My typification of composing strategies is based on an extensive review of the published self- reports of writers. For obvious reasons, such reports tend to be mainly those of ‘literary' writers: novelists, poets, playwrights, essayists and biographers, although I have also assembled a collection of the rather more scattered accounts of academic writers. I have also surveyed 107 academic writers, 86% of whom reported frequent use of one or more of the strategies derived from these sources (Chandler 1992 & 1993).
The metaphor of the writer as 'architect' is prominent in Neo-Classical literary theory, emphasizing, of course, conscious planning and design (Abrams 1953, pp. 166-7, 201). I have not uncovered a direct reference to the strategy as 'architectural' by modern writers who use it (perhaps an indication of how pervasive Romanticism still is), but Gertrude Stein does refer dismissively to writing which comes 'out of an architectural drawing of the thing you are doing' (Ghiselin 1952, pp. 159-60). Of course, no such dismissiveness is intended here.
Academic writers in my study who used this very common plan-write-edit strategy reported that they consciously chose their writing strategies. They were less likely than others to see writing as a way of thinking. Whilst most writers agreed that they wrote better when concentrating on the topic rather than on the way they were writing, frequent users of this strategy showed this tendency more strongly than other writers. All this suggests a 'rationalist' approach. However, they showed an implicit awareness of the role of the unconscious insofar as they exhibited the strongest tendency amongst all groups to think that it helped to leave their evolving texts and to return to them later. They showed less of a sense of writing as intrinsically rewarding than other writers did; they were perhaps among the most pragmatic writers. They showed a slight tendency to be interlinear editors: that is, to add annotations between lines of text on paper. Those who were word processor users showed a far stronger tendency than other writers not to find the size of the word processor screen restrictive.
I encountered examples of the metaphor of bricklaying in my review of writers' accounts of their composing styles. Most explicitly, William Zinsser said: 'I have to get every paragraph as nearly right as possible before I go onto the next paragraph. I'm somewhat like a bricklayer: I build very slowly, not adding a new row until I feel that the foundation is solid enough to hold up the house. I'm the exact opposite of the writer who dashes off his entire first draft, not caring how sloppy it looks or how badly it's written' (Zinsser 1983, p. 97).
In my survey, academic writers who frequently employed a sentence-by-sentence strategy were also very likely to work on a paragraph-by-paragraph basis. Their approach was, of course, largely sequential and the correction of linguistic slips tended to be done mainly as they wrote. They showed a stronger tendency than other writers not to complete a draft in a single session. They showed some tendency to feel that it helped to leave a piece of writing and come back to it later. They usually had a clear idea of what they wanted to say and strongly disagreed that thinking would be difficult without writing. They tended not to agree that the more they were concentrating on the topic rather than the way they were writing the better their writing was. They showed a stronger preference for handwritten letters than did other writers, and tended not to use the word processor (those who did so showed a strong tendency to find the screen size restrictive).
Bricklaying can be a slow process, and Malcolm Cowley referred to writers working in this way (such as William Styron) as 'bleeders' - perhaps because he didn't work this way himself (Cowley 1958, p. 20). Those who employ it often refer to how different it is from the water-colour strategy, and in my survey its use was most closely allied to the use of the architectural strategy. It may be worth remarking that the image of the writer as skilful and intelligent 'artisan' is closely allied with that of the writer as architect in Neo- Classical literary theory (Abrams 1953, p. 166). Many writers who use this strategy may report - because they do only one complete draft - that they do little revision, even though they rework each chunk of text a great deal before proceeding to the next. Such remarks can be misinterpreted by writers who use other strategies, as Tom Robbins revealed in describing his use of the bricklaying strategy:
Tom Robbins clearly distinguished the bricklaying strategy from his idea of what other strategies involve (apparently a combination of the water-colour and oil painting strategies).
Painting done in oils is reworkable over time in a way that painting with water-colours cannot be: in oils, one may paint over details in a way that would quickly become 'muddy' with water-colours. The surface of an oil painting typically has what is referred to as a 'painterly' texture: revealing the marks of the making. Several examples of the metaphor of oil painting appeared in my review of the published self-reports of writers. The novelist Kurt Vonnegut reported: 'Usually I begin with several ideas, start playing with them. They are authentic concerns about things in life that bother me. One way of my dealing with them is in writing. I play with these ideas until they start to feel right. It's something like oil painting. You lay on paint and lay on paint. Suddenly you have something and you frame it... It's like watching a teletype machine in a newspaper office to see what comes out' (Strickland 1989, p. 134). And Alberto Moravia used the same metaphor: 'Each book is worked over many times. I like to compare my method with that of painters centuries ago, proceeding, as it were, from layer to layer. This first draft is quite crude... After that I rewrite it as many times - apply as many 'layers' - as I feel to be necessary' (Cowley 1958, p. 196).
This is a minimal planning and maximal revision strategy. Those in my study who used this strategy frequently showed a strong tendency to write to understand better what they thought. Nearly half of them reported that they did not consciously choose their writing strategies. These writers were, of course, major revisers, and they often deleted a lot too. As one would expect, they were less likely than others to work sequentially. It is interesting that many were in their 20s and 30s, and did not remember being given any guidance about how to approach composition. I wondered whether some writers abandon this strategy as they mature or whether the older generation simply did not grow up using it. Perhaps this pattern reflected dominant educational ideologies in British schools about 10-15 years ago, when 'drafting' was a widespread feature of educational rhetoric (to some extent reflected in educational practice), in contrast to an earlier obsession with pre-planning. There was some tendency for frequent users of this strategy to agree that their writing was better the more they concentrated on the topic rather than on the way they were writing. They exhibited a strong tendency not to mind talking about work in progress, and also to feel that it helped to leave a piece of writing and to return to it later. As for their use of writing tools, they were evenly divided over whether handwriting was too slow for them (other writers tended not to find it too slow). They were much more likely than other writers to be interlinear editors. In my survey the word processor showed up as being most frequently used by oil painters: 79% used one often. They showed a stronger tendency than other writers to report that they felt more productive since they had begun using the word processor, but they showed an overwhelming tendency to review their text on a printout rather than on the screen, generally finding the screen size restrictive.
Muriel Harris suggested that the first drafts of 'multi-drafters' tend to be writer-based rather than reader-based: that is, primarily an aid to the writer's thinking rather than tailored to the needs of readers. Such writers may delete a large quantity of the text which they generate. They may also get lost in their evolving texts and have a strong need to re-read (Harris 1989, p. 187). The oil painting strategy is not confined to literary writers. One leading scientist in the field of biochemistry reported that: 'I evolve a paper out of the mist. It comes in pieces, each piece being smoothed a bit as it comes along. And so it isn't a linear thing starting at the beginning and going to the end, but rather clusters.' Another reported 'writing it several times until I see how I'm going to convey crystallize, and then sort of letting the paper flow... I write the paper and let it come as it comes... My first draft is an enormous, lengthy, amorphous mass... I found myself crossing out... I do a tremendous amount of pruning' (both cited in Rymer 1988, pp. 230, 231).
As in painting in water-colours, this strategy involves an attempt to produce a complete version at the first attempt, with minimal revision. Paintings done in water-colours are typically characterized by a sense of freshness and lightness of touch. James Britton referred explicitly to this strategy as 'painting in water-colours', stressing - as did the poet Stephen Spender (1946) - the difference between this and the oil painting strategy: 'where one pigment may be used to obliterate another.' Britton associated it with his notion of 'shaping at the point of utterance', declaring that 'the initial process must capture immediately as much as possible of the painter's vision' (Britton 1980, p. 65). Such a precipitative approach (in contrast to writing which is more planned, and/or more extensively revised) is often associated with novice writers. Carl Bereiter and Marlene Scardamalia (1987) refer to the uncritical 'knowledge-telling' strategy of novices. In such hands it can result in writer- based prose which is insufficiently adapted to the needs of readers. Apart from inexperience, situational factors (such as deadlines or lack of motivation) can of course lead to the first draft being the final one. However, the water-colour strategy is also the preferred method of many accomplished writers. For literary writers this may reflect an attempt to retain 'spontaneity', 'truth to feeling', or descriptive accuracy. Other writers may simply feel a need to maintain momentum. Some writers refer to complete texts being formed in the mind after a long period of mental 'incubation' or 'germination'. Others refer, as I have indicated, to 'unpremeditated' writing 'dictated' by an inner voice.
For short pieces of writing about which they feel very confident, most writers probably write in this way. For some writers it may be simply an initial strategy for producing a first draft; sometimes only for part of a text. John Steinbeck worked this way because he felt that 'rewrite in process... interferes with flow and rhythm which can only come from a kind of unconscious association with the material' (Plimpton 1977, p. 185). Ray Bradbury similarly reported: 'I do a first draft as passionately and as quickly as I can. I believe a story is only valid when it is immediate and passionate; when it dances out of your subconscious. If you interfere with it in any way, you destroy it' (Strickland 1989, p. 54).
As one would expect, users of this single-draft strategy in my study of academic writers were most unlike those favouring the oil painting strategy in showing a very strong tendency not to do a great deal of revision. They tended to work largely sequentially and showed a very strong tendency to correct any slips mainly as they wrote. They were divided over whether it helped to leave a piece of writing and to return to it later. Most tended to prefer not to discuss work in progress. They also showed a strong tendency not to be interlinear editors. These writers exhibited a stronger tendency than others to make frequent use of the pen or pencil but not of the word processor.
Muriel Harris describes the preferences of 'one-drafters' for beginning with 'a developed focus', generating limited options prior to writing, settling quickly on a plan, making minimal changes to the text, and doing little re-reading. She also suggests that they tend to be intolerant of ambiguity, to need rapid closure, and in general to dislike writing. All this is in strong contrast to the preference of 'multi-drafters' for open-ended exploration, including beginning at an exploratory stage (Harris 1989). William Lutz, an American academic, reported that:
In the case of one of my interviewees (a historian), the use of the water-colour strategy stemmed from his dislike of writing, and reflected a desire 'to get it out of the way as soon as possible' (Chandler 1992, p. 222).
Of the 107 academic writers in my survey, 57 were frequent users of the architectural strategy, 38 made frequent use of bricklaying, 33 used oil painting often, and 20 frequently employed the water-colour strategy. Only 15 of these writers were not frequent users of any of these four strategies. It is tempting to refer to frequent users of strategies as architects, bricklayers, oil painters and water-colourists, and I have sometimes done so here, but this can be misleading because many writers make frequent use of more than one strategy. For instance, the popular British children's writer Enid Blyton seems to have written stories using the water-colour strategy, her autobiography using the architectural strategy, and articles using the oil painting strategy (McKellar 1957, p. 138). On the other hand, a sizeable number of writers may well favour a single strategy: in this sense there may well be some architects, bricklayers, oil painters and watercolourists. Of my 107 academic writers, 45 reported frequent use of only one of these strategies; 35 used two; 9 used three; and only 1 used all four. Where strategies were combined amongst these writers, those most commonly used together were bricklaying and architecture, and those least used together were: oil painting and water-colour; and oil painting and bricklaying. In my survey, all of the four strategies showed up across the academic subject spectrum (arts - social sciences - science). However, whilst the architectural strategy was the one most commonly used in all subjects, the largest proportion of frequent users (65%) was in the sciences. The water-colour strategy was the least frequently used in all subjects but was particularly uncommon in the sciences, where the other uncommon strategy was oil painting.
It has been suggested that the maturity of the writer may be involved in a preference for particular strategies. Elliott Jacques (1970) argued that (amongst literary authors at least) what I call the water-colour strategy is more typical of early adult creativity, whilst the oil painting strategy is more typical of a mature adult mode (from the late 30s). Neither my survey of academic writers nor my review of writers' accounts supported this interpretation.
My framework is not dissimilar to that employed by John Hayes and Linda Flower (1980). However, they interpreted individual differences in composing practices in terms of differences in 'goal setting', an interpretation which would be unlikely to reflect the experiences of those who see their writing in terms of 'discovery'. The architectural strategy is similar to their configuration 4, bricklaying to configuration 1, oil painting to configuration 2 and the water- colour strategy to configuration 3.
I have not dealt here with what we could call writing tactics. These are practices which are more specific to the kind of writing involved. In the context of academic writing, Judith Riley (1984a, b, c) listed three basic tasks in drafting teaching materials at the Open University: a) finding and selecting source material; b) sequencing the material; and c) finding the words. The tactics used varied, in that they chose different starting points or priorities, putting sources first, argument first or writing first. This is perhaps not that different from the writing of fiction, where one may start, for instance, with characters, plot or 'voice'. Riley's metaphors for writing tactics are: packing a box; producing a play; laying a railway track. Box-packers begin by making a contents list, which tends to keep expanding as they read around the topics. The primary task is seen as trying to fit as much material as possible tidily into some kind of structure. Theatre producers begin by assembling a 'cast' of leading authorities, key texts and quotations. The main task is seen as shaping the text into a 'creative whole'. Track-layers begin by planning a route through the material, which might be redrawn several times both before and after attempts at writing the text. Here the main task is to try to meet the needs of the trains (the student readers). She also felt that 'there is not much chance of people changing from one set to another... And why should they? I have no reason to think that any one of these three approaches makes the job much easier than any other or necessarily yields a better product' (Riley 1979, p. 5). This is close to my own view regarding writing strategies. Riley also noted that tactics did not seem tied to particular academic disciplines: all three tactics were found in the social sciences and only box-packers seemed lacking in mathematics. Nor did she find any relationship with years of experience at the task.
(Adapted from my book, The Act of Writing)