UiB : Faculty of Humanities : Department of Foreign Languages

Cultural Identity in Academic Prose:
national versus discipline-specific

(in Norwegian: Kulturell Identitet i Akademisk Prosa:
nasjonal versus disiplinavhengig, or KIAP)

Norwegian

Professor
Kjersti Fløttum
UiB

Postdoctoral research fellow
Torodd Kinn
UiB

Associate professor
Trine Dahl
NHH

Research project at the Department of Foriegn Languages (before August 2007: Department of Romance Studies), University of Bergen,
supported 2002-2006 by The Research Council of Norway.

• 2008
  PhD dissertations:
  Gjesdal, Anje Müller 2008. Étude sémantique du pronom ON dans une perspective textuelle et contextuelle. Universitetet i Bergen.
  
  Vold, Eva Thue 2008. Modalité épistémique et discours scientifique. Une étude contrastive des modalisateurs épistémiques dans des articles de recherche français, norvégiens et anglais, en linguistique et médecine. Universitetet i Bergen.
  
  2006
  MAIN PROJECT FINDINGS
  
  Discipline is more important than language with regard to cultural identities in academic discourse
• Discipline is a more important variable than language to explain similarities/differences between research articles.
• Authors of linguistics articles are more visible and argue more explicitly than authors of economics articles. Authors of medical articles are the least visible of the three disciplines investigated (linguistics, economics and medicine).
• English-writing and Norwegian-writing academic authors are more explicitly present in the text than French-writing authors.
• Medical articles in Norwegian tend to display more explicit author manifestation than medical articles in English and French.
• Comparisons of single articles reveal considerable individual variation in academic discourse, both within one discipline and within one language.
• Various author roles have been identified. The three most important ones are author as researcher, author as arguer and the author as text organiser (or readers' guide).

Recent research has shown that the traditional conception of academic discourse and writing tradition as neutral, objective and devoid of personal traits needs to be reconsidered. The presence of personal traits in texts, however, varies with discipline and language. Such observations represented the point of departure for the KIAP Project. The main research question of the project was whether there exist something that can be called cultural identities in academic discourse, and if so, to what degree these are linked to discipline or national language. The aim was to describe the genre of the research article with a point of departure in certain linguistic features that may point to similarities and differences between articles written in different languages and within different disciplines. In this comparative project, articles in three different languages (English, French and Norwegian) from three different disciplines (linguistics, economics and medicine) have been studied. The investigation is concentrated around "academic voices", manifested in three specific research questions linked to person manifestation: 1) How do article authors manifest themselves in the texts? 2) How are the voices of others reflected? 3) How are the authors' attitudes expressed through the presentation of their own research? Our main hypothesis has been that discipline plays a more important role than language with regard to cultural identities.

Our material consists of an electronic corpus of 450 refereed research articles from recognised journals, published in the period 1992–2003. The total number of words is 3 152 022. The articles are divided into nine subcorpora with 50 articles in each.

In order to answer the three research questions we have investigated the frequency and use of the following six linguistic (rhetorical and argumentative) features (only the English expressions are given here): first person pronouns (I/we), indefinite pronouns (one), metatextual expressions (referring to (parts of) the text itself, such as in this article), negation (not), adversative/concessive constructions (with connectives such as but) and bibliographical references.

In order to get a picture of linguistic practices in relation to the two variables of discipline and language, both quantitative and qualitative analyses have been undertaken. To identify individual differences between single articles, a number of case studies have been carried out. The theoretical framework comprises rhetoric of science, genre theory, enunciative theory and the theory of linguistic polyphony.

Our general results show that discipline is more influential than language with regard to cultural identity. Our main hypothesis was confirmed. This implies that there is greater similarity of form between a medical article in Norwegian and French than between a medical article in Norwegian and a linguistics article in Norwegian. Both discipline and language have an effect on the frequency of all the six features, but for most of them discipline is more important than language. The results also show that there are great individual differences between single articles, even within one discipline and one language. As regards gender, the project has investigated whether this may represent a relevant variable to explain differences between single-author articles. No systematic connections have been found that can be linked to the author's gender.

Below we list the most important findings linked to each of the features:

1) First person pronouns – singular and plural

First person pronouns may be considered as the most direct form of author manifestation. They are clearly more frequent in linguistics and economics than in medical articles. English and Norwegian authors use them more than French authors. Norwegian authors tend to make frequent use of the plural form (vi '='we') in single-author articles.

2) Indefinite pronouns

Indefinite pronouns (one) represent a less direct form of author manifestation due to their vaguer reference. Such pronouns do, however, represent a relevant feature as the reference is always to a more or less indefinite group of people. French and Norwegian in this case display a much higher frequency than English, something which may be explained by the fact that the use of English one is very different from the use of French on and Norwegian man, en/ein. Norwegian medical texts stand out as being frequent users of this pronoun.

3) Metatextual expressions

Metatextual expressions are expressions referring to (parts of) the text itself. They are relevant markers of author manifestation in the sense that the authors through such expressions enter the text in order to guide the reader in some way. The expressions often state what will be done where (e.g. In section 2 we show …). This phenomenon is practically absent in medical articles in the three languages investigated, but it is frequent in linguistics as well as in economics articles. As for language differences, English-writing authors use such expressions more than Norwegian-writing authors, and Norwegian more than French.

4) Negation

To the extent that they are used polemically to refute underlying points of view, negation constructions (not) may say something about author manifestation as well as implicit presence of other "voices", so-called polyphony. Our results show that generally speaking linguists use much more negation than economists and medical authors, and specifically that Norwegian researchers use negation more often than their English- and French-writing colleagues.

5) Adversative/concessive constructions

The conjunction but (French mais and Norwegian men) may be used in many different ways. It is contrastive, but may also have a concessive value. In the latter case the authors first make a concession with regard to something expressed in the sequence preceding but (and this issue may originate in a different voice so that the result is a kind of polyphony). Then this is contrasted with a subsequent sequence presented as the most important one (the author's point of view here and now). In this way concessive constructions contribute to a positioning of the authors. Linguistics is the discipline with the highest frequency of concessive constructions. In terms of language, Norwegian uses them most frequently, something which also corresponds to the findings for negation. Norwegian medicine is a particularly heavy user of concessive constructions compared to English and French medicine.

6) Bibliographical references

Bibliographical references represent an important component in all scientific publishing. In general they are used to document claims and to refer readers to sources for further studies. Through comprehensive standardisation procedures medicine today displays a relatively homogeneous convention. References in medical articles are listed at the end of the text, with numerical reference in the text itself. References in economics and linguistics, on the other hand, are typically inserted with name, year and sometimes page references in the text. Such references may also contain quotes. In this way other researchers' voices are heard (directly) in the text. When taking such discipline differences into account, our findings show that in general medical authors, irrespective of language, use most bibliographical references. Within medicine, English and Norwegian authors have a higher frequency of references than the French.

Our general conclusion is that authors within linguistics and economics are much more overtly present in their texts than medical authors. As regards language, English-writing and Norwegian-writing authors are more visible in the text than French-writing authors.

Academic discourse is argumentative and rhetorical in the sense that authors attempt to persuade and seek acceptance while at the same time trying to position themselves strategically in relation to certain audiences. Such positioning through explicit argumentation follows naturally from the competitive situation characterising today's scientific publication. It is, however, important to state that argumentation and rhetoric are not only to be found in personal statements introduced by a personal pronoun or in statements with polemic negation. Persuasive strategies may of course also consist in presenting facts or observations in "neutral" statements that may appear objective and devoid of personal traces, such as The water in Little Lake is polluted. Medical discourse in all the three languages investigated is less explicitly rhetorical than the discourse of economics, and particularly that of linguistics. This claim is not only supported by our quantitative results; KIAP's qualitative investigations point in the same direction.

The use of first person pronouns shows e.g. that authors take on different roles. The KIAP project has identified at least three different academic author roles. Linguist authors assume all of these in the text, as researchers (we investigate, we find), as arguers (we claim) and as readers' guides structuring the text (in section 3 we present). Economist authors typically restrict themselves to the researcher and readers' guide roles, while the medical authors rarely operate outside the researcher role. The strong adherence to the IMRAD article structure (Introduction, Material/Method, Results and Discussion) probably implies that there is less need to guide the reader through the text than in the disciplines of economics and linguistics where the article structure typically is much more heterogeneous. One reason why medical authors argue less explicitly than e.g. linguists may be that medical research is based on commonly accepted methods and a broad common knowledge base. For linguists, the fact that there is less of a common knowledge base may explain why they have to discuss and argue more for a common understanding of concepts and for delimiting research fields. In addition, research findings in linguistics are often produced through argumentation in the text, as part of the writing process.

As regards the way other researchers, those referred to in the text, are "given the floor", a similar picture is seen. Cited linguists may argue as well as think, while cited economists and medical authors typically are given the researcher role through finding or showing results and observations. This points to explanations which will not be elaborated on here, but which relate to different scientific traditions, different objects of study, different research methods as well as different journal guidelines.

Some limitations

The KIAP Corpus is too limited to serve as a basis for a general study of language use in research articles,. However, our findings clearly point to important trends for the three disciplines and the three languages which are part of the study.

The selection of linguistic features analysed is also limited. An exhaustive analysis of relevant features for academic discourse has not been undertaken, but for the focus adopted in the study, viz. person manifestation, the selected linguistic features are relevant.

The combination of quantitative and qualitative methods applied has made it possible to posit statistically valid findings for individual phenomena, generally, in addition to important observations in individual research articles, specifically.

The results confirm our main hypothesis: To the extent that it is possible to talk about cultural identities in academic discourse, in relation to the phenomena investigated, discipline plays a more important role than language. However, a number of the observed differences cannot be explained by reference to either discipline or language. Variables such as age, academic position, subdiscipline, topic and journal practice may be other possible contributing factors.

For a list of publications and more information on the project, see this site www.uib.no/kiap/. Contact person: Professor Kjersti Fløttum, Department of romance studies, University of Bergen (kjersti.flottum@roman.uib.no).

Summary 2002:

The activities of the KIAP project are based on the following key issue:
Can cultural identities be identified in academic prose, and, if so, are these identities national or discipline-specific in nature?
     We attempt to answer this question through a contrastive study of English, French and Norwegian research articles within the fields of economics,linguistics and medicine.
      This is a linguistically based project, and takes as its point of departure analyses of linguistic features which may be linked to argumentation. More specifically, we study

1. how the article writer manifests him-/herself in the text
2. how works by other researchers are reflected in the article, and
3. how the author's attitudes are expressed, through the presentation and promotion of his or her own research (epistemic axis and evaluative axis)

Examples of relevant features are first person pronouns, meta-text (Ex: In this article I argue that…), the use of bibliographical references (Ex: Brie 1996 maintains that the moon is made of cheese.), and the use of evaluative expressions (Ex: This approach is highly problematic.).
      Such features may be linked to cultural identity. The findings of the KIAP project may hence be used to answer the question of whether the identity of academic writers is primarily national or primarily discipline-specific in nature.
      The project includes manual as well as automatic processing of the material. Both the quantitative and the qualitative analyses are based on data generated from a large electronic corpus comprising 450 research articles. This corpus will also be used for another purpose, viz. to identify features that can be used in automatic summarisation program for argumentative research articles.
     With a point of departure in genre theory and rhetorics of science, the linguistic studies are related to pronominal use, modality, metadiscourse, linguistic polyphony, reported speech and lexico-semantic phenomena.

 Academic Voices in Contrast

 About the project

KIAP Publications

KIAP Corpus

KIAP Events

Project-related publications

Bibliography

Cooperating research groups

Related projects

www.hf.uio.no/inl/sakprosa
www.msh-paris.fr/texto
www.hum.au.dk/romansk/polyfoni
www.hf.uio.no/german/sprik
www.ci-dit.com