Perfective Marking in the Breton Tense-Aspect System

Round 1

Reviewer 1 Report

Overall assessment

This is an interesting paper and a very valuable contribution to the literature. It gives a clear overview of the Breton tense/aspect system, and presents new findings based on a corpus study. The main focus is on the different ways in which past perfectives are expressed in Breton. As I understand it, this turns out to be conditioned mainly be two factors: (i) register (simple past being used in written discourse, and the present and past perfect in oral conversation and oral narration), and (ii) discourse pragmatics (with the past perfect used for “pure narration” (or “strict continuation”) and the present perfect for “pseudo-narration” (or “loose continuation”).

Mainly what I would like to see is an attempt to make greater use of Lascarides and Asher’s theoretical framework to capture the difference between the present perfect and the past perfect in Breton. In the current form of the paper, this distinction is described in rather impressionistic terms with reference to very specific examples from the corpus, but it’s hard to see whether and how this generalizes. I would like to encourage the author(s) to try to formalize the distinction between “strict” and “loose” continuation within L&A’s theoretical framework, and/or to operationalize the distinction more clearly in the corpus study. For instance, are there specific diagnostics or clues in their corpus that could be used to decide whether a certain passage represents “strict” or “loose” continuation? I realize that this may be difficult to do, so this is not an absolute requirement, but any attempt in that direction could strengthen the paper.

I also have a number of clarification questions, which often have to do with terminological issues, which may reflect differences between different descriptive and theoretical traditions. It should not be too hard to clarify these points, and it would help to make the paper accessible to a wider audience. See below for details.

Detailed comments and requests for clarification

Throughout: The abbreviations for the tense forms used in the text are different than the ones in the glosses. For instance, the simple past is abbreviated as “Sp” in the text but as “SPST” in the glosses, the past perfect as “PsPf” and “PST.PRF”, etc. It would be very helpful to the reader to use the same abbreviations in the text and the glosses.

line 28: Why is the simple past called “synthetic” ? (also line 255)

line 64: What exactly is meant by “narrative mode”? Can a working definition be given, or at least some criteria for identifying it?

line 88: “the right boundary” -- of what?

line 96-97: The perfect is here defined in morphosyntactic terms. This works for Indo-European, but how would this carry over to other languages? This is a minor quibble, though, since the next sentence offers a semantic definition. But then why do we need the morphosyntactic definition at all?

line 114: Here (and elsewhere in the paper) the term “imperfect” is used to refer to a particular sub-type of imperfective. Is “imperfect” a traditional term from Breton or Celtic linguistics? It may be good to add a little bit of an explanation of what the difference is between “imperfect” and “imperfective”.

line 136: Explain the “Tense 2 constraint”. I guess this is analogous to the more familiar “V2” or “verb second” constraint? Again, “Tense 2” may be standard term in Celtic linguistics, but a bit of an explanation for a broader linguistic audience may be useful.

line 209: T2 -- I guess this means Tense 2?

line 210: I guess VP here means “verbal particle”, not the usual “verb phrase”. I’d suggest using a different abbreviation, or just writing “verbal particle”.

line 222: What is meant by “situation aspect”? Is it the same thing as what others have called lexical aspect or Aktionsart or Vendler classes?

line 330: What are “non-pluriactional contexts” ?

line 384: What are “inactive actions” ?

line 385: What are “anterior events” ? Anterior to what?

line 390: Maybe the “special non-finite construction” should already be introduced in section 2.

line 434-435: “the Sp is no longer an option, since this is oral language”. As far as I can tell, this is the first time in the article that it is stated clearly that the Breton simple past is not used in spoken language. There are some hints along these lines earlier the paper (e.g., in line 255, where the past perfect is called a “conversational substitute” of the simple past), but it would be helpful (especially to readers who don’t know anything about Breton, like myself) to state explicitly at a much earlier point in the paper (for instance in the introduction) that the simple past is limited to written language. An additional point: the use of the words “no longer an option” suggests a language change here. Is this a reference to the influence from French in the Middle Ages, alluded to on p. 4? Here also being a bit more explicit would be helpful to the reader. Is the claim that Breton used to have the simple past as its only way to express the perfective, but that this was replaced in spoken language by the (present and past) perfect? I would suggest providing this crucial information very clearly right off the bat in the introduction section, rather than in oblique references scattered in various places later on in the paper. It would be helpful to describe in a bit more detail both the diachronic development and the current synchronic situation with respect to register (spoken vs. written language).

p. 12, Table 1: The four categories of the perfect in the table are a bit confusing since they don’t match the usual four-way classification. The literature usually distinguishes four uses (or readings) of the English present perfect: continuative, experiential, resultative, and recent past/’hot news’. The table lumps the latter two together for some reason, and then confusingly adds another category, namely “infinitive”, which is of course not a reading of the present perfect at all, but just the observation that the classification applies only to the combination of the perfect with the present tense, and not to other occurrences of the perfect (e.g., in the past perfect (e.g., had left), or when following a modal (e.g., must have left) or in a gerund (e.g., having left)).

p. 12, Table 2: What does “pseudo-narrative” mean? See also my comment on line 538 below.

line 459: What does “pluriactional” mean? See also line 330. At least giving an example might help here.

line 464-466: Here there’s a casual observation that the English simple past can be used to refer to an “indefinite past event having relevance at the speech time”. I believe this is true, but it would be good to provide some examples, or at least some references to the relevant literature. (For instance the discussion of examples like “John went to Harvard” in Partee 1973 and subsequent literature, or cases like “Boromini built this church” discussed by various authors.) I’d also be curious to see some evidence or references in support of the claim that this is “indicative of a more conversational register.” For all I know this may be true, but has any sociolinguistic study of this been done or is it just an informal impression?

line 482-500: Relating to the previous point, there is some attempt at an explanation here for why the English version of (13) has a present perfect in contrast to the simple past in the English version of (11) and (12). I do not understand this explanation, and it feels a bit ad hoc to me. Does it generalize beyond these examples? Since this is only about English, of course this is not very important for the paper as a whole (the author(s) certainly can’t be expected to have an explanation for all the idiosyncrasies of the English present perfect), but I’m wondering if this could be clarified; or else maybe it could just be cut or relegated to a footnote.

line 514: “Both PrPfs” -- does this refer to the English and Breton PrPf, or to the two Breton present perfects occurring in (14)?

line 517-518: “But in English, the simple mention of a past event, even if it has present consequences, triggers the Sp, contrary to Breton.” -- Which English example(s) is this referring to? (15) and (16)?

line 533: “the Sp is impossible for reasons of register” -- see my comment on line 434-435 above.

line 538: Here again the term “pseudo-narrative” is used without much of an explanation (see also my comment on Table 2 above). Two examples are given (17 and 18), but I don’t understand what makes these “pseudo-narrative”. Why is (17) not simply a resultative present perfect? And (18) looks just like a past perfective. In line 572, it’s observed that the events in (18) are “not strictly ordered temporally”. Is that what is meant by pseudo-narrative? But then how does this apply to (17)? In short, I do not understand what defines “pseudo-narrative” as a category and how it differs from a “real” narrative on the one hand, and from non-narrative modes of discourse on the other hand.

line 596-597: “In short, the proportions are reversed in comparison to corpus B.” Which proportions? PrPf vs PsPf?

line 637-654: I don’t find this discussion of the difference between the PrPf and the PsPf very clear. It seems to be quite speculative and rather ad hoc in reference to features of these specific examples.

line 650: What is “pure narration” and how does it differ from “pseudo-narration”?

line 651-652: “The PrPf suggests a strong connection to the Speech time”. Again this is rather impressionistic.

line 656-662: This is a clear empirical prediction, which is backed up with some statistics. Good! Maybe this could be emphasized more. But it’s not entirely clear why a “strict narrative function” (and what does that mean, exactly?) should result in greater compatibility with specific time adverbials and when-clauses. This makes sense at some level, but is there a theory that predicts this? Could this be explained in terms of Lascarides and Asher’s framework, for instance?

Section 4: Several examples from the corpus are discussed in detail. In and of itself, this discussion is valuable, but it’s never very clear to what extent this is about accidental features of the specific examples, and to what extent this reflects a systematic difference between the PrPf and the PsPf. It all seems rather impressionistic. I would like to believe the authors that these are real generalizations, but is there any way this could be backed up with statistics from the corpus?

line 749-756: Here the crux of the analysis is presented, in the framework of Lascarides and Asher (1993). I wish this part of the paper could be worked out in more detail, for instance by saying more about the discourse relations of Elaboration and (especially) Continuation. A distinction is made between “strict continuation” (associated with the PsPf) and “a looser type of continuation” (associated with the PrPf). This is a promising idea, but it remains quite impressionistic. Could this difference be defined more clearly with L&A’s framework, or operationalized in some way? For instance, what would be some clues that could be used in a corpus to decide whether a passage involves “strict continuation” or “loose continuation”? And is this the same distinction as that between “pure narration” and “pseudo-narration” that was made in the previous section? (I am assuming it is, but this is never stated explicitly anywhere in the paper.)

Section 6: This section is a useful summary and conclusion. Again, I wish the analysis of the discourse pragmatic distinction between the PrPf and PsPf could be sharpened a bit, perhaps by making more explicit use of Lascarides and Asher’s framework, and adapting it for the Breton data.

 

 

minor stylistic/language suggestions

 

line 5: in the aim of --> with the aim of

line 12: past simple --> simple past

line 23: in the aim of --> with the aim of

line 143: we find --> there is

line 170: idem (or use some other equivalent expression)

line 192: it knows no restriction --> they are not subject to any restriction

line 244: To conclude on this section --> To conclude this section

line 452: of a liberal type --> of a more liberal type

line 883: past simple --> simple past

 

 

 

Author Response

First, let me thank you for the very useful comments on my paper. Below is a summary of the main modifications I’ve made, taking your comments into account :

- Essentially, I’ve clarified my use of the Lascarides & Asher framework (esp. in section 4), trying to back it up with corpus data, and essentially combining it with a temporal semantics to explain the difference between the present perfect and the past perfect in Breton narrations. I’ve abandoned the loose vs. strict continuation function, which was too impressionistic.

- about your detailed comments : I’ve corrected all the unclear passages, and added many footnotes to clarify some of the points you raised. I think the distinction between the two perfect forms comes out clearer as in the first version.

- I’ve also corrected the stylistic imperfections you noted.

Reviewer 2 Report

This is an interesting paper. Breton is indeed an underdescribed language and this paper fills a gap. One desideratum would be some basic sociolinguistic information, as Breton has a number of dialects, both traditional and new, with sometimes significant differences among them. I assume that the dialect of the translation is Neo-Breton, and while this probably makes no difference, it would be helpful to have at least a small amount of information on the translation, if possible.

 

The English is very good. There are occasionally very minor, to my ears, non-native uses, but a good editor will find and correct these.

 

I have no quibbles with the argument of the paper as such or with the general experimental setup and methods. Below I give a number of comments and some suggestions that will, I hope, make the presentation clearer and more convincing.

 

I have some questions about the passage (page 2) chosen to illustrate the differences between the English and Breton uses of verb tenses.

• There is a lacuna in the text (according to my English language version): “She stopped to draw a deep breath and then went ranting on. It seemed she had been wanting to say all this for years.” The second sentence is simply left out. Is it also left out in the Breton? I assume so, but it is strange, as the English text is otherwise identical.
• Line 48 in the proofs is left unglossed. It, however, does contain verbs (gouzout a-walc’h and vijes bet), that would seem to be relevant in the present context, so it is unclear why these are not included in the gloss.
• Lines 51-52 are somewhat odd, in that they are not really a translation of the English version, but seem to refer to what is further below in the text. This latter issue is perhaps not something to be addressed here, as examining the translational decisions is not the goal, but it does nonetheless seem relevant that the translator did not translate directly, but rather used the sense of there being a car crash (found later in the passage) rather than there being an explosion.

I should note that I do not dispute the generalizations made (page 2) concerning the passage, but I do think that the points noted here are worth addressing in some way.

 

The use of (non-translated) control texts is good practice, however, the extent to which the texts from the Banque Sonore were used for this purpose is unclear. They are explicitly cited only in examples 20–21. Were they used elsewhere or only here?

Also, it might be good to indicate what types of texts from the Banque Sonore are used. Are the stories by traditional native speakers or native speakers of Neo-Breton. There is considerable tension in (parts of) the community concerning the different varieties, and it may be worth noting the types of speakers. If the texts from the Banque Sonore are of the same variety (i.e. Neo-Breton) as the translation, there is no great cause for concern. If the Banque Sonore texts are rather spoken by a traditional speaker, one might ask further questions about exact comparability.

 

It is sometimes unclear whether some of the statements are generally accepted facts or whether they are anticipating results to be presented later in the paper.

• For instance, ll. 27–33 gives the four correspondences of the English simple past, along with some brief comments. Is PsPf generally accepted as replacing the Sp in oral narrative discourse or is this a finding of the present study? It seems to be a finding, but it is sometimes presented as if it were simply a fact (which it may well become following the publication of this paper). See also ll. 593–4, where it is stated that in dialogues narrating past situations, Breton has the “expected” distribution with many PsPf. Is it expected because this is simply a known fact, or expected given the findings of this paper? See also ll. 255–6.
• 434–5: “Sp in Breton is not an option, since this is oral language”. Perhaps I missed it, but was this stated? The statement (ll. 28–29) that it is used in purely narrative modes of discourse is not quite the same, but perhaps this was indeed specified elsewhere.
• Similarly, ll. 333–4 state “the remaining Sp forms of the English original are rendered by means of the Breton PsPf because the reference to chronologically ordered past situations is explicit”. Is it common knowledge that the PsPf is used with explicit reference to chronologically ordered past situations in Breton, or does this emerge from the present study (cf. also ll. 533–4)? If the former, then a reference to such findings might help. If the latter, then perhaps the passage should be slightly reformulated to make this clearer.

In this final case (ll. 333–4), the following paragraph clarifies, but this is not always so in the article. I would advise checking this carefully. It is indeed important because it will make clearer what exactly this paper contributes. For me at present, this is, at least on some points, somewhat fuzzy.

 

Specific questions:

 

ll. 451–3: Why mention PsPf occurrences if they are not presented in the following tables anywhere?

 

For the diagrams, consistent colors across graphs would be helpful (the imperfect / progressive past in Figure 4 is yellow, while in Figure 2 it is orange).

 

Tables 1 and 2: why the empty column?

 

ll. 502–3: Were you liberal in counting resultative types or are there simply a lot of them? The former suggests you had leeway in your decisions while the latter suggests that this is simply a fact of Breton.

 

ll. 724–7: It is unclear what the purpose of bringing up McCoard and Portner is, since it seems like you do not intend to adopt this line of analysis. Perhaps it is simply a matter of showing your intent / signposting the argument: “Perhaps some version of McCoard and Portner’s extended now function could be used to explain the PfPf use, but an alternate and superior analysis is offered below”.

 

There is no section 5 (though it is stated to exist at the end of section 1). This should be fixed somehow.

Author Response

First, let me thank you for the very useful comments on my paper. Below is a summary of the main modifications I’ve made, taking your comments into account :

- I’ve added substantial information about the dialectal status of Breton, the other corpus used (and how I used it), and the variant of the language on which this study is based ;

- I’ve also clarified the status of the statements I made (whether already known facts or emerging from the sudy).

- I’ve corrected all the unclear passages, and added many footnotes to clarify some of the points you raised. I think the distinction between the two perfect forms comes out as clearer as in the first version.

Reviewer 3 Report

What follows is a list of small corrections and remarks on details.

5 in the aim -> with the aim

23 in the aim -> with the aim

136 Tense 2 constraint: explain what you mean by this.

139  imperfect: ->  imperfect.

244 To conclude on this section -> To conclude this section

314 severed -> separated

323 types of perfects -> typology of perfects

448 Why do you have an empty column in Table 1?

454 Why do you have an empty column in Table 2?

484 In many languages the Present Perfect is the diachronic source of the mirative, expressing surprise. Might this be at stake here as well?

502-3 we were liberal in counting resultative 502 types.  Don't you mean that the resultative in Breton is of the liberal types, as claimed in 452?

533 severed -> separated

 

Author Response

Thanks for your observations. In the revised version I've included all your comments and corrected and clarified what you pointed out.