when trying to pick up heavy objects is the basis of language development.
In 1866 the Linguistic Society of Paris had had enough of this fanciful theorising,
and refused to accept any more papers on the origin of language.
In recent years, however, the study of language evolution has come into its own again
as an exciting and interdisciplinary undertaking that is starting to provide answers
to the questions above. Linguists, psychologists, anthropologists, primatologists,
archaeologists, biologists and neuroscientists are engaging in fierce debate and
conducting ground-breaking work using the latest technologies. You may also have
your own ideas as to the why and the how of the origins of language. Needless to
say, there is no final or correct answer.

1 Ontogenesis vs phylogenesis
The theme throughout this module has been the remarkable and apparently unique
ability of humans to communicate through language, with a particular focus on the
ability of children to acquire language. This study unit looks at language development
from a different perspective, namely, the development of the language ability in the
human species.

Task 7.1

Look up the meaning of the word ‘genesis’ in a dictionary and write it down.
Definitions:
genesis =

148
 If onto- refers to an individual and phylo- refers to an entire phylum or species, what
can you deduce about the meanings of the following words?
ontogenesis =
phylogenesis =

Feedback
You should have noticed that both terms involve ‘the origin and development (genesis)
of something’. Ontogenesis refers to the origin and development of an individual being
(onto-) or the origin and development of a characteristic feature in an individual. Phy-
logenesis, on the other hand, refers to the origin and development of a whole group of
individuals, i.e. a species. The ontogenesis of language then, refers to the origin and
development of the language ability in an individual human being and the phylogenesis
of language refers to the origin and development of language in the human species.

1.1 The evolution of language vs biological evolution

The biological theory of evolution is an attempt to answer the question about the
Anatomy is the origins of all living things. Most biologists, if not most scientists, believe that all
structure of a living things descended from a single common ancestor. All living things, including
living organism humans, are therefore related. According to the theory of evolution, different species
and its parts.
of living things came about because different living environments required different
biological adaptations. Some biological adaptations take place suddenly and
unpredictably, i.e. there is no obvious reason for their occurrence. Such a sudden
adaptation is called a genetic saltation (i.e. a genetic jump in a new direction, giving
rise to a new species). Most adaptations, however, involve gradual evolution over
extended periods of time. But since we cannot go back in time, how can biologists
trace the evolution of living things if the evolutionary process extends over millions
of years? Evolutionists employ two methods to trace the phylogenesis of a species:
they compare the anatomy of different species and they study the development of
infants of a species. Let us take a closer look at these two methods.
Biologists can work out how closely related different species are by looking at shared
characteristics. Such shared characteristics then give a picture of the evolution of
You recapitulate the various species. Human beings can be classified into the following categories
when you repeat by virtue of their shared characteristics:
something or go
over it again. Humans are vertebrates because they have a backbone.
Humans are mammals because they have hair, milk-producing glands and
three separate bones in the middle ear.
Humans are primates because they have finger- and toenails and an opposable
(i.e. a movable) thumb.
Humans are apes because they have no tail, their shoulder blades are at the
back, not on the sides, and they have a Y-shaped pattern on the surface of the
molars (chewing teeth).
A trait is a
particular The second method derives from a very significant, but rather controversial hypothesis
characteristic of in the theory of evolution, namely: ‘ontogeny recapitulates phylogeny’. This is the
something. view that the biological development of an infant mirrors the evolution of the species

LIN2601/1/2017-2023 149
 as a whole. According to this hypothesis, the evolutionary adaptations which define
a species may be revealed by studying the various stages of the development of an
infant of that species.
There is in fact also a third method that is far more sophisticated than the first two,
namely the study of biological relations by means of the genetic code known as
DNA. The DNA of every species is different, but by comparing the DNA of different
species, biologists can establish the relationships between them. Unfortunately, this
method is beyond the scope of this study unit.

In addition to tracing the evolution of a species, say, the human species, evolutionists
There is can also trace the evolution of a specific trait or characteristic of a species. The trait
controversy that we will look at in this study unit is the evolution of a uniquely human ability,
about something namely, human language. In particular, we will consider two contrasting views of
when many
people disagree the evolution of this ability. One of these views claims that the human linguistic
about it. ability evolved in the form of a sudden saltation. The other view claims that the
human linguistic ability evolved gradually as a result of several, seemingly unrelated,
biological adaptations that took place for other reasons and later made language
possible.

2 Problems facing theories of the evolution of language

Any theory of the evolution of the linguistic ability in humans has to try and answer
a number of questions:
1. Is human language merely an extension of animal communication systems or
is it a different system altogether? This is known as the continuity contro-
Homo habilis versy. That is, can human language, though more elaborate and complex
http://media.web. than animal communication systems, be considered as an extension of
britannica.com/eb-
media/39/79539- animal communication?
004-6D93BE53.
jpg Why do all the languages of the world seem to be structured in similar
ways? This is known as the language universals problem. Although the
languages of the world are different enough to make them incomprehensible
to speakers of other languages, they share a significant range of common
structural properties.
2. Why do all children, despite differences in their intellectual abilities, acquire
reasonably effective linguistic competence in their mother tongue at more or
less the same age? This is the nature–nurture problem that we encountered
in study unit 2. The question is, do children acquire language through interac-
An artefact is an tion with their caretakers (nurture) or do they have an inborn ability (nature)
object made by a to acquire language?
human being.
Let us now take a closer look at these controversies.

2.1 The continuity controversy

Humans, as we have noted above, belong to the primate family. The dictionary
definition of a primate is ‘any member of the order Primate, which includes the
monkeys, apes and humans, typically having dexterous hands and feet, binocular

150
 vision, and a well-developed brain’. In terms of shared characteristics, humans are
closely related to the apes (gorillas, gibbons, orang-utans and chimpanzees), but
diverged from the ape family in another family called Hominids. The Hominid
family consists among others of:
Homo naledi (‘star man’ discovered in 2014 in the Rising Star cave in South
Africa – still undated)
Homo habilis (‘handy man’) (2,5–1,6 million years ago)
Homo erectus (‘upright man’) (1,6 million years ago)
Homo sapiens (‘wise man’) (from about 100 000 years ago to modern man)

The only living member of the Hominid family is Homo sapiens sapiens, i.e. modern
humans. However, we know of the other members of this family from fossils and
artefacts such as stone tools.
Human beings are unique among the apes: only humans can communicate by
means of language and only human infants can acquire a language. Although
chimpanzees and humans share 98% of their genes (Foley 1997:48), only humans
have language in the highly structured and elaborate format characteristic of all
languages. Even in the Hominid family, only humans seem to have the ability to
develop language, although scientists disagree on this (cf. Leaky & Lewin 1992).
These observations raise all sorts of questions about the status of human language
in relation to animal communication systems (the continuity controversy).

Just for fun, you may want to listen to the vocalisations of another
primate, the gelada, and read the accompanying article (2013) at
http://www.bbc.com/news/science-environment-22067192

You may recall from study unit 1 that Broca’s and Wernicke’s areas are very
undeveloped in primates compared to humans. Although chimpanzees and other
primates also communicate by means of oral sounds, their communicative system
is primarily composed only of gestures and facial expressions. Humans, on the
other hand, communicate mainly orally, although their communicative system also
includes gestures and facial expressions. So, in a sense, the communicative systems
of chimpanzees and humans are very similar in that they both employ the visual and
auditory modalities (i.e. sounds and signs). However, the human system differs both
quantitatively and qualitatively from the animal system. The range of meaningful
expressions in human language is open-ended and capable of expressing complex and
abstract ideas, whereas the range of meaningful expressions in the chimpanzee system
is highly restricted. Similarly, human language allows scope for creativity, i.e. the
expression of novel meanings, whereas the chimpanzee system is largely determined
and limited by instinctive communicative behaviour.
To summarise: some researchers believe that human and animal communication
systems are unrelated, while others believe that human language is an elaborated
version of animal communication.

2.2 The language universals problem

Many linguists believe that all the languages of the world share a host of common
properties known as language universals.

LIN2601/1/2017-2023 151
 Task 7.2

Consider the list of properties below and decide whether they are universals or not by
ticking the yes or no box:
Yes No
□ □ The sounds of all languages consist of consonants and vowels.
□ □ The words of all languages are divided into syllables.
□ □ All the languages of the world have the categories noun and verb.
□ □ All the languages of the world can express questions, commands, etc.

Feedback
If you ticked all these properties as universals you are right. But these are rather simplistic
universals. By far the most significant universals and by far the most difficult universals
to detect and describe are universals that control the structural patterns of sentences –
the so-called universal syntactic principles. Basically, there are three universal syntactic
principles: (Those of you who did LIN1501 should be familiar with these principles.)
 The classifiability principle, which assigns each word to a category – noun, verb etc.
 The linearity principle, which governs word order.
 The hierarchic principle, which groups words into larger constituents such as phrases.

The questions that any theory of language evolution has to answer are: Where
do these language universals come from? Was there one common ancestor for all
languages or do all human beings have a mental device that specifies the language
universals and therefore constrains the possible nature of our languages? Are these
universals unique to human language or are animal communication systems also
controlled by them?
If the universals are unique to human language, then obviously human language
could not have evolved from animal communication systems. If the universals are
also found in animal communication systems, then apes such as chimpanzees should
be able to master a human language. (There are, in fact, a number of experimental
projects which have attempted to teach chimpanzees English or American Sign
Language – but all with very limited success.)
It should be noted that some linguists dispute the existence of language universals
altogether. Evans and Levinson (2009:431), for example, argue that languages are
actually unbelievably diverse in terms of their structure. For example, there are
languages with less than a dozen speech sounds, languages with 12 dozen sounds,
and sign languages with no sounds at all. They also point out that only about 10% of
the world’s approximately 6 500 languages have proper grammars and dictionaries,
which leads them to conclude that ‘at this stage of linguistic inquiry, almost every
new language that comes under the microscope reveals unanticipated new features’
(Evans & Levinson 2009:432). They argue that not all languages have the major
classes Noun, Verb, Adjective and Adverb, and that there are languages that make
semantic distinctions that we would never think of making. For example, where most

152
 languages make a distinction between singular leg and plural leg-s, the American
Indian language Kiowa has a verb marker meaning ‘of unexpected number’, so it
could be used to refer to any number other than the expected number, namely, one
leg or three legs!
A mutation is a
biological change Languages are certainly incredibly variable, but the questions that need to be answered
that results in a are: Are they infinitely variable (as Evans & Levinson believe)? or Are there some
new species. language universals that constrain the possible variations (as Chomsky and others
believe)?

2.3 The nature–nurture problem

In study units 2 to 5 on developmental psycholinguistics you studied the various stages
of language acquisition in children. All children (even children with disabilities to some
extent), irrespective of their culture or language, go through these stages. In fact, we
may add these stages to the list of language universals above. Furthermore, all normal
children seem to effortlessly acquire the same basic level of linguistic competence at
more or less the same stage despite differences in intellectual capacities.
Do children simply learn their mother tongue by imitating their caretakers’ language
use (an extreme form of the nurture view) or do children have an inborn language
acquisition ability (the nature [nativist] view)? As we saw in study unit 2, the answer
to these questions need not be either/or answers; both of these possibilities may indeed
be part of the language acquisition of children. In fact, unless we accept some form
of the nurture view, we cannot account for the fact that children do indeed acquire
different languages. Similarly, we have to accept some form of inborn ability to
acquire the syntactic universals of their languages, because these universals cannot
be simply imitated or even explicitly taught by caretakers as we shall see further on.
Let us now turn our attention to the phylogenesis of human language and see how different
theories of the evolution of language come to terms with the three problems listed above.

3 The phylogenesis of language in the human species

The two theoretical points of view that we will consider in this study unit are
concerned with both the ontogeny and the phylogeny of the language ability in
the human species (rather than the stages in the evolution of a specific language).
As we saw earlier, evolution is a gradual process extending over thousands of years,
but occasionally there may occur sudden, inexplicable mutations, also known as
saltations. It is exactly on these two characteristics of the evolutionary process that
the two theoretical perspectives on the phylogenesis of language are differentiated.
The one viewpoint claims that a sudden genetic mutation gave rise to the linguistic
ability in the human species; the other claims there was a gradual evolution of
the linguistic ability as a result of a whole range of biological adaptations in the
species (such as walking upright) which, taken together, made language possible.
Gradual evolution thus takes an emergentist perspective – in study unit 2 we defined
emergentism as a theoretical viewpoint which sees language as resulting from the
interaction of other more general abilities which already exist.

For a quick overview of the two theories, you could watch

How did Language Begin? on https://youtu.be/GvRtlH-3Asc

LIN2601/1/2017-2023 153
 Let us now take a closer look at these two different theories, starting off with the
sudden genetic mutation theory.

3.1 The sudden genetic mutation theory

According to this view, some or other ancestor in the human evolutionary line
developed the linguistic ability through a sudden genetic mutation and passed this
ability on to its offspring through its genes. Essentially, this means that within
the primate family (i.e. the family to which humans belong) only humans possess
the linguistic ability, and all human beings possess the linguistic ability in equal
measure. The linguistic version of this view is expressed in the so-called Innateness
Hypothesis – Chomsky’s version of nativism. According to this hypothesis, the
universal properties that characterise language (i.e. those properties common to
all languages) are part of the biological make-up of all human beings. Thus, these
universal properties cannot be learnt or acquired from an external source. Some
proponents of the genetic mutation view even go so far as to claim that innate
linguistic ability is a biologically determined specialised mental organ comparable
to other biological organs such as the heart (Chomsky 1980:185ff ). Notice that
this viewpoint explicitly claims that the innate universals are specifically linguistic
universals (i.e. they are unique to language). However, there are some linguists and
psychologists that support the claim that some human cognitive abilities are innate and
genetically determined, but they do not accept the view that there are independent
innate linguistic universals (see section 5.1 in study unit 2).

You can watch Chomsky outlining his theory of language evolution on

http://bigthink.com/videos/noam-chomsky-on-languages-great-mysteries

The sudden genetic mutation view of the phylogenesis of language has been proposed
as an explanation for a whole range of problematic phenomena, including certain
aspects of syntax that are apparently ‘unlearnable’. For example, how do children
learn which auxiliary should be moved to the front when turning the statement
Anyone who is interested can see me later into a question:
*Is anyone who interested can see me later? (incorrect auxiliary fronted)
Can anyone who is interested see me later? (main auxiliary fronted)

According to supporters of the innateness hypothesis, children would not hear

enough input to be able to learn the correct rule, and yet somehow they manage
to learn this rule. So where does this knowledge come from if it is not learnable?
Within the sudden mutation view it is claimed that this knowledge derives from
the language universals that are innate in the human species. It is claimed that just
as our bodies consist of organs that are species-specific, i.e. genetically determined
and unique to humans, so our minds consist of mental organs, such as a linguistic
ability, that are genetically determined. Chomsky (1980:134), the main proponent
of the sudden mutation view, claims that ‘in certain fundamental respects we do
not really learn a language, rather grammar grows in the mind’.
This grammar that ‘grows in the mind’ has an initial state that comes from the
human evolutionary ancestor with the language gene. In its initial state this grammar
contains all the language universals common to all languages and is therefore called
Universal Grammar (UG) (see also study unit 2).

154
 The sudden mutation view of the phylogenesis of human language accounts for
all three problems that face theories of the evolution of language by postulating a
genetically determined innate linguistic ability or mental faculty in the human species.
This innate linguistic ability places a limit on the variations among languages so
that all languages conform to the same genetically determined linguistic principles.
The ability of children to master these structural characteristics of languages effortlessly
and in a remarkably short period of time is accounted for by the assumption that
every child is born with linguistic knowledge in the form of a Universal Grammar.
This knowledge did not evolve over thousands of years in the evolutionary line
of the human species, i.e. there are no structurally simple languages compared to
structurally complex languages. The linguistic ability of the human species originated
from a sudden genetic mutation in one or other ancestor of the species and was then
genetically inherited by all human beings.
This view of the phylogenesis of language therefore believes there is no evolutionary
continuity between animal communication systems and human language. The
linguistic ability of humans is unique – it is not an extension of, say, the gestures,
facial expressions and vocalisations of chimpanzees. In the sudden mutation view of
language the significant similarities and differences between the bodies and habits of
primates and humans are considered to be immaterial and irrelevant. According to
the sudden mutation view, humans have a language ability and the other primates
do not – there is no continuity between the communication systems of primates
and human language. It is argued that it would therefore not make much sense
to compare the bodies and the habits of various primates and humans in order to
explain why language is possible in humans but not in primates.
Although this view of the phylogenesis of language forms the basis of one of the
dominant linguistic theories of our times, it is not necessarily a satisfactory theory.
In fact, the proponents of this theory simply argue that since there is no alternative
explanation for language universals, the discontinuity between animal communication
systems and human language, and the language acquisition abilities of children, their
theory is currently the only acceptable one to account for these problems. The problem
with the sudden genetic mutation view of the evolution of the human linguistic
ability is that other evolutionary traits of the species are simply ignored. Some of
these other traits may be very relevant for our understanding of the development of
the linguistic ability. In fact, some of these evolutionary developments may indeed
suggest radically different answers to questions about the nature of the linguistic
ability, both in its phylogenetic as well as in its ontogenetic senses. In a sense, the
sudden mutation view forces one to subscribe to the Innateness Hypothesis, i.e. to
view the linguistic ability as a biological organ (such as wings, or hearts, etc.) which
specialises in linguistic knowledge. It also forces one to view language universals as
being genetically determined.
But are there no other alternative explanations? That is what we are going to find
out in the next section.

Adaptation
is the process 3.2 The gradual evolutionary view
of adapting to
fit a changed According to the gradual evolutionary view, the development of the language ability
environment. in humans relates to other seemingly unrelated changes in the ancestry of the human
species. For instance, the evolution of the ability to walk upright rather than on all

LIN2601/1/2017-2023 155
 fours had an effect on the shape of the vocal tract that, in turn, made speech possible.
Such an approach requires that we look beyond language as such, and consider the
evolution of other abilities in the human species. These abilities might have contributed,
in one way or another, to our ability to speak, or to the highly structured nature of
language and the universality of certain properties common to all languages. In this
view, the ability to speak a language may not be a genetically determined ability at
all, but rather the end result of a whole range of biological adaptations in the species
which, taken together, made language possible.
The supporters of the gradual evolutionary viewpoint disagree with the sudden
Adaptation mutation theory and argue that human language would not have been possible
are the changes without significant biological adaptations (evolutionary changes) which make our
that result from bodies quite distinct from the other primates. The phylogenesis of language (including
adapting to
fit a changed written language and sign language) went hand in hand with major anatomical,
environment. neurological, cognitive and social adaptations known as preadaptations or precursors
of the language ability because they took place before the emergence of language.
None of the biological and social adaptations characterising the evolution of the
You are human species has anything directly to do with the development of a language ability
predisposed as such. For instance, most, if not all, the body parts that are involved in speech
to behave in a processing (e.g. the tongue, lips, lungs and vocal cords) have other primary functions
certain way if
something has that are not related to language, for example chewing and swallowing food, regulating
influenced you breathing, coordinating muscle movements and motion, etc. Even though these
and caused you bodily organs and systems are ideally suited for speech purposes, their linguistic use
to behave in that seems to be secondary to these other functions. Nevertheless, the suitability of these
way.
body parts for speech production is quite remarkable. It seems that the evolutionary
changes in these body parts predisposed humans for a linguistic ability.
The fact that adaptations in so many different systems and subsystems are (indirectly)
involved in the phylogenesis of language suggests that the language ability (including
speech, writing, signing) is not a unitary system or single ‘mental organ’. On the
contrary, linguistic ability is widely distributed over many different systems of the
human body, brain and mind – it relates to the anatomy and physiology of the
human body, to the information processing and coordinating abilities of the brain,
to the cognitive abilities of the mind, and to the emotional and social needs of the
human individual.

Task 7.3

Read the following quotes and decide whether they express a sudden mutation view
or a gradual evolution view of the origin of language:
A ‘… the language faculty is an extremely recent acquisition in our lineage, and it
was acquired not in the context of slow, gradual modification of preexisting sys-
tems under natural selection but in a single, rapid, emergent event that built upon
those prior systems but was not predicted by them’ (Bolhuis, Tattersall, Chomsky
& Berwick 2014).

156
 B ‘It would be scientifically improper as well as completely unenlightening to invoke
a mysterious blinding flash somewhere in the early history of our species before
which we were silent and agrammatical beings, and after which we had fricatives,
diphthongs and subordinate clauses’ (McMahon & McMahon 2013:1).

C ‘Our ancestors made an almost unimaginable transition from a non-symbolic,

nonlinguistic way of processing information and communicating information about
the world to the symbolic and linguistic condition we enjoy today. It is a qualitative
leap in cognitive state unparalleled in history. Indeed, as I’ve said, the only reason
we have for believing that such a leap could ever have been made, is that it was
made. And it seems to have been made well after the acquisition by our species of
its distinctive modern form’ (Tattersall 2012:199).

Feedback

Quotes A and C reflect a sudden mutation view of the evolution of language. The clues
are the words ‘single, rapid, emergent event’ in the first extract and the words ‘qualitative
leap in cognitive state’ Extract C. Extract B views a sudden mutation view as ‘scientifi-
cally improper’ and therefore reflects the gradual evolution view.

The following discussion of some of the major biological adaptations that predisposed
the human species for a language ability is necessarily superficial. It would be
impossible to deal with all the relevant adaptations in great detail within the limits
of a single study unit. However, the adaptations that we will consider below are
sufficient to illustrate the gradual evolution of the linguistic ability in humans.
Note that the order in which these adaptations is presented does not necessarily
represent the chronological order in which they evolved. It is in fact impossible to
order these adaptations on a time scale although the order in which the apes, the
hominids and Homo sapiens evolved would be suggestive of the time scale of the
evolution of the various adaptations.

3.2.1 Bipedalism
If bi means two and ped means foot, what do you think bipedalism means?
Hands (or feet)
are prehensile
when they can
be used to grasp I’m sure you managed to work out that bipedalism is the ability to walk upright on
onto things. two legs (rather than on all fours).
Primates are adapted for an arboreal lifestyle, particularly with their prehensile hands
and feet. The prehensile hands and feet of the primates can be seen as an evolutionary
precursor of bipedalism in the sense that the muscular control required for grasping
enables primates to walk upright for limited periods. Zihlman (1982:55) observes
that chimpanzees are ‘jacks of all trades’ in locomotion – besides climbing, hanging,
walking, and running on all fours, chimpanzees occasionally walk bipedally for
short distances. Humans, in contrast, are specialised for walking long distances

LIN2601/1/2017-2023 157
 even while carrying food, tools, or babies. Bipedalism is not only the most striking
distinction between humans and primates, but is in all probability the adaptation
with the widest and most significant anatomical, physiological and behavioural
effects that could be linked to the development of the linguistic ability in humans.
But why did bipedalism evolve?
The apes are restricted to the forests of the African continent. During the Late
Miocene period, which lasted from 23 to 5 million years ago, temperatures began
to fall and forests decreased, leading to an increase in open land. This increase in
savannah created a new ecology in Africa which chased some apes out of the forests
onto the savannahs. The arboreal lifestyle alone was no longer feasible. Some of the
apes had to adapt to a terrestrial lifestyle that required bipedal locomotion. These
were the hominids, who were the direct ancestors of modern humans.
Hominids were relatively large compared to other warm-blooded animals and
Something is therefore required more food. The abundance of animals in the savannah habitat
rudimentary provided an additional food source and so the hominids became hunters as well.
when it is basic or Bipedalism gave them the speed, agility and endurance required for effective hunting,
primitive.
while it also freed the hands. Thus, in addition to the bodily adaptations to cope
with living conditions on the savannah such as bipedalism and the development of
sweat glands, certain behavioural adaptations were also required, such as developing
hunting skills and creating shelters. Both of these activities would have benefited
greatly from the availability of tools, however rudimentary. Once again, bipedalism
seems to be a significant factor – the hands were free to engage in tool making,
rather than walking or swinging in the trees. The range of preadaptations following
from bipedalism is truly remarkable, but none more so than the preadaptations for
the ability to speak and write.

3.2.2 Binocular and colour vision

Binocular vision refers to the ability of both eyes to observe the same visual field
in three dimensions, i.e. in depth. Binocular vision is an evolutionary prerequisite
for the arboreal lifestyle of primates since it is essential for hand-eye coordination in
grasping objects and swinging from branch to branch. The evolution of binocular
vision in the primates correlates with several other significant adaptations that may
be seen as preadaptations, directly or indirectly, for the evolution of language:
 Binocular vision allows a switch from a communication system based on smell
to a visually-based communication system using facial expressions and gestures.
The use of facial expressions for communication requires an increased, more
elaborate set of facial muscles which in turn enable the finely tuned articulatory
movements of speech.
 Binocular vision went hand in hand with an increase in brain size which improved
memory capacity and the ability to make associations (such as remembering food
locations). Colour vision also enables the development of cognitive abilities such
as discrimination, generalisation and categorisation.
 Binocular vision makes the manual manipulation of objects (such as the use of
sticks and rocks to gain access to food sources) possible and creates the potential
for toolmaking.
 Binocular vision creates the potential to form mental images (i.e. pictures in the
mind), which are essential for language.

158
 3.2.3 The vocal tract
Figure 7.1 Vocal tract comparisons: gorilla, human infant, human adult

Gorilla Child Adult

Larynx Larynx Larynx

Task 7.4

Compare the diagrams of the vocal tract of a Neanderthal and a human baby in the
link above with reference to the position of the larynx (the voice box) and the length of
the pharynx (the area of the throat above the larynx). Write down your observations.

Compare the diagrams of the vocal tract of gorilla, human infant and adult human in
Figure 7.1 above with reference to the position of the larynx (the voice box) and the length
of the pharynx (the area of the throat above the larynx). Write down your observations.

LIN2601/1/2017-2023 159
 Feedback
The Neanderthal has a short, horizontal pharynx and a high larynx. The human infant
has a short, curved pharynx and a high larynx, while the adult has a much longer, L-
shaped pharynx and a low larynx, visible in some people as the ‘Adam’s apple’ at the
front of the throat. What is so remarkable about this comparison is that Neanderthal
and human infants are much closer to one another with regard to the high position
of the larynx and the short pharynx than babies are to human adults. In a sense, the
similarities between the vocal tracts of Neanderthals and also gorillas and human ba-
bies suggest that ‘ontogeny recapitulates phylogeny’, i.e. that the evolutionary stages
of a species (phylogeny) can be traced through the study of the developmental stages
(ontogeny) of an individual (Leaky & Lewin 1992:261–2.) The position of the larynx and
the length and shape of the pharynx, among other things, are directly responsible for
the range of speech sounds that makes speech possible. Neither Neanderthals, goril-
las nor human babies can produce this range of speech sounds because of the high
position of the larynx in the throat.

The high larynx in animals and human babies is not accidental. It is crucial for
the prevention of choking. Unlike human adults, gorillas and human babies can
breathe and swallow food simultaneously without the danger of choking because
of the high position of the larynx.
The larynx in human babies begins to move down the throat between 18 months and
2 years and reaches the adult position by the age of about 14 years. The lower position
of the larynx in human adults seems to be a direct consequence of bipedalism. The
shape of the human skull is adapted for walking upright with a distinct curvature
at its base, which seems to have the effect that the larynx is forced lower down in
the throat. Surgical experiments on mice whereby the shape of the base of the skull
was changed in this way had the same effect – the larynx moved lower down the
throat (Lieberman 1991:74). It has also been suggested that the lower position of the
larynx in the throat enhances breathing through the mouth, which may be linked
to the endurance required to survive in a savannah landscape (Lieberman 1991:74).
It seems clear that the lowering of the larynx is motivated both anatomically (through
changes in skeletal structure due to bipedalism) and physiologically (to enable mouth
breathing). The lowering of the larynx, which took place around 200 000 years
ago, created the vocal tract shape and length necessary for the production of a wide
range of speech sounds (Bolhuis et al 2014). But since the changes are motivated
by reasons other than merely the production of speech sounds, the lowering of
the larynx cannot be seen as an adaptation specifically for speech, but rather as a
preadaptation that then made speech possible.

160
 In the following video, Professor WT Fitch gives a short lecture on how
the lowered larynx enabled speech.
https://www.youtube.com/watch?v=eP0Kdkz4iHU

3.2.4 Premature births and longer postnatal dependency

Bipedal locomotion required significant structural modifications in the pelvis of
humans. These structural changes meant that for the survival of both mother and
child, babies had to be born at a much earlier stage of physical and mental maturity.
In fact, we may perhaps say that bipedalism made premature births necessary. Trask
(1999:167), for example, notes that
In contrast with the young of most other mammals (except marsupials), human
infants are born at an extraordinarily early stage of development. A baby deer
can stand and walk within minutes of birth, but a human infant cannot even
turn or crawl until months after birth. At birth, the infant’s skull is not yet
fully formed; the myelin sheaths which insulate the neurons in the brain are
not yet fully developed; even the visual system in the brain is not yet working
properly. Indeed, it has been estimated that human infants really ‘ought’ not
to be born until after eighteen months, instead of the actual nine months.

This ‘premature birth’ requires a much longer period of maternal care after birth.
The close bond between mother and child over an extended period played a major
part (directly or indirectly) in the evolution of language and still plays a significant
role in the development of language in individuals, as we learned in study unit 5.
The close bonding between mother and child in humans helps to satisfy three basic needs
of the helpless human infant: nutrition, protection and emotional needs. Recent research
shows, however, that the dependency of the human infant on its mother is also crucial
for satisfying a cognitive need, the need to acquire knowledge of the environment and
how to relate to it. The longer period of infant dependency may therefore be directly
correlated with the increase in learned behaviour. Research on the close interaction
between infants and mothers suggests that infants ‘have predispositions for learning,
and parents predispositions for teaching’ (Papousek & Papousek 1997:100). During
this early period of development, the phylogenetically evolved communication system
involving gestures, facial expressions and physical contact forms the basis for the
ontogenesis of verbal communication (Goldschmidt 1997:229).

3.2.5 The opposable thumb and toolmaking

The ability of the human thumb to touch the tips of any one of the other fingers
is known as the opposable thumb. Although the opposable thumb evolved in the
primates, it is unclear whether it derived directly from the free hands associated with
bipedalism. Nevertheless, this ability, together with free hands, had a significant
impact on the nature and direction of evolution in the human species. Instead of
adapting to an environment, human beings can alter the environment using tools.
Free hands and the opposable thumb allowed humans not merely to use objects
in the environment as tools (chimpanzees can do this as well) but to manufacture
tools for specific purposes.
The ability to manufacture tools became gradually more sophisticated through the
ages. In particular, the length of the cutting edge that a toolmaker could produce

LIN2601/1/2017-2023 161
 from the same quantity of stone increased dramatically. The hand axes used by
Homo habilis 2 million years ago consisted of lumps of tapered stone with a very
rudimentary cutting edge, compared to Neanderthal stone tools from approximately
100 000 years ago that were much sharper and more effective. By 30 000 years ago,
Homo sapiens sapiens had developed the technology to produce blades from metal
and other materials, making the tools even more efficient.
According to Pinker (1995:352), ‘complex artifacts are thought to reflect a complex
mind which could benefit from complex language’. The level and quantity of
information required by the stone tool technology at a certain point in time entails a
sophisticated information processing system or systems. The increase in information
may therefore be one of the reasons why humans developed both an oral and (much
later) a written linguistic system. Lecours (1995:232), for example, explains the
motivation for the development of speech and writing as follows:
It might be that, about 35 000 (if not a few million) years ago, environmental
pressures exerted on the human brain led it to exploit its genetic potential and
to invent spoken speech; and that, in due time, about 5 300 years ago, other
social pressures – of another nature, although generated to a large extent by
the impact of the first invention – led the human brain to exploit its genetic
potential anew, and to invent writing systems …

3.2.6 Enlarged brain and increased connectivity between regions in the brain
Figure 7.1 A comparison of the brain size of chimpanzees and humans
(www.brainfacts.org)

An increase in the volume of the brain and the connectivity between regions of the
brain played a major role in the evolution of the primates. The increase in brain
size happened mainly in the cortex. As we saw in study unit 1, the function of the
cortex is to integrate all the functions of the nervous system. The cortex receives
and transmits messages to and from all the sensory organs. This part of the brain is
therefore ideally suited for the acquisition of new knowledge and behaviours.
The increase in brain size in the apes seems to stem at least partly from the development
of binocular vision. The elaboration of the visual and association areas in the primate
brain presumably laid the foundation for the extensive connectivity which is one
of the most significant features of the human brain. The extent of connectivity
and interactivity in the human brain is probably the single most important factor
underlying learnt behaviour as opposed to instinctive behaviour. This increase in
the ability to learn is what distinguishes man from all other animals.
Lateralisation of the brain may be the result of this increase in brain size and the
interconnections between regions in the brain. (Remember that lateralisation is
the process whereby one of the two hemispheres becomes dominant for certain
cognitive functions.) Ringo (1991:5 in Foley 1997:60) suggests that ‘big brains need
hemispheric specialisation, because of interconnectedness getting out of hand …’
The increase in brain size and connectivity are essential evolutionary preadaptations for
the linguistic ability. Language requires interaction between various regions in the brain:
the region that controls motor action (required for speech articulation) must be connected
to the auditory system; the areas that control the various senses must be connected to

162
 the system that controls emotions; and all of these regions must be connected to the
cortex where the functions of the nervous system are integrated and holistic concepts
and associations are formed. Without these connections, the conceptual-functional
properties characteristic of language would be impossible to realise.

3.2.7 Summary
In this section we have discussed various preadaptations in the evolution of the human
species that predispose us for language, both verbal and written. The preadaptations
we looked at were bipedalism, binocular vision, the lowering of the larynx, premature
birth, the opposable thumb and toolmaking and increased brain size. We also suggested
that the increased information required to produce sophisticated tools was a possible
motivation for the development of an elaborate communication system such as
language. In the next section we will continue to explore the gradual evolutionary view
and attempt to find an alternative answer for the language universals problem which
served as the major justification for the sudden genetic mutation view of language.

Task 7.5

Summarise your understanding of this section on preadaptations for language in humans

by filling in the gaps in the table below. Which preadaptation has been left out? Add in
an explanatory sentence yourself in the last two lines.
The ability to see in 3-D is known as . This
improved hand-eye coordination and the ability to categorise things and eventually
enabled toolmaking.
A bigger brain, thicker cortex and
led to lateralisation and enabled interactions between the visual, motor and auditory
regions of the brain.
Premature birth and longer postnatal dependency means that
. This led to a longer period of par-
ent–child interaction and an increase in learned behaviour.
The ability to walk upright is known as . This
allowed humans free hands to use for gestures and toolmaking.
The shape of the vocal tract changed due to a

Feedback
The missing phrases as ‘binocular vision’, ‘increased brain connectivity’, ‘babies are
born earlier and are highly dependent on their parents for several years’, ‘bipedalism’,
and ‘lowered larynx’. The missing preadaptation discussed in section 3.2 above was
the opposable thumb, which means that the thumb can touch all the other fingertips
and is thus able to make the fine motor movements required for toolmaking.

LIN2601/1/2017-2023 163
 4 Language universals in a gradual evolutionary view of
the phylogenesis of language
Before we take a closer look at language universals from a gradual evolutionary
perspective we need to clear up a misconception within the sudden mutation
view. The sudden mutation view assumes that the innate Universal Grammar that
specifies universal linguistic principles such as classifiability, linearity and hierarchy
is a unique and independent mental ability of human beings. That is, although
Universal Grammar may be linked to other cognitive abilities of humans, it is an
ability that evolved separately from all other cognitive abilities and is dedicated to
innate linguistic knowledge. In the following task you will see that this assumption
is based on a misconception and that the same principles that underlie certain
linguistic universals also govern other cognitive abilities, such as short-term memory.

Task 7.6

1. Read the following telephone number once and without looking at the number
again, try to recall it from memory (don’t cheat!):
0113659981
Were you able to recall it on the first attempt? I guess not, except for those of you
with photographic memories. Now read the same number again, but written in a
slightly different format. Then try to recall it from memory
011-365-9981
Were you more successful this time? If so, why?

2. Read the following English sentence in a monotone (i.e. do not modify your
intonation at any stage) once and then close your eyes and try to recall it from
memory. (All the punctuation marks have deliberately been omitted.)
Andrew the master of ceremonies while on his way to discuss the program with
me suddenly fell ill
If you read the sentence according to the instructions you may have experienced
great difficulty in recalling it from memory. Breaking up the sequence of words by
means of punctuation marks and proper intonation should make a big difference to
your recalling ability. Read the sentence below and then try to recall it from memory.
Andrew, the master of ceremonies, while on his way to discuss the program with
me, suddenly fell ill.

Feedback
The task above indicates that the short-term memory of human beings is rather limited.
However, this limitation can be overcome by dividing a complex piece of information
into smaller bits. The cognitive process underlying the ability to cope with quantities of
information beyond the capacity of short-term memory is called chunking and human
beings are adapted to perform this chunking intuitively.

164
 4.1 Short-term memory limitations, chunking and language
universals
Let us now see what short-term memory limitations and the cognitive process of
chunking can tell us about language universals.
In 1956 the psycholinguist George Miller published his research on the limits of
short-term memory in the phonological processing of speech in an article entitled
‘The magical number 7, plus or minus two: Some limits on our capacity for processing
information’. Miller found that the ideal number of bits of information that can
be handled by short-term memory is 7, which could be extended to 9. More than 9
pieces of information cause a breakdown in the recall ability of short-term memory.
This finding has been confirmed in many other studies on short-term memory since
then (cf. Loritz 1999). One of the most amazing features of short-term memory is
that its upper limit on recall – 9 bits on the same level – can be exceeded by means
of chunking. By restructuring bits of information into less than 9 chunks or groups,
one can recall rather complex bits of information. This is exactly what we did in the
tasks above. In both cases we had more than 9 bits of information but by means
of chunking we turned the same information into manageable quantities, enabling
short-term memory to handle them.
Like telephone numbers and ID numbers, bits of linguistic information often
exceed the limitations of short-term memory. Thus, even though the human
species has the wonderful ability to produce a range of sounds rich enough to make
verbal communication possible, if the verbal utterances are not structured in such a
way that short-term memory can cope, all our articulatory abilities would come to
nothing. So let us see how linguistic systems overcome short-term memory limitations.

4.1.1 Phonetic chunking

A characteristic feature of human speech is the rapid transmission of speech sounds at
about 15 to 20 per second (Lieberman 1991). If the production of phonetic elements
in the utterance of a word is slowed down, understanding is impeded. We can see
this when young children learn to read by sounding words out letter by letter –
they simply don’t understand what they are reading until they say the whole word,
because the production of the phonetic elements is way below the 15 to 20 elements
per second required for comprehension. Thus, while spelling the letters or sounds
b-o-o-k, they don’t know that what they spell means ‘book’ until they pronounce the
whole string as [bυk]. For the same reason, morse code is not a very effective means
of communication as it is transmitted at such a slow speed that it is hard to interpret.
But, clearly, the 15 to 20 elements per second required by speech processing is a
blatant violation of the maximum of 9 items limitation of short-term memory. How
is this paradox solved in languages? Once again the answer is chunking. A universal
feature of words is that they are made up of syllables (i.e. all languages have this
feature). A syllable is a phonetic string consisting of a vowel (V) and consonants (C).
Different languages have different syllable structures, but all of them have syllables.
Consider the syllable structures of English and Xhosa below.
English: explanation > ex-pla-na-tion > VC + CCV + CV + CVC
Xhosa: amadodana (‘young men’) > a-ma-do-da-na > V + CV + CV + CV + CV
The syllabification of words is a linguistic manifestation of chunking and together
with stress patterns and tones, helps us to process words in short-term memory.

LIN2601/1/2017-2023 165
 4.1.2 Hierarchies as forms of chunking and information processing
You will recall that, of the three syntactic principles that we mentioned earlier, the
hierarchic principle is considered to be the most significant one. We will demonstrate
below that the hierarchic principle is a general cognitive principle rather than a
strictly linguistic principle. Clearly, the hierarchic principle is also motivated by
chunking as a result of short-term memory limitations. But more importantly, it is
also a fundamental principle in the general processing of information, particularly
categorisation.
The demonstration below begins with a very interesting parable. Herbert Simon
(1962), an American scientist, made up this parable to demonstrate how effective
hierarchical chunking could be and why organisms would select hierarchies in their
evolution. I adapted this parable from Sampson (1980).
The parable of two watchmakers – Hora and Tempus
Two watchmakers, Hora and Tempus, produce watches of equal quality.
They use the same number of parts, say 1 000 parts, in each watch that they
respectively make. But their respective watches are differently structured.
Tempus constructs his watches by assembling all the parts in a continuous
process in piecemeal fashion. Thus, he takes a part and adds it to another part,
to which he adds another part and so on. If he is interrupted while making a
watch, the partly constructed watch will fall apart, and he must start at the
beginning when he resumes his watch building. Hora produces his watches
by making sub-assemblies of the various parts, say ten sub-assemblies of a
hundred parts each. If Hora is interrupted, he has to reassemble only the sub-
assembly he was busy with – the other sub-assemblies remain intact. Hora
is a successful watchmaker, Tempus isn’t. Hora produces 4 000 watches for
every one produced by Tempus.

The key to Hora’s success is not the quality of the end product, but rather the fact
that the structure of his watches is hierarchic.
The very same principle that made Hora such an efficient watchmaker, namely
hierarchy, makes human beings exceptionally good managers of conceptual
information. The following task will help you to understand this.

Task 7.7

Organise the following list of objects into a hierarchically structured information set:
lion, dog, kudu, cat, tiger, cow, horse, eland

Feedback
When presented as a list, the conceptual relations between the various elements in the
list are completely obscure. But when the list is restructured in a certain way, conceptual
relations are revealed. You could have organised the elements in the list in various ways
depending on the kind of information you wanted to convey.

166
 For example:
Wild animals Domestic animals
lion, kudu, tiger, eland dog, cat, cow, horse
or
Carnivores (meat eaters) Herbivores (plant eaters)
lion, tiger, dog, cat cow, horse, kudu, eland

How does the general cognitive hierarchic principle relate to the universal linguistic
hierarchic principle? Is it the same principle, or are they altogether different principles?
Complex sentences in all the languages of the world are typically hierarchically structured
and this hierarchical structuring would obviously assist in overcoming the limitations of
short-term memory. However, there is another reason for hierarchical structuring that
relates in a significant way to the efficient processing of complex information.
Consider the sentences below:
This house is a national monument.
This house was built in 1762.
This house was built by Jack.
Jack was a famous architect.
Say there was a stage in the evolution of language when only simple sentences of
the sort illustrated above could be formed by English speakers. A lot of important
information would have been lost. Let us see what would have been lost.
 Can we infer from these sentences that we are talking about the same house in
all three sentences? No, not necessarily.
 Can we infer from these sentences that the individual Jack is the same person in
the last two sentences? Once again, the answer is no.
 Is the fact that the house is a national monument due to the fact that it was built
in 1762 and/or that it was built by the famous architect, Jack? If we cannot link
the meanings of the two sentences in some way or another, we cannot make such an
inference.
 Do we have any reason to believe that Jack lived and practised his profession as
an architect in and around the year 1762? No, there is no justification for such an
inference unless we can link the information in the different sentences.
But there is a way in which we can actually capture all the semantic relations between
bits of information that the four single sentences above could not express.

Task 7.8

Try to combine all four of the sentences above into one sentence by making use of the
hierarchic principle of organising information.

LIN2601/1/2017-2023 167
 Feedback
Your combined sentence could be any one of the following:
This house, which was built in 1762 by the famous architect, Jack, is a national
monument.
This house, which is a national monument, was built in 1762 by Jack, who was a
famous architect.
The famous architect, Jack, built this house, which is a national monument, in 1762.

There are other possible answers, but these sentences clearly show certain crucial
relations between the various bits of information, relations that cannot be expressed in
simple sentences. Since hierarchical organisation is already required by human short-
term memory in order to cope with large quantities of information, hierarchically struc-
tured linguistic expressions simply encode semantic information by the same means.

4.2 Summary

It is clear from our foregoing discussion that the hierarchical structuring of

information is not unique to language but is part of general human cognitive
processes. Furthermore, the linguistic hierarchic principle does not seem to be a
unique principle of an innate Universal Grammar. On the contrary, it seems to be
part of the general cognitive abilities of human beings. The ability to hierarchically
structure information may have evolved in the human species because of short-term
memory limitations. Hierarchically structured information was also required by the
massive growth in information that humans generated by using tools to modify the
environment rather than adapting to an environment.
In the final section of this study unit we will briefly take a closer look at the role of
information in both the ontogeny and the phylogeny of the language ability in humans.

5 Information and the ontogenetic and phylogenetic

development of language
Even though the principle that ‘ontogeny recapitulates phylogeny’ has never been
seriously entertained in evolutionary theories, there is a remarkable correlation
between the ontogenetic and the phylogenetic development of language that seems
to lend some support to this principle. It seems that quantity of information plays
a crucial role in the correlation between the ontogeny and phylogeny of language.
Before we explore this correlation further, let us first look at the role of information
in the phylogenesis and the ontogenesis of language separately.

5.1 Information and the phylogenesis of language

In the case of the apes, all (or most of) the information relevant for survival is in their
genetic make-up on the one hand, and in their environment, on the other hand. The
apes have evolved to live and survive in forests. The kind of information that they
need for these living conditions is innate – it is part of the genetic make-up of their
bodies. Even their communication systems are genetically given. They will never be
able to increase or modify their range of instinctive visual or vocal signals. As we

168
 saw earlier, they simply don’t have the vocal tract necessary to produce a rich variety
of sounds, even if the volume of their brains increased to the level of human beings.

Perhaps more importantly, the apes don’t need language (vocal or written) because
they don’t need elaborate information systems to cope with their environment. Their
environment contains all the information that they need and their bodies are attuned
to this environment. That is why the apes cannot survive if their environment or
ecology (the tropical forests) is destroyed.

In contrast to the apes, human beings, because of their toolmaking ability and
their larger and more interconnected brains, do not have to rely only on genetically
determined information in order to survive. Neither do humans have to rely on
a specific ecology to which their bodies are genetically attuned. Humans, as we
have seen, ‘are free’. They can live anywhere – on the moon, in space, in the polar
regions – without being genetically adapted to do so. Why? Because human beings
can modify their environment and for this the human species has been preadapted.

In the early days of human evolution the modifications in the environment may
have been fairly simple, with simple stone tool technologies, simple shelters, simple
social relations, and so on. In fact, in early hominids much of the information
necessary for survival may still have been in their bodies and in the environment
rather than produced by their brains. But as the tool technologies improved and the
modifications of their environment became more sophisticated, they had to rely more
on information produced by their brains. To manage large and complex quantities
of information you need a code. And language provides such a code. Once again,
the human species is, as we have noted, preadapted for language.
Up to a certain level the quantity and complexity of information can be handled
by an oral linguistic system which relies heavily on memory capacity, for which the
human species is also well adapted with its larger and more sophisticated brain. Since
oral language leaves no traces, we can only speculate on the developmental stages of
oral language in the evolution of the human species. Studies of hominid skeletons
suggest that a vocal tract structure that would make linguistic utterances possible
emerged around 200 000 years ago. Most researchers believe that language emerged
considerably later than this, however, at roughly the same time as the first symbolic
objects such as shell beads (from about 100 000 years ago) and engraved plaques
(from about 80 000 years ago in South Africa’s Blombos cave) (Bolhuis et al 2014).
At some point the modifications of the environment or the sophistication in the tools
needed to modify the environment increase the quantity of information beyond the
limits of even modern humans’ memory capacity. Then we need an information
coding system of a more permanent nature which lies outside our brains. This is
where written language comes into play in the phylogenesis of language. Once again,
humans are preadapted for this need. Apart from the complex brain structures and
visual abilities, the opposable thumb and toolmaking abilities enabled the invention
of writing tools and systems that made the recording of large quantities of complex
information possible.

5.2 Information and the ontogenesis of language

The fact that humans are less dependent than apes on genetic and environmental
information creates a rather special problem for human infants. Where do they

LIN2601/1/2017-2023 169
 get their information from, if, by virtue of their species, neither their bodies nor
a specific habitat can provide the information? The answer is that human infants
have a special environment – their mothers or caretakers. Human babies are largely
dependent on the information in the brain of their mothers or caretakers. The need
for more information increases quite rapidly during this early period of dependence
and there is therefore a very strong urge to acquire language.
As Papousek and Papousek (1997:100) indicate, the urge to learn and the ability to
teach a linguistic communication system may have evolved in the human species.
To some extent, the amount and complexity of the information required during
the various stages of language acquisition correlate well with the various stages of
linguistic development in the human species. In the continuum below we attempt
to correlate the various ontogenetic stages (babbling, the one-word stage, the two-
word stage etc) with (highly speculative) phylogenetic stages in terms of similarities
in the levels of information that the human species and human infants have to deal
with at each developmental stage.

Figure 7.4 A correlation of language development in children (ontogenesis)

and the human species (phylogenesis)

Phylogenesis:

Linguistic stages: simple vocalisations – object and action related one-word utterances
– simple phrases – simple sentences – complex sentences

Levels of information: pre-conceptual – imitative – simple concepts – simple related

concepts – simple causal relations – advanced coherent conceptual complexes

Ontogenesis:
Linguistic stages: babbling/cooing – one-word stage – two-word stage – simple
sentences – complex sentences

Notice how the increase in levels of information correlates with the increase in
levels of linguistic sophistication in the phylogenesis as well as the ontogenesis of
language. For example, babbling is a simple vocalisation and is pre-conceptual in
the sense that babbled sounds do not represent particular concepts. In the early
human ancestors, vocalisations were initially probably similar to babbling in that
they were pre-conceptual and did not link to particular concepts. In the one-word
stage children begin to link words and meanings, using language to express simple
concepts such as objects and actions. Similarly, we would expect early human ancestors
to have begun using single words to express simple concepts in the earliest days of
human language. Later, children develop the ability to combine words, producing
two-word sentences and simple sentences that can express not only single concepts,
but also the relationship between concepts. Early humans may have been the same,
gradually moving from single words to phrases and sentences that could express
relationships between concepts. Our current state of development allows humans to
use language to express sophisticated ideas and advanced complexes of information.

170
 Task 7.9

Test your understanding of the two theories in this study unit by reading through the
following interviews and then answering the questions that follow. The same ques-
tions were posed to two academics in the Department of Linguistics at Unisa, but the
answers are rather different!
Interview question 1
How does human language differ from animal communication systems?
Professor A’s answer:
In my opinion, human language is simply a more elaborate and sophisticated
version of animal communication. For example, both humans and animals such
as monkeys and chimpanzees communicate using oral sounds, gestures and
facial expressions. Of course human and animal systems do differ dramatically
in their capabilities. For example, primates have much smaller brains with fewer
neuronal connections. This means that they are governed largely by instinctive
behaviour and can only use a few, very limited expressions.

Dr B’s answer:
Human language differs both quantitatively and qualitatively from animal com
munication systems. In fact I believe that the two systems are unrelated. The
range of meaningful expressions in any system of animal communication is
very limited, while human language, in contrast, allows us to be much more
creative and express new ideas and insights. The fact that children acquire
language so quickly and effortlessly indicates that humans have an innate
linguistic ability that is unique to our species.

Interview question 2
How does modern human language differ from that of early human ancestors
such as the hominids?
Professor A’s answer:
Well of course that is a difficult question to answer since language leaves no
traces and we have no records of speech from so long ago. I believe, however,
that the early human ancestors used a very simple communication system
using only single words and that over time, the language system became
gradually more complex. I would imagine that as early humans became more
sophisticated in their ability to hunt, to build shelters, and to make tools, that
their spoken language became more complex to keep pace with the demands
of their changing lifestyles.
Dr B’s answer:
There is certainly no evidence that any of the modern languages is structurally
simpler than others – all languages have universal features that make them
equally structur ally complex. This leads me to believe that there was never a ‘sim-
pler’ language used millions of years ago by the hominids or Homo erectus, for
example. My view is that the language ability emerged in the human species through
a sudden biological change that resulted in the species we know as Homo sapiens.

Interview question 3
What methods do you use to do research into the origins of language?
Professor A’s answer:
The method I use is to compare the anatomy of modern humans with the fossil
remains of hominids and modern primates. These differences can tell us a lot
about how and when language emerged in the human species.

LIN2601/1/2017-2023 171
 Dr B’s answer:
My theory is based on simple logic. There are many aspects of human lan
guage that cannot be learnt simply from listening to spoken language. Since
children all over the world do manage to acquire even the ‘unlearnable’ aspects
of language, it is clear to me that this knowledge is already present in our
minds when we are born, and forms part of the genetic make-up that makes
humans unique.
Interview question 4
Are language and cognition two separate systems in the human brain?
Professor A’s answer:
In my opinion, no. Language is intimately related to our anatomy, to our social
environment and to our other cognitive abilities. For example, cognitive tasks
such as categorising and memorising information require an ability to structure
information hierarchically. Language works in exactly the same way, requiring
us to hierarchically structure complex information so that our short-term memo
ries can handle it. This suggests that our language ability is not a special, au
tonomous system, but simply part of our wider cognitive ability.
Dr B’s answer:
????

Questions for you to answer

1. What is the term for the differences of opinion of Professor A and Dr B on the
issue of human and animal communication systems (see Interview question 1)?

2. Which academic supports the sudden mutation theory of the phylogenesis of

language?

3. Which academic supports the alternative to the sudden mutation theory, and what
is this theory called?

4. Which academic bases her ideas on those of Noam Chomsky?

5. Which academic would probably use terms like ‘binocular vision’ and ‘bipedalism’
in her lectures?

6. What would you expect Dr B’s answer to Interview question 4 to be?

Feedback
1. The continuity controversy.
2. Dr B.
3. Professor A supports the gradual evolution theory.
4. As a supporter of the sudden mutation theory, Dr B follows Chomsky’s view of
the phylogenesis of language based on the Innateness hypothesis.
5. Professor A. Supporters of the gradual evolution theory believe that the human
language ability is the end result of a series of preadaptations such as binocular
vision, bipedalism and the lowering of the larynx.

172
 6. As a supporter of the sudden mutation theory, Dr B would answer yes to this in-
terview question. The linguistic hierarchic principle is seen as a unique principle
of an innate Universal Grammar. Universal Grammar is seen as an ability that
is separate from all other cognitive abilities and is dedicated purely to language
knowledge.

6 Summary
We have explored two different theories relating to the evolution of language in the
human species. The dominant theory is still the sudden mutation view based on the
innateness hypothesis, but as we have tried to show you, there is indeed an alternative
in which preadaptations in human ancestry predisposed the human species to invent
language (both oral and written) in order to cope with the information they needed
to make their environment habitable.
We have also considered an alternative explanation to the innateness hypothesis for
universal features common to all languages of the world. The limitations of short-term
memory and the cognitive ability to structure complex information hierarchically
form the basis of this alternative explanation.
There are many possible motivations for the evolution of the linguistic ability in
humans. We focused on one of these, namely, the ability of humans to adapt to
various environments. This ability relies heavily on our ability to generate and
manage large and complex quantities of information. Large quantities of data require
an elaborate coding and communication system (or systems) such as language. In
addition to our ability to learn and to be creative, the linguistic ability is perhaps
the most significant tool enabling humans to conquer and transform the world.

Academic vocabulary building

Choose the word in brackets that is synonymous with the listed vocabulary item
in each case.
purportedly [fundamentally] [allegedly] [generally]
differentiate (between) [distinguish] [compare] [contrast]
significant [indicative] [suggestive] [important]
controversial [contentious] [irregular] [uncommon]
recapitulate [recall] [repeat] [revise]
quantitative [numerical] [countable] [measurable]
qualitative [disposition] [according to qualifications] [valuable]
irrespective (of) [unlike] [disrespectful] [regardless]
inexplicable [unaccountable] [indescribable] [intolerable]
conform (to) [relate to] [comply with] [adapt to]
dispense (with) [do away with] [discount] [disperse]
immaterial [unnatural] [irrelevant] [insubstantial]
irrelevant (for, to) [immaterial] [disinterested] [inapplicable]
correlate (with) [coordinate] [interconnect] [compare]

LIN2601/1/2017-2023 173
 multi- [substantial] [many] [countable]
superficial [shallow] [insignificant] [widely spread]
conducive (to) [conforming], [likeable], [fitting]
chronological [temporal] [ordered in time] [sequenced]
prerequisite [precondition] [requirement] [necessity]
locomotion [sleeping] [walking] [animal movement]
misconception [misunderstanding] [confusion] [misinterpretation]

Further reading

Aitchison, J. 1996. The seeds of speech: language origin and evolution. Cambridge:
Cambridge University Press.
Christiansen, MH & Kirby, S (eds). 2003. Language evolution. Oxford: Oxford
University Press.
Crystal, D. 2010. The Cambridge encyclopedia of language. 3rd edition. Cambridge:
Cambridge University Press. (Chapter 49 is on the origins of language).
Tecumseh Fitch, W. 2010. The evolution of language. Cambridge: Cambridge Uni-
versity Press.

Bibliography

Bolhuis, JJ, Tattersall, I, Chomsky, N & Berwick, RC. 2014. How could language have
evolved? PLoS Biology 12(8):1–6. Available at: doi:10.1371/journal.pbio.1001934.
(accessed 25 January 2016).
Chomsky, N. 1980. Rules and representations. Oxford: Basil Blackwell.
Coppens, Y. 1995. Brain, locomotion, diet, and culture: how a primate, by chance,
became a man, in Origins of the human brain, edited by JP Changeaux & J
Chavaillon. Oxford: Clarendon Press: 104–112.
Corballis, MC. 2014. The gradual evolution of language. Humana: Mente Journal
of Philosophical Studies 27:39–60. Available at: http://www.humanamente.eu/
PDF/Issue27_Papers_Corballis.pdf (accessed 18 February 2016).
Evans, N & Levinson, SC. 2009. The myth of language universals: language di-
versity and its importance for cognitive science. Behavioral and Brain Sciences
32:429–492. doi:10.1017/S0140525X0999094X.
Fagan, BM. 1998. People of the earth: an introduction to world prehistory. 9th edition.
New York: Harper Collins.
Foley, WA. 1997. Anthropological linguistics. Oxford: Blackwell.
Goldschmidt, W. 1997. Nonverbal communication and culture, in Nonverbal com-
munication: where nature meets culture, edited by U Segerstrale & P Molnar.
Mahwah: Lawrence Erlbaum: 229–243.

174
 Leaky, R & Lewin, R. 1992. Origins reconsidered: in search of what makes us human.
London: Little, Brown and Company.
Lecours, AR. 1995. The origins and evolution of writing, in Origins of the human brain,
edited by JP Changeaux & J Chavaillon. Oxford: Clarendon Press: 213–235.
Lidz, J, Waxman, S & Freedman, J. 2003. What infants know about syntax but
couldn’t have learned: experimental evidence for syntactic structure at 18
months. Cognition 89(3):295–303.
Lieberman, P. 1975. On the origins of language: an introduction to the evolution of
human speech. New York: Macmillan.
Lieberman, P. 1991. Uniquely human: the evolution of speech, thought, and selfless
behaviour. Cambridge: Harvard University Press.
Loritz, DA. 1999. How the brain evolved language. Oxford: Oxford University Press.
Maryanski, A. 1997. Primate communication and the ecology of a language niche,
in Nonverbal communication: where nature meets culture, edited by U Segerstrale
& P Molnar. Mahwah, NJ: Lawrence Erlbaum: 191–209.
McMahon, A & McMahon, R. 2013. Evolutionary linguistics. Cambridge: Cambridge
University Press.
Miller, GA. 1956. The magic number seven, plus or minus two. Psychological Review
63:81–97.
Miller, GA. 1981. Language and speech. San Francisco, CA: WH Freeman.
Natural History Museum Publications. 1991. Man’s place in evolution. Rugby: Jolly
& Barber.
Papousek, H & Papousek, M. 1997. Preverbal communication in humans and the
genesis of culture, in Nonverbal communication: where nature meets culture,
edited by U Segerstrale & P Molnar. Mahwah, NJ: Lawrence Erlbaum: 87–107.
Pinker, S. 1995. Facts about human language relevant to its evolution, in Origins of
the human brain, edited by JP Changeaux & J Chavaillon. Oxford: Clarendon
Press: 262–283.
Sampson, G. 1980. Making sense. Oxford: Oxford University Press.
Scupin, R. 1995. Cultural anthropology: a global perspective. Englewood Cliffs, NJ:
Prentice Hall.
Tattersall, I. 2012. Masters of the planet: the search for human origins. New York:
Palgrave Macmillan.
Trask, RL. 1999. Language: the basics. New York: Routledge.
Turner, JH. 1997. The evolution of emotions: the nonverbal basis of human social
organization, in Nonverbal communication: where nature meets culture, edited
by U Segerstrale & P Molnar. Mahwah, NJ: Lawrence Erlbaum: 211–228.
Zihlman, AL. 1982. The human evolution coloring book. New York: HarperPerennial.

LIN2601/1/2017-2023 175
 Glossary

agrammatism the omission of articles and other function words in patients with
Broca’s aphasia.
analytic learning strategy a language acquisition strategy in which input is broken
down into its constituent parts and utterances are built up gradually from one-
word utterances to two- and three-word sentences.
antilocalisationists scientists who claim that the whole brain is responsible for
speech and language.
aphasia a language impairment resulting from a brain lesion caused by a stroke, a
tumour, an infection or a head injury.
Asperger’s syndrome a mild form of autism spectrum disorder associated with good
language skills but problems with socialisation, communication and appropriate
behaviour.
Autism Spectrum Disorder (same as autism) a lifelong developmental disorder as-
sociated with spectrum of symptoms and abilities including social withdrawal,
repetitive behaviour, and language difficulties.

babbling the stage of language acquisition from about six months old during which
children start to explore the abilities of their vocal organs.
baby-talk words simplified words and diminutives used by adults when speaking
to very young children.
behaviourists scientists who believe that learning is a result of imitation, association,
habit-formation and environmental shaping.
binocular vision the ability of both eyes to observe the same visual field in three
dimensions.
bipedalism the ability to walk upright on two legs rather than on all fours like other
primates.
bottom-up word recognition reading and recognising words by breaking them into
smaller units and mapping the letters onto the sounds they represent.
brain stem the structure at the base of the brain that controls vital functions such
as respiration and heartbeat, as well as auditory input.
Broca’s aphasia a language impairment characterised by damage to Broca’s area
resulting in slow, hesitant speech and problems with syntax.
Broca’s area part of the left frontal lobe associated with language production.

caretaker speech (same as motherese and child-directed speech) speech directed

at young children.
cerebellum the lower part of brain which regulates and maintains functions such
as balance and coordination of body movements.
cerebrum the upper part of the brain consisting of two asymmetrical hemispheres.
child-directed speech (same as motherese and caretaker speech) speech directed
at young children.

176
 CHILDES (Child Language Data Exchange System) computerised database of child
language data from thousands of research projects in many different languages
available at http://childes.psy.cmu.edu.
chunking dividing a complex piece of information into smaller bits in order to cope
with quantities of information beyond the capacity of short-term memory.
cognitive theory the theory that language is not separate from other aspects of
cognitive development and that language can only develop if cognitive skills
develop first.
communicative competence the ability to use language to communicate in ways
that are appropriate to the specific culture and context.
connectionist models a theory that views language development as a matter of
developing connections between neurons, strengthening correct patterns and
weakening incorrect patterns so that errors eventually disappear.
continuity controversy the question of whether human language should be con-
sidered as an extension of animal communication or as a completely different
system.
contralateral control the system whereby the left hemisphere controls the right
side of the body, while the right hemisphere controls the left side of the body.
corpus callosum the bundle of nerve fibres that connects the two brain hemispheres.
cortex the convoluted grey matter in the brain that receives and transmits messages
to and from all the sensory organs and controls all voluntary actions.
critical period the period during a child’s early years during which a first language
must be acquired in order for effective language acquisition to take place.
cross-linguistic research research that compares language acquisition in different
languages and cultures.
cross-sectional research studies that compare at least two groups of subjects at one
particular point in time.

developmental psycholinguistics the study of children’s early language as well as the

later language developments that occur during the school years and in adult life.
developmental psychology the study of the biological, social, emotional, and cogni-
tive changes that occur as children develop into mature adults.
dichotic listening test a way of determining which of the two hemispheres is domi-
nant for language using a two-track stereo recording that differs for each ear.
Down syndrome a genetic condition in which a person is born with an abnormal
number of chromosomes that results in specific physical traits, intellectual
disability and language difficulties.
dyslexia a reading disorder in which children have normal intelligence but specific
difficulty in learning to read and comprehend written text.

echolalia repeating the speech of others.

emergentism a theory of language development which sees language as resulting
from the interaction of social experiences and other more general learning
mechanisms which already exist.

LIN2601/1/2017-2023 177
 empiricists scientists who believe that knowledge is acquired through experiences
and information obtained through our senses.
Event-Related Potentials (ERPs) a non-invasive way of measuring what happens
in the brain during language processing using a cap fitted with electrodes to
record voltage fluctuations on the surface of the scalp in response to an external
stimulus.
evolution the natural development and adaptation of plants and animals over time
to become the modern plants and animals we know today.
expansion repetition of a child’s utterance in a grammatically correct form by an
adult, often with elements omitted by the child.
experimental research a controlled way of collecting data by asking subjects to
perform specific activities in a laboratory setting.
expressive children a group of children whose early vocabulary is made up of less
than 50% nouns and who use many pronouns and multiword personal-social
expressions.

fluency stage the reading stage during which children become faster and more skilled
readers who recognise more and more words and can effectively figure out un-
familiar words.
fMRI (same as functional magnetic resonance imaging) a non-invasive way of
obtaining images of the brain (or other tissues) by observing how magnetic
fields pass through it during different tasks.
function words language elements such as articles, prepositions, pronouns, auxiliaries,
conjunctions and plural markers that serve a grammatical function.

genetic saltation a sudden genetic adaptation in a new direction, giving rise to a new
species.
gradual evolution a succession of genetic adaptations over an extended time.
gestalt learning strategy a language acquisition strategy that focuses on whole ut-
terances without analysing them into smaller parts and results in longer speech
utterances at an early age, even if the meaning of each of the parts is not known.

holophrase a single word used by a young child to express a complex idea.

initial reading stage the stage during which children begin to understand how to map
sounds onto letters and learn to recognise common words and read simple text.
Innateness Hypothesis Chomsky’s version of nativism which claims that the uni-
versal properties that characterise language are part of the biological make-up
of all human beings.
input the language children hear.
intellectual disability a deficit in cognitive functions leading to delayed develop-
ment and a mental age that is lower than actual age.

178
 jargon aphasia a language impairment associated with damage to Wernicke’s area
and characterised by meaningless utterances.
joint attention sharing interest in an object or activity with another person.

language acquisition the informal process by which a child learns language in the
preschool years.
language learning the conscious or formal learning process that takes place when
people learn a second (or additional) language after puberty.
language universals problem the question of why all the languages of the world
seem to share a significant range of common structural properties.
larynx the voice box containing the vocal cords.
lateralisation the process whereby one of the two hemispheres becomes dominant
for certain cognitive functions.
linguistic stage the stage from about 1 year onwards when first words emerge and
children start using language to communicate intentionally and purposefully.
linguistics the discipline that studies the structure and function of human language.
localisation the attempt to identify specific parts of the brain that relate to specific
cognitive functions.
localisationists scientists who claim that language functions are located in specific
areas of the brain.
longitudinal research studies that involve repeated data collection from the same
subjects over time.

MacArthur-Bates Communicative Development Inventories (same as MCDI)

a standardised parent report that itemises children’s early words and language
use up to 30 months.
mean length of utterance (same as MLU) an objective way of measuring a child’s
linguistic development by counting the total number of words/morphemes
and dividing by the total number of utterances to obtain the mean number of
words/morphemes per utterance.
metalinguistic development the development of the ability to think and talk about
language, including defining words and correcting one’s own language errors.
morphemes the most basic units of meaning in a language.
morphology the study of the internal structure of words and the rules that are used
to build words.
motherese (same as caretaker speech and child-directed speech) speech directed
at young children.
motor cortex the area of the brain that controls our muscle movements.
multiple-word stage the stage of language acquisition when children start to put
more than two words together and start to learn the syntactic rules necessary
for creating complex sentences.
mutism failure to acquire spoken language.

LIN2601/1/2017-2023 179
 nativists scientists who believe that the ability to learn language is an innate or
inborn human ability.
naturalistic observation a method of data collection using children’s spontaneous
speech in natural everyday situations.
nature–nurture debate the attempt to find out how much of our language ability
is inborn and how much is shaped by environmental factors.
neologisms ‘made-up’ or nonsense words.
neurolinguistics the study of the brain mechanisms that underlie our ability to
acquire and use language.
neurons the brain cells that make up the building blocks of the nervous system and
that receive and transmit impulses.
nonverbal autistic a person with severe autism spectrum disorder who never acquires
language.

one-word stage the period from the child’s first word to the time when the child
starts to put two words together.
ontogenesis the origin and development of an individual or a characteristic feature
in an individual.
opposable thumb the ability of the thumb to touch the tips of any one of the other
fingers in primates.
overextension a lexical error in which a child’s range of referents is greater than that
of adult language.
overgeneralisation a syntactic error in which a child applies rules too widely.

PET-scan (same as positron emission tomography) a brain-imaging technique that

uses a harmless radioactive molecule injected into the bloodstream to observe
which areas of the brain are active during different tasks.
pharynx the area of the throat above the larynx.
phonology the study of the different kinds of speech sounds that occur in human
languages and the knowledge that speakers have about the sound patterns of
their particular language.
phylogenesis the origin and development of an entire species.
poverty of the stimulus the idea that the language children hear does not form a good
enough model for children to use as the basis for working out linguistic rules.
pragmatics the study of the appropriate use of language in different contexts.
preadaptations anatomical, neurological, cognitive and social adaptations that took
place before the emergence of language in the human species which, taken
together, made language possible.
prelinguistic stage the earliest period of language development associated with early
vocalisations like cooing, crying and babbling.
prereading stage initial experiences with print such as listening to stories, playing
with books, learning nursery rhymes, learning the alphabet etc.

180
 primate any member of the order Primate, which includes the monkeys, apes and
humans, typically having dexterous hands and feet, binocular vision, and a well-
developed brain.
protowords idiosyncratic early words invented by the child and used fairly consist-
ently to refer to a particular concept in an attempt to communicate.

rationalists scientists who emphasise the role of reason in acquiring knowledge.

reading to learn the reading stage at which children are sufficiently skilled to be able
to use reading to find out new information and comprehend unfamiliar ideas.
reduplication babbling vocalisations containing repeated sequences of consonants
and vowels such as baba or dadada.
referential children a group of children whose early vocabulary consists of a high
proportion of nouns (more than 50% of their vocabulary).

semantics the study of the meaning of words and sentences.

semi-structured elicitation a method of collecting language data by asking specific
questions or providing specific tasks.
social interactionism the theory that a child learns language by interacting actively
with his or her caretakers and language environment.
sodium amytal test a medical procedure that anaesthetises one hemisphere of the
brain in order to test for hemispheric dominance.
Specific Language Impairment (same as SLI) a language disorder in which children
have normal health, hearing and cognitive abilities but display language skills
below those expected for their age.
Sylvian fissure the deep cleft separating the temporal and frontal lobes of the brain.
syntax the study of the internal grammatical structure of sentences and how the
parts of a sentence are related to each other.

tag question a question formed by adding a short tag onto a statement, e.g. We are
going to the shop now, aren’t we?
telegraphic speech early utterances which omit function words and include only
the words that carry the main message.
top-down word recognition reading and recognising whole words without break-
ing them down into parts.
transcription a written version of spoken data.
two-word stage the stage of linguistic development, usually reached in the second
year, when children start putting two words together to form simple sentences.

underextension a lexical error in which a child’s range of referents for a word is

narrower than in adult language.

LIN2601/1/2017-2023 181
 Wernicke’s aphasia a language impairment characterised by speech that is fluent
and well-structured but often semantically meaningless and incomprehensible.
Wernicke’s area part of the left posterior temporal lobe associated with lan-
guage comprehension.
wh-question a question containing a wh-word like who, what, where that requires
an answer other than yes or no.
Williams syndrome a genetic condition that it is associated with specific physical
traits and a mixture of abilities and disabilities including language difficulties.

Yes-no question a question that requires the answer yes or no and are formed by
switching the position of the auxiliary verb and the subject, e.g. Can she swim?

182