Last updated 26 Nov 07
Abstract
In the French constructions described in (1) and (2), the impersonal pronoun il and the demonstrative ça (ce, cela, c' ) are virtually interchangeable:
(1) "Hé là! Sainte Vierge, est-il possible que le bon Dieu veuille
faire souffrir ainsi une malheureuse créature humaine?"
(Proust M. Du Cote de Chez Swann)
'Hey! Holy Virgin, is it possible that the Good Lord may want to
inflict such suffering on a poor human creature?'
(2) Est-ce possible que durant toute ma vie terrestre, je n'obtienne
jamais un peu de justice? (Bloy L. Journal T. 2)
'Is this [it] possible that during all my life on earth I will never
get a little justice?'
Despite this similarity, most accounts treat il constructions as impersonals and the structures which contain ça as dislocation constructions, and therefore analyze the two pronouns in radically different ways. For example, Jones (1996:128) claims that: "impersonal sentences are constructions in which the subject position is occupied by a dummy pronoun il, which does not refer to anything, while "ce or ça are not impersonal constructions, but dislocated constructions" with the consequence that "the demonstratives are not 'dummy' pronouns, but referential expressions which refer forward to the finite or infinitival clause" (Jones 1996:128).
This position is unconvincing for two reasons. First, the claim that impersonal pronouns do not refer to anything has been called into question by the linguistics community since Bolinger's influential "ambient it" paper (Bolinger 1973). More recently, a number of researchers (Achard 1998, Langacker 1991, 2004, to appear, Smith 2006) have claimed that the impersonal pronouns in different languages (French il for Achard, English it for Langacker, German es for Smith) are indeed referential, albeit very general, expressions. Beyond the details of the specific analyses, these researchers treat the impersonal pronouns as profiling the "field", i.e. the abstract domain in which interactions between participants potentially occur, as the focal figure in the profiled relation. Secondly, although il and ça are quite different in some of their functions, sentences such as (1) and (2) above show that both nonetheless present the semantic characteristics of impersonals (generality of meaning and presentation from the standpoint of a general conceptualizer) when their usages overlap.
In the face of these difficulties, and in an effort to capture the intuitive similarity of sentences such as (1) and (2), an alternative account is proposed which centers on the following positions: First, there is no structural difference between il and ça. Both are referential expressions, even though il is consistently more abstract and general. Secondly, if il is always impersonal, ça has other uses, but in specific constructions (weather verbs and some complement constructions), it exhibits the semantic characteristics which justify the "impersonal" label. Thirdly, impersonal ¸a's characteristics are directly observable in its other uses. Fourthly, il and ça profile different reasons for evoking the scene coded in the complement scene: il constructions pertain to its existence, ça constructions to its categorization/evaluation. Finally, this semantic distinction between il and ça explains their syntactic distribution in cases of mutual incompatibility and semantic overlap alike.
References
Achard, Michel. 1998. Representation of Cognitive Structures: Syntax and Semantics of French Sentential Complements. Berlin and New-York: Mouton de Gruyter.
Bolinger, Dwight. 1973. Ambient it is meaningful too. Journal of Linguistics 9: 261-270.
Langacker, Ronald. 2004. Aspects of the grammar of finite clauses. In Michel Achard and Suzanne Kemmer (eds.), Language, Culture, and Mind, 535-577. Stanford: CSLI.
Langacker, Ronald. To appear. On the subject of impersonals. Unpublished Manuscript.
Smith, Michael. 2006. The conceptual structure of German impersonal constructions. Journal of Germanic Linguistics 17.2: 79-138.
© 2007 Michel Achard
Last updated 26 Nov 07