Corrective and Exhaustive Foci: A Comparison Between Italian and French

Round 1

Reviewer 1 Report (Previous Reviewer 1)

Comments and Suggestions for Authors

Please find my comments in the pdf attached here.

Comments for author File: Comments.pdf

Comments on the Quality of English Language

The paper will benefit from a thorough proofreading. cf. lines 286-288

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report (New Reviewer)

Comments and Suggestions for Authors

See my comments, suggestions and questions in the file uploaded.

Comments for author File: Comments.pdf

Comments on the Quality of English Language

Some minor typos, some of those can be found in the file uploaded.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report (New Reviewer)

Comments and Suggestions for Authors

See the file attached.

Comments for author File: Comments.pdf

Comments on the Quality of English Language

Some typos were spotted, and are included in the file attached above. In some case sentences were not clear and I've asked to rephrase them -- these comments are also in the file.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 3

Reviewer 2 Report (New Reviewer)

Comments and Suggestions for Authors

I think the paper improved significantly, the additions that were made enhanced clarity and filled some gaps.

Some minor comments/observations I still have:

• I wouldn't dare to say your findings challenge "the idea that Italian is less syntactically rigid than French", as author(s) do in the abstract: as I already pointed out in previous revisions, your data can only say that as far as the specific focus type you are investigating Italian does not seem to be less rigid than French. But given that the type of focus you investigated is a 'new beast' you can't draw conclusions on the other types -- and you cannot draw conclusions in the comparison on the two languages. The abstract should make clear that the author(s) are not epanding their findings to other types of focus.
• Author(s) should add the "and also" test to the French example in (6)
• Lines 267-275 (as well as 422-426, 453-458): I'm wondering whether you could also expect something different, namely a decrease in the acceptability of FF given that the two features you are combining are at different points of the scale proposed by Cruschina (2021). Moreover, let me point out that if there aren't experimental data supporting the scale, or at least showing the acceptability of FF in corrective focus, it is hard to draw such predictions (the problematic aspect of your study being that you are skipping a passage and directly investigating a combination of features instead of each of them separately and only later the two combined). Notice also that the references given in several parts of the paper which show that in-situ focus is in any case the preferred strategy (see for instance lines 707-710) are already challenging for your own predictions. In fact, the author(s) themselves add "Thus, the higher ratings expressed for in situ [+corrective, + exhaustive] Foci are not entirely unexpected." It seems to me that this is in contradiction with the predictions made in 1.5.
• Lines 462-464: "we expect sentences with an explicit EM to be rated as more acceptable in both languages, as the EM overtly contributes to encoding the [+exhaustive] feature." But yet again, this is a prediction you can have for an exhaustive focus, are you so sure it has to work in the same way when more than a feature is involved?
• Adding a temptative explanation of why your expectations were not met would have made a nice addition in the conclusions

Typos:

• line 329: Lodovici with capital L + missing 's (Samek-Lodovici's claim that...)
• there's no reference to Figure 7 and Figure 15 in the text
• line 719: while it has no effect
• line 776: unergative should not have a capital letter

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Overall, this article addresses interesting questions regarding the acceptability of different syntactic structures in contrastive focus context, bringing in evidence from two Romance languages. The research questions are well-motivated and the comparison between Italian and French is justified. My major concern is with the author(s)' choice of the statistical analysis and the reporting of their results.

The author(s) chose to carry out ANOVAs on Likert scale ratings. I wonder why they did not use Linear Mixed Effects Regression (LMER) that has been shown to be more powerful than ANOVAs. Considering the large number of participants for the experiment, it makes sense to use LMERs to take account of participant based variation. Moreover, the use of items as a random effect might affect the results for French, where an effect of verb type was reported.

On the other hand, as the data consists of Likert scale ratings, it would have been more appropriate to use Generalized Additive Mixed Models (GAMMs) that have been shown to be more suitable for ratings data. The authors should clarify and explain their choice of statistical analysis for their experiment.

I found that the reporting of results is not correct. The authors keep reporting the presence/absence of 'correlation' between ratings and independent variables. ANOVAs do not test correlation. As mentioned above, authors need to read up on their choice of statistical analysis and interpret and report them correctly.

To control for variation in ratings, the authors decided to normalize them. A reference is needed to justify this decision, followed by an explanation about how to interpret the z-normalized ratings.

Their analysis of verb type x syntactic structure shows that the difference between transitive & unaccusative verbs is independent of structure. How do the authors explain the difference in ratings for verb types in French?

Minor comments:

1. The author(s)' use of 'agrammatical' vs. 'ungrammatical' and the use of * and # to indicate grammaticality judgments needs to be consistent.

2. Table 1 can be moved to an appendix at the end of the article. This comment is also relevant for all the data tables presented in the results section. The author(s) should follow the established guidelines for reporting the results of their statistical analyses. This maybe found with a simple Google search.

3. The author(s) need to present an analysis of ratings for different
EMs in French. This maybe done in just a footnote, if no significant differences are found.

4. The results section is slightly disorganized. I recommend the use of subsections to make it easier to read and understand the results.

5.   Line 395: I think the reference should be to Table 2 and not Table 3.

6. A number of references in the Reference list are missing in the text. The citations should be checked again to make sure that the uncited references are removed.

Author Response

We would like to thank the reviewer for his/her precious suggestion.

Answers to his/her comments are highlighted in green.

Changes and correction in the article are highlighted in green.

 

Review

Overall, this article addresses interesting questions regarding the acceptability of different syntactic structures in contrastive focus context, bringing in evidence from two Romance languages. The research questions are well-motivated and the comparison between Italian and French is justified. My major concern is with the author(s)' choice of the statistical analysis and the reporting of their results.

The author(s) chose to carry out ANOVAs on Likert scale ratings. I wonder why they did not use Linear Mixed Effects Regression (LMER) that has been shown to be more powerful than ANOVAs. Considering the large number of participants for the experiment, it makes sense to use LMERs to take account of participant based variation. Moreover, the use of items as a random effect might affect the results for French, where an effect of verb type was reported.

On the other hand, as the data consists of Likert scale ratings, it would have been more appropriate to use Generalized Additive Mixed Models (GAMMs) that have been shown to be more suitable for ratings data. The authors should clarify and explain their choice of statistical analysis for their experiment.

We clarified and explained our choice, and we further give evidence for the robustness of the models used in section §2.1 (also in notes 11-15).

I found that the reporting of results is not correct. The authors keep reporting the presence/absence of 'correlation' between ratings and independent variables. ANOVAs do not test correlation. As mentioned above, authors need to read up on their choice of statistical analysis and interpret and report them correctly.

We thank the reviewer for this observation. We removed every reference to ‘correlation’, using only terms regarding ANOVA (such us ‘affect’, etc.) through the whole text. Each correction is highlighted in green.

To control for variation in ratings, the authors decided to normalize them. A reference is needed to justify this decision, followed by an explanation about how to interpret the z-normalized ratings.

A reference (Gomez, 2013) has been added, followed by an explanation about how to interpret z-scores.

Their analysis of verb type x syntactic structure shows that the difference between transitive & unaccusative verbs is independent of structure. How do the authors explain the difference in ratings for verb types in French?

We gave a possible explanation in §3.2.1.

Minor comments:

1. The author(s)' use of 'agrammatical' vs. 'ungrammatical' and the use of * and # to indicate grammaticality judgments needs to be consistent.

We changed accordingly.

2. Table 1 can be moved to an appendix at the end of the article. This comment is also relevant for all the data tables presented in the results section. The author(s) should follow the established guidelines for reporting the results of their statistical analyses. This maybe found with a simple Google search.
   
   

We changed the way we reported the results of our statistical analyses following the established guidelines (specifically, we followed the APA style). In turn, all tables from Table 2 on have been eliminated.
As for Table 1, we left it in the main text following Languages guidelines: “Figures, Schemes and Tables should be inserted into the main text close to their first citation”.

3. The author(s) need to present an analysis of ratings for different
   EMs in French. This maybe done in just a footnote, if no significant differences are found.

Since no significant differences were found, the result is presented in note 30.

4. The results section is slightly disorganized. I recommend the use of subsections to make it easier to read and understand the results.

We changed accordingly.

5. Line 395: I think the reference should be to Table 2 and not Table 3.

Since we changed the way we reported the results of our statistical analyses, this reference has been deleted.

6. A number of references in the Reference list are missing in the text. The citations should be checked again to make sure that the uncited references are removed.

We checked and changed accordingly.

 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

see attached .pdf

Comments for author File: Comments.pdf

Author Response

We would like to thank the reviewer for his/her precious suggestion.

Answers to his/her comments are highlighted in green.

Changes and correction in the article are highlighted in green.

 

Review

Summary and Recommendation

This paper presents two acceptability studies investigating the realization of [+corrective,+exhaustive] focus in Italian and French. Corrective focus is defined as replacement of “one entire proposition with an alternative that is explicitly mentioned in the discourse.” Exhaustive focus is taken as selecting a unique referent from a set of alternatives. The latter is implemented by use of the focus marker solo for Italian and seulement for French, plus the construction of ne . . . que in French for in-situ focus.

The study finds that for Italian preferably in-situ focus is used for corrective focus both with or without an exhaustive focus marker, whereas fronting and clefts are not used for this focus type; for French, both in-situ and clefted structures are equally available, whereas focus fronting is strongly disprefered. The results are taken as support for distinguishing two focus heads, one for exhaustive interpretation and one focus head which licenses corrective focus. While I find the empirical data interesting and generally sound, the link to the theoretical conclusions drawn from the results could and should be made more precise.

 

Recommendation: Major Revisions required

As my major concerns are with the analyses of the different structures, the author(s) might want to reconsider the structure of the paper. A possible approach could be to provide different analyses of the focus constructions under consideration from the literature first and what these analyses would predict for the different focus marking strategies in French and Italian. Then they can discuss the data in relation to these predictions. The paper could also make a stronger point about how differences between the two languages can be accounted for.

 

Major comments

• Data: The paper does not provide raw scores, and it is not reported whether data are normalized with or without fillers. This should be reported because it is relevant as the paper takes ratings below zero to mean that native speakers do not accept these structures (p. 10). When calculating the z-scores, the mean for each participant is set to zero, so that those using the scale up to 7 and those using the scale only up to 5 can be compared. In a potential scenario in which all test sentences are rated generally quite good, the mean can be close to the maximum rating, this does not mean then, that the ratings below zero do not accept these structures; they are just relatively worse;

Even though raw scores cannot be 100% reliable (Gomez, 2013) (as a matter of fact it appears that French informants tend not to use the higher part of the scale [5-6], while Italian informants tend not to use the lower part of the scale [0-2]), we provided relevant raw scores in footnote, after each figure.

We specified that data have been normalized excluding fillers in note 11.

Following this comment, we deleted every reference to “grammatical, ungrammatical, acceptable and not acceptable” and reformulated the paper in terms of “more/less acceptable” ore “favored/disfavored” which, we think, is more reasonable for this paper.

 

• Test sentences: It would be very helpful to have all the test sentences and corresponding pictures in an appendix. From the picture and the test sentences, you might have an intervening factor (note though that I am not a native speaker of French or Italian): when looking at the picture and getting the first two sentences, I would think that participants interpret that as matching the picture, after all three mentioned items are somewhat out of the pencil case; so when getting the correction, they need to reinterpret what is meant by the verb take and this in turn might lead to an additional emphasis on the verb;

We provided all the test sentences in Appendix A (for Italian) and Appendix B (for French).

In order to check for any possible intervening factor, the tests were beta-tested with a small group of people who were then interviewed. No intervening factors emerged. Furthermore, we selected different verbs, specifically choosing verb which would not allow possible reading of “contemporaneity” (i.e., inciampare ‘stumble’ on something. In fact, one can only stumble on one thing [at time], when more objects are placed far away from each other, like in pictures taken for the present experiment). Finally, we also conducted a statistical analysis which show that there is not statistical difference between the four verbs selected for each verb type (footnote 10).

 

• While citing Belletti’s work you do not consider her proposal of a focus position within the TP, which might well be relevant for the difference between French and Italian that you find;

This question has been dealt with in footnote 21.

 

• Two focus heads: It is not obvious how the reported results support the assumption of two focus heads. The results show a difference between the sentences with and without the focus marker solo in Italian with ratings increasing when solo is present. For French the opposite effect is observed. Nevertheless, the paper proposes two focus heads for both languages, with essentially the same kind of licensing mechanism, independent of the presence or absence of the focus particle.
• Ordering of FocP » Ei: The argument for ordering based on the ban of crossing paths is not convincing. First, licensing is not movement, so I would rather expect you to find similar restrictions in the domain of agreement. Second, why should the exhaustivity operator license only the FM solo / seulement and not the whole phrase?

We really thank the reviewer for comments 4 and 5 which definitely allowed us to improve the quality of our paper. While the analysis in terms of crossing path has been removed, we provided cross-linguistic evidence. Furthermore, previous literature and an ad hoc prosodic analysis further corroborated our analysis. Specifically, this question has been dealt with in section 3.1.4.

 

• Exhaustivity in focus vs. exhaustivity of only: there is a long-standing discussion of the differences of exhaustive focus (in clefts) and exhaustivity of only (see originally Horn 1981, but also the discussion in Beaver & Clark 2008); in your analysis exhaustivity is licensed by the same head; does that mean you assume that these are the same operators? If not, what are the differences?
• French vs. Italian: The paper does not offer an explanation of the differences between French and Italian concerning the presence of FM. I could imagine that it helps to think about how the interpretation with and without the FM are arrived at. I would think that the relevant alternatives that are construed in (8) and (9) might be different for the focus marking vs. only strategy.

Alternatives with only: {Marco took the pen; Marco took the pencil; Marco took the rubber} only asserts that just one of these alternatives is true and the others need to be discarded; (btw. does this still fall in the class of corrective focus?)

Alternatives without FM: {Marco took the pen and the pencil and the rubber} Correction of this proposition: Marco took the pen;

If that is true then the difference between Italian and French is that for the latter the construal of the second set of alternatives is more plausible, which might be related to the fact that French wh-questions are frequently answered with clefts;

In the present paper we specifically offer a syntactic analysis of Corrective and Exhaustive foci in Italian and French. However, comments 6 and 7 have been taken into considerations in footnote 33 and through the whole paper.

 

• Criterial Freezing? For French clefts, the paper proposes that the clefted constituent moves from Spec,EiP to Spec,FocP. Shouldn’t the phrase be frozen (see Rizzi’s Criterial Freezing) in Spec,EiP?

We really thank the reviewer for this comment, which definitely allowed us to improve the quality of our paper. This question has been dealt with in section 3.1.4.

 

Typos and other issues

1. 1, Abstract “Contrastive Focus implies the correction of one entire proposition with another alternative” Shouldn’t that be corrective focus?

We changed accordingly.

 

1. 1, Abstract Kiss 1998 ->É. Kiss 1998 please change throughout and in references;

We changed accordingly.

 

p.2, last paragraph it exists -> there exists

We changed accordingly.

 

1. 12, ex. 10’ There is a trace ti in TP which does not have an antecedent in the structure

We changed accordingly.

 

1. 19, 1st paragraph “Starting from Den Dikken et al. (2000) . . . ” -> analsyes that relate clefted constituent and cleft clause in a predication structure can already be found in Delahunty (1982, 1984) and Heggie (1988);

We changed accordingly.

Author Response File: Author Response.pdf