STUDENT WRITING: LANGUAGE AND LOGIC Bruce Jeans Deakin University, Victoria Abstract: Somewhere in their programs many undergraduate, and most graduate, students are required to write a research report of some kind. Sometimes this is a relatively simple literature review. Often it is a report of a small-scale empirical study. These reports have two major components: one of these is the content of the report and the other is the style of writing that the student has used. This paper argues that, at graduate level particularly, writing style and content are difficult to distinguish. Despite the admonitions and advice of academic staff many students still have great difficulty in preparing a well-formed and accurate research report. This paper will discuss typical student difficulties and the attitudes of academic staff towards these difficulties. Introduction This study is part of a larger study examining the scholarly activities of students preparing research reports. Although these scholarly activities are of many different kinds in the end the student will prepare a research report of some kind. The quality of the student's work is usually judged by the quality of this report. But just how is this judgment made? Are we as academic staff relatively homogeneous in our views about the criteria to be used? Do we have roughly the same views the relative importance of structure and content? Do we have different expectations for undergraduate and graduate students? What account do we or should we take of contextual variables as for example in the case of a lower income student whose mother tongue is not English? These are interesting questions in their own right but they also bear on the processes of higher education and on how we think about the quality of teaching in higher education. The data reported and the comments made in this paper reflect the outcomes of a pilot study and of many conversations with academic in a number of countries - English-speaking and non-English speaking. The pilot study involved 204 academic staff in five universities. Experience had suggested that there might be characteristic, discipline related attitudes to student writing and so an attempt was made to categorise the respondents as belonging either to Art, Arts or Sciences. Art included domains such as photography, two and three dimensional creations, dance and music. Arts included education, social science, philosophy and literature. The sciences category included the physical sciences, mathematics and various combinations of these. A specific purpose questionnaire was designed. It included five biodata items including age and gender, thirty strength of opinion items, eleven limited choice items and one open-ended item. The example sentences used in the strength of opinion items were drawn mainly from research reports prepared by Education students in graduate programs in Australian universities. The following item is a typical strength of opinion item in which respondents were asked about the acceptability of certain sentences: The ozone is becoming studied by scientists. 0_____________________________5_____________________________10 Not at all alright Absolutely Item 023 is typical of the limited choice items: 023 Which, if any, of the following do you expect in a review of the literature? (Please tick the numbers on the left hand side). i chronological organisation ii organisation around major paradigms iii other logical forms of organisation ______________________________ iv analytical comment v synthetic comment vi identification of research paradigms vii identification of methodological weaknesses viii identification of areas for further research ix other _____________________________________________________________ The estimated time to complete the questionnaire was twelve to fifteen minutes. Multiple copies of the questionnaire were then mailed to colleagues who were asked to distribute them to staff members and to coordinate the return of the completed questionnaires. The return rate was 69%. This was considered satisfactory for the pilot study. Language and logic The relationship between language, cognition and logic has long been a matter of debate. In its simplest form the argument is that the language we speak has a determining effect on the way that perceive and think. From this perspective, native speakers of non Indo-European languages for example live in a different world to that of a speaker of an Indo-European language. At the turn of the century this certainly seemed true. The Indian, Polynesian and Asian cultures that the anthropologists wrote about were so very different. To some considerable extent this perception was created and encouraged by the anthropologists who emphasised differences rather than samenesses. One of the major difficulties was that the world views appeared to be so different that it was easy to generalise from this to the determining effects of language. There was some evidence that language did have an effect. It is undoubtedly true that languages differ in their environmental lexicon. However, they vary in understandable ways. The fact that one language has ten words for rice and another has five reflects the differential importance of rice in the two cultures but it doesn't imply differences in logic. Languages also differ in structure. The importance of word order differs between language groups. Does this imply that there are underlying differences in logic? Probably not. The fact that French sometimes has noun before adjective as in moulin rouge does not mean that French people perceive objects first and qualities next or that English people perceive qualities before objects. What about languages like Mandarin? This language differs in many ways from Indo-European languages but there is little evidence to suggest that native Mandarin speakers perceive the external world in some fundamentally different way to native English speakers. If you think about a group of people living in Papua New Guinean group in isolated conditions and having no contact with other people it is not surprising they have one way of looking at the world and that they would have no reason to think that there is any other way of doing so. In an intellectual sense the European cultures were like this but their isolation was intellectual rather than geographic. Some years ago I was struck by the European usage of scientific in connection with Education. Surely these people don't believe that Education is a science! No, they didn't. What they did mean was that Education can be studied scientifically. A minor example but it illustrates the subtlety of language usage. Nevertheless there is some real effect of language on cognition. Whilst it is still not clear if this effect is different for different linguistic systems it can be demonstrated that at least some features of language predispose us to perceive and react to external stimuli in different ways. The difficulties of translating from one language group to another are well known. The simplest explanation for these effects is that languages facilitate habitual ways of dealing with the world but do not determine how we can deal with the world. The debate about this fundamental proposition has simmered since the turn of the century. More recently it has surfaced in derivative forms. For example, the argument has been made that some aspects of English usage reinforce a male-oriented worldview. I don't think that anybody has suggested that this has been deliberate. Rather it is a case of the way we look at the world having an effect on the language we use. If we change some of the vocabulary we would at least be recognising a degree of language bias and hence recognise the role of women. I have not seen it written or heard it argued but implicitly it seems to have been assumed that if we change the language we can change the way that we think about ourselves. There are other examples. The idea that the construction of self can be investigated by examining written material reflects the language-logic hypothesis. This example illustrates the three elements - language, logic and worldview. For a long period it was taken for granted that if your worldview was different your logic must be different and that this different logic would be reflected in your linguistic system. More recently I think that there has been some acceptance of the view that if you make certain, reproducible, observations and move from these observations to general description or explanation without introducing self-evident contradictions then the process ought to be described as logical and or rational. It would be odd indeed to argue that whole groups of people think and live irrationally. Similarly, I think it is unlikely that there are gender differences in logic but not necessarily in the assumptions that one feeds into the logic process. All that I have said applies to spoken language. Can these arguments be generalised to written language? This question is very much more complex because we have three elements; the way people think, the way they speak and the way that they write. I think that all of these are inter-related. The inter- relationships are complex. We are not sure about how thinking is connected and speaking. For example, is thinking as linear (temporally) as speaking? Under what circumstances does writing constrain the expression of thought and under what circumstances does writing facilitate the development of thinking? It is difficult enough to infer cognitive structure and content from speech structure and content - it is even more difficult to infer cognitive structure and content from the structure and content of a person's writing. Yet, this is the very essence of our work in higher education. For this reason I think that we need to assume that the language-logic, or Whorfian, hypothesis can be extended to include written language. Changing the language I have mentioned the view that the attempt to de-man-ise the English language is an acknowledgment of the equal male-female contributions to our culture. I have also indicated that such changes probably affect the habitual ways that we think about the distribution of power in our culture. Is this sufficient argument to move on and say that we might be able to change the way our students think by changing the way that they use language? Although it seems a fairly long bow to draw I think that the answer is yes - at least in part. It is actually quite difficult to make sustainable generalisations. Apart from any conceptual complexity, teachers vary a great deal in the criteria they use to assess student writing. These criteria depend on the purpose of the writing, the student's age/ability, and on other factors such as ideologies and -isms. For example, many school teachers believe that, in the early years of schooling, the child's ideas and creativity are more important than the spelling and grammar used to express them. Some teachers go further and argue that ideas are nearly always more important than grammar and spelling. When you consider students at university there are some extra dimensions. For example, computing, mathematics and the physical sciences have very comprehensive and standard non-natural symbolic systems that are not suited to the social sciences. These symbol systems are international and intelligible to students whose mother tongue is not the language of instruction. Not all of scientific writing use non-natural language and so the matter of the degree to which written language and logic might be related still has to be considered. The essence of writing in the social sciences is not the correct or accepted use of standardised non-natural symbol systems. Rather, it is the clearly and tightly argued case for a proposition. If this is so then I think that it is clear that the structure of the writing is very, if not critically, important. It is not just a matter of standard spelling and conventional grammar because the expression of a logical argument is built up over a series of connected sentences. It might be argued that non-standard spelling is not too critical provided that there is no significant ambiguity involved. It might even, with less strength, be argued that sentence structure is not too important provided that the essential premises or propositions are clear. But can we make a reasonable argument that the relationship between sentences is not too important provided that we can understand the gist of the argument? To what extent should we infer logic from student writing? Is it enough that students convey a general impression that they understand specific ideas or should we require them to write exactly what they mean? Is there a general teaching-learning principle involved here - can we change their ideas by changing their writing? In fact, isn't this our job? When we are assessing student assignments we can look at the content - ideas, facts, hypotheses etc. We can also look at technical matters such as spelling and grammar, sentence length and structure of paragraphs; and for certain students we might also look at the style that the student has used. By style I mean those conventions that distinguish one kind of writing from another - the novel from the scientific paper. We can do these things but it seems that academic staff vary widely in the importance they place on these matters. For example, the Sciences group rated the acceptability of The ozone is becoming studied by scientists significantly higher than the Arts group. Similarly the sentence This is not the key to open the door with was significantly more acceptable to science staff than to arts staff. It may be that science staff are more likely to be concerned with meaning than with grammatical structure. Perhaps they are more willing to impute meaning than non-science staff. Student writing: some examples The examples that I have used were taken from student writing - assignments and research papers. The first two examples are composites to save space and to illustrate particular points. Additional context may have made them easier to interpret and several respondents indicated that they relied on context to derive specific meaning. This teacher has some difficulty sequencing the maths lessons and had to talk to a curriculum consultant but when they were talking together one of the grade five boys asked for some information on ecology and so we had a talk about that before we finished for lunch. What do you think is the key idea in this sentence? This sentence is 47 words in length. It contains several joining words. Much of it is irrelevant to what appears to be the main thrust - the teacher's difficulty. It moves from this teacher to we. My impression is that the person writing the sentence starting out using an impersonal style and moved to a personal narrative without being aware of the transition. You will notice an abbreviation. I think that this kind of writing approximates written speech. It does not however, contain the hesitations and repetitions that characterise speech. When academic staff were asked about preferred sentence length slightly less than half (90) the sample thought that it was irrelevant. 82 people thought that sentence length should be between fifteen and thirty words and a majority of these preferred a maximum of twenty words. The questions yet to be answered, or that have only been partly answered, include whether not the service (excluding the employment of integration aides) should be extended, without further reduction, to the outer metropolitan areas, not including Berwick, before we finalise the budget. This sentence is 43 words in length. It is notable for the number of included or embedded clauses. One can understand what is meant although if we delete all the embedded material we would have The questions yet to be answered before we finalise the budget. I don't think that this is acceptable. This seems to be another example of relatively unfiltered thinking/speaking. A similar item was included in the questionnaire. The overall level of acceptability of embedded clauses was 3.4 on a scale of 0 (not at all acceptable) to 10 (absolutely). Despite the low mean there were people who rated the acceptability as 10 (and some who rated it as 0). The relatively high standard deviation of 2.7 reflects this variability. In the following examples the problems are less obvious and might be considered as personal preference. This is due to the teachers being very busy. The ...is.....-ing structure is quite common in student writing. The cause and effect relationship is quite clear. However, I would prefer something like This is because teachers are very busy. The acceptability of this construction was 2.9 (sd 2.3). Clearly, the sample found this structure low in acceptability but, as one might expect, there some responses in excess of 8. The ozone is becoming studied by scientists. This sentence is a variation of the previous one. It also contains the ...is ....- ing structure. It is not difficult to understand what the writer means. It is however not particularly elegant. The sample acceptability rating was 1.2 (sd 1.2) with a minimum of 0 and a maximum of 4. There was more agreement about this item than the previous item although they have similar structure. Next to the storeroom is a small office. The acceptability of this structure was 6.0 (sd 2.7) with a minimum of 0 and a maximum of 10. These examples are sufficient to illustrate typical student writing and the reactions of academic staff. One strategy that students sometimes use to simplify their writing and hence avoid critical comment is to reduce answers to point form. Sometimes this is appropriate but often it is not. This technique has the virtues of simplicity and economy but it lacks elegance and may not be particularly accurate because one frequently needs some context to help interpret the point. Of course, one can argue about the elegance criterion but at the very least it is preferable to inelegance. Student views about their writing My experience has been that students vary widely in their reactions to the comments we make about their assignments. In general, however, I find that students, particularly graduate students, want us to evaluate the functional or content component rather than the stylistic or structural component. Students will accept an allocation of a few percent of the marks for spelling, grammar and presentation but there is considerable resistance to anything like equal weighting for structure and content. The situation is even more complex with mature students and post- graduate students. There is quite strong resentment if one suggests changes to grammar, structure etc. Even though any deduction of marks is minor compared to the frequency of `errors' students often take the suggested corrections as a much more general comment on their professional competence. This concerns me a good deal. One could of course stop correcting the language and simply focus on the content. In the particular case of teachers the exemplar use of language is, in my view, fundamentally important. Academic staff views I stress that this is essentially a pilot study and hence limited in its generalisability. It seems that academic staff are willing to accept quite wide variations in spelling, structure and style. All of the sample drew attention to spelling mistakes in some way and many corrected at least some of them. There is for example, no dominant view about sentence length. Providing that the meaning is reasonably clear sentence length is not a concern. It is hard to disagree with this view but I think that longer sentences are more likely to be obscure than shorter sentences. Some staff prefer one idea in a sentence - some don't mind how many ideas there are. Some like sentences organised into paragraphs - some don't. Most of the sample corrected at least some grammatical mistakes but there were some who paid no attention to grammatical structure and some who corrected all grammatical mistakes. There was a greater tendency to simply draw attention to grammatical mistakes without specifying the nature of the mistake but this was by no means universal. Views on abbreviations, Latin or otherwise, vary widely, for example, around 12% of the sample did not accept Latin abbreviations. Of those who did accept such abbreviations 50% said that their incorrect use would influence the grade that they would recommend for a dissertation or thesis. There was substantial acceptance of the idea that graduate students ought to be directly taught how to write a thesis. The mean for this item was 6.6 (sd 2.2). This could be done in a number of ways. One way would be to have systematic undergraduate policy in which later year assignments and research papers are structurally more complex than those in earlier years. There is no universal criteria for undergraduate and graduate writing. Some of the sample (25%) made no allowances at all while others (50%) clearly distinguished between undergraduate and graduate writing (applying more complex criteria to the latter). Some 55% of the sample thought that structured writing was highly correlated with logic thought and about 25% of the sample made major allowances for students whose mother tongue was not the language of instruction. However, about 20% made no such allowance. Standards - professional journals. A paper in a professional or learned journal is written in a particular and some would say peculiar style. These papers are not simple statements of the author's views. Invariably there are supra-structural constraints. The overall length, for example, might be anywhere from 1,000 to 5,000 words. There is usually a requirement that all sources cited in the paper be referenced in standard form at the end of the paper. If there are data in the paper one's conclusions need to be consistent with the data. There will be many other conventions, the passive voice, third person, grammatical and lexical conformity and so on. Only a small proportion of our graduate students write papers for learned journals. It would probably help the student if we were more proactive in this matter and assisted students to find appropriate journals to which they could submit papers for publication. Conclusion There has been a considerable change in the emphasis on the formal structures of written language in our high schools. Much more importance has been placed on communication and creativity. Direct rote and rule learning of grammar and spelling have been largely replaced by incidental and indirect learning. How one assesses this change depends on one's perspective but for those high school students who go on to university lack of skill in formal or professional writing is likely to cause some difficulty. This is not an argument that high schools should teach more spelling and grammar. If, however, we think it is desirable that students write clearly, unambiguously and economically we need to be conscious of the gap between the reality and the vision. We might also want to conduct our teaching and learning so that the gap is reduced systematically. It is clear from these data that academic staff are quite heterogeneous in the criteria that they use to assess students' written reports. The data in this study suggest the criteria that wee bring to the task of assessment are relatively personal and not particularly related to discipline or to country. There was some evidence that academic staff in the sciences are more likely than social science staff to impute meaning in the non- formulaic writing of their students and are less likely to draw attention to, or correct mistakes in spelling and grammar. Art staff seem to be rather similar to sciences staff. Whilst these differences might be statistically significant there are very substantial individual differences and the student in any faculty is going to experience a wide variety of criteria. In one sense this doesn't matter too much because part of the experience of being in a university is learning to accept diversity. On the other hand estimates of a student's performance should not be too idiosyncratic. We could try allocating a proportion of marks for style and presentation but unless we make it a significant proportion we are simply indicating to the student the relative unimportance of these factors. The majority of academic staff have had little or no formal teacher education and so it not so obvious as to where and how they learn what criteria to use and how to apply them. Some of these things are learnt in collegiate discussion and some are no doubt absorbed by constant exposure to academic writing in journals. One's ideas of appropriate grammar and spelling may even be carried over from high school. With the advent of mass higher education and with the increase in overseas students in Australian universities we should not simply ignore the variations in criteria indicated by the data in this study. Whether one decides to act on data such as reported here depends on the importance that one places on the quality of student writing. Even with the limited data available from this study I believe the issue is important enough for School and Departmental heads to at least promote discussion and the exchange of ideas. It is also clear that a substantial research study is warranted. PO Box 423 Victoria, 3280 AUSTRALIA