On the Subject of English

The Linguistics of Language Use and Learning

Henry Widdowson

• Publisher: De Gruyter Mouton
• Year: 2019
• ISBN:
  9783110617108
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: Linguistics
• Index: Languages & Linguistics

The globalized use of language calls into question conventional ways of thinking in linguistics, applied linguistics and language pedagogy.
This book critically examines this thinking from an historical, at times satirical, perspective and proposes an alternative conceptualization.
The first section defines a number of key concepts about communication which are taken up in subsequent sections and shown to be relevant to the different but related areas of language study. Issues about the relationship between linguistics and applied linguistics set the scene for a discussion of the nature of discourse, and then how this bears on the understanding of the globalised use of English as a lingua franca.The final section considers the implications of this perspective on communication for how the subject of English language teaching might be redefined.
The book is relevant for anyone who sees the need for a critical consideration of established concepts in linguistics and language pedagogy.

Section 1: Theoretical bearings

Preamble

As indicated in the introduction, much of the discussion in the papers in subsequent sections is informed by certain concepts, and conceptual distinctions (expounded more fully in Widdowson 2003, 2004) which represent my particular theoretical take on the nature of language use and learning. These are explained in this section as a set of personalized encyclopedia entries. This section is intended to serve as a kind of prospective priming or pretextual setting (see below) for the reading of the particular papers that follow, or as a kind of extended glossary that can be retrospectively be referred as and when the need might arise.

Text & discourse (revised from an entry in Hogan 2010)

A fundamental distinction needs to be made between an act of communication and the language which is used to enact it. The distinction becomes clear by considering the ambiguity of the word “speech”, as it occurs, for example in the title of the film “The King’s Speech” where the term refers both to the speech impairment of the sovereign, and to the public address he has to make. On the one hand, reference is made to the features of the language that is produced, to the physical textual manifestation of the spoken medium. On the other hand the reference is to how this is a discoursal realization of a mode of communication. The word used as a count noun would conventionally signal the second sense and not the first, a difference effaced in the ambiguous fusion of the film title. So speech as spoken text, as linguistic manifestation, has to be distinguished from a speech, an instance of spoken discourse.

But speech as spoken text does not, of course, only involve the physical articulation of speech sounds, but the production of forms that encode semantic meaning at lexical, morphological and syntactic levels of language. So text can be defined generally as the language that people produce in the process of communication. It is the linguistic expression of intended meaning, the overt trace of covert communicative purpose. In the spoken medium, this expression takes the form of audible signals produced by one or more participants, which are ephemeral and leave no trace unless recorded. In the written medium, the text is durable and is itself a participant record of intention. Spoken text is of its nature incomplete and dependent in that it is accompanied by other, paralinguistic, expressions of intended meaning. Much of it, as in conversation, is jointly produced on line by reciprocal interaction. Developments in digitalized communication now allows for such reciprocal interaction in the written medium, as in the online exchanges in text messaging, where aspects of the spoken mode are retained by replacing face-to-face paralinguistic signals by such semiotic devices as emoticons.

Text has no independent existence for language users: it is produced only to realize some communicative purpose or other, and is a reality to recipients only to the extent that they pragmatically activate it. As a linguistic product, however, it can be analysed in dissociation from the communicative process that gave rise to it. Over recent years, vast quantities of text have been collected in corpora and analysed by computer. Most of this has been written text for the obvious reason that this can be easily scanned and stored in its original form. Spoken text can only be analysed in this way in a derived version, if it is first transcribed, that is to say transformed into a kind of writing.

Text analysis by computer, as now extensively carried out in corpus linguistics, reveals properties of language usage in detail not immediately accessible to intuition. It provides profiles not only of the frequency of occurrence of lexical and grammatical forms, but patterns in their co-occurrence. It reveals idiomatic regularities in usage over and above those required by grammatical rule (Sinclair 1991a). Such analysis is often referred to as discourse analysis, and indeed the terms text and discourse are often taken to be synonymous, both referring to actual and attested language behaviour or performance as distinct from the abstract knowledge of the language code, or competence. Since competence has generally been defined as a knowledge of sentences, this has led some scholars to suggest that text analysis and discourse analysis are terminological variants, both referring to the study of language beyond the sentence.

There are difficulties about this conflation of text and discourse. In the first place, if texts are to be defined as naturally occurring usage, they often take the form of single and separate sentences, and even of language below the sentence. Examples would be public notices like KEEP LEFT, WAY OUT, DANGER, PRIVATE and so on. Though we can analyse these notices in terms of their formal properties as sentences or sentence constituents, this is not how we experience them as uses of language. We identify them as texts not because of their form, but their function, because we recognize that they are the expression of an intention to communicate. This seems obvious in the case of these simple minimal texts, but the same would apply to any texts whatever their linguistic form: food labels, recipes, menus, book reviews, newspaper articles and so on. We identify textuality by recognizing intentionality.

This does not mean that we recognize what the intentions are that are being expressed. I can identify a piece of language as a text, even as a type of text, without being able to understand what is meant by it – a public notice in a foreign language, for example, or a complex set of instructions in my own.

This is where a distinction between text and discourse becomes crucial. Discourse can be defined as the underlying meaning of the message: what the first person text producer means by the text, and what the second person makes of it. Texts are only a partial record of the intended discourse since, in their design, assumptions will be made about a shared context of knowledge and belief, and recognition of purpose on the part of presumed recipients that do not need to be made linguistically explicit. In many cases, like the public notices mentioned earlier and other text types of a basic utilitarian kind, these assumptions will be readily ratified, and then intention and interpretation will correspond so closely that it can lead to the mistaken supposition that pragmatic meaning is inscribed in the text itself.

The meaning that is inscribed in texts is semantic. Texts are made out of the semantic resources that are encoded in a language and as such will always provide indicators of the pragmatic intention of the discourse they textualize, but the extent to which these indicators can be acted upon will vary considerably and will to a large extent depend on how far the actual recipients correspond with the presumed recipients the text producer had in mind when designing it. Of course there is always likely to be some correspondence between the intended discourse and that which is derived from the text, or otherwise no communication would take place at all. The semantics of the text will always provide a basis for, and set limits on, pragmatic inference. But since a text is necessarily an incomplete record of what its producer means to say, the meaning inferred will always be approximate. It will also depend on how far recipients are prepared to be co-operative: they may choose to disregard intention indicators and derive a discourse from the text to suit purposes of their own. In some cases, especially with literary texts, intentions may be difficult to infer from textual evidence, and may indeed be considered irrelevant to interpretation. What the text means to the receiver then overrides whatever the producer might have meant by it.

So what discourse is interpretatively derived from a text depends on how the interpreter relates the text to a context of familiar knowledge and belief.

Context & co-text (revised from an entry in Hogan 2010)

The term context is used to refer very generally to the extra-linguistic circumstances in which language is produced as a text, and to which the text is related, the setting in which the language is used, for example, and the participants involved. But such circumstances are many and indeterminate and only when they relate to the text in the realization of discourse do they count as context. Many circumstantial features may have no bearing whatever on what meaning is intended by a text or how it is interpreted. The question is: how does one establish which attendant circumstances are contextually significant and which are not.

The importance of taking context into account as a matter of principle in the definition of meaning has been long established. Early in the last century, the anthropologist Bronislaw Malinowski argued that an understanding of how language functions as “a mode of action” depends on establishing a relationship with its “context of situation” (Malinowski 1923). Subsequently, the linguist J. R. Firth reformulated the notion as “a suitable schematic construct to apply to language events” (Firth 1957). This construct makes mention of “the relevant features of participants” and “the relevant objects”, but leaves unanswered the key question of how relevance is to be determined.

Context is a selection of those extra-linguistic features that are recognized by the language user as relevant in that they key into text to achieve communication. One set of criteria for determining relevance can be found in the conditions for realizing pragmatic meaning as proposed in the theory of speech acts (Searle 1969). A piece of text, the uttering of a particular linguistic expression, for example, can be said to realize a particular illocutionary force to the extent that situational features are taken to satisfy the conditions that define the illocution. The recognition of relevance comes about because language users are familiar with such conditions as part of their extra-linguistic socio-cultural knowledge.

But familiarity with illocutionary conditions is only one kind of socio-cultural knowledge that is brought to bear in the recognition of contextual significance. The world we live in is made familiar by projecting two kinds of order on to it: linguistic encoding on the one hand, and socio-cultural convention on the other. Communication involves an interaction between them: we make texts with the first with a view to keying them into the second. Socio-cultural conventions take the form of schemata: customary representations of reality in various degrees, culture-specific, modes of behaviour and thought which are socially established as normal. Contexts are features of a particular situation that are identified as instantiations of these abstract configurations of experience which are realized and recognized by users as discoursally relevant. These schematic constructs are not, however, static and fixed, since once they are engaged they can be extended and changed. Though communication depends on some schematic convergence to get off the ground at all, it can then develop its own creative momentum. Although context is generally understood as an extra-textual phenomenon, apart from text but a crucial concomitant to it, the term is also often used, misleadingly, to refer to the intra-textual relations that linguistic elements contract with each within text. An alternative, and preferable, term for this is co-text.

Co-textual relations occur between linguistic elements at different levels. At the morpho-phonemic level, for example, Labov shows the tendency for segments of spoken utterance to vary according to the phonetic and morphological environment in which they co-textually occur, and is able to specify variable rules for their occurrence. These are distinct from other variable rules that Labov postulates which have to do with contextually motivated variation – where speakers, intentionally or not, adjust their pronunciation in relatively formal situations in approximation to prestige social norms (Labov 1972).

Co-textual relations at the lexico-grammatical level have attracted particular interest over recent years in the field of corpus linguistics. Computers now provide the means for collecting and analysing vast quantities of text and for identifying in detail what regularities of co-textual patterning occur. One such pattern is that of collocation, the frequency of occurrence of one word in the environment of another. But co-textual patterning extends beyond the appearance of pairs of words in juxtaposition and is also manifested in word sequences of relative degrees of fixity. The identification of such co-textual relations has led to the recognition that text is essentially idiomatic in structure (Sinclair 1991a).

Cohesion & coherence

Whereas contextual relations have to do with the pragmatics of discourse, co-textual relations of this lexico-grammatical kind have to do with the semantics of text, with the inter-connection and mutual conditioning of encoded meanings which provide a text with its internal cohesion. As exemplified in Halliday and Hasan (1976), there are a number of linguistic devices that can be identified as having a cohesive function. One example is where one or more semantic feature is copied from an antecedent expression and carried over to those that follow. Thus, a pronoun like she would link cohesively with a noun phrase like the woman in white occurring earlier in a text in that it copies the features of singular and female. It should be noted however that the co-textual link of cohesion, being semantic, does not guarantee that the appropriate pragmatic reference will be achieved. There may be more than one antecedent to which the copying expression may semantically relate, or even if the semantic link is recognized, it may fail to indicate the referential connection because this depends on extra-textual schematic assumptions. In such cases one can only make sense of the text as discourse by invoking extra-textual contextual factors. Co-textual cohesive links, therefore, do not themselves result in referential coherence, which is contextually dependent and a matter of pragmatic interpretation. A use of language may be co-textually cohesive as text but contextually incoherent as discourse, and vice versa.

Pretext

Acts of communication occur in contextual continuity. Speech act theorists tend to describe the pragmatics of language use in terms of separate utterances each with its own propositional content and illocutionary force. But although this may be methodologically convenient, particularly since it allows for a correlation with sentences, it nevertheless is bound to misrepresent to some degree the indeterminate and cumulative nature of communicative process itself. The texts that people produce, in speech or writing are the realizations of a discourse that is related to the continuity of social and individual experience. Acts of communication are essentially expressions of that contextual continuity. They are projections from the past in that the discourse that first person speakers/writers (P1) textualize is informed by their own knowledge of the world based on previous experience including assumptions about what it is appropriate to say on a particular occasion. And they anticipate the future in that their purpose is to act upon a second person recipient (P2) in one way or another. So the textual realization of discourse intentions presupposes a pretext in two senses. On the one hand, they draw on a knowledge of preceding communicative contexts, knowledge which P1 assumes to be shared. On the other hand they are designed to serve a perlocutionary purpose. Whenever we engage in communicate activity, it is to have some effect on our interlocutors – to make a favourable impression on them, get them to see things as we do, to have them act or think in some way. In this respect, the primary purpose of communication is perlocutionary – to bring about some effect on the second person recipient. If this were not the case, it is hard to see why we would bother to communicate at all. We always have a communicative pretext.

And just as pretext is central to the textualization of the discourse intentions of P1, so it is also central to P2’s discourse interpretations of text. Whatever a P1 might intend to mean by a text, whatever effect it might be designed to have, is variously interpreted depending on the extent to which P2’s previous contextual experience corresponds with that of P1 and crucially how far P2 recognizes and ratifies P1’s pretextual purpose. Communication involves the approximate reconciliation of the disparity between what P1 intends to mean by a text and what the text means to P2. Where there is close pretextual correspondence, such reconciliation will be easy to achieve but in other cases, negotiation will be needed to bring the different interlocutor positions into whatever degree of approximate convergence is taken to be appropriate to the occasion. This negotiation of position in spoken interaction will in many cases be subject to constraints that restrict the individual’s room for manoeuver, most obviously in socially sanctioned unequal encounters where interlocutors have pre-assigned roles they are required to conform to. In writing, where no such overt negotiation is possible, P1 has to design texts which incorporate negotiation by proxy by presupposing and anticipating the interpretative positioning of the reader.

Co-operative & territorial imperatives

Since pra...

Table of contents

• Title Page
• Copyright
• Contents
• Acknowledgements
• Introduction: Only connect!
• Section 1: Theoretical bearings
• Section 2: Linguistics and applied linguistics
• Section 3: The analysis and interpretation of language use
• Section 4: English as a lingua franca
• Section 5: Linguistics in language learning and teaching
• References
• Name Index
• Subject Index