Carlier, Anne, Veronique Lagae, and Celine Benninger, eds., (2000) Passe et
parfait, Cahiers Chronos 6, Rodopi, Amsterdam and Atlanta, Georgia, v+142
pages, ISBN 90-420-1211-0 (paperback).
reviewed by Robert I. Binnick, University of Toronto at Scarborough, Canada.
This volume, the latest in the series of "Cahiers Chronos," presents the
proceedings of the third Chronos colloquium, held at Valenciennes on 29-30
October, 1998. As the title indicates, all the papers concern the past or
the perfect of French, with two exceptions (on, respectively, Latin and
Serbo-Croatian). Six of the ten papers form a core group concerning
temporal order in discourse and specifically the contribution of the French
imparfait (imperfect tense) and passe simple (simple past tense) to the
construction of temporal order. These papers defend or criticize the view,
propounded in Kamp and Rohrer (1983), that the passe simple, in and of
itself, serves to advance narrative time, while the imparfait does not. A
theme running through many of the papers in the volume is the atypical use
of past tenses--the perfective imparfait, the imperfective passe simple, the
passe compose (complex past tense, in form a present perfect) used as a
1.0 BOOK SUMMARY
(A considerably longer, detailed summary of the individual papers in this
volume is to be found at
1. Jacques Moeschler, "L'ordre temporel dans le discours: le modele des
inferences directionalles" ["Temporal order in discourse: the model of
Moeschler distinguishes forward inference (inference en avant) and backward
inference (inference en arriere), which he identifies roughly with,
respectively, the discourse relations "narration" and "explication" (Asher,
1993; Asher et al., 1995). In narration the order of events as recounted is
that in which the events occurred, while in explication there is a causal,
not a sequential, relation between the events recounted, and they may be
presented with an inversion of their order of occurrence.
Directional inferences are calculated in a process in which real-world
knowledge plays a role. Directional inferences are partly the result of
directional traits associated with conceptual information and procedural
information (encoded in non-lexical categories such as negation, tenses,
connectives, etc., and giving instructions regarding the manner of the
construction of mental representations). The complete representation of an
occurrence is calculated by combining the two types of information.
Contextual hypotheses validate the directional hypotheses implied by
linguistic expressions and cancel those inferences when the contextual
information is contradictory to the direction of time inferred from the
linguistic material. The passe simple forces a forward inference, as in
(1), while the pluperfect forces a backward one (2).
(1) Max poussa Jean. Il tomba. 'Max pushed John. He fell.'
(2) Max poussa Jean. Il etait tombe. 'Max pushed Jean. He had fallen.'
The passe compose is ambiguous, only real-world facts causing one to prefer,
out of context, one or the other reading. But the connective et 'and'
forces a forward inference (3) with both passe simple (3a) and passe compose
(3b), while parce que 'because' forces a backward inference with both tenses
(3a) Max poussa Jean et il tomba. 'Max pushed Jean and he fell.'
(3b) Max a pousse Jean et il a tombe. 'Max pushed Jean and he fell.'
(4a) Jean tomba parce que Max le poussa. 'Jean fell because Max pushed him.'
(4b) Jean est tombe parce que Max l'a pousse. 'Jean fell because Max pushed
Some directional traits are strong, others weak (conceptual information
always bears a weak trait, while procedural may be strong or weak). If a
strong trait is present, it gives the direction; in its absence, the
temporal direction is inferred from the context. Propositional procedural
information is stronger than morphologically encoded information, so in case
of a conflict between a connective and a tense, the connective wins (as in
4a). Context has priority over any directional trait, however.
2. Carl Vetters and Walter De Mulder, "Passe simple et imparfait: contenus
conceptuel et procedural" ["Passe simple and imparfait: conceptual and
Vetters and De Mulder argue that tenses have conceptual content, and,
supporting the aspectual theory against Moeschler's, that the content of
certain expressions is heterogeneous, at once conceptual and procedural
(Kleiber 1994, 1997).
To the idea that temporal semantics is superfluous , they counter that
tenses do have their own temporal values and are not interchangeable in
texts. The temporal interpretation of utterances does not ignore theof
tenses, nor does it depends solely on pragmatic principles.
Nor are tenses are neutral in regard to order. In examples with inverse
order, the English simple past is rendered by the passe compose (5); the
passe simple normally imposes a sequential interpretation (6). (7), given
its sequentiality, is pragmatically difficult, implying that Jean dropped
the glass AFTER it broke.
(5) Le verre s'est casse. Jean l'a laisse tomber. 'The glass broke. Jean
(6) Jane me quitta (e1). Elle tomba amoureuse (e2) de quelqu'un d'autre.
'Jane left me. She fell in love with somebody else.' [e1<e2]
(7) ?*Le verre se cassa. Jean le laissa tomber. 'The glass broke. Jean
Those who argue that the content of tenses is purely procedural ignore
their aspectual values. Moeschler concerns himself solely with
lexically-supplied aspectual information, with"Aktionsart" (modes d'action),
placing aspect solely on the lexical level because affixation in his model
conveys only procedural information. However, our authors present four
arguments for considering tenses as playing a role in the conceptual
representation of occurrences via their aspectual values.
The most important of these is that there exist links between lexical and
grammatical aspect; the lexical aspect (Aktionsart) of a situation can be
reinterpreted by tenses (Moens and Steedman 1988, Vet 1994, Swart 1995,
Steedman 1998). Moeschler argues that it is the procedural character of
tenses that is responsible for these reinterpretations. But for Vetters and
De Mulder, such phenomena constitute a manifestation of a grammaticalized
aspectual content that cannot be described as an epiphenomenon of a more
fundamental procedural content.
Alternatives to the aspectual account focus on the opposition of
background/foreground, point of view, or bounding (bornage). These
approaches, it is argued, essentially restate the aspectual distinction in
3. Louis de Saussure, "Quand le temps ne progresse pas avec le passe simple"
["When time does not advance with the passe simple"]
Saussure's article concerns those unusual cases in which the passe simple
seems not only not to advance narrative time, but to reverse it. Saussure
argues that in fact such apparent cases actually always involve a type of
inclusion and not really an inversion of order.
Saussure argues that retrogression is impossible with the passe simple.,
despiute examples such as (9).
(9) Socrate mourut empoisonne. Il but la cigue. 'Socrates died of poison.
He drank hemlock.'
But Saussure points out that not just any such sequence is possible.
Despite the stereotype that passengers get off the plane after it lands,
(10) is at best marginally acceptable. Yet (11), where there is a causal
relation (perhaps the passengers had had reason to fear an accident) as
opposed merely to a stereotype, is fully acceptable. Yet even such a causal
relation does not in general suffice, as shown by the marginality of (12).
(10) ?Les passagers descendirent. L'avion atterit. 'The passengers got off.
The airplane landed.'
(11) Les passagers descendirent joyeusement. L'avion atterit sans encombre.
'The passengers got off happily. The airplane landed without incident.'
(12) ?Max tomba. Paul le poussa. 'Max fell. Paul pushed him.'
Saussure concludes that examples such as these do not involve regression.
Rather, the speaker of such an utterance asks that the addressee represent
the two events as involving a complex relation organized in representations
in which "drink hemlock" enriches the denotation of "die" and "eat
mushrooms" enriches "fall ill"; the two utterances refer in reality to two
aspects of one and the same complex event. "Die" can incorporate the cause
of death, while "expire" cannot, explaining the difference between (13) and
(13) ?Socrate mourut. Il but la cigue. 'Socrates died. He drank hemlock.'
(14) *Socrate expira. Il but la cigue. 'Socrates expired. He drank
On this view, what sanctions an example such as (15) is the possibility of
constructing in some way a representation of a combined concept of 'fall
because of the foot knocking against'.
(15) Max tomba. Son pied heurta une pierre. 'Max fell. His foot knocked
against a stone.'
4. Bertrand Verine, "Pour une interpretation aspectuelle des tiroirs du
passe: deux insertions cotextuelles du zeugme [passe simple et imparfait]"
["Towards an aspectual interpretation of the 'tiroirs' of the past: two
contextual insertions of the zeugma passe simple--imparfait']
Verine starts by observing the coordination with 'and' of a proposition in
the imparfait and another in the passe simple when the events recounted are
consecutive. Bres and Verine (1998) refer to such combinations as
Although the imparfait cannot create a new reference point, the event
recounted in the imparfait in (16) (the rabbits� grazing on the grass) is
undoubtedly consecutive to its context (the rabbits leaving their burrows).
(16) Des lapins sortirent de leurs terriers et broutaient le gazon.
(Flaubert, Bouvard et Pecuchet) 'Rabbits left (passe simple) their burrows
and grazed (imparfait) on the grass.'
The passe simple, which causes the process to be seen as an integral whole,
is able to create a new reference point: consequently it represents the time
implied by the verb as ascending in orientation (in the sense of Guillaume
1955/69), flowing from the past to the future. The imparfait, on the
contrary, focuses on the conversion of the accomplishment into the
accomplished and marks neither the initial nor the final point of the
process, representing the time implied by the verb in a descending
orientation, from future to past.
Verine proposes then that it is these aspectual values that determine the
use of each form in discourse: it is because the passe simple constructs an
ascending representation of the time implied by the verb that it is mostly
associated with the effects of perfectivity and progression; similarly, it
is because the imparfait gives the implied time a descending representation
that it is principally associated with the effects of imperfectivity and
non-progression. It is this opposition of the ascending orientation of the
passe simple and the descending one of the imparfait that often renders
their coordination with et 'and' contradictory, and which explains the
associated sense effects.
5. Jacques Bres, "Un emploi discursif qui ne manque pas de style:
l'imparfait en cotexte narratif" ["A discourse use that doesn't lack style:
the imparfait in narrative context"]
Bres examines the long-standing problem in French aspectology of the
perfective use of the imparfait such as (17) in which the imparfait
unexpectedly advances narrative time. This use is variously termed the
narrative, perspective, impressionist, or picturesque (pittoresque)
imparfait, or the "imparfait de rupture" ("imperfect of breaking").
(17) ...et le soir, a 22h30, le reverend Jackson etait appele a la Maison
Blanche ou il passait deux heures a parler et a prier avec Hilary et
Chelsea. (Le Monde, 21/8/98) '...and in the evening, at 10:30, Reverend
Jackson was (imparfait) called to the White House, where he spent
(imparfait) two hours talking and praying with Hilary and Chelsea.'
Bres is opposed both to treatments such as that of Tasmowski-De Ryck
(1985), in which the imparfait de rupture is assigned a different meaning
from that of the "normal" imparfait, and to those, such as the ones by
Simonin (1984), Grosselin (1999), and Vog�e (1999), in which the basic
meaning of the imparfait is "deformed" into an aoristic value.
The sense effects of the imparfait de rupture are seen as the product of
the representation of time implied by the imparfait and a context with which
it is incompatible. Accordingly Bres goes on to analyze the various
"stylistic effects" attributed to this type of imparfait--narrative meaning,
rupture, picturesque or impressionist meaning, perspective, acceleration or
deceleration--as consequences of the meaning of the imparfait in particular
The attribution of narrative meaning to the imparfait, for example,
follows from its use in a textual function--foregrounding--normally that of
the passe simple. The essence of narrativity, he argues, lies in "putting
into ascendance" the time represented by the events recited; on the
macrostructural level, a narrative moves from a more distant past to a less
distant past. In the case of (18), it is the narrative context which gives
the meaning of ascendance, not the imparfait itself. The remaining
"effects" of the imparfait can be similarly explained.
(18) Je la reconduisis a la barriere. (...) Deux jours apres elle revenait.
Et elle me prit dans ses bras. 'I saw (passe simple) her to the barrier.
... Two days later she came back (imparfait). And she took (passe simple)
me in her arms.'
6. Bertrand Sthioul, "Passe simple, imparfait et sujet de conscience"
["Passe simple, imparfait and the subject of consciousness"]
Sthioul notes that sequences of passe simples and passe composes are
improved by the addition of a term reflecting subjectivity, for example an
expletive (cf. 19a, 19b).
(19a) ?Paul sortit. Dehors, il fit froid. 'Paul went out. Outdoors, it
(19b) Paul sortit. Dehors, il fit bigrement froid. 'Paul went out.
Outdoors, it was damn cold.'
Examples such as (20) serve as counter-examples to the thesis of the
temporal non-progression of an utterance in the imparfait (Kamp and Rohrer,
1983) and as witnesses to the presumed capacity of the imparfait to report
the thought of the character of a narration. For Molendijk (1990) and Vet
(1991) such a sequence is possible because the telic process implies a new
state, which defines a new temporal period inside of which the second
process comes to be situated. Another type of explanation (Larochette,
1980; Tasmowski-De Ryck, 1985; Dowty, 1986), considers that the utterance
reports the thoughts of a person, so that the temporal reference of the
imperfect can be determined from the time of consciousness of the person
(20) Pierra alluma la lampe. La lumiere donnait maintenant a la piece un
air de tristesse desolee. 'Pierre lighted (passe simple) the lamp. The
light now gave (imparfait) the room an air of disconsolate sadness.'
This latter account Sthioul accepts, minus the implicit consequence that
the imparfait has an inherent capacity to describe a thought distinct from
that of the speaker. He believes that the determination of the point of
view according to which the situations or events described are envisaged
depends on a complex pragmatic process which is only indirectly tied to the
meaning of verbal morphemes. These effects are not produced directly by the
meanings of tenses, but indirectly by the process of assigning the temporal
reference of the process in question.
Traditionally it was assumed that a tense itself marked a focalization; the
passe simple indicated the consciousness of the narrator, while the
imparfait marked a focalization. On the contrary, says Sthioul, it is not
possible to directly determine from the tense the type of perspective
adopted. It is not tenses, but the idea of movement contained in the term
filer 'fly by' that, at the beginning of L'Education sentimentale, forces
the reader to infer that the events are seen through the intermediary of a
focalizer identifiable as a passenger on the boat on which Frederic embarks
(21) Enfin, le bateau partit. Et les deux berges, peuplees de magasins, de
chantiers et d'usines, filerent comme deux rubans que l'on deroule.
(Flaubert, L'Education sentimentale) 'Finally, the boat left (passe
simple). And the two banks, populated with stores, yards, and shops, flew
by (passe simple) like two ribbons that one unrolls.'
Nonetheless he does not think that one should abandon the idea of a link
between the choice of verbal morpheme and the determination of a narrative
perspective. Rather [like Moeschler], he accepts that tenses are markers
which communicate procedural information, in the sense of the relevance
theory of Sperber and Wilson (1989).
He hypothesizes that one can describe the basic signification of a tense
with the help of formulas indicating the relations between variables. By
default, the addressee is led to consider that the variables correspond to
moments of time and so to construct relations of temporal and aspectual
nature (antecedence, inclusion, etc.) between a state of fact, one or more
periods of reference and the moment of speech. In case such an
interpretation is blocked, the addressee has to assign to one of the
variables another value, that of a "moment of consciousness", and then to
postulate that the utterance reports a thought distinct from that of the
speaker at the moment of speech.
Sthioul concludes that the effects of subjectivization can be inferred from
utterances of various tenses, without any direct link established between
the category and the effect produced. Focalizations, whether internal or
external, are born in effect from the necessity of reconciling, in the
process of determining the temporal reference of the process, certain
constraints of contextual order with constraints tied to the instructions
pertaining to the tenses used. In the case of the imparfait, it is a matter
of knowing if the point of perspective corresponds to the period of
reference furnished by the context. For the passe simple, the question is
one of knowing if the inchoativity of the process expressed refers to the
event itself or to the coming into awareness (prise de conscience) of that
7. Sylvie Mellet, "Le parfait latin: un praeteritum perfectum" ["The Latin
perfect: a preterite perfect"]
The Latin perfect has usually been treated as ambiguous in the same way as
the French passe compose: both a past tense and a present perfect. This
analysis differs from that of the Roman grammarians themselves, who
characterized the form as a "praeteritum perfectum". Mellet argues for an
analysis in line with the terminology of the Romans themselves, that is, one
treating the perfect as past in tense (preterite, praeteritum) and
perfective (perfectum) in aspect.
Mellet argues that in discourse the Latin perfect always has a past value,
but its interpretation is dependent on various elements in the context. The
interpretation of resultant state is a contextual inference.
8. Marcel Vuillaume, "Heureusement que Pierre n'est pas venu demain!"
Vuillaume discusses the use of the passe compose in a future sense. He
cites a number of conditions on this future use of the passe compose. First,
it is limited to telic verbs. Secondly, negation seems not allowed in
But in the context of heureusement que 'it's fortunate that', these
conditions are radically altered. There is no problem about the occurrence
of atelic verbs (22), and negation, far from being impossible, is in fact
(22) Heureusement que tu n'as pas voyage demain! 'It's fortunate that you
won't have travelled tomorrow.'
(23) *Heureusement que tu as voyage demain! 'It's fortunate that that you'll
have travelled tomorrow.'
What he shows is that the passe compose with the future can occur in this
construction only when the action has already taken place, is not going to
take place again in future within the relevant period of time, and the
occurrence of the event in the past serves to preclude its occurrence in
future. Thus an example such as (24), in which repetition is not precluded,
is at best odd.
(24) ?Heureusement qu'il n'a pas plu demain! "It's fortunate that it hasn't
He concludes that there is a conflict between two constraints, a semantic
one to the effect that a future adverb cannot co-occur with the passe
compose of an atelic verb, and a pragmatic one to the effect that one should
not presuppose what one knows to be false, and consequently in an example
such as (22), should not have recourse to the future (the use of which would
presuppose that the trip should take place again). This conflict is
resolved by a metarule giving priority to the pragmatic principle and
leading thus to the use of the passe compose.
9. Paul-Louis Thomas, "Le plus-que-parfait en serbo-croate (bosniaque,
croate, montenegrin, serbe) dans une approche contrastive avec le fran�ais"
["The pluperfect in Serbo-Croatian (Bosnian, Croat, Montenegrin, Serbian)
and French in a contrastive approach"]
Grammars of Serbo-Croatian say little about the pluperfect beyond noting
its replacement by the perfect. This article aims to specify the conditions
necessary for its appearance in an utterance, whether there are cases where
the perfect cannot be substituted for it, and whether it can be used equally
in the perfective and imperfective aspects.
The pluperfect of French is not always translated by the Serbo-Croatian
pluperfect. The necessary condition for the use of the pluperfect in
Serbo-Croatian is the presence on the level of the utterance, or at least in
the context, of another process anterior to the moment of utterance in
relation to which the process in the pluperfect is presented as past: the
process expressed by the pluperfect is presented as accomplished in relation
to a process at another past time. Thus "il m'a dit que son oncle l'avait
invite" 'he told me that his uncle had invited him' is not translated with
the pluperfect unless it is plain that the process in the pluperfect is
completed before the relevant past process; however, "il m'a dit que son
oncle l'avait invite et qu'il avait accepte l'invitation" 'he told me that
his uncle invited him and that he had accepted' does elicit the pluperfect.
(The pluperfect is not obligatory, however; it can be replaced by the
perfect.) The pluperfect is also used, as in French, for background
Where the pluperfect is obligatory is in cases where the perfect would
imply the wrong order of events. For example, in a sequence of independent
propositions with verbs in the perfective perfect, the chronological order
of events matches that of the propositions, as in the case of the French
passe simple. The pluperfect allows one to establish a different
10. Denis Creissels, "L'emplois resultatif de etre + participe passe en
Creissels discusses the construction exemplified by (25) and relates it as
an expression of a resulting state to the corresponding expression of an
(25) Les coupables sont arretes depuis trois jours. 'The culprits have been
arrested for three days.'
The construction in (25) resembles both the passe compose and the passive,
but is neither; it is a resultative construction. Creissels rejects
alternative analyses, in which the participle here is opposed as an
adjectival form to the verbal form in these other constructions, arguing
that the participle in those constructions also has adjectival properties
(occurring with tres 'very' and being represented by the pro-form le 'so').
He examines the combination of etre 'be' with past participle and concludes
it can have five different interpretations.
He goes on to consider lexical constraints on the resultative construction.
First, only telic verbs are allowed. Second, where adjectival forms
corresponding to the verb exist, the resultative is precluded. Thus sec
'dry' blocks "la lessive est sechee" 'the wash is dried'. There do in
addition seem to be verbs with which the resultative is simply not used.
1.1 CRITICAL EVALUATION
Three oppositions characterize the pair passe simple/imparfait, and each
has been seen in its turn as the central distinction between the two.
First, the passe simple is perfective, the imparfait imperfective. Second,
the passe simple is a foregrounding tense, used to construct the main,
narrative line of narrative discourses and texts, while the imparfait is a
backgrounding tense, generally used to give background information,
description, etc. Finally, the passe simple advances narrative time while
the imparfait does not.
Each of these three contrasts has been viewed at one time or another as the
source of the other two. It might be said, for example--and it has often
been said--that the reason why the passe simple foregrounds occurrences is
that it forms the backbone or main narrative line in narration, which after
all, is at the heart of that particular enterprise, and the reason why it
forms the backbone is that it advances narrative time, and the reason for
this in turn is that it is perfective.
This leaves at least two questions unanswered, however. First, why are
there aspects at all? To say that aspect exists to serve the purposes of
discourse coherence is, I think, to enter into a vicious circle, for this
immediately raises the question of just how aspects do this, and the obvious
answer is by dint of providing different views of situations--but then how
do they do that, and why? Well, they do it by structuring discourse. No
wonder some scholars have rejected the aspectual distinction as the egg in
this particular cycle of egg-and-chicken and declared it simply a chicken
"comme les autres", in short, an epiphenomenon. Considering myself some
sort of a semanticist, I would like to believe that discourse-pragmatics
follows semantics, pace Hopper, Thelin, Thieroff, and (I think) Waugh, but
then: why ARE there aspects? Of course, to cede priority to either
grounding or discourse cohesion is to raise the question in another form,
and at a different level (or stage) of analysis.
A second issue has to do with atypical uses of the French tenses (but of
course similar problems arise in regard to other languages such as Ancient
Greek, and are amply attested in the literature) such as the narrative
imparfait, the backgrounded passe simple, etc. For the most part such uses
have been consigned to the junkheaps of "contextual meaning", "pragmatics",
etc., on the pretext that after all, the interpretation in context of a
category or form is quite a different matter from its basic semantic value.
But this raises significant questions concerning the respective roles in
interpretation of semantic values (basic meanings, senses), pragmatics
(such as presuppositions, implicatures), and context, both in the form of
"co-text" and real-world knowledge. A semantics-oriented theorist would
choose to believe that linguistic expressions "usually", "basically", or
"generally" mean what they mean--there is a sort of semantic default--and if
they don't, well, there's something special going on. A pragmatics-oriented
scholar might equally choose to reject this view, given how often the
semantic value is no determinant (or at least not the sole or the prime
determinant) of the interpretation of that expression or form in situ, and
under what a wide range of conditions the "basic meaning" fails, and
conclude instead that an expression's "semantic value"--if there be such a
thing at all--is only one of the ingredients that go into interpretation,
and perhaps a minor one at that.
In the present volume these concerns are concretized in terms of two
questions: first, how DO we interpret tense forms in discourse, and second,
what do we want to say about atypical uses such as the notorious "imparfait
de rupture"? Clearly, a strongly semanticist position such as Kamp and
Rohrer's (1983) that particular tenses inherently either advance or do not
advance narrative time is untenable. Why is it that in an example such as
(26) we understand the last sentence in the way we do, as representing the
consciousness of a protagonist and not of the narrator? Precisely because
the passe simple does NOT advance narrative time. For if it did, we could
interpret this sequence of sentences simply as a sequence of occurrences in
objective time: first he set out, then he went along the road, finally there
was a tree in the middle of the road. If the passe simple does not advance
narrative time, well, just what tense does?
26. Paul se mit en route. Il parcourut une dizaine de kilometres. Tout a
coup, il y eut un arbre au milieu de la chaussee. 'Paul set out. He
traversed a dozen kilometers. Suddenly there was a tree in the middle of
Most of the authors in this volume reject even a weaker version of the
semanticist thesis while not going completely over to a strong pragmaticist
position. They would not want to attribute the interpretation of tenses in
situation either to the tenses themselves (because of the numerous
counter-examples of atypical uses) nor to pragmatic factors alone (because
there do seem to be default values, because the whole point of aspects seems
difficult then to discern, and because there do seem to be causal links
between tense choice and interpretation. But none of the present papers,
though they are all fascinating and instructive, seems successfully to
provide a global account of past tense use in French.
Asher, N. 1993. Reference to Abstract Objects in Discourse. Dordrecht:
_____ et al. 1995. "De l'espace-temps dans l'analyse du discours."
Bres, J. and B. Verine. 1998. "D'un zeugme verbo-temporel: l'appariement [PS
et IP]." In A. Englebert et al., eds., La ligne claire: de la linguistique
a la grammaire. Melanges offerts a Marc Wilmot a l'occasion de son 60e
anniversaire, Paris, Brussels: Duculot, 175-185.
Dowty, D. 1986. "The Effects of Aspectual Class on the Temporal Structure of
Discourse: Semantics or Pragmatics?" Linguistics and Philosophy 9.37-61.
Grosselin, L. 1999. "Le sinistre Fant�mas et l'imparfait narratif."
Cahiers de praxematique 32.19-42.
Guillaume, G. 1951/69. "La representation du temps dans la langue
fran�aise." In Langage et sciences du langage, Paris: Nizet and Quebec:
Presses de l'Universite Laval, 184-207.
_____. 1955/69. "Epoques et niveaux temporels dans le systeme de la
conjugaison fran�aise." In Langage et sciences du langage, Paris: Nizet
and Quebec: Presses de l'Universite Laval, 250-271.
Kamp, H. and C. Rohrer. 1983. "Tense in Texts." In R. Baeuerle et al.,
eds., Meaning, Use, and Interpretation of Language, Berlin: De Gruyter,
Kleiber, G. 1994. Anaphores et pronoms, Louvain-la-Neuve: Duculot.
_____. 1997. "Sens, reference et existence: que faire de
l'extra-linguistique?" Langages 127.9-37.
Larochette, J. 1980. Le langage et la realite, vol. II, Munich: Wilhelm
Moens, M. and M. Steedman. 1988. "Temporal Ontology and Temporal Reference."
Journal of Computational Linguistics 14.15-28.
Molendijk, A. 1990. Le passe simple et l'imparfait: une approche
reichenbachienne, Amsterdam and Atlanta: Rodopi.
Simonin, J. 1984. "Les reperages enonciatifs dans les textes de presse." In
A. Gresillon and J.-L. Lebrave, eds., La langue au ras du texte, Lille:
Presses universitaire de Lille, 133-203.
Sperber, D. and D. Wilson. 1989. La pertinence. Communication et cognition,
Steedman. 1998. "The Productions of Time." Ms., Dept. of Computer and
Information Science, University of Pennsylvania. Available at http:
Swart, H. de. 1995. "Contraintes aspectuelles et reinterpretation
contextuelle." Semiotiques 9.89-116.
Tasmowski-De Ryck, L. 1985. "L'imparfait avec et sans rupture." Langue
Vet, C. 1991. "The Temporal Structure of Discourse: Setting, Change, and
Perspective." In S. Fleischman and L. R. Waugh, eds., Discourse Pragmatics
and the Verb, London and New York: Routledge, 7-25.
Vog�e, S. de. 1999. "L'imparfait aoristique, ni mutant, ni commutant."
Cahiers de praximetique 32.43-69.
Reviewer: Robert I. Binnick, Professor, Linguistics and East Asian Studies,
University of Toronto, Canada. PhD in Linguistics. Author, Time and the
Verb (1991, Oxford University Press, New York and London). Currently has
project on the Bibliography of Contemporary Research on Tense, Verbal
Aspect, Aktionsart and Related Areas, with Web page at http:
Robert I. Binnick
Division of Humanities
University of Toronto at Scarborough
Scarborough, Ontario M1C 1A4