Voice in Scientific English: A Corpus-Based Approach to the Use of Active and Passive Voice in Medical Research Articles in English

Abstract

[eng] The notion of voice has attracted significant attention in academic writing (John 2005, Herrando-Rodrigo 2019, 2022, inter alia) as it is conceived as a compelling device to ascertain how specialised discourse writing is shaped. The interrelatedness shown between voice, authorial (in)visibility and (im)personality makes it a hard concept to delimit, as each of them may partially entail any of the others. As impersonality, at least in medical writing, has traditionally been related to the extensive use of the passive voice, taking grammatical voice as a starting point to operationalise the notion of impersonality seems well-justified. The corpus-based study presented in this thesis aims at quantifying a set of active and passive structures recurrent in RCTs. They are quantified through corpus searches using SQS and CQPS. The high frequency of these structures, after scrutinising a set of real RAs, allows their inclusion within the restricted set of lexicogrammatical patterns typically occurring in medical register. Quantitative and qualitative analyses are carried out to determine the extent to which active and passive structures primarily relate to (im)personality, and secondarily, to authorial (in)visibility. This study concludes with a proposal to scientifically categorise the structures analysed in the corpus-based study through a Cartesian coordinate system, in which mainly two distinct parameters or axes are used, namely personality (x-axis) and grammatical voice (y-axis). The corpus-based study is complemented with a close textual analysis, whose main purpose is to attempt to disentangle the complex and multi-layered dimension of impersonality by exploring the extent to which the strategies of (im)personalitsation described interact to textually characterise medical discourse as being considerably impersonal. Through a manual classification of the lexicogrammatical patterns analysed, I attempt to assess the degree of impersonality they carry with them by analysing the extent to which these instances relate to authorial presence (or absence) in the text. This study is also aimed at assessing the pragmatic functions these patterns perform within the ii text as far as authorial (in)visibility and (im)personality are concerned. Through the exploration of the interplay between these patterns, an evaluation of how such interaction affects the transmission of knowledge and the degree of impersonality it transmits has also been done. The gradation of structures from more personal to more impersonal-like is represented through a continuum, so that the overall picture of impersonality and authorial visibility (also including information on grammatical voice) in RCTs is visually shown. The results of both studies decipher the interrelatedness between the notions of impersonality, authorial (in)visibility and the use of grammatical voice in medical discourse using a mixed-method approach, in which quantitative data are subsequently qualitatively analysed. The effect other lexicogrammatical patterns exert over this impersonal perception is also seen as fundamental, as they may be used to either reinforce this impersonal reality or counteract it. To explore the correlation between these patterns and three distinctive analytical dichotomised parameters, namely personality vs. impersonality, active voice vs. passive voice, and authorial visibility vs. authorial invisibility, has been crucial to unravel the complex notion of impersonality, and both a Cartesian coordinate system and a cline have proved to be excellent methods to do so.