Neal Norricks' paper provides a lucid and convincing account of the extent to which ordinary language is far from ordinary and of how what is conventionally seen as literary language pervades many everyday language events. His work in this tradition is oriented largely to research in conversation analysis and is highly innovative because within the field of applied linguistics in general the range of work along parallel lines, though rich, is mainly focussed on written text (e.g. Cook, Language Play, Language Learning). In this regard I was especially impressed with Neal Norrick's attention to prosody and with his cogent demonstration that it is crucial to a poetics of conversation (see also parallel work on English and German data by Couper−Kuhlen and Guenthner).

Norrick covers a wide range of creative patterns in his analysis of conversational data and I find his analyses perceptive and revealing. He is in this paper particularly alert to uses of humour (see Norrick for a much fuller study) and bases most of his examples on naturally−occurring narrative events recorded mainly within family and generally within domestic settings. It is crucial to Norrick's method that he is able to utilise his own knowledge of the participants and of the context of the recordings as in so doing he overcomes one of the main difficulties in analysing such data: the danger of ascribing intentions and uses of language to speakers who may have not intended such effects as well as the danger of interpreting listeners' or co−conversationalists' responses in ways which may distort the data. In [→page 292] fact, Norrick is especially convincing when writing about methodology and his sensitivity to the need for constant alertness to contextual factors is eloquently displayed. I took away from his paper a clear sense of the extent to which literary uses of language in conversation are regularly co−produced and are not in any simple sense the work of a single, verbally gifted speaker, that, therefore, creativity in everyday talk is a natural social and interpersonal activity and that literary uses of language are not a capacity of special people but a special capacity of all people. In this regard Norrick's work is profoundly democratic.

My own work in this field complements that of Neal Norrick, though I have given less attention to prosodic factors and more attention to lexico−grammar and 'figures of speech' (ironically features which are all too rarely investigated in 'speech'). I too have worked with a corpus of naturally−occurring conversation (in my case the 5 million word CANCODE corpus)¹⁾ and I and my close colleague and co−researcher Michael McCarthy have adopted throughout a mainly sociolinguistic perspective on the data. The CANCODE data have been recorded in a wide variety of different social contexts so in addition to narrative we have examples of service encounters, work−place meetings, people engaged in tasks such as cooking or dressing a shop−window together, colleagues delivering a formal report at a group meeting as well as data from more intimate and family encounters. One conclusion we are drawing is that creative literary language use may not be not limited to any one social occasion but appears to pervade all areas of our corpus. However, the more familiar and informal the social context, the more likely it may be that such uses of language are co−produced by speakers or activated by an individual speaker. Our conclusions underline the importance of Norrick's findings that creativity occurs where risk is reduced, that is, when participants in a speech event feel relaxed and socially at ease with one another (see, in particular, Carter, Investigating English Discourse chs 6 and 8; Carter and McCarthy). Like Cook and Chiaro, we find ourselves stressing the elements of (re)creation or play more extensively, acknowledging the [→page 293] extent to which all literary forms depend on existing forms for their (intertextual) effects.

There is space only for a brief illustration from our corpus and not possible to cite the full version of the extract used because it runs to several minutes of recording. Analysis of the transcript accounts for what we call "professional and transactional information provision." The extract is selected for its contrast with the more intimate encounters explored by Neal Norrick and to underline that word play is pervasive and does not only occur in more intimate domains.

[Contextual information: The primary purpose of the meeting is an examination of the legal particulars of documents relating to Credit Security. The extract here is taken from the end of the meeting: <S01> manager: male (55); <S 02> company representative: male (40s); <S 03> company representative: male (40s). Speakers <S02> and <S03> report to the manager, speaker <S01>. The extract here occurs when the meeting is coming to a close]

<S 03> But the release now of savings is going to be an issue all right isn't it.
<S 02> Yeah.
<S 01> Yes.
<S 02> How is it approved. And can the board delegate that authority to somebody. To to release erm can, yeah that's right. Can the board delegate it?
<S 03> Well I [unintelligible] Well my reading of that would say that that is quite specific.
<S 02> Yeah.
<S 03> You don't know whether there's provision for the appointment of loans officers and credit officers and all this kind of.
<S 02> Mm.
<S 03> I wouldn't. There doesn't seem to be anything there except to say that the board must approve this.
<S 02> But but in accordance with the registered rules.
<S 03> [unintelligible]
<S 02> That's the only pos=, so it's, the question is thirty two three B. What's the inter=, can that, can the board delegate its authority under that section Geoff.
<S 01> Yeah.
<S 02> Thirty two three B.
<S 03> Or I wonder is that a limit according to the registered rules. Monitoring of it.
[→page 294]
<S 01> [unintelligible]
<S 03> [whistles]
<S 02> I know.
<S 01> I used to [unintelligible].
<S 03> [unintelligible]
[laughter]
<S 01> I used to think I was a pair of curtains but then I pulled myself together.
[laughter]
<S 01> I used to think I was being ignored but nobody still talks to me.
<S 02> [laughs]
<S 01> Cowardly [unintelligible] this morning. What was in the tea? It's that s=, it's that bloody foreign coffee [laughs] that's what is it.
<S 03> Foreign coffee?
<S 01> That's that foreign coffee [unintelligible].

After a long period of time in which documents are pored over and during which time the main purpose of the exchanges has been to transmit or obtain information, the meeting finally erupts in a kind of carnivalesque spirit, in which the speakers take a holiday from information transfer and joke and banter their way through to the end of the formal proceedings of the meeting (ll. 20−28). The business done, it seems, they are free to play with words and the labels for what is in their immediate environment. The speakers pun on the idea of curtains being 'pulled together' (a phrase which of course also means 'to put oneself in a better or more positive frame of mind') and this is in turn creatively extended when the same speaker jokes on the fact that in spite of this no−one talks to him. Other speakers then feel free to joke on stereotypes of what is 'foreign' being of inferior quality, referring in the process to an inability to distinguish between tea and coffee.

It is clear that context and interaction type restrict opportunities for such uses of language and in this particular instance an increase in creativity seems to coincide for all the speakers with points of release from their institutional roles. There are numerous similar instances in our corpus of a cline from informality to formality with creative humour and word play and joke telling being used for purposes of topic−switching, for use by a (work−related) superior to make others feel at [→page 295] ease and for language play and (re)creation. It is interesting, too, that the uses of humour are not unconnected with the institutional power of the speakers (note here that humour is initiated by the company manager and only subsequently picked up and developed by other speakers). Relationships between the gender of the speaker and their interaction with creative utterances are, however, still being explored. For example, it seems that female speakers sometimes use creatively marked language as a means to break into male−dominated talk. Indeed, in our data word play appears to be more common in female talk than in male talk. Men prefer rehearsed joke telling and often rather stagy sexual puns or, as indicated in the above extract, are often initiated by conventional locutions or preformulated sayings, e.g., "I used to think I was a pair of curtains but then I pulled myself together."

In fact, when compared with women, men may not be particularly spontaneously creative in talk, although women do appear to very successfully manage sexual banter, especially in all female company (Eggins and Slade). The Nottingham CANCODE team is also currently exploring how non−literal hyperbolic speech acts such as "Why don't you just cut my throat?" are used for humorous effect and how these speech acts are distributed according to different social and gender roles.

As I emerged from reading Neal Norrick's excellent paper I am left, however, feeling that there is still much to do to reverse existing paradigms in both linguistic and literary studies: namely, the deeply embedded paradigms that literary language has to be motivated against a background of non−literary language and that non−literary language is therefore by default of less value to us in reading the language of the world.

For example, in at least the following domains there remains much research to do: further studies of the talk functions of conventional poetic parts of speech (Cameron; Clift); fuller studies in relation to problem−posing and problem−solving practices in the work−place, not least in the areas of HIV and psychotherapeutic counselling where [→page 296] creative language choices can create paradigm shifts in awareness and perception and in the relationship between professional and patient (Candlin et al.; Garbutt; Ragan); more contextually−appropriate theories of value, especially aesthetic value (Armstrong); further cross−lingual and cross−mode studies building on data such as email⁄chat−room corpora but also looking more closely at the subtle creative relationships between the 'creative' and the 'critical'—that is, using poetic language for antagonistic, non−collaborative purposes (Rampton; Boxer and Cortes−Conde); taking fuller research cognisance of the different ways in which creativity is contextually and culturally shaped in and through language in different parts of the world (Fabb chs 9 and 10; Lubart).

In his work and not only in this paper, of course, Neal Norrick has, however, provided this research community with templates and insights for further exploration as well as a model for how literature and language, areas of work in poetics so often kept separate, may be brought into greater and more mutually beneficial synthesis.