Romance and Croatian in Contact: Non-Clitic Auxiliaries in Istro-Romanian

Round 1

Reviewer 1 Report

The article provides a good and interesting picture of the auxiliaries in Istro-Romanian. The authors describe the grammatical properties showing that they cannot be considered as clitics and are very similar with their counterparts in old Romanian. The Romanian diachronic issues are carefully discussed and compared with the synchronic state of Istro-Romanian (since this variety lacks of attestations older than the 19^th century). The authors claim that the language contact with Serbo-Croatian contributed to the preservation of some archaic properties in Istro-Romanian that are now lost in Romanian.

A revision taking into account the comments present in the attached pdf is needed. The following points, in my opinion, could also be reconsidered.

1. Glosses in original examples need to be consistent throughout the paper: for example, the Romanian and Istro-Romanian examples are glossed following the Leipzig Glossing Rules, while the Serbo-Croatian in (12), (18) (that should be renumbered (19)), (19) (actually (20)), (20) (actually (21)) follow minimal glossing. If the authors want to maintain this distinction, it should be stated clearly in the text.

2. The article does not mention the distinction between Northern and Southern Istro-Romanian, two mutually comprehensible but distinct varieties that diverge in verb inflection. The authors should consider to insert a note saying that with the term Istro-Romanian, they refer to all the possible varieties present on the territory. Since this distinction is not present in the text, the reader could suspect that some grammatical properties could be present only in one of the two varieties. The authors should double-check the empirical phenomena in the sources.

3. The authors should check also the transcription of their Istro-Romanian examples. First of all, a note about the different transcription methods is necessary since they use a range of different authors in their corpus: for example, the palatal [c] is sometimes transcribed as <chi>, sometimes as <t”>. Regarding the latter example, the form vet^u is present in (6c) while the source presents vet”. This is maybe the result of a poor OCR recognition, as it affects a lot of other examples (see the comments in the pdf). An alternative solution would be the standardization of the examples following one method of transcription.

4. The corpus could be easily augmented consulting the site https://www.vlaski-zejanski.com/en/, where there are a lot of texts with translations provided in English.

5. The scholars that approached Istro-Romanian usually consider the Croatian (or Serbo-Croatian) as the Source Language of some grammatical properties. However, since the settlement of the “Istro-Romanians”, they were in contact with the Cakavian variety. Unfortunately, though, there are very few corpora that the authors could consult to gather more information about Cakavian. A reference that may be useful is Kalsbeek (1998) on the Cakavian variety of Orbanici, 50 kilometers from the first “Istro-Romanian” village.

Regarding the use of “Croatian” and “Serbo-Croatian”, the authors could consider the introduction of the conventional abbreviation BCS (Bosnian-Croatian-Serbian) or SBC or BCSM, since the examples present in the study are usually from the Serbian area.

Bibliographical references present in this review and in the annotated pdf:

Filipi, G. 2002, Istrorumunjski lingvistički atlas/Atlasul Lingvistic Istroromân/Atlante Linguistico Istrorumeno, Pula: Znanstvena udruga Mediteran.

Kalsbeek, Janneke 1998. The Čakavian Dialect of Orbanići near Žminj in Istria. Studies in Slavic and general linguistics 25. Amsterdam - Atlanta: Rodopi.

Vrzić, Z. & R. Doričić 2014, Language contact and stability of basic vocabulary:

Croatian words for body parts in Vlashki/Zheyanski (Istro-Romanian) «Fluminensia»

26(2): 105-122.

Comments for author File: Comments.pdf

Author Response

To begin with, we are grateful to the reviewer for their comments and suggestions, which have helped us improve the paper in many respects. Thank you very much for the bibliografical recommendations as well.

1. We have regularised the glosses and we have checked the transcriptions of the Istro-Romanian examples (points 1 and 3). Also, we have inserted a footnote which clearly states that we have preserved the transcription conventions used in these sources.
2. We have introduced the distinction between Nothern and Southern Istro-Romanian, and pointed out that the phenomena analysed in the paper show the same behaviour in both varieties. (point 2)
3. As recommended, we have consulted Kalsbeek (1998), but we were unable to find data (or a discussion) of interest for the particular phenomena discussed in our paper. We have however introduced a footnote pointing out that Istro-Romanians were in contact with the Čakavian variety; on our recent fieldwork in Istria we have discovered that the Istro-Romanians we have consulted as informants no longer speak Čakavian; one of the informats actually explicitly stated that she gave up speaking the local dialect in favour of standard Croation because she was marginalised in school for speaking the dialect.
4. We have also adopted the conventional abbreviation BCS (Bosnian-Croatian-Serbian) for the phenomena that characterise all these varieties.
5. Thank you very much for the suggestion of augmenting the corpus with https://www.vlaski-zejanski.com/en/. For this particular paper, we have chosen to restrict ourselves to the traditional written corpora mentioned in the sources (which have been extensively analysed); we will make good use of this resource in our future research.
6. And, finally, thank you very much for all the suggestions in the attached document.

Reviewer 2 Report

Romance and Croatian in contact: non-clitic auxiliaries in Istro-Romanian

1 General observations

The paper Romance and Croatian in contact: non-clitic auxiliaries in Istro-Romanian presents novel empirical phenomena from Istro-Romanian, which have also not been explored formally. The discussion centers around the fact that Istro-Romanian auxiliaries present characteristics that are indicative of a non-clitic status. For example, as shown in the paper, they can license verbal (VP) ellipsis - context (2e) in the paper inserted here under (1), where the perfect auxiliary licenses the deletion of the embedded verbal material (the past participle).

(ex. 2e in the paper)

As is well known, this phenomenon is severely restricted in standard Romanian (as in the example we provide in (2)), and in fact most Romance languages. This makes the Istro-Romanian data worth of further investigation so as to pin down their nature, the limits of variation, and to best model them theoretically.

The authors point out the important issue of surface string identity (row 678) between old Romanian, Istro-Romanian, and Croatian. Language contact between the Romanian variety spoken in Istria and Croatian (which exhibits verbal ellipsis licensed by auxiliaries) is assumed to have acted as a catalyst, generating convergence in Istro-Romanian. This has allowed the preservation of various archaic properties, among which the non-clitic status of auxiliaries, whose traces are still visible in some old Romanian data (as the authors show) but have been lost subsequently, as the Romanian auxiliaries have advanced on the grammaticalization cline. Figure 1 is particularly instructive.

The topic discussed is of interest and we agree with the authors that it definitely requires more attention than it has received. The topic is related to non-trivial issues in the grammar, such as the nature and limits of ellipsis, the notion of clitic, the process of grammaticalization, variation under language contact, etc. Also, the paper contains discussion and analysis of Eastern (and Daco-) Romance dialects (in particular, Istro-Romanian), which have been rather neglected in formal accounts. The analysis is supported by the data, and the paper also shows a good use and coverage of the bibliography. Taking all these aspects into consideration, the general conclusion is that the paper is suitable for publication, once some (minor) revisions are made, as indicated on the .pdf itself. However, these changes should not take extensive time to implement and we are positive the authors can take care of them very fast.

2 Other observations

We are including here some further comments about style and grammar. As indicated on the paper itself, there are various places where the prose appears to be too dense. Maybe those parts can be further simplified and clarified so that they are crystal clear to all types of readers, even those who are not specialists in Romanian syntax but would like to learn more. If possible, more background on Croatian auxiliary-licensed-ellipsis would be very useful and also some more examples from Romanian would made the discussion even clearer too. Also, it would be useful to have a native speaker proof read the paper. Regarding the data themselves. we have briefly checked some of the examples in some of the corpora mentioned. In some places there are orthographic differences - maybe the authors could indicate where the text has been adapted or/and some other changes that have been made. This would avoid any further confusion, especially for readers who are not specialists in the topic and do not know yet all the conventions in this area.

Thank you very much for bringing this interesting paper to our attention. We wish the author(s) best of luck with their research. This is very nice work.

Comments for author File: Comments.pdf

Author Response

Thank you very much for your evaluation and for all your observations and suggestion; we are genuinely grateful!

1. We have tried to accommodate all the comments in the attached document; we have tried to provide answers to most of the issues raised; however, for some problems we have been limited by the existing bibliography (for example, auxiliary-licensed VP-ellipsis in Croatian/BCS is very poorly documented in the literature).
2. We have added a note concerning the transcription conventions and we have elaborated on the interlinear glosses to make the examples more transparent.

Reviewer 3 Report

This paper presents some very interesting data concerning the nature of the auxiliary system of Istro-Romanian, and argues that this system preserves certain archaic (Daco-)Romance features, possibly due to contact with Croatian. However, there is one major analytic, theoretical flaw running through the paper. This flaw concerns what is meant by "clitic auxiliaries". It is claimed that the Istro-Romanian auxiliaries are not clitics; the fact that they license VP-ellipsis is supposed to support this. This is very reasonable in the light of extremely well-known English facts like "I will eat my dinner and John will/*'ll too". However, it does NOT follow that "most of the other Romance languages" (p. 2) have clitic auxiliaries; there is little independent motivation for treating Modern French or Italian auxiliaries as clitics, for example. Non-clitic aux may well be necessary but not sufficient for licensing VP-ellipsis. It is also stated several times that Rivero's Long Head Movement is not head-movement, but Note 8, p. 11, states that in fact it is, following Vicente (2007). It is head-movement (movement of a minimal category) to a Specifier position. If aux-raising were truly phrasal movement (as stated at the top of p. 12), we would need an account of how the VP-complement of the aux raises out of AuxP/TP/IP before the remnant phrasal category raises, and this is not addressed. Furthermore, this isn't necessary if the proposal in Note 8 is adhered to. This all leaves the reader somewhat perplexed. It seems as though the notions of "clitic" and "head" are being confused. But an auxiliary can perfectly well be a head without being a clitic; cf. non-contracted aux in English for example. All of this needs clarifying; in this connection a full analysis of the range of {Aux, V, clitic} orders seen in (10) is necessary; this may clarify what is being assumed about auxiliaries. Finally, the claim that contact with Croatian may have caused an archaising system to survive is not unreasonable, but -- again -- the nature of Croatian aux needs to be clarified: some of them (e.g. the 3sg copula "je", seen in (19a) and (20)) are known to be second-position clitics, while there are two series of several auxiliaries, a clitic one and a non-clitic one. The fact that Croatian has clitic auxiliaries and most Modern Romance languages don't makes the contact claim implausible; perhaps it's more likely that contact with Croatian has caused the second-position (Tobler-Mussafia) nature of some Istro-Romanian clitics, including perhaps some auxiliaries, to be retained.

Author Response

Thank you very much for the comments, especially those concerning the formal analysis (naturally, we have considered all the comments and tried to accommodate them accordingly). With respect to the formal analysis, we acknowledge that we have oversimplified the story by introducing a distinction between heads and phrases with respect to verb movement – an oversimplification which is downright incorrect! Thus, when looking at V-to-C movement in old Romanian (and at the limited instances of V-to-C movement in modern Romanian) we observe that the correct generalisation is that clitics cannot undergo movement to C (the case of auxiliaries), unless they are incorporated into the structure of the moving verb (the case of clitic aspectual adverbs, which undergo movement to C along with the lexical verb). The difference between stressed and unstressed auxiliaries in Croatian/BCS also points to this fact (this is indicated in a footnote). We believe that the non-clitic nature of Istro-Romanian auxiliaries is what allows them to undergo movement to C (it is not accidental that Istro-Romanian auxiliaries may occupy clause-initial positions, satisfying Wackernagel/Tobler-Mussafia – active to a certain extent in Istro-Romanian). We acknowledge that the fact that a property that is strictly morphophonological in nature has effects on verb movement is at odds with how interfaces (and the general architecture of the language faculty) are conceived in minimalism; but the empirical facts strongly point to this analysis. In a footnote, we have speculated that the cause of this might be the fact that clitics can not access specifiers, under the premises that Vicente’s account is adopted. (We have also clarified the analysis of Long Head Movement, clearly specifying that by LHM we refer to a word order pattern first extensively described by Maria Luisa Rivero). However, we did not insist too much on the formal analysis, as the main goal of the article is represented by the novel empirical data. 

Round 2

Reviewer 3 Report

The paper is much improved compared to the earlier version and can now be published.