Le sujet en langue des signes française

Impersonal constructions are a highly debated issue in the studies of spoken languages, especially regarding the description of the Romance and Slavic languages. Impersonal constructions are linguistically distinguished from personal constructions, i. e. verbal forms associated with a referential argument in the subject position. The present study is devoted to impersonal constructions with a non-referential subject, while taking into account a very specific language : French Sign Language (LSF). While spoken languages are said to be visual-gestural and audio-vocal languages, signed languages are characterized by their exclusive visual-gestural modality, involving different features of their linguistic system, such as iconicity or – what is at issue here – the referential use of space. In sign linguistics, spatialization refers to the signer’s ability to introduce the reference to an entity linked with a specific location in space. For example, this referential strategy allows him or her to point at a location to refer back to the argument that was linked to it earlier in the discourse (Lillo-Martin, 1986 ; Bahan et al., 2000 ; Meurant, 2008 ; Risler, 2013). Therefore, we aim to study how a loss of referentiality can be expressed in a visual-gestural language. In the recent past, linguistic descriptions of sign languages have provided evidence of agent-defocusing structures, such as the passive (Janzen et al., 2001 ; Guitteny, 2005) or the impersonal constructions (Barberà & Quer, 2013 ; Kimmelman, 2016). Hence, on the basis of previous findings on impersonal structures in spoken languages, and taking into account studies describing the syntactic functioning of signed languages, this article analyses personal and impersonal structures in LSF. Different items lead our path of reflexion. Based on identification features of the subject in spoken languages, we define in this article the properties of the subject in LSF. With regard to the literature, we distinguish primarily two paths for analysing the subject in sign language. On the one hand, French studies reveal that thematic roles are predominant in the analysis of LSF, whereas the syntactic functions of the phrasal constituents are rarely set out. However, the interpreted sentences illustrate in general the choice of an active perspective : the agent corresponds to the syntactic function of subject. We can conclude from these analyses that, with agreement verbs (Padden, 1988), the starting point of the verbal movement is analysed as marking the agent-subject position. On the other hand, the study of Meir et al. (2007) offers different analysis possibilities, by assigning the syntactic function of the subject to the signer’s body, regardless of its thematic role. In light of these studies, our article highlights possible interpretations of personal and agent-defocusing constructions in French Sign Language, and raises the question whether the subject function is marked on a specific location or not. In order to conduct this study, we have chosen to work on empirical data, extracted from semi-spontaneous and narrative corpora, since they favor referential structures and anaphora. The analysis consists of two stages. The first part of this study is devoted to the analysis of basic sentences in LSF, in which the personal subject is specified, in order to define the features of the subject in this language. This analysis raises a syntactic specificity of sign language : like other signed languages (Lillo-Martin, 1986 ; Bahan et al., 2000 ; Kimmelman, 2017), LSF accepts the argument position of the subject, and/or the object, to be lexically empty. In the context of this study, we focus on the subject argument and note that the subject position is not systematically filled by a lexical component. Our analysis suggests that the arguments are mentioned by the expression of anaphora, in the predicative form : the syntactic mark of the subject argument on a spatial location allows the spatial instantiation of the argument in the inflection. Moreover, the observation of personal references allows us to note the role the body plays in the argumental structure, encoding the properties of the subject argument. Following Meir et al. (2007), we analyse this corporal locus as marking the syntactical function of the subject. However, we suggest that the syntactic function of the subject is marked by the corporal locus, irrespective of the type of the verb. The second part of this study is devoted to structures with an unspecified human subject and compares them to the personal constructions analysed in the first part. The comparative approach does not show syntactic distinctions between these personal and impersonal structures. Indeed, we observe both lexical and empty subject constructions with personal as with non-specified subjects. Finally, with respect to the role of the body, this comparative study allows us to observe that the body encodes the properties of the first argument regardless of its referentiality. Then, in the absence of syntactic differences between personal and impersonal references, only the semantic level allows us to notice the distinction. In personal as well as in agent-demoting constructions, this assignation of the subject function to the body calls into question the interpretation of the starting point of agreement verbs’ movement. Indeed, if the end point of the verbal movement – which represents the patient – is on the signer’s body, and is assigned the function of subject, it may consequently be a structure defocusing the agent. This assumption is corroborated by the observation of the use of specific spatial markings, on a high and lateral space, emphasizing the process by not referring to any specified agent. Similar spatial markings have been studied for Catalan Sign Language (LSC) (Barberà & Quer, 2013), giving us an interesting perspective to guide our analysis of this phenomenon. Moreover, previous French studies on the role of space in the actantial distribution in LSF describe the use of a high lateral space, on each side of the signer’s head, for the reference to an indefinite agent (Millet, 2004, 2006).These considerations in relation to the assignation of the subject syntactical function to the signer’s body or to the starting point of the verbal movement led us in our questioning of the passive or the active perspective. Though it is too soon to settle the question, this analysis suggests that a distinction between impersonal and passive constructions could be drawn. Thus, the questioning raised by this study will be the basis for further research on defocusing agent strategies and their semantic tendencies in LSF, with the aim of distinguishing impersonal and passive constructions in this language