An Introduction to Applied Linguistics

particular social and historical contexts. As Miller (1984) argues,
similarities 
in form and function are seen as deriving from the similarity in the social 
action undertaken. Thus, texts are looked at not only according to the textual 
regularities they display but also according to what class, gender and ethnic bias 
they incorporate, what discursive practices are constructed in the text, and, as a 
consequence of this, what social practices they reflect.
This new conception of genre in critical discourse analysis sees genres as both 
social and textual categories and no longer as fixed and immutable, but as dynamic 
and changing. As with the systemic accounts of genres, genres in critical discourse 
analysis are seen not only as a reflection of social reality but also as constructing 
social reality. Genres, therefore, not only arise out of the social context but, in 
fact, they shape the social context.
Grammar and Discourse: Spoken and Written 
Differences
In written discourses, writers can rely on readers to process the text in a logical 
and commonsense way. So, if a subject is not repeated in a co-ordinated clause, 
the reader simply assumes that the same subject applies:
We stood and gazed at the sea.
(Understood: We stood and we gazed at the sea.)
But because spoken discourse is usually so tied to its immediate context (unlike 
written texts which are often produced at one time and place to be read at 

---

64 An Introduction to Applied Linguistics
another), speakers usually have even less need to refer to everything that is in the 
context and can take for granted that listeners will know what is being referred 
to. This is often reflected in very short, reduced turns in informal conversation, 
where items normally required by the grammar of writing are simply absent. For 
instance, it is a general rule of English, that verbs in interrogative clauses must 
have a subject, and yet subjects and auxiliary verbs are often absent in questions 
directly referring to the listener(s):
Hi, Nigel, been working?
(Understood: ‘Have you been working?’)
A: 
Anybody want soup?
B: 
No thank you.
(Understood: ‘Does anyone want soup?)
Statements, too, often occur without a subject where the subject is obvious or 
may be assumed to be known:
Turned out well in the end.
(Understood: ‘It turned out well in the end.’)
Countable nouns sometimes occur without articles of any kind:
A: Nice 
restaurant.
B: 
Yes, it is, isn’t it.
(Understood: ‘It’s a nice restaurant.’)
These common features of spoken discourse mean that a grammar written 
solely on the basis of written texts, where such phenomena might be rare or 
completely absent, is incomplete. Equally, some structures which are common 
in writing may be very rare indeed in everyday conversation, for example 
non-finite -ing clauses in English. At the beginning of this chapter we used this 
sentence to introduce types of discourse:
As the day outside the home draws to a close, the members of the household will again 
come together, quite possibly sitting down for a joint meal with enough time to 
review the day and dream about the future.
This type of -ing clause is very rare indeed in informal conversation. So, once 
again, we can say that a grammar that fails to make the spoken–written distinction 
may be incomplete or even misleading, giving the impression that structures are 
equally common in speech and writing. A discourse grammar, since it derives 
its description from real contexts of use rather than from isolated or invented 
sentences, will necessarily be interested in the spoken–written divide wherever it 
is relevant. Carter and McCarthy (1995) give further examples of typically spoken 
grammatical phenomena, and argue that language teaching should take note of 
the differences, especially where skills are separated into speaking or listening 
skills and writing or reading skills, in syllabuses, materials and language testing. 
At least one major new descriptive grammar now offers wide-ranging information 
on spoken and written differences (Biber et al., 1999).
Lexical Patterns in Spoken Language
Discourse analysts are interested in how speakers’ and writers’ use of lexis creates 
patterns over longer stretches of text beyond the sentence (see Chapter 2, Grammar, 

---

65
Discourse Analysis
and Chapter 6, Corpus Linguistics). Here, we shall focus on spoken texts. Speakers 
make their lexical choices, and listeners receive and interpret them. It is clear from 
actual discourse contexts that the fixedness of meanings that we associate with 
dictionary definitions is open to negotiation, and that lexical meaning emerges 
from context, rather than being entirely
pre-ordained. For example, agreed 
meanings can be signalled and confirmed by repetition, where the listener repeats 
the speaker’s lexis:
Speaker 1: 
And then we went down to San Diego. Santa Barbara.
Speaker 2: 
California. Lovely.
Speaker 1: 
Yeah. Oh.
Speaker 2: 
Yeah.
Speaker 1: 
It was really beautiful.
Speaker 2: 
It’s a beautiful place.
However, a notable feature of conversation is the way speakers often trade 
approximate synonyms, rather than repeating one another, especially when 
exchanging subjective meanings, in an attempt to converge on agreed 
interpretations. In the next extract, lovely and so nice are matched:
Speaker 1: 
Alice where did you get that skirt? Cos I want one like that.
Speaker 2: 
Isn’t it lovely?
Speaker 1: 
It’s so nice.
Speaker 3: 
In Top Shop.
This phenomenon is known as ‘relexicalization’ (McCarthy 1988). Repetition and 
relexicalization enable conversational participants to converge on meanings, to 
negotiate them in particular contexts. Sometimes, more than one pair of lexical 
items is involved, and complex chains of lexical interaction may be observed, 
involving both repetition and relexicalization in the same stretch of talk:
Speaker 1: 
Ooh. Look at the sky.
Speaker 2: 
Oh that’s lovely.
Speaker 1: