Frequency, iteration and tense use in French

advertisement

Frequency, iteration and tense use in French

Arie Molendijk, University of Groningen

0. Introduction

In French, both the P(assé) S(imple) and the IMP(parfait) may occur in sentences containing a frequency adverb: toujours ‘always’, souvent ‘often’, parfois ‘sometimes’, etc. Sentences containing an iterative adverb ( plusieurs fois ‘several times’, deux fois ‘twice’ etc.), on the other hand, allow the use of the PS while excluding the IMP. At least, this is true if these sentences are supposed to have an iterative reading, i.e. a reading implying that the repetition referred to by the iterative adverb is non-cyclic.

De Swart (1991) explains these facts by relating the aspectual difference between the PS and the IMP (perfectivity vs. imperfectivity) to what has been said by searchers like Ducrot (1979) and Kleiber (1987) about the difference between iteration and frequency (boundedness vs. unboundedness). We will see, however, that De Swart’s argumentation raises a problem: it does not account for the fact that sentences containing a frequency adverb allow the use of the

PS (section 1). This phenomenon (and other linguistic facts as well) can easily be explained if the difference between frequency and iteration is characterized in terms of the number of eventualities reported by the sentence containing a frequentative or iterative expression.

Frequentative sentences introduce only one eventuality (of type ‘state’) into the domain of discourse, whereas iterative sentences (i.e. sentences that are supposed to have an iterative, non-cylic reading) mention several (identical) eventualities that follow each other in time

(section 2). This claim, which helps us to understand some interesting linguistic facts from different languages (section 3), correctly predicts the use of the tenses in frequentative and iterative sentences (sections 4 and 5).

1. Frequency, iteration, PS and IMP

According to Ducrot (1979), Kleiber (1987), De Swart (1991), and others, the difference between frequency and iteration can be characterized as in (1):

(1) Frequency:

Iteration: no counting counting

unbounded quantity.

bounded quantity.

Arie Molendijk: Frequency, iteration and tense use in French

The fact that frequency does not imply counting, whereas iteration does, can perhaps be related to the following observation: an expression like de suite 'at a stretch', which can be used to provide information about the (exact) number of times something happens (happened, will happen), easily combines with iterative adverbs, but not with frequentative ones:

(2) Frequency: *parfois de suite stretch’ (lit.) etc.

'sometimes at a stretch', *souvent de suite 'often at a

Iteration: trois fois de suite 'three times at a stretch', plusieurs fois de suite

'several times at a stretch' (lit.) etc.

Now, De Swart argues that boundedness and unboundedness are intimately related to perfectivity and imperfectivity, respectively. So there is a natural explanation for the fact that iterative adverbs easily occur in perfective contexts, but not in imperfective ones, since boundedness implies perfectivity while conflicting with imperfectivity. According to this line of reasoning, the PS alla is correct, in an example like (3):

(3) Charles alla (PS) / (*)allait (IMP) plusieurs fois au cinéma

'Charles GO (PS / (*)(IMP) several times to the movies' because there is no conflict between the boundedness of the adverb and the perfectivity of alla .

The IMP allait , on the other hand, is unnatural (in an iterative, non-cyclic reading) because of a conflict between the boundedness of the adverb and the imperfectivity of allait .

1

But this kind of reasoning raises some problems. It is not in accordance with the fact that not only the IMP, but also the PS, can be used in states, despite the unboundedness implied by this aspectual class. This is shown by (4) and (5):

2

1

The IMP allait is correct in a context implying a cyclic reading: ‘à cette époque-là, Charles allait au cinéma plusieurs fois par semaine’. But we are only interested here in sentences having an iterative (noncyclic) reading

2

Depuis … is imperfective, whereas

à partir de

… is perfective. According to Martin (1971), this explains why the IMP is normally used in sentences containing depuis … , whereas the PS normally occurs in sentences containing

à partir de …

Molendijk (1990) explains these facts in temporal terms.

As the presence of depuis … indicates that the sentence is supposed to expresses global simultaneity, the

IMP is used. And as à partir de … expresses temporal succession, it naturally combines with the PS.

2

Arie Molendijk: Frequency, iteration and tense use in French

(4)

(5)

Charles était mon meilleur ami (depuis ce jour-là)

'Charles BE (IMP) my best friend (since that day)'

(A partir de ce jour-là), Charles fut mon meilleur ami

'(From that day on), Charles BE (PS) my best friend'

More specifically, one would not expect frequency adverbs to combine with the perfective PS, since we are dealing there with 'unboundedness (frequency) + boundedness (PS). But frequency adverbs do not only occur in IMP sentences, as in (6), but also in PS sentences, as shown by (7):

3

(6) Charles allait souvent au cinéma (depuis ce jour-là)

'Charles often GO (IMP) to the movies (since that day) '

(7) (A partir de ce jour-là), Charles alla souvent au cinéma

'(From that day on), Charles often GO (PS) to the movies'

In the next sections, I will show that there is a straightforward way of explaining the use of the tenses in these sentences. To illustrate my point of view, I will use two types of examples: frequentative sentences containing souvent 'often' and iterative sentences containing plusieurs fois

'several times'. I assume that what I will say about these sentences also applies to sentences in which occur other types of frequentative or iterative expressions.

2. Frequency and iteration: another difference

Consider (8):

(8) Charles va souvent au cinéma

'Charles often GO (PRES) to the movies'

This example refers to something that is supposed to be simultaneous with t

0

: the moment of

3

For depuis … + IMP and

à partir de …

+ PS, see preceding note.

3

Arie Molendijk: Frequency, iteration and tense use in French speech. Now, what holds at t

0

, in a case like (8), is not (necessarily) something that is actually happening, since (8) may be truthfully uttered while Charles is at home, for instance. So there must be something else that is true at t

0

. It is easy to see that what is true at t

0

is the habit of going to the movies. And as (8) refers to just one habit, it introduces only one eventuality into the domain of discourse, not several (identical) separate eventualities. This has already been pointed out by several linguists (see, for instance, De Swart 1991). In this regard, (8) can be compared to a sentence like (9):

(9) Charles fume trop

'Charles SMOKE (PRES) too much' which must also be analyzed as reporting one habit rather than as mentioning more than one eventuality.

What I have said about (8) applies to all sorts of sentences containing a frequency adverb. I do not think the tense of the sentence, the presence of certain adverbs and the like is of any importance here. So my claim is that not only (8), but also a sentence like (10) introduces just one eventuality into the domain of discourse:

(10) Entre 1992 et 1994, Charles alla souvent au cinéma

'Between 1992 and 1994, Charles often GO (PS) to the movies'

This claim is supported by the following observations. Frequentative sentences are semantically very close to sentences of type state (see De Swart 1991). The latter type of sentences can be characterized as in (11):

(11) Truth condition for states:

If a sentence S of type state is true at an interval or a moment of time X , then S is true for each moment of time t

X .

(11) involves that an example like (12):

(12) Pendant 2 ans, Charles habita à Paris

'For 2 years, Charles LIVE (PS) in Paris

4

Arie Molendijk: Frequency, iteration and tense use in French can be analyzed as in (13):

(13)

Charles habiter à Paris

'Charles live in Paris' is true for every moment of time t included in (pendant) 2 ans 'for 2 years'

Note that this kind of condition does not necessarily apply to sentences of another aspectual type.

So, in an example like (14):

(14) Pendant 2 heures, Charles travailla dans le jardin

'For 2 hours, Charles WORK (PS) in the garden'

Charles working in the garden does not necessarily hold at each moment included in the interval referred to by the adverb. This sentence may be truthfully uttered even if the eventuality mentioned in (14) was interrupted once in a while. Now, what is interesting here is that the truth condition given in (11) also applies to an example like (10), which can indeed be analyzed as in

(15):

(15) Ex. (10):

Charles souvent aller au cinéma 'Charles often go two the movies' is (supposed to be) true for every moment of time t included in entre 1994 et 1996 'between 1994 and 1996'

What I have said clearly supports the idea that frequentative sentences behave like nonfrequentative states, whatever the tense of the sentence, the presence of certain adverbs etc. This is in agreement with my claim according to which frequentative sentences introduce only one eventuality into the domain of discourse.

Now let us turn our attention to sentences containing iterative adverbs. None of what I have said about frequency applies to iteration. For instance, sentences containing an iterative adverb cannot refer to the moment of speech:

4

(16) (*)Charles va plusieurs fois au cinéma

4

At least, this is true for sentences having a non-cyclic reading.

5

Arie Molendijk: Frequency, iteration and tense use in French

'Charles GO (PRES) several times to the movies'

(cf. (8), which is perfect). The mere fact that (16) is not acceptable (in an iterative reading) seems to indicate that iterative sentences do not just introduce one eventuality into the domain of discourse, but several (identical) ones. In this regard, it must be stressed that each single eventuality that we are intuitively tempted to describe as such can be presented as ongoing at t

0

, whatever the Aktionsart of the sentence. This is shown by (17):

(17) State / activity / accomplishment / achievement ongoing (true) at t

0

:

Charles est malade (state) ‘Charles BE (PRES) ill’

Charles travaille dans le jardin (activity) ‘Charles WORK (PRES) in the garden’

Charles chante une chanson (accomplishment) ‘Charles SING (PRES) a song’

Charles sort de la maison (achievement) ‘Charles LEAVE (PRES) the house’

So defending the idea that iterative sentences introduce just one eventuality into the domain of discourse would force us to explain why 'iterative eventualities' would be an exception to the rule illustrated by (17). It seems more natural to assume that iterative sentences mention several

(identical) eventualities (that necessarily follow each other in time). Note here that, as a general rule, it is not possible to present a sequence of eventualities that follow each other in time as actually occurring at t

0

. At least, this is true for non-narrative contexts:

(18) (*) Charles construit une grange et en peint la façade

'Charles BUILD (PRES) a barn and PAINT (PRES) the front'

So (18) and (16) are unnatural for exactly the same reason: they mention several eventualities that are supposed to follow each other in time.

My treatment of iteration implies that truth conditions of the type given in (11) do not apply to iterative sentences. And indeed they don't. A sentence like (19):

(19) Entre 1994 et 1996, Charles alla plusieurs fois au cinéma

'Between 1994 and 1996, Charles GO (PS) several times to the movies' cannot be characterized as in (20):

6

Arie Molendijk: Frequency, iteration and tense use in French

(20) Ex. (19):

False statement:

Charles aller plusieurs fois au cinéma

'Charles go several times to the movies' is (supposed to be) true for every moment of time t included in entre 1994 et

1996 'between 1994 and 1996'

Let us examine now what claims can be attached to what has been said in this section.

3. Frequency and iteration: claims

The observations that have been made so far seem to imply that the claims given in (21) can be made:

(21) Frequency:

Frequentative sentences introduce only one eventuality into the domain of discourse. This implies that, in terms of the number of eventualities introduced by frequentative sentences, there is no difference between these sentences on the one hand, and sentences of type state that do not imply quantification at all, on the other hand. (Frequentative sentences are of type state, see preceding section). So, from a syntactical point of view, one may expect these two types of sentences to behave similarly.

Iteration:

Iterative sentences introduce more than one eventuality into the domain of discourse. As these eventualities necessarily follow each other in time, iterative sentences are supposed to behave syntactically like sentences that mention a

(temporal) succession of eventualities.

If my treatment of frequency and iteration is correct, then the claims given in (21) should not only be true for French, but for any language in which frequency and iteration can be linguistically expressed. This idea is corroborated by several facts that can be seen as additional evidence in favor of my treatment of frequency and iteration. For instance, not only in French, but

7

Arie Molendijk: Frequency, iteration and tense use in French also in a language like English, both frequentative sentences and regular states may combine with a certain kind of durative adverbs. They begin with for in English, with pendant in French, see

(22):

(22) Both frequentative sentences and 'regular states' combine with for .../pendant...

:

English:

- For 2 years, Charles often went to the movies

- For 2 years, Charles was my best friend

French:

- Pendant 2 ans, Charles est souvent allé au cinéma

- Pendant 2 ans, Charles a été mon meilleur ami

But neither type of sentences combines with durative adverbs beginning with in...

(English) or en...

(French). So we cannot replace for 2 years by in 2 years , nor pendant 2 ans by en 2 ans , in

(22). The opposite is true for both iterative sentences and sentences mentioning a sequence of eventualities. They combine with in ... and with en , as shown by (23):

(23) Both iterative sentences and sentences referring to a sequence of different eventualities combine with in.../en...

English:

- In one day, Charles went several times to the movies.

- In one day, Charles built a barn and painted the front

French:

- En une journée, Charles est allé plusieurs fois au cinéma

- En une journée, Charles a construit une grange et en a peint la façade

But they do not co-occur with for...

and pendant...

: it is not possible to replace in one day/en une journée by for one day/pendant une journée , in (23).

These facts seem to suggest that the claims given in (24) can be added to the ones given in (21):

8

Arie Molendijk: Frequency, iteration and tense use in French

(24) Tense use in frequentative sentences:

In different languages, frequentative sentences behave like regular states. So it is reasonable to admit that the possibility of using a given tense form in frequentative sentences can be explained by the rules that apply to tense use in sentences of type state.

Tense use in iterative sentences:

In different languages, iterative sentences behave like sentences referring to a sequence of eventualities following each other in time. So one may assume that the possibility of using a given tense form in iterative sentences can be explained by the rules that apply to tense use in sentences referring to a temporal succession of different eventualities.

These claims would explain why, in a language like English, PROG(ressive) can be occasionally used in frequentative sentences, whereas this form never occurs in iterative sentences. Indeed, the claims given in (24) imply the existence of a neat analogy between the two sentences of (25) on the one hand, of (26) on the other hand:

(25) 'Regular states/frequentative sentences do not exclude PROG:

You are being a bad boy

Charles is always teasing her!

(26) Sentence mentioning a (temporal) sequence of different eventualities and iterative sentences exclude PROG:

*Charles was building a barn, then he was painting it

*Charles was teasing her several times

But let us rather concern ourselves with French. Applying the claims given in (24) to French yields natural explanations for the use of the PS and the IMP in French sentences implying quantification.

4. The use of the PS and the IMP in French frequentative and iterative sentences .

9

Arie Molendijk: Frequency, iteration and tense use in French

If what I have said in (24) about frequency and iteration is valid, then the possibility of using the

IMP, in frequentative sentence like (27), the PS being excluded, can be directly related to the fact that the IMP imposes itself in a sentence like (28) (regular state):

5

(27) Depuis ce jour-là, Charles allait/*alla souvent au cinéma

'Since that day, Charles often GO(IMP/*PS) to the movies'

(28) Depuis ce jour-là, Charles était/*fut mon meilleur ami

'Since that day, Charles BE (IMP/*PS) my best friend'

And the obligatory use of the PS in an example like (29) follows from the fact that the PS imposes itself in (30):

6

(29) A partir de ce jour-là, Charles alla/*allait souvent au cinéma

'From that day on, Charles often GO (PS/*IMP) to the movies'

(30) A partir de ce jour-là, Charles fut/*était mon meilleur ami

'From that day on, Charles BE (PS/*IMP) my best friend'

As for the use of the tenses in iterative sentences, the PS is obligatory in an example like (31) (at least, in a context implying a non-cyclic reading):

(31) Charles alla /*allait plusieurs fois au cinéma

'Charles GO (PS/*IMP) several times to the movies' just because the same is true for an example like (32) (non-cyclic reading)

7

:

(32) Charles acheta / (*)achetait un journal, puis il alla / (*)allait au cinéma

'Charles BUY (PS/*IMP) a newspaper, then he GO (PS/*IMP) to the movies'

5 For an explanation, see footnote 2.

6

Idem.

7

See the comment on (33) for an explanation.

10

Arie Molendijk: Frequency, iteration and tense use in French

More generally, the possibility of using the IMP and the PS in frequentative sentences naturally follows from the fact that both tense forms can also be used in regular states. And the fact that the

PS imposes itself in iterative sentences derives from the obligatory use of the PS in sentences referring to a temporal succession of eventualities.

This implies that our 'story' about tense use in frequentative and iterative sentences will not be complete until the questions given in (33) are answered:

(33) Why are both the IMP and the PS possible in (non-frequentative) sentences of type state? And why is the PS obligatory in sentences referring to a temporal succession of eventualities?

Restrictions concerning the length op this paper do not allow me to treat these questions in detail.

Let me just make the following remarks. The use of the PS in sentences mentioning a sequence of eventualities following each other in time seems to follow directly from the fact that this tense form is essentially perfective. So it is typically used in sentences expressing temporal succession, as has been pointed out by many grammarians and linguists. Answering the first question of (33) seems to be a more delicate matter. In particular, the possibility of using the perfective PS in states, which imply unboundedness, may raise some problems. Several explanations have been proposed here. They all imply that something that is essentially unbounded can be nevertheless be presented as bounded by specific linguistic means. And unbounded entities can be presented as being bounded. (See, for instance, Moens 1988, Molendijk and De Swart 1998)

8

. But things are often more complicated here than one would expect. For instance, English does not seem to have problems in presenting run 10 miles as unbounded, as in (34):

(34) Charles was running 10 miles

8

This is also true for languages that do not have specific perfective and imperfective verbal forms. For instance, in Dutch, it is possible to ‘convert’ an unbounded situation like mijn beste vriend zijn ‘be my best friend’ into a bounded one by adding an adverbial like twee jaar lang ‘for two years’ to the sentence referring to this situation: twee jaar lang was hij mijn beste vriend ‘for two years, he was my best friend’.

11

Arie Molendijk: Frequency, iteration and tense use in French while native speakers of French have difficulty in accepting a sentence like (35) (in a non-habitual reading):

(35)

??Charles courait 10 kilomètres

'Charles RUN (??IMP) 10 kilometers'

If there are no ontological differences involved here, why is the English sentence much better than the French one? This is a question that I propose to examine in later work.

5. Conclusion

In the preceding sections, I have argued that the use of the PS and the IMP in frequentative and iterative sentences can be directly related to the following fundamental difference between frequency and iteration: frequentative sentences behave like regular states introducing only one eventuality into the domain of discourse, whereas iterative sentences behave like sequences reporting several eventualities following each other in time. These facts explain why both the PS and the IMP can be used in frequentative sentences (depending on the context, of course), whereas iterative sentences only allow the use of the PS.

12

Arie Molendijk: Frequency, iteration and tense use in French

References

Ducrot, O.

1979 'L'Imparfait en français'. Linguistische Berichte 60: 1-23.

Kleiber, G.

1987 Du côté de la référence verbale: les phrases habituelles . Bern: Peter Lang.

Martin, R.

1971 Temps et aspect. Essai sur l’emploi des temps narratifs en moyen français , Paris.

Moens, M.

1988 Tense, aspect and temporal reference, Ph.D. thesis, University of Edinburgh

Molendijk, A.

1990 Le passé simple et l’imparfait: une approche reichenbachienne . Amsterdam – Atlanta:

Rodopi

Molendijk, A.

1997 'Kwantificatie, aspect en temporaliteit: een vergelijking van het Nederlands en het Frans'.

Tabu 27: 67 - 86.

Molendijk, A. and De Swart, H.

1998 'Frequency and Tense Use in French'. Belgian Journal of Linguistics 12.

De Swart, H.

1991 Adverbs of quantification: a Generalized Quantifier approach . Diss., University of

Groningen.

13

Download