Three is markedly different in form and purpose from the others. It presents short accounts of a series of empirical studies that were undertaken by the author, often with specialist collaborators, in order to explore and evaluate the claims made in Wray (2002b). Each study focused on one situation in which formulaicity played a role in communication, and in which that role was made more than usually evident by virtue of ‘pushing the boundaries’ of normal language use. Little theoretical discussion is offered in the accounts themselves, because Part Four goes on specifically to consider how the studies inform the five key questions listed above. Rather, Part Three can be viewed as a repository of brief factual descriptions of the features of each study, so that, in Part Four, it is not necessary to engage in lengthy digressions to fill in detail when discussing the emergent issues.
Chapter 10 reports how formulaic language was used in a machine translation device in the British Post Office. Chapter 11 looks at how software helps people with cerebral palsy hold conversations by storing turns prepared in advance. Chapters 12 and 13 outline studies of how language learners responded when challenged to memorize prefabricated sentences for later use. In Chapter 14 a court dispute is described: it centred on a formulaic dialect expression that was interpreted literally, turning it into a racial slur. Chapter 15 examines how, in a television comedy sketch, the impression of naturalness was affected by the fidelity with which the actors adhered to the original script.
Part Four returns to the five questions outlined above, drawing on the contextualization and theory from Parts One and Two, the studies reported in Part Three, and also other lines of evidence, including Harry Potter’s spells, flag signals and military bugle calls, language teaching in Papua New Guinea, Aviation English, communication in pre-modern man, cryptic crosswords, actors’ attitudes to their script, and the Cultural Revolution in China.
Finally, Chapter 22 reviews the results of ‘pushing the boundaries’, before demonstrating how an understanding of how formulaic language operates can result in practical applications. The chapter ends with some ideas for future research projects on formulaic language.
2
Conceptualizing formulaic language
Introduction
This chapter sets the scene for the exploration of the boundaries of formulaic language. Since boundaries are defined relative to a central ground, the main task is to lay out that ground. First, some observations about the nature of formulaicity are offered, as a means of introducing illustrative examples and demonstrating some of the typical linguistic features associated with it. Next, a theoretically-grounded definition of formulaicity is given. This definition reflects the conclusions drawn from the lengthy evaluation of evidence for various types of formulaic language undertaken in Wray (2002b).
Third, three key conceptual claims about the nature of formulaic language are presented. Each conceptual claim captures a different aspect of the theory developed in Wray (2002b).2 Inevitably one cannot, in a brief digest, re-present all the arguments and considerations that led to those conclusions, and the interested reader is urged to follow up on the references provided, to see why a particular view was taken. The first claim is that the mental lexicon is heteromorphic. This is an observation about how we process formulaic language. The second is that the content of the lexicon is determined through ‘Needs Only Anaslysis’, and it relates to how we learn formulaic language. The third claim, that morpheme equivalent units enable the speaker to manipulate the hearer, addresses the question of what we use formulaic language for.
Orientation: what is formulaic language?
While most linguists accept that there is such a thing as formulaic language, consensus about what it is exactly is severely limited. Underpinning the notion of formulaic language is the sense that certain words have an especially strong relationship with each other in creating their meaning – usually because only that particular combination, and not synonyms, can be used. For example, if something is beyond consideration we might say it is ‘out of the question’ but we cannot rephrase the description as ‘*external to the query’. Theories differ considerably regarding what it means that some words have a particular bond with others. Some accounts, including the one in this book, envision the string of words to have its own separate identity as an entry in the mental lexicon. Others focus on how we might create ‘rat runs’ in our processing, so that certain formulations are quicker and easier to produce than others with the same meaning. Others still are interested in identifying rules and constraints on them that will generate the formulaic expression. These different theories draw the line between what is formulaic and what is not formulaic in different places.
At the uncontentious end are what are often termed the ‘true’ idioms – a set of not all that frequent but particularly evocative multiword strings that express an idea metaphorically, such as ‘kick the bucket’, ‘red herring’, and ‘raining cats and dogs’. It is rare for a theory of language not to give such expressions a special status that pairs the wordstring with a holistic meaning, though even here attempts can be made to see whether parts of the meaning can be mapped onto parts of the form (see ‘What goes into the lexicon’, below). At the highly contentious other end are collocational associations such as, in academic prose, ‘fully developed’ and ‘highly complex’ (Biber, Johansson, Leech, and Conrad 1999: 565). Here, it is not that people doubt that the attraction between the words is particularly strong, but that theories differ as to how one best accounts for that attraction.
Between the two extremes come other types, including expressions that seem entirely detached from normal meanings and/or do not reflect the normal rules of grammar, for example, ‘by and large’, ‘come a cropper’; expressions that are metaphorical but, arguably, are less distant from the literal meaning than idioms, such as ‘set one’s store by’, ‘take stock’, ‘watch one’s back’, ‘take it’s/their toll’; expressions in which one or more words have a peculiar meaning, such as ‘do someone in’; a wide set of idiomatic turns of phrase, such as ‘do let’s v’; ‘you’d better v’; ‘not very well’ (in the sense of ‘ill’); and – central to some theoretical models but excluded from others – ’partly-fixed frames’ or ‘constructions’, complete phrases or clauses partly realized with specific lexical material and partly left open for interchangeable items, for example, ‘the end of the N’; ‘as a result of NP’; ‘the way in which CLAUSE’. These frames are often viewed as carrying their own meaning and pragmatics.
Interacting with the models of language patterning and use are certain assumptions about the ‘word’ as the fundamental building block of language – the implications of this view are explored in Chapter 5. Seeing the single word as a consistent unit of language, and thus viewing a two-word string as two units joined by a grammatical rule, gives a special status to what, in some cases, is a rather arbitrary result of historical practices. Why, for instance, is ‘out of two units when ‘into’ is one? Some words, including ‘into’ and also ‘although’ and ‘cupboard’, evidently used to be two words, and it is easy, in retrospect, to imagine that they probably had some properties associated with a single word before they actually became united into one in writing. But unless a pair of words is being written in different ways, a sign of transition (for example, ‘in(-)depth’, ‘what’s( )more’,
