Relative Clauses in Native Lower Sorbian and the Relativizer how
Round 1
Reviewer 1 Report
Comments and Suggestions for Authors
The paper is very interesting, the data are clear, the hypothesis is clear; however, I have a concern related to the formal analysis: the idea that that initial k- is deteled due to the movement of an empty operator in SpecCP, although used to explain similar cases in other languages, is not very convincing; I would also mention the possibility of a phonological process, since in many examples in the paper there is a final -ka sillable in the word preceeding the ka complementiser.
Also, before presenting the corpus data, it would be useful to have a table with the inflectional forms of the relative elements.
Please revise the paragraph 2 on page 2 (line 51 - the comma)
Author Response
Comment 1: I have a concern related to the formal analysis: the idea that that initial k- is deteled due to the movement of an empty operator in SpecCP, although used to explain similar cases in other languages, is not very convincing; I would also mention the possibility of a phonological process, since in many examples in the paper there is a final -ka sillable in the word preceeding the ka complementiser.
Reply: I added a new endnote where I discuss this idea (endnote 22) and argue that k-deletion applies regardless of the shape of the preceding word. If I misunderstood the alternative, please let me know.
Comment 2: Before presenting the corpus data, it would be useful to have a table with the inflectional forms of the relative elements.
Reply: I added a new table containing the inflectional paradigm for 'kak' (table 5), because inflectional forms otherwise play no role in this paper. If more specific inflectional information is wanted, please let me know.
Reviewer 2 Report
Comments and Suggestions for Authors
# Review of *Relative clauses in Native Lower Sorbian and the relativizer *how
The article presents a syntactic analysis of relative clauses in Lower Sorbian, a Slavic language spoken in eastern Germany. The main focus of the article is on the relativizer *ak* 'how', which is claimed to be the most common relativizer in Native Lower Sorbian (the variant of the language which is the main data source for the article). The main claim of the article is that *ak* is a deleted version of the complementizer *kak* 'how'. The deletion is claimed to be triggered by empty operator movement, as formalized in (26).
I am not a syntactician, but I found the article to be well-written and clear. I have a few questions and comments that I hope the author(s) can address in a revision and that I hope will help improve the article:
1. The rule (26) is somewhat controversial since the usually adopted view is that modularity bans the direct interaction between syntax and phonology. There is just one comment on this in the last section of the article. I think the author(s) should address this issue more explicitly and either provide more evidence for the rule or discuss alternative analyses.
2. If *ak* relatives are the main type of relative clauses in Native Lower Sorbian, they should be able to be used both restrictively and non-restrictively, unlike in other Slavic languages where e.g. *čto* (Russian) or *co* (Czech) is preferentially used in non-restrictive contexts.
3. In Section 7, the author(s) show how *ak* differs from both German and Czech. But is that empirical difference related to the theoretical claim of the article? If not, the author(s) should discuss this more explicitly. Connected to this: if *ak* is the default relativizer in NLS, it is also not surprising that it is compatible with a wider range of head nouns than in other languages (Czech, e.g.).
In sum, I think the article is a valuable contribution to the linguistic literature and I consider the article to be publishable after the author(s) address the minor issues I raised above.
Comments for author File: 
Comments.pdf
Author Response
Comment 1: The rule (26) is somewhat controversial since the usually adopted view is that modularity bans the direct interaction between syntax and phonology. There is just one comment on this in the last section of the article. I think the author(s) should address this issue more explicitly and either provide more evidence for the rule or discuss alternative analyses.
Reply: I have added a new subsection in section 5 (pp. 13-14), where I address this issue. Many thanks for bringing it up! I hope it is sufficient for the purpose of this paper (certainly much more can be said about the interaction between syntax and phonology).
Comment 2: If ak relatives are the main type of relative clauses in Native Lower Sorbian, they should be able to be used both restrictively and non-restrictively, unlike in other Slavic languages where e.g. čto (Russian) or co (Czech) is preferentially used in non-restrictive contexts.
Reply: Thanks for bringing up this issue. From my experience with native speakers, my intuition is that it is indeed also used in non-restrictive RCs. I could only find one example in the corpus, which I added to the examples in the section on Czech, since there I discuss that ak-relatives have a much broades usage than jak-relatives in Czech. I thought this to be a suitable place. If it turns out that my intuition is wrong, I suspect that this wouldn't be a big problem. Also in English, the relativizer 'that' is banned from non-restrictive RCs, yet it is is the dominant strategy to form RCs. If there is something I misunderstood, please let me know.
Comment 3: In Section 7, the author(s) show how ak differs from both German and Czech. But is that empirical difference related to the theoretical claim of the article? If not, the author(s) should discuss this more explicitly. Connected to this: if ak is the default relativizer in NLS, it is also not surprising that it is compatible with a wider range of head nouns than in other languages (Czech, e.g.).
Reply: I revised the introduction to section 7, in order to make transparent why I discuss these two types in connection to ak-relatives. I hope this revision removed the opacity of section 7 of the previous version.
