The research on how writers actually think about their texts as they produce them is typified by observational and retrospective accounts. In observational studies, it is usual to use 'protocol analysis' as a technique, where writers are asked to comment on what they are doing and thinking about as they are writing (e.g. see Cotton and Gresty, 2006). Retrospective accounts are given in response to questions after the writing session is over. Sometimes, writing sessions are videotaped to aid subsequent analysis. Interviews and questionnaires are also commonly used in retrospective studies to ask writers about their writing procedures. Table 1.1.4 shows the level of detail described in some of these studies.
Studies using these methodologies lead to the conclusion that what drives writing is very much:
(i) who the text is being written for;
(iii) how much of the text has been already produced (Hayes, 2006).
Within these constraints, writing is often characterised as a hierarchically organised, goal-directed, problem-solving process. Writing, it is said, consists of four main recursive processes — planning, writing, editing and reviewing. These activities, however, do not necessarily occur in the fixed order suggested. Writers move to and fro in accordance with their individual goals of the moment — although, naturally, more time is spent on planning or thinking at the start, and on editing and reviewing at the end.
Studies of the teaching of writing have shown that instruction in each of these activities leads to better performance (e.g. see Graham, 2006). However, some authors, such as Peter Elbow, think that it is misleading to think of
Table 1.1.4 Multiple and overlapping thought processes when writing
While I am writing, my mind is either simultaneously engaged in or rapidly switching between processes that perform all or most of the following functions:
• monitoring the thematic coherence of the text;
• searching for and retrieving relevant content;
• identifying lexical items associated with this content;
• formulating syntactic structures;
• inflecting words to give them the necessary morphology;
• monitoring for appropriate register;
• ensuring that the intended new text is tied into the immediately preceding text in a way that maintains cohesion;
• formulating and executing motor plans for key strokes that will form the text on screen;
• establishing the extent to which the just-generated clause or sentence moves the text as a whole nearer the intended goal; and
• revising goals in the light of new ideas cued by the just-produced text.
These processes cannot all be performed simultaneously. Attempting to do so ... would result in overload and writing would stop. The fact that I am writing this at all, therefore, is testament to the writing system's ability to co-ordinate and schedule a number of different processes within the limited processing resources afforded to it by my mind.
Adapted, with permission, from Torrance and Galbraith (2006), p. 67, and the Guilford Press.
writing as moving in separate stages from planning through writing and editing to reviewing. Elbow advocates writing some appropriate text first, not worrying too much at this point about spelling and syntax, and then repeatedly editing and refining the text to clarify what it is one wants to say (e.g. see Elbow, 1998). There is room, of course, for both positions. It can be helpful to think about the sequence and the structure of a paper (or book chapter) before one begins to write it, but one need not necessarily start at the beginning. And it can be equally helpful to let the thoughts pour out when writing a particular section, before revising it. In my view, the actual product determines the process, but the processes involved can be many and varied.
Numerous investigators have tried to distinguish between writers in terms of the ways that they think about their writing and their procedures. As we have already seen, computer-based tools can be used to measure different aspects of style (or readability). Microsoft's Office program, for instance, provides measures of word, sentence and paragraph lengths, the percentage of passives used, and various measures of readability (such as the Flesch RE score). Another program, Pennebaker's Linguistic Inquiry and Word Count
(Pennebaker et al., 2001), calculates the percentage of words used in any one text in any one of seventy-four different linguistic categories. Some of these separate categories can be grouped, for example, into emotional words (e.g. 'happy', 'sad', 'angry'), self-references (e.g. 'I', 'we') and cognitive words (e.g. 'realise', 'think', 'understand').
Studies using these measures have confirmed that individual writers have distinct styles or 'voices'. My colleagues and I, for example, once showed that three highly productive writers maintained similar writing styles over a period of more than thirty years, despite the many changes in the technology that they had used over this period (Hartley et al., 2001). Indeed, 'forensic linguistics' is a discipline that specialises in detecting changes in authorship (e.g. in a witness's statement) by using computer-based stylistic measures (e.g. see Coulthard, 2004).
So, although all the articles in a particular journal may look much the same, different writers will have used different methods to achieve this uniformity. Indeed, as noted above, one of the ways that manuscripts differed, before the advent of word processing, was in their physical appearance. Stephen Spender, the poet, distinguished between writers he labelled 'Beethovians' and those who he labelled 'Mozartians', and, if you have ever seen an original (or facsimile) manuscript of either of these composers, you will know exactly what he meant. A score by Beethoven is full of crossings out and looks an incomprehensible mess. A score by Mozart is, by contrast, neat and pristine. Beethoven, it can be argued, working from earlier sketches in his notebooks, was struggling to get it right. Mozart had it right already in his head and just copied it out:
When I proceed to write down my ideas, I take out of the bag of my memory, if I may use that phrase, what has been previously collected into it in the way that I have mentioned (above). For this reason the committing to paper is done quickly enough, for everything is, as I said before, already finished; and it rarely differs on paper from what was in my imagination.
(Excerpt from a letter attributed to Mozart, in Ghiselin, 1980, p. 35)
In modern terminology it is more common to distinguish between writers who are 'pre-planners' (Mozartians) and 'revisers' (Beethovians). Indeed, several studies distinguished between academic writers in terms of these two separate categories before the advent of word processing. Others, however, placed them along a spectrum — from pre-planners to revisers. Thus, for example, Torrance et al. (1994) described postgraduates in the social sciences who:
(i) extensively pre-planned their writing and then made few revisions (planners);
(ii) developed their content and structure through extensive revisions (revisers); and
(iii) both planned before they started to write and revised extensively as part of their writing process (mixed).
Torrance et al. found that their postgraduate planners reported higher productivity than did both the revisers and the mixed groups. Table 1.1.5 provides quotations from fully fledged academics to illustrate what these different kinds of writer say. It is not necessary, of course, to stick to one particular method. John Le Carré, for example, in a radio broadcast, reported using a storyboard method for planning three of his novels but letting the plot develop for others.
Some research with adolescents suggests that writing and changing what you want to say as you go along (revising) lead to better writing than planning the writing in advance and then writing it out (planning). However, more recent research along these lines suggests that there might be further individual differences here. Kieft (2006), for instance, found in one of her studies that 15 to 16-year-old students who were high self-monitors — i.e. those who frequently evaluated their text as they were writing — did equally well whether or not they were taught to revise through multiple-drafting or to produce an outline first. However, those who were low self-monitors did better when they were taught to produce an outline first.
Other investigators have used fancier names for describing different kinds of writer. Nonetheless, they are arguing essentially the same thing — that there is a variety of writing styles based along a spectrum from pre-planning at the start to revising at the end. Thus Chandler (1995), for example, distinguished between 'architects' (planners in advance), 'oil-painters'
Table 1.1.5 Quotations from academic writers
I like to write a plan. I produce section headings and fairly detailed jottings about what these will contain, and then follow them through.
I write very much in sections at a time, from the beginning to the end.
I do plan my writing, but I usually find that in the process of writing the plan might take a new direction. I will then 'go with the flow'.
I usually pre-plan it, although on the occasions when I have just let it 'flow' it seems to have worked quite well.
Cut and paste was invented for me. I start off with headings ... I then start shifting things around.
I have ideas in the back of my mind, but I only really know what I want to say as I write them down. That drives me into more reading and re-reading of my texts.
Reproduced from Wellington (2003), pp. 22-3, with permission of the author and the publishers.
(changers and revisers), 'bricklayers' (one step at a time) and 'water-colourists' (who aim to complete the text at the first attempt).
The architect strategy is typically the 'plan, write and revise' strategy discussed above. Architects make detailed plans and stick to them. Oil painters may think of new ideas while they are writing. They tend to produce drafts and print them out while they are working. This allows them to read and to revise. A characteristic refrain of these writers is, 'How do I know what I am going to say until I can see what I have said?'. Sharples (1999) classifies the novelists Frederick Forsyth as a water-colourist and Beryl Bainbridge as a bricklayer.
I am inclined these days to the view that new technology has made it more difficult to categorise and describe differences in the ways that writers go about writing. Word processors allow writers to write how they like at whim, and to vary their approaches. But writing is still a complex business, however, even with word processors. The writing strategies described above in Table 1.1.5 do not begin to approach the fine detail of what is actually required. Table 1.1.4 gives a better picture.
Was this article helpful?