A cross-linguistic discourse analysis of the Perfect

Abstract

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.

Full text

Page 1

UNCORRECTED PROOF
+ Models
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
Abstract
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.
www.elsevier.com/locate/pragma
Journal of Pragmatics xxx (2006) xxx–xxx
1
2
3
4
5
6
7 8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
* 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: h.deswart@nias.knaw.nl, henriette.deswart@let.uu.nl.
0378-2166/$ – see front matter # 2006 Published by Elsevier B.V.
doi:10.1016/j.pragma.2006.11.006

Page 2

UNCORRECTED PROOF
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:
(1)a.
b.
Sara left the party.
Sara has left the party.
E,R-S
E-R,S
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
EnglishPresentPerfect.TheobservationthatthePPisgenerallyincompatiblewithadverbsthat
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
(2b)1:
(2)a.
b.
*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:
(3)a.
b.
*When John has seen (PP) me, he has got (PP)/got (SP) frightened.
When John saw (SP) me, he got (SP) frightened.
Atfirstsight,thefactthattheReichenbachiananalysisoftheEnglishPresentPerfectmightbetaken
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
+ Models
PRAGMA 2553 1–34
31
32
33
34 3535
3535 363637 373838
363636
36373839 3939 404041414242
373737
404142
383838
4344
45
393939
46
4040 40
47
4141 41
48
42 4242
49
43
50
51
52
53
54
55
56
57
58
5960 60
60 6061616262
61 6161
61 626363 636464 6565
62 6262
64 65
636363
6667
68
6464 64
69
6565 65
70
66
71
72 7373
7373
74 747575
74 7474
74
75 7676767777 7878
75 7575
7778
76 7676
79 80
81
77 7777
82
78 78 78
83
79
84
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.

Page 3

UNCORRECTED PROOF
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.
KampandReyle(1993:598–601)treatthePerfectasanaspect,ratherthanatense(seesection1.2
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
considerationinthispaperarerathersimilar,andinvolveacombinationoftheauxiliary(haveorbe)
withapastparticiple.InFrench,thisconstructioniscalledthePasse ´ Compose ´ (PC)(6b),inDutchit
iscalledtheVoltooidTegenwoordigeTijd(VTT)(6c),andinGermanthePerfekt(Perf)(6d),sowe
can provide the following translations of the English Present Perfect (PP) in (6a):
(6)a.
b.
c.
d.
He has eaten.
Il a mange ´.
Hij heeft gegeten.
Er hat gegessen.
[English PP]
[French PC]
[Dutch VTT]
[German Perf]
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):
(7)a.
b.
c.
d.
He ate.
Il mangeait.
Hij at.
Er aß.
[English SP]
[French IMP]
[Dutch OVT]
[German Pra ¨t]
H. de Swart/Journal of Pragmatics xxx (2006) xxx–xxx3
+ Models
PRAGMA 2553 1–34
84
85
86 8787
87 8788 8889 89
888888
88899090 9091 919292
89 8989
9192 9393 9394949595
90 9090
9495
96 97
98
91 9191
92 9292
99
9393 93
100
949494
959595
101102102
102 102103103
96
103 103103
103
104105
106
104
107
108
109
110
111
112
113
114
115
116
117
118119119
119 119 120120 121 121122 122
120120120
120121 122123 123123124 124125125126126
121 121121
124 125126 127127127 128128129 129 130130
122122122
128129130 131131131 132132 133133134 134
123 123123
132 133134
135136
137
124124124
125125125
138
126 126126
139
127127127
128 128128
140141141
129129 129
142142 142
130 130 130
143 143143
131131 131
144144144
132 132132
145 145145
133133 133
146146146
134
134134
147147147
135
148148148
149 149149
150 150150
151151 151
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,
seeVerkuyletal.(2004),Molendijketal.(2004),andreferencestherein.Aswewillseeinsections3and4,thePerfective/
Imperfective contrast is not driving the translations.

Page 4

UNCORRECTED PROOF
function of the English and Dutch Perfect constructions. Therefore, we work out an analysis in
two parts.
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:
(8)a.
b.
c.
d.
*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.
[English]
[Dutch]
[French]
[German]
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:
(9)a.
b.
*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).
[English]
[Dutch]
[French]
[German]
c.
d.
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
+ Models
PRAGMA 2553 1–34
158
159
160
161
162
163
164
165
166167167
167 167168168169169 170170
168168168
168 169170171 171 171172172 173173174174
169169169
172 173174175175 175176176177177178178
170170170
176177 178179 179 179180180 181181 182182
171 171171
180 181182
172172 172
183184
185
173173 173
186
174174 174
187
175 175 175
176 176 176
177177177
188 189189
189189 190190191191192 192
178178178
190190 190
190191192193193193 194 194195195
196197
179 179179
191 191191
194195
180180180
192192192
198 198198199199200 200201201
181 181181
193193193
199200201202 202 202203203204204 205 205
182 182182
194 194194
203204 205
183
196
208
209
195195195
206207
197
210
198198 198
211
212
199 199199
213
200 200200
214
201201 201
215
202202202
216
203 203203
217
204204204
218
205 205205
219
206
220
221
222
223
224
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).
5TherelevanceofthisobservationforGermanhasbeenhighlightedbyLo ¨bner(2002).ForFlemishspeakers,theDutch
example (9b) may also be acceptable; for speakers of standard Dutch from the Netherlands, it is not.

Page 5

UNCORRECTED PROOF
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
(i)E-R,S
(ii)
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
(i)E-R,S
(ii)
E@X where @ is any temporal relation, and X is an event.
(12)Semantics of the French PC and the German Perfekt
(i)E-R,S
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
+ Models
PRAGMA 2553 1–34
224
225
226
227
228
229
230
231
232
233
234
235
236 237 237
237237238238
238238238
238239 239239240240 241241
239239239
240241242242242243 243244244
245
240240240
243244
241241241
246246246
247247 247248 248
242 242242
248 249249 249250250251251
243 243243
250251 252252 252253 253254254
244244244
245
253 254255 255255
256256 256257257
246 246246247247247
257
258258258259 259260260
248248248
259 260
249249249
261262
263
250250250
264
251251 251
265
252252 252
266
253253253
254 254 254
267
268
255255 255256256 256
269
257257 257
270
258258 258
271
259259 259
272
260260 260
261
274
275
273
276
277
278
279
280
281
282
283