Evaluative adjectives in academic writing in the humanities and social sciences

Abstract

This study deals with evaluative adjectives in French academic writing in the field of humanities and social sciences (linguistics and economics) through a corpus study of various kinds of texts (research articles, theses, course books). Although not as much attention has been paid to adjectives in French as to other parts of speech, I believe that this category plays a prominent role in argumentation and persuasion and can shed light on the rhetorical strategies used by an author. Following Kerbrat-Orecchioni's (1980) typology of subjective adjectives, I performed a corpus study on evaluative adjectives referring to scientific nouns such as scientific artifacts, scientific observables, relations, and qualities in order to study disciplinary variation and to identify the most common patterns. The results show that axiological evaluation is not very common, in contrast to more "neutral" evaluative types such as novelty, importance, time, comparison and complexity. Firstly, in order to convince the reader, authors seem to avoid very subjective evaluation in scientific writing. Secondly, recurrent associations are often cross-disciplinary and exhibit strong selectional preferences between nouns and evaluative adjectives: for example, temporal adjectives generally refer to scientific artifacts while axiological adjectives mainly refer to results. Thirdly, contrary to my expectations, evaluative adjectives of all semantic types are more numerous in economics than in linguistics, and especially those expressing importance and novelty, something that seems to highlight the importance of authorial self-promotion in this discipline. Finally, the use of evaluative adjectives seems closely linked to rhetorical strategies: they are especially numerous in introductions (and in conclusions in economics), where they are used mainly to justify and promote the author's work.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 17 citations exist upstream.