An Introduction to Applied Linguistics

VP?’; offering, ‘Would you like X?’ and so on; these notional syllabuses developed 
at a time when linguistic interest had begun to shift to the communicative 
properties of language (Widdowson, 1979).
Various teaching approaches also draw on insights from these differing approaches 
to grammar. Approaches influenced by formal theories such as generative grammar 
tend to view language learning as rule acquisition and, therefore, focus on 
formalized rules of grammar. Those that evolved from functional considerations, 
known as communicative language teaching, view language as communication 
and tend to promote fluency over accuracy, consequently shifting the focus from 
sentence-level forms to communicative functions, such as requests, greetings, 
apologies and the like.
More recently, some applied linguists have argued for an approach that 
draws not on one or the other, but on both (Rutherford and Sharwood Smith, 
1988; Widdowson, 1989). Widdowson (1989) is particularly insistent that it is 
a mistake to concentrate solely on functional considerations while ignoring 
form altogether. He observes, for instance, that just as approaches that rely too 
heavily on achievement of rules of grammar often lead to dissociation from any 
consideration of appropriateness, so approaches which rely too heavily on an 
ability to use language appropriately can lead to a lack of necessary grammatical 
knowledge and of the ability to compose or decompose sentences with reference 
to it. There is, he says, ‘evidence that excessive zeal for communicative language 
teaching can lead to just such a state of affairs’ (Widdowson, 1989: 131). What is 
needed is an approach that provides a middle ground in that it neglects neither.
Newer linguistic theories that attempt to combine form and meaning (though 
they give less attention to appropriate use) are cognitive grammar (Langacker, 1987) 
and construction grammar (Fillmore, Kay, and O’Connor, 1988). Constructions 
integrate form and meaning at various levels of complexity from the morphology 
of words to phrases to clauses. An oft-cited example is that of English verb-
argument constructions (Goldberg, 2006). For instance, many English verbs enter 
into a pattern called the ‘ditransitive’ or ‘double object’ construction, in which 
the indirect object precedes the direct object following the verb. This construction 
entails the meaning ‘X causes Y to receive Z’, as in ‘Sam mailed Paul a letter’. 
When newer verbs enter into this construction, they inherit the semantics of the 
construction and force us to interpret the sentence in the same way. For example, 
‘Paul faxed Sam a reply’. It is important to note that, contrary to formal grammar, 
construction grammar takes the position that certain words fit certain patterns. 
In other words, it is not the case that any word will fill any slot in a construction.

---

22 An Introduction to Applied Linguistics
Pedagogical grammarians Celce-Murcia and Larsen-Freeman (1999) give 
strong support to the view that, in language teaching, a formal or functional 
approach should not be taken to the exclusion of the other. In fact, these authors 
recommend adopting a three-prong approach, including meaning as a separate 
dimension, along with those of form and function. They recognize that grammar 
is not merely a collection of forms ‘but rather involves the three dimensions of 
what linguists refer to as (morpho)syntax, semantics, and pragmatics’ (Celce-
Murcia and Larsen-Freeman, 1999: 4). They illustrate the importance of all three 
dimensions by means of a pie chart divided into equal and interconnected parts 
labelled ‘Form’, ‘Meaning’ and ‘Use’ (Figure 2.1). They feel this chart is useful 
as a conceptual framework for teaching grammar as it serves as a reminder that 
learners need not only to achieve a certain degree of formal accuracy, but that 
they also need to use the structures meaningfully and appropriately as well (see 
also Larsen-Freeman, 2001; 2003).
MORPHOSYNTAX/FORM
How is it formed?
SEMANTICS/MEANING
What does it mean?
PRAGMATICS/USE
When/why is it used?
Figure 2.1 Interconnected dimensions of grammar
Type versus Token
In terms of descriptive grammars, there still remain questions about what it is, 
exactly, that should be described. Descriptions of language will also have different 
outcomes depending on whether they account for types of linguistic elements 
in the abstract, or for tokens of linguistic elements as they actually occur in 
contexts of use. Descriptions that deal with forms in the abstract describe a range 
of category types, but those that deal with actual tokens (instances) of language 
use reveal more than category types: they also reveal the relative frequency of 
forms and their habitual co-occurrence in different contexts. Whereas a type 
description might present a broad array of structures and give each equal weight, 
a token description ‘might well reveal that some of these were of rare occurrence, 
or restricted to a realization through a limited range of lexical items, almost 

---

23
Grammar
exclusively confined to certain contexts, or associated with certain meanings’ 
(Widdowson, 1990: 75).
With the development of computers and computer analysis of language, token 
descriptions are now possible on a massive scale, and such descriptions have 
revolutionalized the way we view language (see Chapter 6, Corpus Linguistics). A 
well-known example is the COBUILD Bank of English Corpus, which contains 
more than 500 million words (mostly from written texts). Sinclair (1985) notes 
that type descriptions lacking attested data do not provide an adequate source of 
reference for language teaching. Instead, he believes that language for pedagogical 
purposes should be a projection of what actually occurs as recorded by the 
computer analysis of text.
Projects based on analyses of this and other corpus studies have produced various 
dictionaries and grammars, including the Collins COBUILD English Grammar 
(1990) of which Sinclair was Editor-in-Chief; Biber, Johansson, Leech, Conrad, 
and Finegan (1999); Hunston and Francis (2000); and Carter and McCarthy 
(2006). These grammars attempt to make statements about English, as attested by 
an analysis of patterns of words in linguistic corpora.
Discourse Grammar
Corpus studies have also led to an increased interest in analyses of ‘discourse 
grammar’, that is, analyses of the functional roles of grammatical structures in 
discourse. Here we are using discourse to mean the organization of language at a 
level above the sentence or individual conversational turn – that which connects 
language at the suprasentential level. In addition to the discourse context, there 
is also the influence that the non-linguistic co-text plays in the deployment of a 
speaker’s grammatical resources.
Speakers and writers make grammatical choices that depend on how they 
construe and wish to represent the context and on how they wish to position 
themselves in it (Larsen-Freeman, 2002). For example, speakers use the past 
perfect tense–aspect combination in English, not only to indicate the first of 
two past events, but also to give a reason or justification for the main events of 
the narrative. These events are not the main events themselves but, rather, are 
felt to be an essential background to what happened (see McCarthy and Carter, 
1994; Hughes and McCarthy, 1998). The italicized structure in the following 
excerpt, from an illustration given by Hughes and McCarthy (1998), occurs in a 
conversation between two young women who are talking about mutual friends 
from their days together at Brunel University.