In this paper I shall assess the function of evaluative procedures in literary analysis and the role of linguistic analysis in the evaluation of a text’s aesthetic ‘worth’. For the purposes of such an investigation, I shall take it as axiomatic that there is such a thing as ‘literature’, despite the fact that it is notoriously difficult to locate ‘literariness’ in formal characteristics alone. As regards the term ‘linguistics’, there are difficulties too, but I propose here that linguistic analysis is based on the employment of a metalanguage in the description of the workings of the object language, whether those workings are directly shown as constitutive elements, as in morphology, or are sub-textual inferences, as in pragmatic meaning, to give two examples. By ‘evaluation’ I mean broadly the assessment of literary merit in terms of aesthetic texture; and there are two questions that relate most pertinently here. The first is whether there are objective criteria available for the assessment of literary worth; the second is in what ways the discipline of linguistic analysis (whatever we mean by that) might contribute to the formulation of a poetics of value. There are also two areas of analysis that are most immediately apparent – two sites where the object language is located: published works of literature, and creative writing by students in higher education courses. The latter, of course, might in some cases become also the former; increasingly creative writing courses are not merely about self-expression but about exploring opportunities in the market place.

My focus will be largely on creative writing in the classroom or seminar room, particularly in higher education. However, the problems, which are evident in the analysis of creative writing in the academy, are not so very different from those evident in the wider literature industry. In terms of critical approach, though, there is a great deal of difference, for the methodologies and metalanguages of the professional academic literary analyst are generally not those employed by a community of writers discussing and evaluating each other’s works. It is this sense of a community of writers – essentially a single stratum of writer-critics - which seems to characterise creative writing in the academy. The question that concerns me here is whether generalised abstract evaluative criteria are available in the community and whether, if they exist, they are applicable in the academy in the evaluation of students’ work. In tackling this issue I am conscious, as I suspect many are, of being pulled in two different directions. On a theoretical level and taking a broad historical sweep at the development of literary taste, I am inclined to believe that no such criteria are available and there are no objective, transcendent ways of evaluating texts. Such a sceptical position was the initial impetus for this paper. I ought further to make it clear that I am not simply a global cultural relativist, casting the net of suspicion over all attempts at objectivity and its attendants essentialism and generality. If not entirely attainable, objectivity is nevertheless a much-maligned necessary fiction. The other directional pull comes from critical practice, not critical theory (and this is an instance where the two do not happily coincide). As a linguist, literary critic, and (sometime) creative writing teacher, I exercise critical judgement over students’ creative writing - judgement that seems to fall in line with most others in the department, and with other departments across the country.

At some level, then, assessment of creative writing in the academy works. Successful creative writing courses have blossomed in the UK in recent years, and rather longer in the United States. The admission that creative writing in the academy is a success might sound an obvious and even patronising thing to say, but it is by no means a straightforward issue. It seems obvious that as professionals, we know what is good and what is bad, but it is at least a curiosity that what appears to work in the academy - i.e. that there is rough agreement on what is good or bad (in individual examples assessed) - obviously does not work in that other, more public, world, where reviewers are much less likely to agree. Is the academy just a nicer place to be than the big, bad world of reviews? Is there any sign of objectivity in this mass of broad agreement and wild disagreement?

Many would be unhappy, of course, with the term ‘objective’, while being unsure of what to replace it with. It might be the case that ‘objective’ is simply an unattainable ideal and that ‘plausible’ and ‘workable’ are more applicable. If this is so, then there are both practical and theoretical implications. If we find that there are no objective criteria applicable in the evaluation of textual worth, it need not sound the death-knell either for literary studies or creative writing courses, for it may be, as is frequently argued by writers, that the same problems exist in the evaluation of the academic essay in that objective criteria are equally hard to find. However, this does not so much rescue creative writing as condemn the academic essay to the same fate. Many creative writing teachers, particularly those who teach mainly ‘straight’ literature courses, attempt to dissolve the argument by suggesting that there are no purely objective criteria available, but that it does not matter because the same situation obtains in the assessment of conventional analytical essays. This line of argument – essentially that the academic essay and the creative writing piece are subject to at least analogous assessment criteria - is seriously flawed for one simple reason. In the academic essay there is a body of knowledge and its expression and exposition. In creative writing there is no body of knowledge - at least no body of knowledge that bears any relation to the content of the work. I’m thinking here of the relation between, say, historical ‘knowledge’ and the role of history in any creative text. It is not part of the text’s worth that the historical knowledge represented be accurate, unless, of course one routinely gets historical dates wrong in a text which is supposed to be ‘historically based’. Even then there are clearly many aspects of the writing which errors of this kind do not affect – the accuracy or inaccuracy of such material does not in any crucial way affect the aesthetic texture of the writing. In the academic essay accuracy is a specific criterion of value.

Perhaps one might argue that the formal characteristics of any particular genre represent a body of knowledge analogous to that represented by the ‘subject’ of an academic discipline. In the writing of a sonnet, therefore, that knowledge is represented by a demonstration of a correct replication of those characteristics that inhere in sonnets. The problem here is that formal or technical accuracy is not a guide to what is ‘good’ literature, although it may form part of our evaluation. In genres much less defined by set formal characteristics – prose fiction, for instance – those characteristics are even harder to define, let alone use as a measure of quality. I shall return to formal criteria in the discussion of linguistic approaches to evaluation.

Let us briefly consider the contemporary role of creative writing in the academy. First, it is undoubtedly enormously popular with students. Second, creative writers are likely to be an intrinsic part of those English departments that include creative writing, rather than just the odd visiting writer or writer-in-residence. In addition to undergraduate courses with large bits of creative writing, there are Masters courses, and even more recently the opportunity to submit a doctoral ‘thesis’ comprising part critical work and part creative. An increasing number of academics in English departments, too, are also creative writers. Never, it seems, have writers and critics been so close.

One of the problems with this Edenic view of the relationship between writers, critics and academics is that the thing that people on the outside would consider to be ‘what we do’, i.e. evaluate literary texts, is not explicitly taught in any obvious or structured sense. Although micro-evaluation continually goes on between writer and writer in the seminar, and between writer and tutor in the tutorial, explicit questions of value and evaluation form a tiny, often unacknowledged part of the undergraduate critical experience. In creative writing classes it is rare that explicit, generalised criteria are applied in analysis or evaluation, save those broadly assumed, such as ‘originality’, ‘linguistic ingenuity’ or ‘avoidance of cliché’ – a negative characteristic which in our post-modern world becomes a perilous edict. I am not talking here about criticism made of an individual work with whatever critical tools that come to hand. As I have suggested, criticism of creative writing in the form of ‘critical feedback’ either by peers or by teachers, is a crucial part of the creative writing experience. But nowhere is the generalised, essential vision of critical judgement expressed. There may be a "common pursuit of true judgement", to use F. R. Leavis’s (1952) phrase, but how such judgement is arrived at is notoriously difficult to spell out. This is not to say, however, that creative writing courses have no criteria, and that the whole thing is just a free-for-all. My contention is that we can evaluate students’ work through intimate knowledge of both the product and the process – the nuts and bolts of the writing experience – but when we step outside to posit generalised objective or quasi-objective criteria we find that they do not, indeed they cannot, work.

Most popular works of criticism, for example reviews, focus on the value of the textual object. Traditionally the academy has not done so - or at least evaluation has not been an explicit part of a pedagogical programme. F. R. Leavis’s programme (1952 and passim) following I.A. Richards was always both implicitly about value (specifically the value of literature in society) and the ways in which we can be critics and evaluate a text’s worth. One thing is for sure: students do not leave the University with the skills of a critic in the popular sense; they do not leave following three years work in the ‘common pursuit’ of ‘true’ judgement. The question remains as to whether the ad hoc critical practice which is part of the creative writing experience is to be preferred precisely because it has no essentialist criteria, or that the criticism undertaken is based on an implicit essentialism after all, but it is not spelled out.

There is a fairly long tradition that assumes that there are no specific criteria of evaluation that can be fixed a-priori (Lenz 1965: 7). Criteria have to grow out of the texts themselves and be formed by a group of powerful individuals into a cohesive cluster exemplified in canonical works. If no privileged criteria exist then evaluation becomes the responsibility of a group of individuals who share the same aesthetic and moral vision. In this case, what are essentially a-posteriori claims are abstracted and generalised into a-priori ones. This kind of approach is, broadly speaking, ‘intrinsic’ in that it appeals to criteria only on the text’s own terms and the ground of judgement derived from the nature of the text itself. . A similar, Aristotelian view is that a work shall be judged good to the extent that it fulfils the intrinsic imperatives of the kind to which it belongs. Both of these conceptions of value can be characterised as ‘generic-intrinsic’ in that they posit essential generic qualities and evaluate according to how much the text under consideration conforms to them. The problem with these kinds of intrinsic approaches to aesthetic worth is that they are notoriously circular. We have to know what a sonnet is before we can appeal to its intrinsic qualities; and what is a sonnet is only a generalisation of sonnets already written, for there is no perfect sonnet in some Platonic heaven.

A final variation on the intrinsic approach, very much that of the Anglo-American New Criticism, is the stipulation that a text is ‘good’ to the extent that it achieves what it is trying to do. E.D. Hirsch (1976) agrees in part with this view, but refuses to give all the credit to the text. For the New Critic the text itself displays (or fails to) its creative intention; for Hirsch there is no need to look for absolute certainty or objectivity in the assessment of a text’s worth. Such certainty is an illusion, although we can believe in something quite close to it. For Hirsch, objectivity is a straw man (nothing is truly verifiable), but relativism is a nonsense, too. The problem is that any assessment of ‘what a text is trying to do’ necessarily involves the kind of objectivity that Hirsch suggests we can do without. How do I know if a text has succeeded in doing something unless I am absolutely sure of the objective? Hirsch states:

There is of course a further, trivial problem for the writer: if I set out to write a cliched, third-rate novel I can probably succeed without too much difficulty and will therefore have produced ‘good’ literature. Hirsch does acknowledge this, however, and admits therefore that evaluation based on successful outcome of aims or intention cannot form a privileged criterion of literary value. But there are two very important issues attached to Hirsch’s claims. First, there is the claim that no interpretive statements can be made concerning literature if ‘certainty’ were a necessity. If this is the case it might throw us back to our earlier problem, that the instability of literary evaluation is matched by the instability of interpretation in literary essays.

The relationship between interpretation and certainty in literary analysis is an important one, and philosophers of language in particular have considered the ontological status of literary utterances. Much less has been said about the status of interpretive utterances. That Dorothea Brooke marries Casaubon is a ‘fact’, however trivial, about Middlemarch . That she is foolish for so doing is an interpretive statement, but it is not simply the case that there is an objective fact and a subjective impression or judgement of that fact. Through a close examination of the literary artefact certain judgements about structure, form and meaning can be made, despite the vagueness and ambiguity of language. But in the evaluation of the text, the literary object disappears, for there is nothing that we are trying to find out. Instead we can only ‘guess’ at the author’s intention and, as Hirsch points out, we may be correct. Or we may be incorrect, of course. The best that this kind of evaluative procedure can offer us, therefore, is a guess at what the author was trying to do before assessing whether he or she actually did it or not.

A different approach to literary evaluation is expressed in the so-called ‘extrinsic’ model of literary value, where a text is seen as ‘good’ on the grounds of its relationship with something outside it. This essentially Platonic conception of literature suggests that good literature is literature that is good for society: moral and exemplary. Formal characteristics of a text do not impinge upon the notion of value here at all, although it could be argued that ‘perfection of form’ is more likely to be admired by society. Another version of the extrinsic model is the idea that a text or other aesthetic creation is worth whatever people will pay for it: it has not intrinsic worth. For our purposes here, these arguments can be bracketed off, although this is not to say, of course, that literature is neither a function of the marketplace nor a part of the superstructure of society.

The whole issue of evaluation is more pressing at the present because the academy is in the grip of a culture that demands explicit pedagogical and assessment criteria. The problem for creative writing is that literary critics have shied away from evaluation in pursuit of formalist or historicist objectives, and creative writing teachers are much more likely to perform ‘in house’ evaluation, relevant to the particular task in hand, in response to the individual student’s work. Although Karl Popper (1965) maintained that in any investigation use is always made of undefined terms, the current culture has no use for them. However, to formally define every aspect of critical investigation is neither desirable nor possible. Certainly in the current obsessive culture of explicit and visible criteria matched with discrete learning outcomes, a dose of Popper is to be welcomed. But can we really define what is going on in assessing the value of a text and be explicit about our criteria and assume that those criteria are not merely ephemeral? Can that seeming most formal of disciplines, linguistics, be marshalled to provide some solid criteria that can be disseminated throughout the community? Cannot the literary text simply be the analogous object-language that the metalanguage of linguistics dissects? This kind of relation, between literary object and linguistic meta-language, has of course a long history (unfortunately most often referred to as ‘stylistics’ or ‘stylistic analysis’), but its history is once again one of formalist, not evaluative, criticism. Similarly, students’ literary-critical essays are concerned with a body of knowledge: there is an object and a metalinguistic enterprise. Criteria can be constructed on the basis of the relationship between that body of knowledge and those metalinguistic elements. In creative writing there is no object: the criteria are all about the manipulation of those analogous metalinguistic signs. Perhaps, then, there is a connection at last: linguistics is the analysis of sign-manipulation, the assessment of creative writing is the analysis of the manipulation of aesthetic texture. In both acts there is no ‘object’ beyond its expression. If they are not the same act, then they are at least analogous.

Let us investigate the formal apparatus of linguistics a little further. Despite its manifestly formalist origins , ‘literary linguistics’ has on occasions suggested that it might be the way to put aesthetic evaluation on a more sure footing. In its early days (see Ohmann1966, Widowson 1975, Leech, 1969 etc,) it sold itself as an antidote to ‘impressionism and relativism’, and although many have given up the quest for an objective poetics of value, it is by no means a closed issue. Some literary linguists feel that linguistics at least offers the best hope for objectivity. Michael Toolan (1996) states, though never really demonstrates, that stylistics can provide a way of evaluating literary excellence; that is, there is something in the methodology of linguistic analysis that can reach the aesthetic core of the literary text (something that is lacking, by implication, in other critical or evaluative approaches). The claim reminds us of the debate between Roger Fowler and F.W. Bateson in the early nineteen seventies (Fowler 1971) . Fowler, herald of the new, precise formalist-linguistic criticism with its admixture of British empiricism and continental theory, took issue with Bateson, defender of the "ineradicable subjective core" of the literary experience (which no amount of linguistic or quasi-linguistic terminology could help to locate or describe). I suspect that most students, and a fair percentage of staff, in English departments, are sympathetic with Bateson’s view. Even more recently Joanna Thornborrow, despite certain caveats, states that stylistics is as good as we can get in terms of an objective way of assessing the value of a literary artefact. Included in her list of key aspects of stylistics is:

• The discussion of texts according to objective criteria rather than according purely to subjective and impressionistic values (Thornborrow 1998: 4)

Such ‘objective criteria’ might be, for instance, the syntactic organisation of the text, which is seen as neither subjective not ephemeral, but uncontroversially and manifestly there in the text. Although it is not explicitly connected with the issue of evaluation one can visualise a possible analogous relation. If certain features are unambiguously there in the text, and if a text’s worth is a product of such features, then the more precisely we can describe them the closer we get to the objective aesthetic truth. Unfortunately, a text is not simply a function of its formal features - it is not simply a formal arrangement of signs. A text is a site both of meaning and form, and is transformed into discourse when construed by a reader or interpreter. Any greater ‘accuracy’ in the description of formal features – even semantic features – does not lead us into sounder critical judgement. Perhaps the way forward, then, is to take into account the reading process, the ‘process of construal’.

Such a process can be analysed within a pragmatic framework. Broadly speaking, pragmatics, as the relation between signs and interpreters and users (Morris 1946) locates the linguistic sign in a network of meaning-functions including context, the role of the addressee and cognitive domain. Pragmatic linguistic theories have been appropriated for the discussion of literary value, albeit on a very simple basis. Relevance theory is a case in point. The theory should not give rise to an interpretive methodology; it is a theory that attempts to account for text processing. Its basic premise is simple: in processing text, addressees or readers go for ‘optimal relevance’ the swiftest and most manifest interpretation possible. It is important to recognise that this is not a conscious decision played out in reading or interpreting, but a cognitive process to which we all are subject. Adrian Pilkington, at the close of a paper on relevance theory and metaphor states:

The problem with this approach as a way forward to evaluation is that it is completely counterintuitive. The common ground between some modern literary theorists and pragmaticists is the conception of the text as partly created by the reader. Immanent textual value, however, can only exist if the text is construed as stable and objective. If we switch our perspective to the reader – as many critics before the relevance theorists have done – all sorts of problems arise including old issues disguised as new ones.

For the relevance theorist textual complexity is manifested in a range of weak contextual effects and greater processing effort. A contextual effect is when some new information is processed in the light of what is already known. Pilkington states:

This is a cognitively-focused version of a simple truth that complex texts have a range of possible meanings open to them and in some senses are difficult to decode. However, it does assume a communication model of language in that the ‘default’ activity of the addressee is to get to the implied meaning of an utterance with the minimum of effort. This sounds like common sense, but reduces all language-interpretation to a narrow cognitive act. Pilkington suggests that by its very nature, the literary text requires greater processing effort, manifesting a greater range of weak implicatures (compare such a text to a note to the milkman – this has a single strong implicature of the request for milk). The weight of meaning is now firmly back with the text and Pilkington’s analysis does not take into account the different types of interpretive acts or reading experiences possible. The literary text (and, in fact, all texts) can be (and indeed often is) read in a manner quite contrary to the principle of relevance: go for the most difficult and stubborn interpretation; guarantee vagueness, obscurity.

Now, perhaps the theory can still be tied in with the more formal aspects of linguistics. We might say, for instance, that formal complexity is more likely to give rise to a range of weak implicatures . Putting aside for a moment the question of what precisely formal complexity might be we could posit, say, a certain syntactic complexity as one possible criterion. The later Henry James, the later James Joyce, T.S.Eliot, and the modernists would easily fit in here. It is obvious, though, that formal difficulty as one mobilisation of weak implicatures is not a sound criterion for assessment of aesthetic value, for formal complexity is not equivalent to the kind of cognitive complexity that the relevance theorists and others wish to pin down, even if ‘complexity’ itself were a valid expression. Clearly, the kind of complexity evident (or not so evident) in the poetry of William Carlos Williams is quite different from that of Finnegans Wake. The manifestation of a range of weak implicatures cannot, therefore, be attributed solely to some arbitrary formal characteristic. If ambiguity or obscurity are markers, too, of a range of weak implicatures there is no necessary link to value. Indeed, as Umberto Eco(1979) has shown, there is something about the texts of Joyce and other modernists which is ‘closed’ rather than ‘open’ in that they prescribe their readerships with a regulated literary competence. Indeed, to press this further, one might say that because literary texts are generically prescribed, the range of contextual effects is smaller. I might not know what a lyric poem ‘means’, but I know the kinds of things that it is trying to do.

Let us look specifically at assessment and evaluation within the academy. In this case I am abstracting some generalised criteria which I see implicit in creative writing courses. Let us also return to those formal aspects – that is, aspects of the text that can be assessed (to a degree) without reference to content . These formal aspects of writing are taught and assessed, and it seems that we might be on safer ground here. After all, if you don’t get the iambic pentameter right, your blank verse is not going to be very good. But what exactly is ‘getting it right’? Is it not absurd to say that we cannot objectively tell the difference between this:

Midwinter spring is its own season
Sempiternal though sodden towards sundown,
Suspended in time, between pole and tropic.
When the short day is brightest, with frost and fire,
(T.S. Eliot, "Little Gidding" Four Quartets, 1944)

And this:

All hail to the Rev. George Gilfillan of Dundee,
He is the greatest preacher I ever did hear or see.
He is a man of genius bright,
And in him his congregation does delight.
(William McGonagall, "Address to the Rev George Gilfillan")

Possibly. Even if it were the case that we were absolutely sure of our criteria for evaluation, we are not dealing with Eliot, an historical figure, public and academic ‘poet’ who is very much part of the canon of English literature, and McGonagall. We are dealing with (mostly) lower-middle and middle-class students, all with roughly the same life-experience, all with roughly the same writing skills – if we take a very broad picture. This is an absolutely cricual point. If we say there is a difference between Eliot and a first year undergraduate creative writer it does not logically imply any difference between kinds of first years. Of course, it seems counterintutitive to say that there is little variation in students’ writing, but in one stratum it is clear that this ia so. There is what might be called ‘good’ writing and what might be called ‘bad’ writing, by reasonably common consent (Eliot versus McGonagall), but there are vast amounts of student writing which can not be pushed to either of these extremes. This work is precisely the kind of work teachers of creative writing have to deal with for the greater part of their critical lives. It is here that workable criteria for evaluation seem unavailable.

If creative writing is taught, then there must be a body of knowledge to teach and skills to acquire. The status and function of that body of knowledge are quite different to that of other knowledges, for they are observed only when instantiated in a creative act. Certainly on a formal level this is true: we can teach poetic form, the mechanics of speech and thought representation and other aspects of literary writing. The problem is that these forms are never simply ‘there’ decontextualised from the other aspects of writing in which they are embedded, so one can never analyse pure ‘form’. Further, writing seems to be a skill that is ‘acquired’. Is this notion of acquisition the same ‘acquired’ as in ‘an acquired taste’ or even the ‘acquired’ of ‘language acquisition’? The problem again is that this term suggests something that is not learnt but simply ‘picked up’. I believe this to be partly the case with writing. Just how do you ‘get better’ at writing – that is, not formally better, but ‘aesthetically better’, whatever that phrase may suggest? The only criterion I have found almost universally agreed upon is stated by I.A. Richards, and is borne out by studies of children’s writing by Katharine Perera (1984). Richards states:

Notwithstanding the potential vicious circle of having to find out what is ‘good’ – it seems to be agreed that to get better as a writer one needs to read examples of good writing. In this way, creative writing skills are said to be ‘acquired’. This throws us back once again to our initial evaluative criteria based on authority, rather than principle: a body of canonical works is established by a powerful group of individuals and implict principled criteria arise from the circulation of such works.

Creative writing in the academy is characterised by the view that writing is a continuous process of refinement and self-reflection, of editing and revising. If this were not the case, then the whole notion of the improvement of skills and thus the production of ‘better’ writing, would be misguided. As a pedagogical vision, this seems to make absolute sense: we could not possibly teach if there was no sense that something could be learned. However, this conception of writing is only one of a range of possibilities, and it very much a product of social and historical forces. One might just as easily build a writing philosophy upon ‘instant’ writing and non-reflection. The other crucial point is that in terms of creative writing, the writer him or herself is often not the best person to reflect upon and change the work. As Northrop Frye demonstrated in 1957, a writer is often a poor critic of his or her own work. This is not just a statement of contingency - it is a theoretical impasse from which there is no escape.

Finally, I suggest that the way we see creative writing is very much a product of its times. There are no eternally valid privileged criteria for evaluation and assessment of creative writing. Perhaps a compromise solution can be reached – indeed, such a solution already seems to be working within the academy, and was expressed long ago by Welleck and Warren:

My conclusion about the criticism and evaluation of creative writing is a pragmatic one: ad hoc discussion and evaluation of writing, without solid, objective criteria but with assumed common aims and objectives is the only way that creative writing can continue. In this sense, writers and tutors support each other until the marketplace looms, although increasingly courses are geared towards the act of publication. Creative writing in the academy is characterised by the sharing of ideas, self-reflection and a sense of common identity; and these are not bad things to have in a climate of fragmentation of disciplines, obsessively observable learning outcomes and knowledge as currency.

[If you would like to submit a response to this piece , please e-mail s.l.earnshaw@shu.ac.uk. If you would like to e-mail the author direct, the address is: . Readers' comments will be posted in December and the author's response to these comments will be posted in January 2001].

.

References

Eco, U. (1979) The Role of The Reader: Explorations in the Semiotics of Texts New York: Hutchinson.

Fowler, R. (1971) The Languages of Literature: Some Linguistic Contributions to Criticism London: Routledge and Kegan Paul.

Fry, N. (1954) Anatomy of Criticism Princeton: Princeton University Press.

Green, K. (1997) "Butterflies, Wheels and the Search for Literary Relevance" Language and Literature 6.2 132-38.

Hirsch, E.D. Jr. (1976) The Aims of Interpretation Chicago: University of Chicago Press.

Jakosbon, R. (1957) "Closing Statement: Linguistics and Poetics" in Chatman, S. and Levin, S (eds) (1967) Essays on the Language of Literature Boston: Houghton Mifflin 296-322.

Lenz, J.W. (ed) (1975) On the Standard of Taste and Other Essays Indianapolis: I.U.P.

Morris, C. (1946) Signs, Language and Behaviour New York: Brazilier.

Leavis, F.R. (1952) The Common Pursuit Oxford: Clarendon Press.

Leech, G. (1969) A Linguistic Guide to English Poetry London: Longman.

Ohman, R. (1966) "Literature as Sentences" in Chatman and Levin (eds) (1967) 231-238.

Perera, K. (1984) Children’s Writing and Reading: Analyzing Classroom Language Oxford: Blackwell.

Pilkington, A,. (1990) "A Relevance-Theoretic View of Metaphor" Parlance 2.2 102-117.

Popper, K. (1965) Objective Knowledge London: Routledge and Kegan Paul.

Richards, I.A. (1919) Practical Criticism Cambridge: Cambridge University Press.

Thornborrow, J. (1998) Patterns in Language London: Routledge.

Toolan, M. (1996) Language in Literature London: Arnold.

Welleck, R. and Warren, R. (1949) Theory of Literature London: Jonathan Cape.

Werth, P. (1999) Textworlds: Representing Cognitive Space in Discourse London: Longman.

Widdowson, H. (1975) Stylistics and the Teaching of Literature London: Longman.

COMMENTS

If you would like to send comments to Working Papers, please e-mail s.l.earnshaw@shu.ac.uk Comments will be published at the end of January, and the authors' replies at the end of February.

If you would like to contact Keith Green directly, his e-mail address is: k.m.green@shu.ac.uk