PRAGMA 2553 1–34
A cross-linguistic discourse analysis of the Perfect
Henrie ¨tte de Swart*
NIAS, Meijboomlaan 1, 2242 PR Wassenaar, The Netherlands
Received 3 August 2005; received in revised form 4 May 2006; accepted 24 November 2006
Since Reichenbach (1947), the Present Perfect has been discussed in relation to the Simple Past. The
Reichenbachian characterization E-R,S has led to the view that the English Present Perfect, with its
restrictions on modification by time adverbials and its resistance to narrative structure is the Prototypical
Perfect. If the Pluperfect is different, or if counterparts of the Present Perfect in other languages behave
differently, that is because they are less Prototypical Perfects. In this paper, we argue that the most
important cross-linguistic differences do not require a different sentential semantics, but should rather be
explained in terms of different discourse level properties. We investigate Perfect constructions in four
languages: English, French, Dutch and German. We argue that all four are Reichenbachian Perfects, and
have very similar aspectual properties. Moreover, they introduce the same discourse configuration of
Elaboration. However, they differ in the additional constraints imposed upon the possible relations
between the event time E and other times or events in the sentence or the surrounding discourse. These
differences imply that we can use a Present Perfect construction to tell a story in French and German, but
not in English or Dutch.
# 2006 Published by Elsevier B.V.
Keywords: Perfect; Tense; Aspect; Discourse; Narration; Translation
1. Semantics of the Perfect: temporal, aspectual and discourse properties
In this section, we define a cross-linguistic semantics and pragmatics of Perfect constructions.
We focus on temporal structure in section 1.1, on aspectual properties in section 1.2, and on the
discourse configuration introduced by the Perfect construction in section 1.3.
Journal of Pragmatics xxx (2006) xxx–xxx
* Permanent address: Department of French/Uil-OTS, Utrecht University, Kromme Nieuwegracht 29, 3512 HD
Utrecht, The Netherlands. Tel.: +30 253 6204/30 70 512 2700; fax: +30 253 6167/30 70 51 17162.
E-mail addresses: email@example.com, firstname.lastname@example.org.
0378-2166/$ – see front matter # 2006 Published by Elsevier B.V.
known that the Pluperfect, which corresponds with the Reichenbachian structure E-R-S combines
1.1. Semantic rules for the Perfect in English, French, Dutch and German
In order to account for the relation between the Simple Past (SP) in (1a) and the Present
Perfect (PP) (1b), Reichenbach (1947) proposes an analysis in terms of three points:
Sara left the party.
Sara has left the party.
In (1a) as well as in (1b), the event E of Sara’s departure is located before the speech time S on
the time axis. The main difference between the two sentences is that (1b) does not only look at
the past, but maintains the importance of S. (1b) tells us that Sara left with the result that she is
not at the party at the moment, whereas (1a) only reports the leaving. In order to capture this
difference in perspective, Reichenbach (1947) introduces the notion of reference time R. In the
case of the SP, R coincides with E, which yields the structure E,R-S: the event E coincides
with (,) the reference time R, which precedes (-) the speech time S. For the PP, Reichenbach
proposes the schema E-R,S, such that E precedes the reference time R, which coincides with
the speech time S.
The Reichenbachian structure E-R,S has been used to explain various properties of the
locatethe eventEintime(2a)hasbeenrelatedtotheclaimthatlocatingtime adverbialsmodify
R, rather than E. Given that R coincides with S, we expect the Present Perfect to be compatible
with deictic adverbials only, which seems to be confirmed by the contrast between (2a) and
*Sara has left at six o’clock.
Sara has left this afternoon.
Furthermore, we expect the PP to be an inappropriate tense to tell a story, for narrative contexts
require the perspective to shift to the sequence of events, rather than to stay at Boogaart (1999)
uses the occurrence of a tense in a subordinate clause introduced by when as a criterion for
narrative use. The observation that the PP cannot be used in this context (3a), whereas the SP can
(3b) confirms that the former is not a Narrative tense, but the latter is2:
*When John has seen (PP) me, he has got (PP)/got (SP) frightened.
When John saw (SP) me, he got (SP) frightened.
to explain the restrictions of this tense on locating time adverbials and its infelicitous use in
narrative contexts is a nice result. However, the analysis raises problems for other cases. It is well
H. de Swart/Journal of Pragmatics xxx (2006) xxx–xxx2
PRAGMA 2553 1–34
3637 3839 39394040 41414242
38 38 38
65 65 65
75 7676 7677777878
1The sentence is not ungrammatical in all varieties of English, but most standard analyses of the Present Perfect are
based on these grammaticality judgments.
2Of course,thereis muchmore to sayaboutthe semantics andpragmaticsofthe EnglishPresentPerfect,butwe arenot
in a position to provide a full literature review here. We refer the interested reader to Portner (2003) for a recent overview
of relevant properties and analyses.
with locating time adverbials which can relate to either the reference time R or the event time E,
cf. the ambiguous (4):
(4) Sara had left at six o’clock
a. At six o’clock, Sara had already left.
b. (At some point in time it became clear that) Sara had left at six o’clock.
According to Hornstein (1990), the ambiguity of (4) is the result of an attachment of the locating
time adverbial at E (4a) or at R (4b). In this approach, it remains unclear why modification of E is
blocked with the Present Perfect in (2a), though. Furthermore, it does not explain why the Past
Perfect is easily used in narrative contexts such as when-clauses:
(5) When John had crossed the street, he entered a shoe store.
below for details). They propose that the Pluperfect is ambiguous between a Perfect in the Past
and a Past in the Past. Although this solves the problem of the Pluperfect, it implies that the
success of the ‘prototypical’ interpretation of the Reichenbachian analysis is restricted to the
Present Perfect, and does not extend to other Perfect tenses. Independent evidence in favor of the
strong Reichenbachian interpretation might come from a cross-linguistic analysis of the Perfect.
However, our study of the French, Dutch and German counterparts of the Present Perfect shows
that these constructions do not pattern exactly like the English Present Perfect.
From a morpho-syntactic point of view, Perfect constructions in the four languages under
withapastparticiple.InFrench,thisconstructioniscalledthePasse ´ Compose ´ (PC)(6b),inDutchit
can provide the following translations of the English Present Perfect (PP) in (6a):
He has eaten.
Il a mange ´.
Hij heeft gegeten.
Er hat gegessen.
In all four languages, the Perfect construction contrasts with a Simple Past construction, called
the Imparfait (IMP) in French (7b), the Onvoltooid Verleden Tijd (OVT) in Dutch (7c), the
Pra ¨teritum (Pra ¨t) in German (7d).3Again, we can illustrate with a set of translations from the
English Simple Past sentence in (7a):
[German Pra ¨t]
H. de Swart/Journal of Pragmatics xxx (2006) xxx–xxx3
PRAGMA 2553 1–34
120 120 120
128129130 131131131 132132 133133134 134
3The Simple Past tense system in French is more complex, because there is a contrast between the Perfective Passe ´
Simple (‘il mangea’) and the Imperfective Imparfait (‘il mangeait’). There is an extensive literature on this contrast that
we will ignore in this paper, because Camus does not use the Passe ´ Simple in his novel L’e ´tranger. For an overview of
relevant issues,and an evaluationof the relations between Passe ´ Simple, Imparfait and Passe ´ Compose ´ in modernFrench,
Imperfective contrast is not driving the translations.
function of the English and Dutch Perfect constructions. Therefore, we work out an analysis in
The similarities between the morpho-syntactic paradigms cannot hide the fact that there are
important differences in meaning and use between the Perfect and Simple Past forms in the four
languages, which raise the question how they fit in with the Reichenbachian analysis outlined for
English. With respect to the two main criteria of modification by time adverbials and narrative
use as witnessed by the occurrence in narrative when-clauses, none of the three languages
behaves like English. The three Perfect constructions other than the English PPall combine quite
easily with locating time adverbials, as illustrated in (8b), (8c) and (8d), which are literal
translations of (2a), repeated as (8a)4:
*Sara has left at six o’clock.
Sara is om zes uur vertrokken.
Sara est partie a ` six heures.
Sara ist um sechs Uhr abgefahren.
Furthermore, the Dutch Present Perfect form blocks narrative use in when-clauses (9b), just like
its English counterpart did in (3a), repeated here as (9a). However, the French PC (9c) and its
German counterpart (9d) freely occur in subordinate clauses introduced by ‘when’, as witnessed
by the well-formedness of the translations of (9a) in (9c) and (9d)5:
*When John has seen (PP) me, he has got (PP)/got (SP) frightened.
*Toen Jan me heeft gezien (VTT) is hij bang geworden (VTT)/
werd (OVT) hij bang.
Quand Jean m’a vu (PC), il a eu peur (PC).
Als Johan mich gesehen hat (Perf), hat er Angst bekommen (Perf).
Of course, we can assume that Dutch, French and German are exceptional, and the English PP
is the only tense that provides a perfect illustration of the Reichenbachian schema E-R,S.
Nerbonne (1985) adopts an ambiguity analysis for the German Perfekt. Along similar lines,
Vet (1992, 1999, 2001) proposes that the French PC is ambiguous or polysemous, and has both
a Present Perfect and a Simple Past tense use. These proposals are similar to Kamp and Reyle’s
analysis of the English Pluperfect. However, the ambiguity approach leaves Dutch as an
intermediate case. The Dutch VTTallows modification by time adverbials, but in the standard
language, it is not used for narration: does that mean that it qualifies as a Perfect or as a Simple
Past tense? We conclude that an ambiguity or polysemy analysis is not as clearcut as it looks at
first sight, and might not, in the end, be the best way to account for the semantic and pragmatic
differences observed. In this paper, we therefore propose a radically different approach. On the
one hand, we want to maintain the characterization of the English PP, the Dutch VTT, the
German Perfekt and the French PC as Perfects that obey the Reichenbachian schema E-R,S.
On the other hand, with Lo ¨bner (2002) wewant to emphasize the relevance of the Narrative use
of the French and German Perfect forms, that contrast with the more restricted discourse
H. de Swart/Journal of Pragmatics xxx (2006) xxx–xxx4
PRAGMA 2553 1–34
166 167 167
167167168168 169169170 170
168169 170171 171171172 172 173173174174
172173174175175 175176176 177177178178
176 177178179179179180180181181 182182
190191192 193193193194194195 195
198198 198199199 200200201201
199200201202202 202203203204 204205205
194 194 194
4French examples go back to Vet (1980) at least. For Dutch, this has been observed by Boogaart (1999). For German,
this has been pointed out by Herweg (1990), Klein (1999) and Musan (2002).
example (9b) may also be acceptable; for speakers of standard Dutch from the Netherlands, it is not.
just like many other tenses (e.g. the Passe ´ Simple and Imparfait, cf. de Swart, 1998). Let us
therefore turn to the aspectual nature of the Perfect.
We claim that all four Perfect constructions are instantiations of the Reichenbachian Perfect
schema E-R,S, but the Dutch and English Perfects are subject to additional constraints. The
English PPblocks anytemporal relation whatsoeverwith the eventtime E. As a result, it does not
allow modification by time adverbials (relation with another time) or a narrative use (relation
with other events). The Dutch VTT resists temporal relations between E and other eventualities,
but not other times. As a result, it allows modification by time adverbials, even though it cannot
be used in narrative contexts. The French PC and the German Perfekt are not subject to any
further constraints, which guarantees that they freely combine with time adverbials and can
establish discourse relations with other events (and thus occur in narrative contexts). However,
they remain Perfects in the sense that the event in the past is viewed from the speech time S,
according to the Reichenbachian structure E-R,S. Accordingly, we propose the following
semantic rules for the Perfect in English, Dutch, and French/German:
(10) Semantics of the English PP
E@X where @ is any temporal relation, and X is either an event or a
moment other than R or S.
(11) Semantics of the Dutch VTT
E@X where @ is any temporal relation, and X is an event.
(12)Semantics of the French PC and the German Perfekt
These rules allowus to maintain the Reichenbachian schema for the Perfect in all four languages.
They give a weak interpretation of the Reichenbachian analysis in which the restrictions on the
English PP do not follow from the schema E-R,S itself, but are formulated as additional
constraints on the Perfect construction. This allows counterparts of the Present Perfect in other
languages to be less constrained Perfects.
Independent evidence in favor of this analysis comes from an analysis of the English, Dutch
and French Pluperfect. For all three Pluperfect constructions, we can maintain the
Reichenbachian schema E-R-S. Without any further constraints, we would then derive the
properties of the English Pluperfect illustrated in (4) and (5) above. Thus, we do not need to posit
an ambiguitybetween Perfect inthe Past and Past inthe Past.Wewill notelaborate thispoint, but
concentrate on the cross-linguistic analysis of the Present Perfect.
The Reichenbachian schema of the Perfect focuses on the temporal location of the eventuality
E, and of the reference time R, the time from which the eventuality is viewed. It does not say
anything about the aspectual nature of the Perfect. The question whether the Perfect is a tense or
an aspect has been much debated in the literature. In this paper we attempt to give an integrated
analysis, under the assumption that the Perfect has both temporal and aspectual characteristics,
1.2. The aspectual nature of the Perfect
An aspectual definition of the Perfect has been proposed by Kamp and Reyle (1993), and has
been adopted by de Swart (1998). This definition is tense neutral (it generalizes over the Present,
H. de Swart/Journal of Pragmatics xxx (2006) xxx–xxx5
PRAGMA 2553 1–34
237 237238 238
238239 239 239240 240241241
240241242242242243 243244 244
241 241 241
248249 249249 250 250251251
250251 252252252253253254 254
255255 255256 256256
The same reviewer would also like to see an extension of the analysis to other Perfect forms, cf:
(iii)Quand Pierre est rentre ´, Marie est sortie.
When Pierre has returned, Marie has left.
(iv) Quand Pierre sera rentre ´, Marie sera sortie.
When Pierre will have returned, Marie will have left.
As the reviewer points out, (iii) can establish either a succession relation, or an overlap relation, whereas in (iv) there is
only an overlap between two result states. An account of this contrast in my framework would require a definition of the
discourse semantics of the future. Unfortunately, this is outside the scope of this paper.
Past and Future Perfect), and it assumes that the Perfect operates on an eventuality e and
is located in time by the tense operator (Present, Past or Future), so this analysis confirms that the
perspective on the event reported in the Present Perfect remains at the speech time S. In this
framework, the semantic structure of a sentence like (13a) is as given in (13b):
Mary has met the president.
[PRES [PERF [Mary meet the president]]]
DRT condition: e
The Perfect is an extensional operator that operates on an eventuality e and introduces the result
state softhateventualityasimmediatelyfollowinge.Inthe semanticrepresentationofDiscourse
Representation Theory (DRT), this is written as the condition e
and s ‘‘abut’’, i.e. they touch on the time axis (so there is no temporal ‘‘gap’’ between them), but
theydo not overlap. Both e and s are asserted, butit is the result state that is locatedin time by the
tense operator, so this analysis confirms that the perspective on the event reported in the Present
Perfect remains at the speech time, and thus respects the Reichenbachian analysis E-R,S.6The
representation of (13a) in DRT is spelled out in Fig. 1.
s (13c). This means that e
H. de Swart/Journal of Pragmatics xxx (2006) xxx–xxx6
PRAGMA 2553 1–34
288 289 289
290291292 292292293293294 294
Fig. 1. DRS representation of ‘Mary has met the president’.
6As an anonymous reviewer points out, there are problems with the notion of result state in view of examples such as
(i) and (ii):
(i)Jules Ce ´sar a conquis la Gaule entre 58 et 50 av. J.-C.
Julius Cesar has conquered Gaul between 58 and 50 before Christ.
(ii)La voiture a fro ˆle ´ le trottoir.
The car has grazed the sidewalk.
According to the reviewer, the result state of Gaul occupied by the Romans stopped 1500 years before the speech time, so
that no result state holds at S for (i). Events of grazing do not leave traces, so that (ii) only refers to a past event, and does
not assert any result state. I don’t think these examples invalidate the application of Kamp and Reyle’s analysis to French,
Dutch and German, but obviously, the notion of result state is weaker in the narrative use of the Passe ´ Compose ´ than
elsewhere, without disappearaing altogether though. Unfortunately, limits of time and space do not permit me to address
this issue here. The interested reader can consult Portner (2003) and references therein for extensive discussion of the
notion of result state.
andEnglishtranslations,the temporalvalue hasalmostdisappeared infavorofthe argumentative
flavor: the proposition follows as a logical next step in the argumentation, rather than as the next
event in a temporal order. The difference is a matter of degree, so the observation is a subtle one.
However, the fact that (56) above is an example where the English translator left out the alors of
not necessary for the temporal structure of the English text. The Dutch translator maintains the
c. De concierge boog zich naar haar toe, praatte tegen haar, maar zij schudde het
hoofd, mompelde iets en zette met dezelfde regelmaat haar gesnik voort. Daarna
kwam de concierge naast mij staan. Hij nam naast mij plaats. Na geruime tijd
lichtte hij mij in, zonder mij aan te zien: (...) p. 70
Time adverbials overrule the current reference time and introduce a new reference time (Kamp
and Reyle, 1993; de Swart, 1999). According to de Swart (1999), this leads to a ‘‘break’’ in
narrative structure. Instead of letting the story ‘‘tell itself’’, the frequent use of time adverbials
forces us to jump from one reference time to the next. It is quite conceivable that the translators
have intentionally kept as many as possible of the time adverbials in the French original in order
to emphasize the lack of narrative character of the novel. After all, a story told in the PC does not
‘‘tell itself’’, because the PC is not inherently a Narrative tense. Camus uses scenarios and strong
rhetoricalrelations to create bits and pieces ofnarrativediscourse, butthe alienating natureof the
novel is mostly due to the fact that the PC induces this fragmented, disconnected series of events,
as has been pointed out by Weinrich (1973:268–269). In our analysis, this follows from the
rhetorical structure of Elaboration created by the Perfect. The Dutch and English translations
have difficulty preserving the character of the French original, because they cannot tell the story
in the Perfect, and need to resort to inherently Narrative tenses like the OVT and the SP.
Maintaining the time adverbials in the translation is one way of disrupting the overly strong
narrative flavor of these tenses. This claim is confirmed by the observation that time adverbials
that are left out in the translation create more ‘‘fluent’’, more ‘‘narrative’’ bits of text, as
illustrated by (55) and (56) above, because the current reference time is notresetbetweenthe two
4.3.2. Temporal and argumentative role of time adverbials
In the second case, the time adverbial plays a temporal role in the French original, but a more
argumentative, rhetorical role in the English/Dutch translation. The following is an example:
(61)a. J’ai demande ´ deux jours de conge ´ a ` mon patron et il ne pouvait pas me les
refuser avec une excuse pareille. Mais il n’avait pas l’air content. Je lui ai me ˆme
dit: ‘‘Ce n’est pas de ma faute.’’ Il n’a pas re ´pondu. J’ai pense ´ alors que je
n’aurais pas du ˆ lui dire cela.
I asked my boss for two days off and he couldn’t refuse under the circumstances.
But he didn’t seem pleased. I even said, ‘‘It’s not my fault.’’ He didn’t answer.
Then I thought maybe I shouldn’t have said that.
Ik heb twee dagen vrij gevraagd aan mijn baas; die kon hij mij niet weigeren met
een zo geldige reden. Maar hij was er niets mee ingenomen. Ik zei zelfs tegen
hem: ‘‘Het is mijn schuld niet.’’ Maar hij gaf geen antwoord. Toen bedacht ik dat
ik hem dat niet had moeten zeggen.
The time adverbial alors in (61a) has both a temporal and an argumentative flavor. In the Dutch
H. de Swart/Journal of Pragmatics xxx (2006) xxx–xxx30
PRAGMA 2553 1–34
1520 15211521 1522 1522
1548 1549 1549
1555 1555 1555
The PC indicates the bounded nature of the process in (63a), and assez longtemps measures the
sentences, we can assume temporal succession because of the contrast between sitting in silence
and getting up. The two activities are incompatible, so they cannot be true at the same time, and
we infer that the getting up follows the sitting in silence. Both the English and the Dutch
translation in (63b, c) insert a time adverbial which makes the temporal relation of succession
explicit. Cleary, the presence of the measurement phrases for quite a long time and een tamelijk
adverbial, but again, the rhetorical, argumentativevalue of the expression toen is predominant in
4.3.3. Interaction of tense system and time adverbials
In the third case, the time adverbial is necessary in the French original for the temporal
unfolding of the story. It is also necessary in the Dutch/English translation, but for a different
reason, because of a different interaction between tense and time adverbial. The following is an
(62)a. Il m’a regarde ´ de ses yeux clairs. Puis il m’a serre ´ la main qu’il a garde ´e si
longtemps que je ne savais trop comment la retirer.
He looked at me with his bright eyes. Then he shook my hand and held it for so
long that I didn’t quite know how to take it back again.
Hij zag mij met zijn heldere ogen aan. Daarna schudde hij mij de hand, die hij
zo lang vasthield dat ik niet meer wist hoe ik haar terug moest trekken.
Puis in the French original (62a) forces a relation of narration between two quantized events
(cf. 31). The English and Dutch translations looked at me and zag mij aan are not necessarily
bounded in character, because their lexical aspect indicates a process, and there is no gramma-
tical aspect imposing boundaries. As a result, the temporal relation between the two sentences
would be overlap. Insertion of an adverbial then or daarna forces the succession reading, and
allows the activity in the first sentence to be interpreted as a process which takes the next eventas
its right boundary. The role then and daarna play in this context is reminiscent of the role puis
that puis imposes a right boundary to the inherently unbounded situation described by the
sentence inthe Imparfait.As such,puis isindispensableinsuchcontexts.Ifthenanddaarnahave
the same bounding role in contexts like (62b) and (62c), we can defend the view that they are
indispensable in the English and Dutch translations, but they fulfill different role than puis in the
This analysis is confirmed by the observation that some cases in which the translation has
inserted time adverbials where they were absent in the French original are of this type. An
example is (58), repeated here as (63):
(63)a.Nous sommes reste ´s silencieux assez longtemps. Le directeur s’est leve ´ et a regarde ´
par la fene ˆtre de son bureau.
We sat in silence for quite a long time. Then the warden got up and looked out of
the office window.
Een tamelijk lange poos bleven wij zitten zonder iets te zeggen. Daarna stond de
directeur op en keek door het raam van zijn werkkamer.
H. de Swart/Journal of Pragmatics xxx (2006) xxx–xxx31
PRAGMA 2553 1–34
1586 15861586158715871588 1588
16201620 16201621 1621 16221622
translation, the alternation between Perfekt and Pra ¨teritum mirrors the contrast between Passe ´
Compose ´ and Imparfait, even though the Pra ¨teritum is obviously not an Imperfective tense form
like the Imparfait. We expected that the Dutch and English translations would make less use of
time adverbials, because they use inherently Narrativetenses, which should not require extensive
use of time adverbials. This hypothesis was not verified by the translations we consulted, and we
developed an explanation for our findings in terms of the different functions a time adverbial
plays in the interaction with the tense and aspect system in each language.
lange poos is not sufficient to interpret the first sentence as denoting a quantized event. They still
have the flavor of an unbounded process, and the transparency of the SP and the OVT leaves it at
that. In order to impose a right boundary on this process, the translators introduce time adverbials
which force succession in time.
We conclude that there are good reasons why the time adverbials Camus introduced to force
progress in time are maintained in the English and Dutch translations, even though these texts do
not narrate the storyin the Present Perfect form. The first and most important function of the time
adverbials in combination with the English Simple Past and the Dutch OVT is to disrupt the
narrative character by jumping from one reference time to the next. Two other functions are to
support the argumentative force of the text, and to introduce boundedness in cases where lexical
aspect does not provide that. So the time adverbials are needed in the translations, even though
they may have different functions than in the source text.
In this paper, we have defended the idea that an appropriate cross-linguistic analysis of
Perfect constructions can only be established if we take the discourse level into account. The
Reichenbachian structure E-R,S needs to be complemented with an aspectual dimension, and
with a rhetorical analysis in which the Perfect establishes a structure of Elaboration, where
the speech time (or the utterance situation more generally) provides the topic. The
Elaboration structure implies that the sentences in the Present Perfect are connected by means
of the rhetorical relation of Continuation. If the eventualities reported in the Perfect are free to
enter temporal relations with other times and events, narrative use becomes possible, as in
French and German. If the eventualities cannot enter temporal relations at all (as in English)
or block rhetorical relations with other events (as in Dutch), discourse use is much more
We carried out a discourse analysis of the French Passe ´ Compose ´ to support our view that even
this very liberal Perfect construction is not an inherently Narrative tense, because it maintains its
orientation towards the speech time. A study of the first two chapters of Camus’ novel L’e ´tranger
shows that temporal structure is the result of connectives, lexical information contributed by
adverbs and verbs, presuppositions and implicatures, and rhetorical structure (Occasion, scripts/
The cross-linguistic analysis of Perfect constructions was put to the test by the study of
English, Dutch and German translations of L’e ´tranger. The alternation between the Passe ´
Compose ´ andtheImparfaitfindsapartialtranslationinthecontrastbetweenSimplePastandPast
Progressive in English. Although, as predicted, the Dutch translation is somewhat more liberal in
its use of the Perfect than English, it mostly tells the story in the Simple Pasttense. In the German
H. de Swart/Journal of Pragmatics xxx (2006) xxx–xxx32
PRAGMA 2553 1–34
de Swart, Henrie ¨tte, 1998. Aspect shift and coercion. Natural Language and Linguistic Theory 16, 347–385.
de Swart, Henrie ¨tte, 1999. Position and meaning: time adverbials in context. In: Bosch, P., van der Sandt, R. (Eds.),
de Swart, Henrie ¨tte, 2003. Coercion in a cross-linguistic theory of aspect. In: Francis, E., Michaelis, L. (Eds.), Mismatch.
Form-function Incongruity and the Architecture of Grammar. CSLI Publications, Stanford, pp. 231–258.
de Swart, Henrie ¨tte, Molendijk, Arie, 1999. Negation and the temporal structure of narrative discourse. Journal of
Semantics 16, 1–42.
Of course, a full analysis of the Passe ´ Compose ´, and its counterparts in English, Dutch and
German would require the study of other materials besides the text and translations of just one
novel. We are fully aware of the limitations of our corpus, butwe hope this paper providesa good
starting point for such a broad empirical study. All in all we conclude with Kamp and Rohrer’s
dimension to our understanding of natural language.
Anagnostopoulou et al. (2001).
I am grateful to audiences in Lyon, Paris, Groningen and Konstanz for constructive
remarks on presentations of the research reported here. I am especially indebted to Mark
Steedman for his suggestion to study translations of Camus. The hospitality of the linguistics
department of Konstanz University created the right environment for me to work on the
German translation, so thank you Miriam Butt. I would like to thank Arie Molendijk for
collaboration on the original study of the discourse structure of L’e ´tranger, Co Vet for general
support of my research on tense and aspect, Kyle Wohlmut for discussion of the English
translation, and two anonymous reviewers for taking the trouble to write a careful report for
Journal of Pragmatics.
Kenstowicz, M. (Ed.), Ken Hale: A Life in Language. MIT Press, Cambridge, MA, pp. 189–238.
Asher, Nicholasl, et al., 1995. De l’espace-temps dans l’analyse du discours. Se ´miotiques 9, 11–62.
Asher, Nicholas, 1996. Mathematical treatments of discourse contexts.
Colloquium. ILLC, Amsterdam, pp. 21–40.
Free University, Amsterdam.
Borillo, Andre ´e, et al., 2004. Tense connectives and discourse structure. In: Corblin, F., de Swart, H. (Eds.),Handbook of
French Semantics. CSLI Publications, Stanford, pp. 309–348.
Bybee, Joan, Perkins, Revere, Pagliuca, William, 1994. The Evolution of Grammar. Tense Aspect and Modality in the
Languages of the World. The University of Chicago Press, Chicago.
Bras, Myriam, Le Draoulec, Anne, Vieu, Laure, 2001. French adverbial Puis between temporal structure and discourse
structure. In: Bras, M., Vieu, L. (Eds.), Semantic and Pragmatic Issues in Discourse and Dialogue: Experimenting
with Current Theories. Elsevier, Amsterdam.
Bras, Myriam, Le Draoulec, Anne, Vieu, Laure, 2003. Connecteurs et temps verbaux dans l’interpre ´tation temporelle du
discours: le cas de puis en interaction avec l’imparfait et le passe ´ simple. In: Mellet, S., Vuillaume, M. (Eds.), Modes
de Repe ´rage Temporel. Rodopi, Amsterdam.
de Saussure, Louis, 1996. Encapsulation et re ´fe ´rence temporelle d’e ´nonce ´s ne ´gatifs au passe ´ compose ´ et au passe ´ simple.
Cahiers de Linguistique Franc ¸aise 18, 219–242.
In: Proceedings of the 10th Amsterdam
H. de Swart/Journal of Pragmatics xxx (2006) xxx–xxx33
PRAGMA 2553 1–34
She published an Introduction to natural language semantics (Stanford, CSLI Publications, 1998), The semantics of
Incorporation (in collaboration with Donka Farkas, Stanford, CSLI Publications, 2003), and articles in Journal of
Semantics, Natural Language and Linguistic Theory, Lingua, Linguistics and Philosophy and elsewhere.
de Swart, Henrie ¨tte, Molendijk, Arie, 2002. Le Passe ´ Compose ´ narratif: une analyse discursive de L’e ´tranger de Camus.
In: Laca, B. (Ed.), Temps et Aspect: de la Morphologie a ` l’Interpre ´tation. Presses Universitaires de Vincennes, Saint-
Denis, pp. 193–211.
Glasbey,Sheila,1995.When,discourserelations andthe thematicstructureofevents. In:Proceedings oftheworkshop on
tense, space and movement, vol. 5, Toulouse, pp. 91–104.
Herweg, Michael, 1990. Zeitaspekte. Die Bedeutung von Tempus, Aspekt und Temporalen Konjunktionen. Deutscher
Universita ¨ts Verlag, Wiesbaden.
Hornstein, Larry, 1990. As Time Goes By. MIT Press, Cambridge.
Kamp, Hans, Rohrer, Christian, 1983. Tense in texts. In: Ba ¨uerle, R., Schwarze, C., von Stechow, A. (Eds.), Meaning,
Use and Interpretation of Language. De Gruyter, Berlin, pp. 250–269.
Kamp, Hans, Reyle, Uwe, 1993. From Discourse to Logic. Kluwer, Dordrecht.
Klein, Wolfgang, 1999. Wie sich dat deutsche Perfekt zusammensetzt. In: Klein, W., Musan, R. (Eds.), Zeitschrift fu ¨r
Literaturwissenschaft une Linguistik, vol. 113. pp. 42–85.
Landeweerd, Rita, 1998. Discourse semantics of perspective and temporal structure, Ph.D., Groningen University.
Lascarides, Alex, Asher, Nicholas, 1993. Temporal interpretation, discourse relations and commonsense entailment.
Linguistics and Philosophy 16, 1–32.
Lo ¨bner, Sebastian, 2002. Is the German Perfekt a perfect perfect? In: Katz, G., et al. (Eds.), Proceedings of Sinn &
Bedeutung 7. pp. 255–273.
Moens, Marc, 1987. Tense, aspect and temporal reference, Unpublished Ph.D. dissertation, University of Edinburgh.
Moens, Marc, Mark, Steedman, 1988. Temporal ontology and temporal reference. Computational Linguistics 13, 15–28.
Molendijk, Arie, de Swart, Henrie ¨tte, 1999. L’inversion causale en Franc ¸ais. Travaux de Linguistique 39, 77–96.
Molendijk, Arie, et al., 2004. Meaning and use of past tenses in discourse. In: Corblin, F., de Swart, H. (Eds.),Handbook
of French Semantics. CSLI Publications, Stanford, CA, pp. 271–308.
Musan, Renate, 2002. The German Perfect: Its Semantic Composition and its Interactions with Temporal Adverbials.
Nerbonne, John, 1985. German Temporal Semantics: Three-dimensional Tense Logic and a GPSG Fragment. Garland,
Pustejovsky, James, 1995. The Generative Lexicon. Cambridge University Press, Cambridge.
Reichenbach, Hans, 1947. Elements of Symbolic Logic. The Free Press/Collier–Macmillan, New York/London.
Smith, Carlota, 1991. The Parameter of Aspect. Kluwer, Dordrecht.
Verkuyl, Henk, 1972. On the Compositional Nature of the Aspects. Reidel, Dordrecht.
Verkuyl, Henk, et al., 2004. Tense and aspect in sentences. In: Corblin, F., de Swart, H. (Eds.), Handbook of French
Semantics. CSLI Publications, Stanford, CA, pp. 233–270.
Vet, Co, 1980. Temps Aspects et Adverbes de Temps en Franc ¸ais Contemporain. Droz, Gene `ve.
Vet, Co, 1992. Le passe ´ compose ´: contextes d’emploi et interpre ´tation. Cahiers de Praxe ´matique 19, 37–59.
Vet, Co, 1999. Le passe ´ simple, le passe ´ compose ´ et les re `gles d’interpre ´tation discursive. In: Ple ´nat, M., et al. (Eds.),
L’emprise du Sens. Structures Linguistiques et Interpre ´tations. Me ´langes de Syntaxe et de Se ´mantique Offerts a `
Andre ´e Borillo. Rodopi, Amsterdam, pp. 323–336.
Vet, Co, 2001. Deux cas de polyse ´mie: le Passe ´ Compose ´ et le Futur pe ´riphrastique. In: Kronning, H., et al. (Eds.),
Langage et Re ´fe ´rence: Me ´langes Offerts a ` Kerstin Jonasson a ` l’Occasion des Ses Soixante Ans. Actas Universitatis
Upsaliensis, Uppsala, pp. 686–697.
Weinrich, Harald, 1973. Tempus. Besprochene und Erza ¨hlte Welt. Verlag W. Kohlhammer, Stuttgart.
Henrie ¨ttedeSwart(Ph.D.Groningen,1991)wasaffiliatedto StanfordUniversitybeforeshebecameprofessorofFrench
linguistics and semantics at Utrecht University in 1997. Her area of research is cross-linguistic semantics. Her work
focuses on tense and aspect, referential properties of noun phrases, quantification and negation in sentence and discourse.
H. de Swart/Journal of Pragmatics xxx (2006) xxx–xxx34
PRAGMA 2553 1–34