September 2007

2007
Volume 11, Number 2

Top

Grammar, Meaning and Pragmatics: Sorting Out the

Muddle
Michael Swan
<swanmic googlemail.com>

Abstract

The term "pragmatics" is commonly used in two quite different senses. In

linguistic discourse, "pragmatics" refers to the strategies (exploitation of
shared knowledge, assumptions about communicative intent, etc.), by
which language users relate the dictionary/grammar meaning of
utterances to their communicative value in context. "Pragmatics" in this
sense deals with what is not encoded in language, and applies to all
language use. In language teaching, on the other hand, "pragmatics"
generally refers to the encoding of particular communicative functions,
especially those relevant to interpersonal exchanges, in specific
grammatical and lexical elements of a given language. Confusion between
the two senses leads to the common and mistaken claim that all the
structures of a language encode two levels of meaning, "semantic" and
"pragmatic," both of which must be learnt for communicative
competence. A further common claim, that earlier language teaching
failed to consider pragmatic aspects of language, is equally unfounded.

Two Kinds of Meaning

Since the early days of the Communicative Approach, language teachers have been
told that they have to pay attention to two kinds of meaning: the "semantic"
meanings of words and structures which can be found in dictionaries and grammars,
and the "pragmatic" values which these linguistic elements take on when they are
used in communication. In an influential paper published in the early 1970s (Hymes,
1971), the sociolinguist Dell Hymes put forward the view that "communicative
competence" involves knowing not only dictionary/grammar meaning, but also the
rules that determine the appropriacy or otherwise of utterances in context. This line
of thought was welcomed for several reasons. Language teachers at the time were
dissatisfied, as language teachers usually are, with their learners' inability to convert
their knowledge of linguistic forms into successful language use, and the idea that
they could solve the problem by teaching something called "communicative
competence" was an attractive one. The construct offered an engaging alternative to
the purely formal type of language competence investigated by Chomskyan
linguistics. It chimed well with the concerns of many applied linguists, who were then

TESL-EJ, Sept. 07 Swan 1

 turning their attention from language on the page to language between
people--discourse analysis was the new syntax. It added some intellectual kudos to
applied linguistics by forging a link with the work of linguistic philosophers such as
Searle and Austin, who were also looking, broadly speaking, at what people do with
language. And, quite simply, it fitted in with the "spirit of the age," which was deeply
concerned with interpersonal dynamics. So Hymes' pronouncement that there are
"rules of use without which the rules of grammar would be useless" (1971, p. 278)
became something of a mantra for the applied linguists of the time.

One of the major reasons for questioning the adequacy of grammatical

syllabuses lies in the fact that even when we have described the
grammatical (and lexical) meaning of a sentence, we have not accounted
for the way it is used as an utterance . . . . Since those things that are not
conveyed by the grammar are also understood, they too must be governed
by 'rules' which are known to both speaker and hearer. People who speak
the same language share not so much a grammatical competence as a
communicative competence. Looked at in foreign language teaching
terms, this means that the learner has to learn rules of communication
as well as rules of grammar. (Wilkins, 1976, pp. 10-11)

Despite the difficulty of clarifying what exactly might be meant by teaching "rules of
use" or "rules of communication" (see Swan, 1985 for discussion), the notion that
teachers and learners need to concern themselves with two levels of meaning has
remained prominent in pedagogic thought and writing. Celce-Murcia and
Larsen-Freeman, in a widely used course on pedagogic grammar, explain that
"Grammatical structures not only have a morphosyntactic form, they are also used
to express meaning (semantics) in context-appropriate use (pragmatics)" (1999, p.
4). Ellis (2005), discussing the need for learners to focus on meaning, similarly
distinguishes two senses of the term: semantic and pragmatic. Doughty and
Williams, discussing the rationale for the task-based teaching of structure, say:

[W]e recognize that the term meaning, which is often equated only with
its lexical component, in fact subsumes lexical, semantic and pragmatic
meaning. To be more accurate, we note that focus on form includes
forms, meaning and function (or use) . . . . We suggest that the degree of
effectiveness . . . of focus on form ultimately depends on the level of
integration of the learner's attention to all three aspects of form,
meaning and function in the TL. (1998, pp. 244-5)

Problems With the "Two Levels of Meaning" Notion

The idea that we should be teaching two kinds of meaning is so familiar that we can
easily fail to see how problematic it actually is. Stated in general terms, the claim
does indeed seem quite plausible. We know very well that the exact significance of an
utterance in communication can be different from the apparent meanings of the
words and structures involved. What do you think you're doing? is probably not a
simple enquiry about the hearer's mental processes. I thought we might go out for a
drink refers to the present, not the past. Problems arise, however, when we try to
focus on what exactly is meant, in practice, by saying that in general the structures of
a language have both semantic and pragmatic meanings, and that these can be
taught. Larsen-Freeman, while encouraging teachers to deal with both 'meaning and

TESL-EJ, Sept. 07 Swan 2

 use' in grammar (2003, pp. 35-42), concedes that for some structures the two can be
difficult to tease apart. They certainly can. Plural morphology encodes reference to
more than one entity: is this meaning or use? Whichever it is, where is the other? The
English future perfect refers to anteriority in the future; if this is meaning, what is
the use of the tense? The modal must can express both certainty and obligation; but if
one of these is meaning and the other use, which is which? Some grammatical
elements, indeed, do not seem to have any kind of meaning. Attributive adjectives
normally precede nouns in English and follow them in French, but it is hard to assign
any semantic value to this structural fact. English quantifiers are followed by of
before another determiner (some of those people), but not before a bare noun (some
people); again, it seems unrealistic to ask what this "means." And what is the
"meaning" of gender agreement in French, or of the verb-second rule in German?

Why?

If the idea that grammatical structures intrinsically have two distinguishable kinds of
meaning, "semantic" and "pragmatic," is fundamentally flawed (as I think it is), why
has it persisted for so long? I believe that this has a good deal to do with a
widespread and continuing confusion about what exactly is meant by "pragmatic."
Although the term "pragmatics," relating loosely to the study of "how we do things
with language," is pervasive in discussions of language teaching, it can be very hard
to pin down exactly what people mean by it, or how it relates to syntax, lexis and
semantics. Essentially, I think that this is because the term actually has two very
different kinds of reference that are often tangled up one with the other. Let's call
these "Pragmatics A" and "Pragmatics B."

Pragmatics A: What Is Not Encoded

When we encode an utterance, our hearer or reader can use dictionary/grammar

knowledge to decode it to the point of establishing its meaning in a kind of
general-purpose sense. But (as we are constantly reminded) the dictionary/grammar
meaning of any utterance underdetermines its meaning in context: its "value", or the
role it plays in the ongoing communication. To understand a sentence like Your
driver will be here in half an hour, a hearer needs to feed a good deal of extra
information into the utterance: the fact that in this instance the variable your refers
to the hearer him/herself; the exact identity of the driver in question; the location of
here; and the time frame within which in half an hour has to be calculated. None of
this information is encoded in the grammar and semantics of the sentence itself. The
sentence I still haven't forgiven her for the thing about the hedgehog and the music
stand depends for its interpretation on shared knowledge that is in no way expressed
by the language forms used. Correct interpretations of utterances can indeed take us
a very long way away from their surface encodings. In specific situations the
following sentences, for example, might be used to convey the messages shown in
brackets (or other very different ones), and be successfully understood as doing so.

Your coat's on the floor. ('Pick up your coat.')

Jane's got her exams on Friday. ('I can't come to lunch.')
Let's not have a repetition of last time. ('Don't get drunk and start flirting with
Melissa.')
It's Wednesday. ('Put the trash out.')

TESL-EJ, Sept. 07 Swan 3

 The linguistic discipline known as Pragmatics (what I am here calling "Pragmatics
A") takes this kind of problem as its principal subject matter. Textbooks on
pragmatics (e.g., Levinson, 1983; Yule, 1996; Horn & Ward, 2004) concern
themselves centrally with the principles used by speakers/writers and
hearers/readers to bridge the gap between code meaning and context-determined
meaning. Such works typically discuss, among other things:

the way context, shared knowledge, and familiarity with conventional

schemata, routines and genres all contribute to meaning, and are taken into
account in framing and interpreting utterances.
the "co-operative principle" underlying successful communication, whereby
speakers and writers normally avoid saying too much or too little, give true
information, say what is relevant, and aim at clarity of expression (the maxims
of quantity, quality, relation and manner identified by the philosopher Grice).
"implicature"--the ways in which the flouting of these maxims can be
interpreted (e.g., damning with faint praise, showing off, exaggeration for
emphasis, irony, metaphor).

By definition, Pragmatics A is concerned primarily with what is not encoded, and its
analytical categories are applicable to every act of communication: there is surely no
sentence which has an absolute universal value totally independent of the personal
and situational context in which it is uttered.

Pragmatics B: What Is Encoded

When language teachers talk about "pragmatics," however, they are not generally
thinking about the branch of linguistics discussed above, or the topics that this is
mainly concerned with. By virtue of being efficient communicators in their mother
tongues, students already know how to relate code to context so as to determine the
communicative intention of a given utterance, and the strategies and principles
involved are to a great extent universal and language-independent. Certainly, there
are some culture-specific differences--for example, one culture may value silence,
indirectness or the overt expression of respect more than another. But by and large,
Pragmatics A, dealing as it does with what is not encoded, is outside the scope of the
language classroom.

"Pragmatics," for language teachers, is to do with what is encoded. Languages do

not leave their speakers to grapple unaided with the problem of bridging the gap
between the dictionary/grammar meanings of utterances and their precise value in
communicative contexts. All languages provide ways of reducing the problem by
labelling, in general terms, the typical communicative roles that utterances can take
on. So we can encode linguistically the fact that that we are asking a question, or
expressing doubt, or adding information, or showing respect, or making an
objection, or exaggerating. It is these language-specific features--what I am calling
"Pragmatics B"--that are of direct concern to language teachers. Pragmatics B deals
with questions like the following:

How are common speech acts encoded in the target language? How does one
ask questions, make requests, express respect, invite, interrupt, etc?
Are there cross-language differences in the distribution of speech acts? Does L2

TESL-EJ, Sept. 07 Swan 4

 mark a logical contrast overtly while L1 leaves it to be interpreted from the
context? Does L1, unlike L2, have an answer form that indicates that the
speaker has asked a stupid question? Does one language go overboard for
respect while the other only makes a few gestures in that direction? Do
speakers necessarily have to use politeness markers in one of the languages,
but not the other, when asking a question; or is there a default interrogative
structure that will do unless one needs to be especially polite (as in English)?
What are the false friends? Does a negative question work for making requests
in L1 but not in L2? Does the equivalent of Have you eaten? mean "Have you
eaten?", or does it mean 'How are you?' or something else? Is the equivalent of
Please used in L2, but not in L1, when offering something or replying to
thanks?

Pragmatics B is not easy to delineate clearly. The kind of topics listed above are often
grouped in a general way under the heading "doing things with language." But where
does one draw the line? If making requests, issuing invitations and enquiring about
health are "doing things with language," then so presumably are defining things,
predicting the weather and talking about computers. Another common definition of
pragmatics in language teaching characterises it as being concerned with "the choices
that users of a particular language make when using the forms of the language in
communication" (Celce-Murcia & Larsen-Freeman 1999, pp. 4-5). But of course all
use of language involves choices: we do not use the same words to talk about cars as
to talk about flowers, or the same structures to refer to the past as we do to refer to
the future.

Whatever the problems of definition, though, language teachers and pedagogic

writers find it convenient in practice to use the term "pragmatics" mainly with
reference to two areas where the roles that utterances can take on are perhaps
especially in need of linguistic labelling. These are:

1. The construction of spoken exchanges, where interpersonal aspects of

communication involving respect, face etc are often key issues, so that
languages tend to have quite complex ways of framing utterances to ensure
that speakers' intentions are not misinterpreted.
2. The construction of written text, where failure to interpret the flow of
argument may cause misunderstanding, and where the use of devices such as
structuring conventions, anaphoric elements and discourse markers can
facilitate interpretation.

This, then, is the domain of Pragmatics B: not (as with Pragmatics A) the whole of a
linguistic system, but the subset of linguistic structures which encode the particular
types of meaning just referred to.

The Confusion

The common claim that all structures have two kinds of meaning, both of which we
need to teach, seems to derive from a simple confusion between these two different
versions of "pragmatics". Certainly, we can agree that all utterances have not only
dictionary/grammar meanings but also separate context-determined values in use
(Pragmatics A). However, we cannot legitimately import this generalisation into a

TESL-EJ, Sept. 07 Swan 5

 discussion of pragmatic encodings (Pragmatics B). Context-determined values are
features of utterances; they are not encoded in the structures used to create these
utterances, and do not therefore give rise to structure-specific 'rules of use' which can
be taught. English passives, for example, can be used to imply agency without stating
it directly, and this can be exploited for the pragmatic purpose of making veiled
criticisms. But, as Batstone points out we cannot relate this insight very closely to
specific contexts. While the sentence: The windows haven't been cleaned for months
may be uttered with the intention of causing embarrassment or hurt, "it would
appear odd to assert that the passive can be used to cause embarrassment or hurt in
domestic disputes concerning the cleaning of windows" (1994, p. 14).

A crucial point is that, as pointed out above, those pragmatic encodings that can be
taught (and which are therefore relatively context-independent) are not by any means
found in all the elements of a language. Pragmatics B, unlike Pragmatics A, has
limited and partial scope. This is easy to see with lexis. While some words and
expressions encode centrally pragmatic functions (for example Please, Dear Sir or
Feel free), others encode no pragmatics at all (for example dishwasher, marinate or
in time). Yet others have both pragmatic and non-pragmatic functions: this can be
used not only to indicate physical or temporal proximity, but also to clarify the
linking of items in text; certainly, as well as conveying definiteness, can label a
concessive move in an argument (Certainly, she did some good work at the
beginning. But . . . ). The situation is exactly the same for grammatical structures.
Imperatives perform a variety of pragmatic functions, labelling their associated
utterances as being for example commands, requests, or invitations. In contrast,
plural morphology in English nouns has no pragmatic significance: its function is
purely semantic. Plural morphology in French pronouns and verbs, on the other
hand, can encode either semantic meaning (reference to more than one thing or
person), or pragmatic meaning (where using the second-person plural to address a
single person expresses respect by metaphorically aggrandising the addressee).

If not all structures can be exploited for purposes that we might reasonably call
pragmatic, there are certainly very many structures which can be, as research by
construction grammarians and others is making increasingly clear--see for example
Goldberg (1995), Celce-Murcia & Larsen-Freeman (1999), Green (2004), Kay (2004).
Even where structures do have more than one kind of potential function, however, the
distinction between "meaning" and "use," or "semantic" and "pragmatic" meaning
seems somewhat artificial, and it is not always easy to operationalize. As pointed out
earlier, the two main senses of English must do not seem to fit neatly into the two
slots. For another example, take the past tenses in the following sentences:

1. I saw Oliver yesterday.

2. Only 18? I thought you were older.
3. If I had time I'd do a lot more reading.
4. I think it's time we went home.
5. If you had a moment, I wouldn't mind a bit of help.
6. How much did you want to spend?

We can if we wish describe the tense in 1) as having a "literal meaning," referring as

it does simply to past time, and the tense in 6) as involving a "pragmatic use" of the
past form for polite distancing. But what about the others? It seems difficult to
assign the various intertwined nuances of time reference, hypotheticality, and

TESL-EJ, Sept. 07 Swan 6

 interpersonal indirectness to one or other category. In the case of such polyvalent
items, it would be simpler, surely, to abandon the notion of a two-part division
between "meaning" and "use," and simply say that the structures have several
different functions.

Conclusion

Looking at the conceptualisation of pragmatics in language teaching that developed

during the communicative revolution of the 1970s and that is still very much with us,
I cannot help feeling that, in the words of the old cliché, what was true wasn't new
and what was new wasn't true. It has always been considered important for students
to learn how the language they are studying encodes the most important pragmatic
functions, even if the heavily structure-based pre-communicative language courses of
forty years ago were not always very good at teaching these systematically. The
interest of the last few decades in various dimensions of "language in use" has
certainly expanded our knowledge of how languages work, and our concern to
approximate real-life language use in the classroom has greatly improved our
teaching methodology. However, it is irritating to be told, as one often is, that the
pragmatic dimension of language, "language in use" or "grammar as choice," is a
recent discovery, unknown before the advent of the communicative approach, or even
(as is sometimes implied in contemporary academic and pedagogic writing) neglected
before the researches of the author and his or her colleagues.

Most grammars have focused on structure, describing the form and

(sometimes) meaning of grammatical constructions out of context. They
have not described how forms and meanings are actually used in spoken
and written discourse. (Biber et al., 2002, p. 2)

(Note the interesting suggestion that in the bad old days we did not teach even
semantic meaning most of the time, concealing from our students the potentially
inflammatory information that plural nouns refer to more than one entity, that past
tenses are prototypically used to refer to past time, and that forms like older and
more beautiful express comparison.)

Here is Gurrey, writing about language teaching in the mid-fifties:

[I]t is comparatively easy to memorize grammatical forms, but difficult

to master the usage of those forms. And it is of no value at all to know all
the Tense forms of a Verb, unless one knows also which Tense to use. It is
more useful to know when to say "Have you finished that letter?" and
when to say "Did you finish that letter?" than to know the Principal Parts
of the Verb 'to finish'. (1955, p. 71)

Billows, writing a few years later about how a teacher might approach the simple
present tense, gives an impressively long list of the various kinds of reference that the
tense can have in communication (1961, pp. 166-7). If it is felt that this is not exactly
"pragmatics," then consider the language functions covered in lessons 1-8 of a typical
structure-based course of the 1960s (Candlin, 1968). They include: greeting, enquiring
about health, leave-taking, thanking, expressing regret, eliciting and giving
information, offering, requesting goods and services, proffering, self-identification,
asking for more precise information, confirming what has been said, exhortation,

TESL-EJ, Sept. 07 Swan 7

 identifying and naming, agreeing to carry out instructions, and enquiring about
plans.

And here is the very first lesson of another popular English course of the 1960s
(Alexander, 1967):

MAN: Excuse me.

WOMAN: Yes?
MAN: Is this your handbag?
WOMAN: Pardon?
MAN: Is this your handbag?
WOMAN: Yes, it is. Thank you very much.
(Text accompanied by illustrations of a woman leaving her handbag on a
train)

It would be hard to find a better demonstration of how to integrate the teaching of

structure and pragmatic use at beginners' level.

As a glance at any history of language teaching (e.g., Howatt & Widdowson, 2004)
will show, "language in use" has been taught, well or badly, since languages were first
studied. The recent construct of a distinct "communicative competence," separate
from ordinary language knowledge and skills, that can be taught by focusing on how
all the forms of a language "are actually used in spoken and written discourse," is in
my view a chimera, based on a confused understanding of what is meant by
"pragmatics". It is certainly important to make sure that students understand the
various ways--pragmatic or not--in which the principal structures of a language can
be used, and that they become proficient in these uses. But to approach this goal by
encouraging teachers to try to identify and teach as separate items the "meaning"
and "use" of all grammatical structures is in my view to send them on a wild goose
chase.

About the Author

Michael Swan is a freelance writer specialising in English language teaching and

reference materials. His academic interests include descriptive and theoretical
grammar, cross-language influence, and the relationship between applied linguistic
theory and language teaching practice. He is a Visiting Professor at St Mary's
University College, Twickenham.

References

Alexander, L. (1967). First things first. Harlow: Longman.

Batstone, R. (1994). Grammar. Oxford: Oxford University Press.

Billows, F. L. (1961). The techniques of language teaching. London: Longman.

Candlin, E. F. (1968). Present day English for foreign students (4th edn.). London:

TESL-EJ, Sept. 07 Swan 8

 London University Press.

Celce-Murcia, M. & Larsen-Freeman, D. (with H. Williams) (1999). The grammar

book. Boston, MA: Heinle & Heinle.

Doughty, C. & Williams, J. (1998). Focus on form in classroom second language

acquisition. Cambridge: Cambridge University Press.

Ellis, R. (2005). Principles of instructed language learning. System 33/2, 209–224.

Goldberg, A. E. (1995). Constructions: A construction grammar approach to

argument structure. Chicago: University of Chicago Press.

Green, G. M. (2004). Some interactions of pragmatics and grammar. In Horn, L. &

G. Ward (eds.), The handbook of pragmatics, (pp. 407–426). Oxford: Blackwell.

Gurrey, P. (1955). Teaching English as a foreign language. London: Longmans, Green

and Co. Ltd.

Horn, L. & Ward, G. (eds.) (2004). The handbook of pragmatics. Oxford: Blackwell.

Howatt, A. P. R., (with H. G. Widdowson) (2004). A history of English language

teaching. Oxford: Oxford University Press.

Hymes, D. (1971). On communicative competence. In Pride, J. & J. Holmes (eds.),

Sociolinguistics, (pp. 269–293). London: Penguin.

Kay, P. (2004). Pragmatic aspects of grammatical constructions. In Horn, L. & G.

Ward (eds.), The handbook of pragmatics, (pp. 675–700). Oxford: Blackwell.

Larsen-Freeman, D. (2003). Teaching language: From grammar to grammaring.

Boston, MA: Thomson Heinle.

Levinson, S. (1983). Pragmatics. Cambridge: Cambridge University Press.

Pride, J. & Holmes, J. (eds.) (1971). Sociolinguistics. London: Penguin.

Swan, M. (1985). A critical look at the communicative approach (1). ELT Journal
39/1, 2–12.

Widdowson, H. G. (1978). Teaching language as communication. Oxford: Oxford

University Press.

Wilkins, D. (1976). Notional syllabuses. Oxford: Oxford University Press.

Yule, G. (1996). Pragmatics. Oxford: Oxford University Press.

TESL-EJ, Sept. 07 Swan 9

 © Copyright rests with authors. Please cite TESL-EJ appropriately.

TESL-EJ, Sept. 07 Swan 10