The Grammar of Genres and Styles

From Discrete to Non-Discrete Units

Dominique Legallois, Thierry Charnois, Meri Larjavaara, Dominique Legallois, Thierry Charnois, Meri Larjavaara

• Publisher: De Gruyter Mouton
• Year: 2018
• Topic: Languages & Linguistics
• eBook ISBN:
  9783110592849
• Subtopic: Linguistics
• Index: Languages & Linguistics

The book provides new findings about the grammar of genres and styles. It combines new methods with different kinds of empirical material, from social reports to live TV sports commentaries or 16th century newspapers, in English, French, Latin and Spanish. The study of non-discrete units suggests new ways of seeing the linguistic variation between genres and styles and the ways in which belonging to a genre predetermines linguistic choices.

Dominique Legallois, Thierry Charnois and Meri Larjavaara

The balance between quantitative and qualitative literary stylistics: how the method of “motifs” can help

Note: We are grateful to the ANR Programme franco-allemand en Sciences humaines et sociales (FRAL) – Projet PHRASEOROM – for financial support.

Dominique Legallois, University of Paris 3 Sorbonne Nouvelle

Thierry Charnois, University of Paris 13

Meri Larjavaara, Åbo Akademi University

Abstract:In this chapter, we study 60 novels by twelve 19th century French authors, aiming to show the complementarity between a stylistics of identification (based on stylometry techniques), and a stylistics of characterization (adopting a qualitative approach). The two stylistics belong to very different traditions, but by taking motifs as units of analysis, it is possible to identify some of the lexico-grammatical patterns typical of each author. The study presents in detail the method of extraction of patterns (motifs) and proposes many examples of stylistic features, especially among the novelists Hugo, Balzac, Flaubert, and Gaboriau. Very often, these features have not been identified by traditional stylistics.

At a time when computational methods of extraction of the morphosyntactic specificities of a given author or a literary genre are becoming more and more accessible thanks to new computational tools, linguists and specialists in stylistics and literary discourse can no longer remain unaware of the contribution of quantitative data. Admittedly, a quantitative approach is only one aspect of the stylistic analysis of texts, but this criterion has to be taken into account since statistically overrepresented features of language in texts can be considered as linguistic singularities, peculiar to an author, to a text or to a genre. Some analyses aiming to identify the linguistic specificity of a corpus give extremely interesting results, as evidenced by the small but growing number of monographs or articles demonstrating the approach: for example, in the French textometric tradition, Brunet (1985) on E. Zola, Magri (2009) on the comparison between travel writing and fiction, Kastberg Sjöblom (2006) on J.M.G. Le Clézio, or in other traditions, Burrows (1987) on J. Austen, Hoover (2007) on H. James, Ho (2011) on J. Fowles, Oakes (2014) for various applications, etc.

However, computational stylistics remains marginal for “academic” literary specialists in stylistics, despite the current development of digital humanities. The reasons for this marginal position may stem from the complexity of the methods used, the technical nature of analyses, or a difference of “culture” between text practitioners. There may however be other, deeper reasons than these: namely, the fact that statistical and automated analyses are invariably paradigmatic. They concern discrete descriptors, that is to say, units (lexical forms, lemmata, morpho-syntactic categories, punctuation marks) with no direct syntagmatic relations between one another, even if correlations can always be calculated. The contextual interpretation of these descriptors remains vague, therefore, since the main focus is commenting on tables of quantified units, rather than utterances. Highlighting the specificity of a descriptor is of course essential, but to grasp the role of the descriptor in its textual environment is a step that is still too rarely taken. This is the limitation of the discrete approach and may explain why its adoption by specialists in stylistics has been rather lukewarm so far.

The main aim of this contribution is to show how the study of an important area within stylistics, namely the characterization of an author’s style, can benefit from a new method in corpus linguistics, the discovery of sequential patterns or “motifs”, i.e. contiguous strings of word forms/lemmas/POS tags. Motif analysis can be viewed as complementary to discrete approaches and constitutes a more powerful paradigm than other non-discrete analyses such as lexical bundles or clusters because motifs combine annotations at different levels of abstraction.

As will be seen from the analyses given below, this type of complex unit, while not claiming to be exhaustive since a style cannot be reduced to a set of lexico-grammatical patterns, nonetheless provides a more accurate vision of the stylistic characteristics of an author.

The study is based on a corpus of 60 novels by 12 19th-century French writers (see below). The aim is to identify the syntactic motifs favored by each author.

In the first part of the chapter, we examine the issue of the definition of style proposed in a recent article by Herrmann and colleagues (Herrmann et al. 2015), in which the quantitative approach features prominently. Through a critical discussion of this study, we show that there are in fact two major trends in stylistics, and even two types of stylistics: a stylistics of identification, which has its roots in the stylometry of the nineteenth century, and a stylistics of characterization. The first one (essentially quantitative) is a stylistics based on discriminatory features; the second (essentially qualitative) is interpretative. In order to ensure that quantitative and qualitative approaches are complementary, it is necessary to take not only discrete units, but also non-discrete units such as motifs into account in the analysis.

The second part deals with the presentation of the method of motifs. We formulate a definition of the term and we introduce three possible statistical methods of analysis: an endogenous method, an exogenous method, and a combination of the two.

In the thi...

Table of contents

• Cover
• Title Page
• Copyright
• Contents
• Grammar of genres and styles: an overview
• Text types, audience and thematic organisation in the recent history of English
• Reference chains and genre identification
• Taking into account coherence relations to describe a textual genre: methodology and application to the discourse of tourist attraction guides
• Linguistic features of genre and method variation in translation: a computational perspective
• Approaching French theatrical characters by syntactical analysis: a study with motifs and correspondence analysis
• Towards a topological grammar of genres and styles: a way to combine paradigmatic quantitative analysis with a syntagmatic approach
• The balance between quantitative and qualitative literary stylistics: how the method of “motifs” can help
• Live TV sports commentaries: specific syntactic structures and general constraints
• Bursts of written language as performance units for the description of genre routines
• Index