Nadja Mifka-Profozi, PhD Second Semester 2013 Wednesday 12am 2 pm (March 20th)

What is aptitude for language learning? Early research on language aptitude Measuring aptitude: MLAT, PLAB Language analytic ability as part of aptitude complexes Evidence from research into aptitude complexes What is working memory? Reading span and speaking span tests Investigating effects of CF in relation to IDs

A cognitive factor /variable that affects L2 acquisition/ L2 learning: ability or propensity of an individual to learn a second or foreign language with ease It has been known for long that some individuals can learn a second language faster and with more success whereas others may experience considerable difficulties Audio-lingual and grammar-translation methods gave rise to the research on aptitude. On the other hand, CLT ignored it and abandoned for several decades

The first attempts to test language aptitude early in the 20th century, but the most complex work undertaken in the 1950s and onwards Research by Carroll and his colleagues in the late 50s and early 60s remained influential until today proved as the best predictor of achievement in both first and second language Carroll and Sapon (1959) authored the aptitude battery known as MLAT later only slightly changed but remained in use until today Probably the most robust language tests to date

MLAT - Modern Language Aptitude Test (four components) 1. Phonemic Coding Ability 2. Grammar Sensitivity 3. Inductive Language Learning Ability 4. Rote learning for foreign language materials, later changed to Associative memory

PLAB - Pimsleur Language Aptitude Battery designed by Pimsleur (1966) for younger learners (junior high school students) More emphasis on auditory component but lacks the test of grammatical sensitivity and has less strength in the memory component

 Ability to code unfamiliar sound so that it can be retained

Phonem ic Coding Ability

Gramm atical Sensiti vity

Capacity to identify functions of words in sentences

Capacity to induce rules and create new sentences

Capacity to form links in memory

Words in sentences (examples from MLAT) 1. The lake was dotted with SPEEDING boats. Sometimes (A) the very (B) best method for good (C) learning (D) is constant practice (E).

2. SEVERAL were absent from the meeting. In spite of (A) the many(B) proposals which(C)were made, only(D) one (E) could be adopted.

The role of aptitude as a cognitive ID variable cannot be ignored In spite of critiques [the anti-egalitarian flavour (Skehan), association with outmoded methodologies such as audio-lingualism, grammar-translation methods] research into language aptitude can help to identify and apply the best suited teaching methodologies IMPORTANT: Aptitude is not monolithic, so it is better to talk about aptitude complexes

Generally high correlations with achievement in L1 and L2, higher than anything else, equal to motivation (r = 0.4 0.6) Thus, aptitude and motivation two major ID factors in FL learning YET, there are some differences depending on learning conditions High correlations with standardised IQ tests and fluid intelligence BUT aptitude is not equal to intelligence it is entirely language related (Skehan)

Corresponds to Carrolls (MLAT) sub-tests: one related to the capacity to perform a topdown analysis, the other to the ability to extract abstract rules based on the given words and to apply these rules to unknown group of words (bottom-up process) High correlations with measures of fluid intelligence Tests designed to measure such an ability often use artificial languages

Glossary

Kau = dog Meu = cat Kau meud bo=the dog is chasing the cat Kau meud bi=the dog was chasing the cat So = watch Ciu = mouse

Pa = we, us Xa = you Pasau meud bo = Our dog is chasing the cat Pa meud bo = we are chasing the cat Paxbo = we are chasing you Pa meud bor = we arent chasing the cat

1. The cat was watching the mouse.

2. Your cat is chasing the mouse.

Meud ciu so Meud ciu si Meu ciud so Meu ciud si

b.
c. d.

Xacu meud bo Meuxa ciud bo Xaseu ciud bo Ciuxa meud bo

Wesche (1981) studied how to adapt instruction to account for IDs in aptitude-methodology matched to the type of learner (a) analytic or (b) memory oriented Reves (1983) - measured aptitude of L1 Arabic students learning: a) English formally b) Hebrew informally Aptitude scores predicted equally well in each situation Aptitude may be even more important in informal contexts as learners are left without assistance (Skehan 1998a)

Robinson(1995, 1997) examined the correlations between aptitude measures and performance in four conditions: 1. Explicitly instructed 2. Rule-search (inductive) 3. Implicit (no instruction) 4. Incidental (meaningrelated)

Significant correlations with the aptitude measures for all conditions except for the incidental one (where focus was on meaning). The study was replicated with the Samoan language the same results)

Harley & Hart (1997) showed that predictive qualities of different aptitude components change with age: early immersion students had stronger correlations with the memory components, older learners stronger correlations with language analysis sub-tests DeKeyser (2000) replicated Johnson and Newports study (1989) with Hungarian immigrants to US: a strong negative correlation between age of arrival and level of attained proficiency after the age of around 17 No correlation between aptitude and proficiency up to the age of 17 Correlation of 0.60 between aptitude and proficiency after this age

Until the late 1950s memory was treated as a single unitary faculty Brown (1958) in UK and Peterson&Peterson (1959) in US observed that small amounts of information would be rapidly forgotten if not rehearsed. They proposed the short-term (STM) memory system that operated on different rules than LTM In the late 60s and 70s the patients with amnesia were studied evidence for clear separation between LTM and STM (a short-term store, of limited capacity)

Baddely and Hitch (1974) proposed the three-part model of working memory, having conducted experiments with students using a digit span task simultaneously with the tasks demanding reasoning, learning or comprehending (processing and storage of information) Working memory for language may be one (if not the) central component of language aptitude (Myake&Friedman, 1998) Working memory is a temporary storage system that underpins our capacity to think (Baddeley, 2003)

Baddeley and Hitch (1974) the model in which temporary visual and verbal stores are controlled by an attentional system
Visuo-spatial sketchpad CENTRAL EXECUTIVE (Attention control) Phonological loop (Articulattory loop)

Phonological loop important for vocabulary acquisition, phonological awareness correlates with reading performance in 1st and 2nd language Visuo-spatial sketchpad: less easy to study, plays a role in our visual and spatial knowledge Central executive: the most important complex, but the least understood component of WM

Baddeley and Logie (1999) proposed that executive is purely an attentional system with no storage function. The fourth component is added: the episodic buffer assumed to be a temporary storage system that creates integrated representations based on information from perception Central Executive
Visuo-spatial sketchpad Phonological loop

Episodic buffer

Visual semantics

Episodic

LTM

Language

Human memory comprises a set of separable systems working memory is one of these providing a temporary storage to keep important information in mind while comprehending, thinking or doing. Working memory is limited. It is simultaneously a constraint during processing and the gateway to long-term memory, therefore impacting on potential change in long term memory

Working memory capacity determines how well individuals can use context to both comprehend and produce words. Reading Span Test subjects read aloud a set of sentences and then at the end of the set they have to recall the last word of each sentence (Daneman and Carpenter, 1983) Speaking Span Test subjects are presented with increasingly longer sets of unrelated words, at the end of the set their task is to use each word to generate a sentence containing that word (Daneman and Green, 1986)

Quasi-experimental classroom research comparing the effects of recasts and clarification requests (oral CF ) on the acquisition of French past tenses Two experimental groups and a control group Design: pre-test, treatment, immediate post-test, delayed post-test 4 weeks later Treatment tasks: information-gap, picture-based tasks, meaning-oriented, designed to elicit the use of target structures Results: overall superiority of recasts over clarification requests, for both structures, tested in written and oral mode

TEST

GAINS

LANGUAGE RECAST (N=16)

ANALYSIS (r) CL.REQ. (N=17) .44* (p < .05) .43* (p < .05) . 58** (p < .01)

Oral PC

Short-term Long-term

.41 .62** (p < .01) .41

Oral IMP

Short-term

Long-term Written PC Short-term

.07 .07
.06 - .05 - .58** (p < .01)

. 63** (p < .01) .67** (p < .01)

TEST

GAINS

WORKING RECAST (N=16)

MEMORY (r) Cl.REQ. (N=17) - .19 .03

Oral PC

Short-term Long-term

. 42 . 17

Oral IMP

Short-term
Long-term

. 34
. 07 - .61** (p < .01) - .37 - .13 - .28

- .02
. 06 .52* (p <.05) .42* (p < 05) .19 .07

Written PC

Short-term Long-term

Written IMP

Short-term Long-term