Contenu du sommaire : Les proformes indéfinies ; indéfinition et prédication

We show in this paper that all the uses of QU- words, as indefinite pronouns as well as connective words, can be explained from a lexical and a predicative constraint ; the first constraint, of semantic nature, excludes any determinate use inisolation, the form being interpreted as a free variable ; the second constraint imposes, even in polarity contexts, a double prédication resulting in obligatory morphological extensions, even in the pronominal uses, these extensions being more or less grammaticalised.

The paper makes assumptions about the grammatical category of the form que in French and the syntactic relations in which are involved the constituents it introduces. In all its uses in declarative sentence forms it is analysed as a complementizer, contrasting with its pronominal status in interrogatives. As for the syntactic relations involving the whole constituent that que introduces, it is shown that they are of three types : – a pragmatic marker linking an utterance to various “discourse patterns” – a marker of conventionalized discourse routines resulting in “quasi-modifiers” – a member of a specific type of “correlative” construction.

In this paper, a unified description of the introducers of ‘completive subordinate clauses' in French (as in Je crois qu'il va pleuvoir ‘I think that it is going to rain') is given, taking into account the que P / ce que P variation (as in Je suis heureux que P / Je suis heureux de ce que P ‘I am happy that S'/ literally ‘I am happy of this that S'), and linking the ‘completive' que to other uses of que (interrogative, relative, ...), within an overall theory of Qu-words (described in other papers). The basic hypothesis is that ‘Qu- words' in French, like their English counterparts ‘Wh- words', build a word class whose main characteristics is to introduce variables, in various manners corresponding to their different uses (interrogative, indefinite, subordinative, ...). No ‘complementizer' is considered : que, in its function of introducing ‘completive subordinate clauses', is a full-fledged ‘Qu- word', namely a pronoun used in a highly abstract way.

The present study aims at characterizing the status of the element que in equative constructions, as represented by the examples (1) and (2) : (1) a) Robert est aussi intelligent que Maria. (2) a) Robert boit autant qu'il mange. This use of que seems to be located at the fringe of the already widely described properties of the indefinite proform (PI) in interrogative or relative clauses. This use can obviously not simply be reduced to an instantiation of the conjunction. The analysis of the functioning of equative que will lead to three conclusions concerning its status and its role in the connection of clauses : a) Que will be characterized as a scalar adverbial, restricting the range of the equality/ superiority/ inferiority with the degree of the standard it introduces. b) Que operates within a correlative relation and connects a secondary predication to an element of the predicative structure of the prior predication, but without reclassifying it. c) If a mutual dependence of the two components of the marker is needed to express the equative function, que is still syntactically governed and semantically directed by the first component (aussi/autant).

The authors study the proform quoi when it is anaphoric of a predicative phrase (Il déjeuna. Après quoi il partit). In this structure, this proform loses several properties – what shows a grammaticalisation in progress. The authors make the assumption that quoi functions as a relative pronoun or an indefinite pronoun without value of subordination. Five groups emerge : In the first one, the demonstrative ce is essential as antecedent to the pronoun quoi, with a predicative value (Ce à quoi il parvint). In the second one, ce is optional (Ce en quoi il a raison / Ce à quoi elle répliqua que P) ; it can have a predicative value with the preposition en but not with the preposition à. In the third and the fourth groups, ce is impossible : in the third group, the group prep. + quoi (à quoi) assumes an argumentative function whereas in the fourth group, the group prep. + quoi (après quoi) is an adverbial phrase. At last, in the fifth group, quoi can be analysed as an indefinite pronoun : the utterances it introduces accept a different modality from the previous sentence, i.e. the interrogative modality.

Estelle Moline accounts for a broad panel of comme-constructions. She shows how the presence of the comme morpheme steadily correlates with the suspension of assertion. She first describes the cases where comme corresponds to the qu- manner proform : in such cases, no assertion is possible within a simple clause, and embedded clauses trigger an indefinite presupposition. She then deals with constructions where comme is not a qu- manner proform : here, in complex clauses, the presupposed content of the embedded clause is definite ; in simple clauses, comme generally shows a kind of distance-taking from the assertion.

Où que P and n'importe où, non specifying pro-forms, occupy regularly positions which are more satellite than those of forms such as qu- que P and n'importe qu-. The aim of this study will be to describe how relations between n'importe où or où que P and verbal nucleus interfere with their semantic functioning : concessive, negative polarity or free choice effects. Où que P seems to prefer the value of concessive conditionals without denying the use of polarity or the use of free choice. In the latter situation we essentially observe a more condensed integration. On the other hand, n'importe où functions generally as a free choice marker, but it accepts exceptionally a polarity application in all contexts supporting a downward monotone orientation. In this case, it prefers a more loose connection with the main clause or with one of its constituents.

In this paper, we suggest that French morpheme quand is a “proform”, and not a conjunction, in some of these contemporary uses. This result is based on several distributional properties and on some attested examples from varied sources : spoken French, Web, literature, etc. With this corpus-based approach, we also question the relevance of analogy between parts-of-speech, syntactic relations and semantic values.