Lingvisticæ Investigationes - Volume 18, Issue 2, 1994

Plus-que-parfait: past anterior or past perfect?Following the author's investigation into the use of plus-que-parfait, passé antérieur and passé surcomposé by means of a gap-filling exercise administered to native speakers (Engel forthcoming), it appears that plus-que-parfait (PQP) is not a rival to the other two past anterior forms, despite their relative infrequency, i.e. it cannot replace them in particular contexts. Furthermore, PQP is not the automatic choice in seemingly appropriate PQP environments. It has also been noted by various scholars (such as Majumdar & Morris 1980) that PQP is increasingly taking on a past punctual function, which is of some interest considering its affinity to imparfait, i.e. the non-punctual past. This paper reexamines data from the author's corpora of newspaper texts, looking at the role of PQP, in order to establish its particular function in modern journalistic texts. Finally, an overview of the various shifts in the French tense system is presented, paying particular attention to the past and past anterior forms, that area which has witnessed the most change.

The phenomenum of "false liaison" (linking errors) in French has almost never been studied in itself. A considerable number of examples from different sources are examined from a lexical, morphological and syntactic perspective. If many instances can be satisfactorily reduced to phrasal affixes (Miller 1991), the remaining cases fall under two types of explanations, analogy and "liaison à distance" (linking from a distance) (Pichon 1935), each of which alone is not sufficient, like the postulation of plural and verbal markers (Kaye & Morin 1982), to cover all the data, and a comparison of the consequences of these accounts is offered. Accounts on syntactical grounds do not make strong predictions, if at all, on this phenomenum, since false liaisons seem to be only loosely restricted. It is suggested that an explanation invoking analogy is possible on a morphological basis for many cases, and that some cases of liaison at a distance may be due to performance error involving the syntactic production stage of an utterance.

This paper presents an analysis of about 400 idiomatic expressions that belong exclusively to the regional variant of French spoken in Belgium. The model adopted is that of M. Gross's lexicon-grammars: each expression is listed as a lexical entry and classified according to its syntactic structure, in one of the classes that were proposed by M. Gross for standard French idiomatic expressions.Our research shows, on the one hand, that all the classes of standard French are represented in Belgium French and, on the other, that certain subtypes are particularly productive in Belgium while others seem to be less common than in standard French.

Relative Clauses (RCs) have been described and classified according to many different criteria (Sect. 1). This article deals with the distinctions that can be observed within the variety of English RCs in terms of phonological shape, syntactic structure, semantic content, presuppositions and thematic structure of the utterance. A classification of RCs is proposed, which is based upon criteria of all different levels (Sect. 2). The aim is to provide a way to characterize as precisely as possible the functional features of any RC in correlation with its formal features. The classification is then applied to an analysis of the often stated equivalence between RCs and other kinds of linguistic constituents, leading to the recognition of the different types of RCs that are functionally equivalent to adjectives, participles, prepositional phrases and coordinated sentences (Sect. 3). A further application proposed is the comparison between the functional and formal features of English and Italian RCs (Sect. 4).