Psychology of Writing Process Research Paper

1.    Writing As Thinking

Writing  involves systems for formulating, executing, and monitoring text. The seminal model of the writing process proposed  that  formulation involves planning content and generating sentences based on knowledge drawn  from  the  writer’s  long-term  memory  (Hayes and Flower  1980). Planning  a text entails generating ideas,   organizing   them,   and   setting   goals   to   be achieved  in  the  text.  Translating ideas  or  sentence generation  includes the subprocesses  of speech, such as selecting appropriate lexical entries and assembling the  constituents in the  proper  order.  Reviewing  the text involves reading and editing operations that detect faults at multiple levels of text structure,  ranging from local mechanics  to  the  coherence  of the  whole text. The writer must monitor whether the processes are successful and decide which writing process to engage at a given moment.

If one  focuses on  the  reflective  cognition  seen in planning and reviewing, then writing looks like thinking. Composition studies in rhetoric and education  in the 1970s stressed the similarity of writing to problem solving. Later composition  studies shifted the focus to social cognition  and stressed that  thinking  is situated within a community  of other writers and readers.

1.1    Problem Solving

Planning,  translating, and  reviewing are constrained by the  task  environment and  the  writer’s long-term memory.  The  topic,  the  type  of  text,  the  intended readership, and the text produced  thus far constitute elements of the task  environment, as do the writer’s tools, collaborators, and deadlines. The task environment   and   long-term   memory   together   pose   the problems of content (what to say) and rhetoric (how to say it) that must be solved in order to produce  a text.

Mature  writers  seek solutions  to both  the content and rhetorical  problems  posed by the writing task at hand, using a strategy called knowledge transforming (Bereiter  and  Scardamalia 1987).  The  writer  gains insight  into  the  ideas  being  formulated as a  consequence  of  expressing  them  as  text,  actively  transforming  what they know about  a topic as a result of trying to write about  it. Unlike the reflective problem solving of mature writers, young children fail to elaborate  their  thinking  through the  act  of writing. They  instead  engage in knowledge  telling. The  task environment automatically cues children to retrieve an idea from long-term memory and then translate it into sentences. Writing continues in this manner until retrieval fails to provide relevant material. During adolescence,  knowledge  transforming begins  to emerge as a complement  to knowledge telling.

Data  based on verbal protocols  and other forms of introspection indicate  that  adult  writers  recursively attend  to  planning,  translating, and  reviewing  processes. Although planning  generally occurs more frequently  in the beginning stages of a first draft  and reviewing  later,  all  three  operations may  be  called upon  at any point.  One effective strategy  for coping with the multiple demands of writing on working memory  is to  delay  concerns  about   how  ideas  are expressed  and  focus on planning.  This increases the number  of ideas generated  during planning  (Glynn et al. 1982). Another  effective strategy  is to prepare  an outline  of ideas and  their  organization before  composing  a  first  draft.  Advance  outlining  permits  the writer to focus on translating ideas into sentences and can increase the fluency and quality of writing at least for short texts on familiar topics.

There are individual differences in the strategies preferred by adult writers (Kellogg 1994). Some prefer to plan and review in advance in careful detail, perhaps creating diagrams,  outlines, and notes, before starting a first, relatively polished draft. This has been called by several names, such as the Planner or Mozartian strategy.   Such  writers   transform  their   knowledge about  the topic by tackling the content and rhetorical problems   as  part   of  prewriting.   Others   prefer   to discover ideas and develop their thinking on a topic by actually writing about  them in a rough draft and then later doing extensive revision. This goes by the name of   Discoverer,    Reviser,   or   Beethovian    strategy. Surveys of faculty and graduate students have revealed that planning carefully and revising little is associated with higher  productivity in academic  writing.  However, it is unknown  whether  particular kinds of texts are generated  more readily with a discover and revise approach (e.g., creative writing), or why some writers prefer  one  approach to  another  while others  adapt their  strategy  to  the  task  at  hand  (e.g., personality traits).

Writing   exposes   not   only   the   content   of   our thoughts but also our ability to think, remember, and articulate. Avoidance  of writing  can  be a means  of coping with the anxiety and other emotional  reactions to expressing thoughts as public symbols. In writer’s block,   text   production  halts   altogether,   although reading  and  thinking  about  the  task  may  continue. The prewriting phase can be lengthy and a useful time for reflection for any writer. But in writer’s block, the prewriting  activities  become  a way to  procrastinate and  no  longer  serve any  useful  purpose.  Procrastination, excessive worrying, impatience, perfectionism, and evaluation  anxiety characterize  writer’s block.

1.2    Social Cognition

Besides an act of problem solving, writing is inherently a social act. Texts are created to be read within a discourse community  (Rafoth and Rubin  1988). The social aspects may be explicit, as in collaborative writing,  or  they  may  be implicit,  as  in the  way  an imagined reader shapes the writer’s ideas and manner of expression.

Expert  writers  conform  their  texts  to  the  expectations of their readers. This point is well illustrated  in the  publication style requirements of  the  American Psychological Association.  The style demands an allegiance to the empirical method  and so defines the psychologist’s task  as a writer.  Conclusions  must  be hedged so as to avoid stepping beyond the limits of the data,  creating  an  air  of  uncertainty atypical  in  the humanities.  Disagreements  are  couched  in terms  of differences about  appropriate methods  or interpretations of data, never in the personal terms at times seen in literary criticism.

It is often through revision that  writers make their texts  accessible to  readers.  Immature writers  fail to take their audience into account and revise effectively. Reading  and  editing  the text are highly effortful  for young  writers,  in part  because  of working  memory limitations.  The lack of thorough revision also stems from a failure to establish clear plans in the first place. Without   explicit  goals  that   include   the  needs  of potential  readers,  young  writers  cannot  detect problems  in meeting  those  goals.  Revising  becomes little more than editing errors in spelling, punctuation, and other writing mechanics. Besides needing explicit representations of  what  they  wrote  and  what  they intended, writers further need a representation of how the text appears to their readers. Skilled writers can see their work from another  person’s perspective. Taking the perspective of another  individual  is an important milestone in social cognitive development in many domains,  including written language production.

2.    Writing As Language Production

Writing  is an act of language  production, too.  However,  speech  production has  received  far  more  attention   in   psycholinguistics   than   writing,   largely because the field of linguistics stressed the spoken over written word during the twentieth century. Psychological studies of literacy are also lopsided in that reading  outweighs  writing (for example,  see Reading Skills). This asymmetry derives from the methodological  difficulties of studying  language  production compared  with comprehension.

In  experiments  on  language  comprehension, the inputs to the system can be controlled  and the outputs carefully measured.  What  someone  says or writes in response to an experimenter’s prompt  is, by contrast, unpredictable. Rather  than experimentation, the natural observation of speech errors or slips of the tongue guided theories of speech production, and slips of the pen have likewise informed  us about  written spelling. Observations of  pauses  in  handwriting—both their location  and  duration—have provided  insights  into writing  as language  production. Another  alternative to  experimental  manipulations has  been  the  use  of neuropsychological case studies that  correlate  lesions in brain structure with difficulties in speaking, writing, or both.

From   slips  of  the  tongue  and  neuropsychology, much is known about the translation of ideas and their motor  execution  in speech (for example,  see Speech Errors, Psychology of). Five levels of processing must be distinguished  in sentence production. The message level corresponds to the planning  of the ideas or propositions to be conveyed. Writers  may start  with only inchoate feelings or images that must be mapped onto  propositional representations suitable  for  sentence  production. Functional processing  reflects the translation of propositions into  the words  or lexical items along with the grammatical roles to be played by various phrases. Positional processing continues translation by retrieving the lexical (e.g., phonological segments)  and  grammatical morphemes  (e.g., inflections),  and  positioning  these  constituents in a hierarchical syntactic structure. Next, the phonological component completes translation in speech by specifying all aspects of the sound structure of the sentence, including its prosodic  features such as stress patterns. Unlike speech, writing requires a graphemic representation  (Caramazza 1991). Graphemes can be computed  either from the visual appearance or orthographic code of a whole word or by applying rules for converting  phonology  to orthography in spelling out the word. The order of graphemes for spelling a word and  the motor  programs for controlling  the musculature  in typing or handwriting must be specified. In handwriting, for example,  the size of the letters and their allographic  form must be selected.

In  crafting  a written  sentence,  more  care may  be given to correctness and precision than in speech. Whereas   sentence   fragments   are   characteristic  of speech, written language is expected to be well formed. The motor  execution process, particularly with handwriting, is markedly slower than typical rates of speech production, and  the  writer  is freed  from  the  social awkwardness   of  pauses,   allowing   for   both   more intensive formulation and editing than would be found in most speech. The production of a single sentence in writing,  then,  is  an  intricate   affair,  dependent   on multiple   neurological   structures.   The   inability   to express a thought in writing, agraphia, usually occurs to some degree as a consequence  of significant  brain damage, regardless of its location. Specific disorders of central  language  systems, as in aphasia,  or of motor systems, as in Parkinson’s  disease, cause agraphia, as does  widely  spread   pathology,   as  in  Alzheimer’s disease.

But written language production entails more than the production of a sentence, more than appropriately arranged letters  and  words.  A  text  is made  up  of multiple  sentences  that  must  be coherent  at  both  a local level and a global level (for example, see Memory for Text).  Local coherence is achieved by cohesive ties that link one sentence to another. To illustrate, anaphoric pronouns form a link between the pronoun and the noun to which it refers in the previous sentence (e.g., The  dog chased  the  thief  down  the  street.  He barked  fiercely all the way.). Local cohesion is necessary but not sufficient, however. All the sentences in a text must relate to one another  so that the text is also globally coherent. An extended text must be organized both  at the level of its microstructure and its macrostructure.

3.    Writing As Memory

Without  knowledge,  writers  cannot  succeed at  their task. Diverse kinds of knowledge must be available in long-term memory, retrieved when needed, and maintained  in working  memory  as the writer plans,  edits, and  so  on.  In  particular, writers  must  know  much about  their topic, content  knowledge,  and  about  the requirements of a well-structured text, discourse knowledge.

3.1    Long-Term Memory

Content  knowledge guides the generation  of informative and interesting ideas. To develop a theme, writers must be able to retrieve ideas from long-term memory throughout the  course  of  prewriting,   drafting,   and revising. An inadequate stock of content knowledge is one reason why student  writing fails and professional writing  succeeds.  Just  as expertise  in a domain  improves a reader’s ability to comprehend  a text, it also improves   their   ability   to   write   in  an   interesting manner.  For  example,  a  high  degree  of  knowledge about  baseball enables writers to generate many more game relevant propositions in a narrative  about  a half inning of play (Voss et al. 1980). Although less knowledgeable writers are able to generate macrostructure   propositions well, they  fail  to  supply  the necessary  detail  for  the  microstructure of  the  text. Including  sufficient detail improves the reader’s comprehension  and opinion  of the text.

Discourse  knowledge  guides  the  selection  of  relevant  ideas and  the organization of these ideas. The writer must know how to link clauses together  using conjunctions,  how  to  use  pronouns  so  that   their reference is unambiguous, and how to choose between the  use of the  same  word  or  a synonym  in linking clauses.  These  and  other  cohesive  devices must  be learned  for  a  writer  to  produce  text  that  is locally coherent.  Although it is often emphasized that people acquire  the syntactic  rules needed to support  spoken language before a child enters school, learning to use cohesive links among clauses in a written text develops throughout the primary  grades.

To  create  global  coherence  in narrative  or  other kinds  of text,  each of the individual  sentences  must help   to   develop   the   story,   theme,   or   argument. Narrative schemas are acquired at a young age and are heavily practiced throughout childhood  in listening to and  telling stories.  Knowing  the typical  sequence of events in a story helps the writer to select ideas that need to  be included  and  provides  an  organizational format for expressing them. Teachers draw on the writer’s  knowledge  of discourse  by giving narrative assignments, and students show their strongest writing skills in the narrative  mode. In contrast, schemas for argumentation and persuasion  are weakly developed in all but the highest ability college students,  and the quality  of the average  student’s  expository  and  persuasive writing is correspondingly poor (Britton  et al. 1975).  Regardless   of  the  type  of  discourse   being written,  the more one knows about  the requirements of  global  coherence,  the  better  prepared   one  is to actually produce  an effective and memorable  text.

3.2    Working Memory

Representations must be not only retrieved from long-term  memory,  but  they  must  be  maintained in  an active state for writers to plan, translate, and review. The details of how verbal, visual, spatial, and central executive  components of  working  memory  support writing processes are still unclear. But it is known that if working memory resources are diverted away from writing, then performance suffers. For example, when adult  writers  try  to  maintain  a set of five words  in working  memory  and  concurrently transcribe  orally presented sentences in French, they make subject–verb agreement errors (Fayol et al. 1994). The limitations of working  memory  are especially problematic for children learning to write. Planning  suffers because they often forget the ideas recently generated and so repeat them  again.  Sentence  generation  suffers because  the number  of linguistic units that may be coordinated is limited. Furthermore, handwriting or typing  are not as effortless in children as they are in adults. Children must divert working memory resources from formulation  and monitoring to motor  execution.

Finally, the extensive interactions among planning, translating, and reviewing processes found in mature writing are highly effortful. For example, in planning a text,  an experienced  writer  sets rhetorical  goals that are shaped  by the intended  audience. Such high-level planning has consequences for the selection of ideas to be included  and  even the choice of words  in a given sentence.  It is not  the case that  the writer  plans  the ideas  of  the  text  and  then  automatically translates them into sentences. The formulation and monitoring processes typically require working memory resources even when motor execution has been automatized. Given   its   demands,   learning   to   write   effectively requires  instruction and  practice  throughout  childhood   and  young  adulthood.  Whereas   learning   to speak  occurs  universally  to  a remarkable degree  of mastery  by five years of age, the demands  of literacy greatly extend the language acquisition  period.

Bibliography:

1. Bereiter C,  Scardamalia M  1987 The  Psychology  of  Written Composition. Lawrence Erlbaum,  Hillsdale, NJ
2. Bock K  1996 Language  production: Methods  and  Psychonomic Bulletin and Review 3: 395–421
3. Boice R  1985  Cognitive   components  of     Written Communication 2: 91–104
4. Britton J, Burgess T, Martin  M, McLeod A, Rosen H 1975 The Development   of   Writing      Macmillan   Education, London
5. Caramazza A 1991 Issues in Reading, Writing, and Speaking: A Neuropsychological   Kluwer   Academic,   Dordrecht, The Netherlands
6. Fayol M,  Largy  P, Lemaire  P 1994 When  cognitive  overload enhances   subject–verb   agreement      The   Quarterly Journal of Experimental  Psychology 47A: 437–64
7. Garrett M F 1980 Levels of processing in sentence production. In: Butterworth B (ed.) Language production. Academic Press, London, pp. 177–220
8. Glynn S M, Britton B K, Muth  D, Dogan  N 1982 Writing and revising persuasive documents: Cognitive demands. Journal of Educational Psychology 74: 557–67
9. Hayes J R,  Flower  L S 1980 Identifying  the  organization of writing      In:   Gregg   L W,  Steinberg   E R   (eds.) Cognitive Processes in Writing. Lawrence Erlbaum,  Hillsdale, NJ, pp. 3–30
10. Kellogg R T 1994 The Psychology of Writing. Oxford University Press, Oxford, UK
11. Lorch M P 1995 Disorders of writing and spelling. In: Kirshner H S (ed.)   Handbook  of Neurological Speech  and Language Disorders. Marcel Dekker,  New York,  295–324
12. Madigan R,   Johnson   S,  Linton   P  1995  The  language   of psychology:  APA  style as epistemology.  American  Psychologist 50: 428–36
13. McCutchen D 1996  A  capacity  theory  of  writing:  Working memory  in composition. Educational Psychology  Review 8: 299–325
14. Rafoth B A, Rubin  D L (eds.) 1988 The Social Construction of Written  Ablex, Norwood, NJ
15. Traxler M J, Gernsbacher M A 1993 Improving  written  communication through perspective-taking. Language and Cognitive Processes 8: 311–34
16. Voss J F, Vesonder G T, Spillich G J 1980 Text generation and recall by high-knowledge and low-knowledge individuals. Journal of Verbal Learning and Verbal Behavior 17: 651–67