Some Observations on the Cantonese Lexical Suprafixes

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This excellent article presents a rather exhaustive discussion of suprasegmental tonal affixation. It is primarily descriptive and empirical, and provides a detailed summation of all the relevant forms of suprafixation in Cantonese. It discusses both synchronic and diachronic aspects of the suprafix feature and reviews competing studies about its origin and development. The article fully engages with all the relevant scholarship and provides ample and apt illustrations and examples. A full and careful reading found no issues that need improvement. All findings and conclusions are well argued and well supported. The appendices provide useful supplementary data. Overall, the article is an important and highly satisfactory study that will be an important reference on the subject, one that this reviewer will frequently refer to in the course of future work.

 

This reviewer was happy to see the use of tone marks on all pīnyīn in the article, including on place names and other proper nouns. That is a fine practice that should be more widely followed!

 

Also, this reviewer can provide an update regarding Simmons 2023 that is cited on line 591 and in the References section. It will be published soon and can currently be listed as follows:

 

Simmons, Richard VanNess 史皓元. Forthcoming. “Historical phonology beyond the rime books: a preliminary exploration of the phonological path of ér suffix morphology 韵书以外的历史音韵学：儿尾词法的音韵轨迹.” Forthcoming in the inaugural volume of 《汉语研究》Studies in Chinese Language and Linguistics, a journal published by The Institute of Dialects and Culture, Nanjing University 南京大学方言与文化研究所, expected to be out in late 2024.

 

 

Author Response

I would like to thank Reviewer 1 for the encouraging words.

Comment: "Also, this reviewer can provide an update regarding Simmons 2023 that is cited on line 591 and in the References section. It will be published soon and can currently be listed as follows: [...]"

Response: Thank you very much for the update in the publication. I have changed the item in the bibliography accordingly.   

Reviewer 2 Report

Comments and Suggestions for Authors

Major Comments

 

The article is well-written, consolidating rich data relevant to the topic of Cantonese lexical suprafixes. The description provided on this topic is in-depth and precise, and the analysis is rigorous. This topic has been previously discussed mainly within the context of traditional Chinese linguistics. It is commendable that the author(s) have thoroughly discussed it in English and within a contemporary linguistic framework.

 

The only major concern I have about this article is regarding its originality. The greatest new contribution of this article to the existing literature seems to be the inclusion of data from Pinghua. However, this new data leads to a conclusion that is very similar to the commonly accepted one regarding the origin of Cantonese lexical suprafixes. If this manuscript were submitted as a discussion or review rather than an original research article, I would recommend accepting it without any reservations. However, as far as I know, it is submitted as an original research article. I would recommend the author(s) to rewrite the article to highlight/improve its originality.

 

Minor Comments

 

1. In Section 3.3, the author(s) may also comment on or mention the proposal by Zhū Xiǎonóng (2004) that suprafixes originate from child-directed speech.

2. Much of the content in Section 1.1 may be more suitable for endnotes (content about transcription/presentation issues) or could be trimmed (content about “Cantonese”).

3. On page 5, the place name “fat8 lan4-1 sɐi1 法蘭西 ‘France’ (archaic)” is regarded as involving a lexical suprafix. However, as it is a transliteration, the tone 1 in “lan1” could be the original tone.

4. On pages 6 and 7, the author(s) state that “The tone 2 lexical suprafix can be applied to a host syllable in any tone other than tone 1/7 (high level) or tone 2 (high rising)” and that “The tone 1 suprafix can be applied to a host syllable in any tone other than tone 1/7 (high level).” However, since the lexical suprafix tones 1 and 2 have the same forms as the original tones 1 and 2, respectively, is it possible to argue that they are compatible underlyingly, but we cannot hear the difference between with and without the lexical suprafixes in today’s Cantonese?

5. In endnote 5, the tone marks of the Chinese “four-corner” system are not displayed properly in the file.

Author Response

I would like to thank Reviewer 2 for the many useful comments. 

Major comment: "The only major concern I have about this article is regarding its originality. The greatest new contribution of this article to the existing literature seems to be the inclusion of data from Pinghua. However, this new data leads to a conclusion that is very similar to the commonly accepted one regarding the origin of Cantonese lexical suprafixes. If this manuscript were submitted as a discussion or review rather than an original research article, I would recommend accepting it without any reservations."

Response: I would like to raise two separate issues.
Issue 1: The journal classes publications into three types: article, review, and commentaries. A "review" is one that offers a comprehensive analysis, with no new, unpublished data presented. An "article" provides a substantial amount of new information. (I don't blame the reviewer for not having been told by the publisher about what they mean exactly by "article" versus "review"; I also had to check.) My manuscript contains a substantial amount of new information in section 2, including the appendix. Although only a relatively small amount of new information is presented in section 3, the proposal theorised from the new data represents a small but significant addition to the existing framework. 

Issue 2: The manuscript has two rather different main sections. Section 2 talks about what happened within Cantonese, and section 3 talks about what happened in Yuè and Pínghuà dialects in general. The reviewer has objections of the manuscript being classed as "original research" based on the contents of section 3. In section 3, while there are some original contents, it is indeed true that the conclusions that I came up with are largely similar to the existing mainstream conclusions, and most of section 3 acts more as a review of the existing opinions. As stated in section 3, one of the aims of section 3 is to propose a small extension to Kwok (2016)'s cline of grammaticalisation. As the reviewer can understand how much work is involved in each single step in Kwok (2016)'s cline of grammaticalisation, my proposal of one step already involves a lot of original work, as shown in §3.2. I object to the reviewer's characterisation that my manuscript lacks originality. Not only was the original work in §3.2 majorly downplayed / side-glanced, the original work in the entire §2 (especially §2.2 and §2.3) was ignored, including the eleven tables in the appendix.

That said, I do not blame the reviewer; I only blame myself for not having expressed myself in an adequate manner. In fact I would like to thank the reviewer for these comments, from which I have learnt immensely. I have now added statements highlighting which parts contain original research, and which parts only add slightly to existing framework. Also highlighted is the fact that one of the aims of section 3 is to bring to the Anglophone audience important works written in Chinese like Kwok (2016).

Minor comment 1: "In Section 3.3, the author(s) may also comment on or mention the proposal by Zhū Xiǎonóng (2004) that suprafixes originate from child-directed speech."
Response 1: I agree with this suggestion, but it is difficult to incorporate this into the main text. I have added an endnote (endnote 11) talking about this point. Zhū (2004) is also mentioned elsewhere in the manuscript.

Minor comment 2: "Much of the content in Section 1.1 may be more suitable for endnotes (content about transcription/presentation issues) or could be trimmed (content about “Cantonese”)."
Response 2: This I cannot agree with. What is obvious to you as a specialist would be unclear to many readers. In addition, that this journal mandates endnotes instead of footnotes means that many readers would miss reading these important pieces of information that are crucial to the rest of the article. As for the discussion on "Cantonese", I haven't met two specialists who are in total agreement on what range of Yuè dialects should be considered "Cantonese". I have to state clearly what I mean by "Cantonese" in the main text of this article.

Minor comment 3: "On page 5, the place name “fat8 lan4-1 sɐi1 法蘭西 ‘France’ (archaic)” is regarded as involving a lexical suprafix. However, as it is a transliteration, the tone 1 in “lan1” could be the original tone."
Response 3: This is a valid point. I have delete fat8 lan4-1 sɐi1 法蘭西 ‘France’ in the manuscript.

Minor comment 4: " On pages 6 and 7, the author(s) state that “The tone 2 lexical suprafix can be applied to a host syllable in any tone other than tone 1/7 (high level) or tone 2 (high rising)” and that “The tone 1 suprafix can be applied to a host syllable in any tone other than tone 1/7 (high level).” However, since the lexical suprafix tones 1 and 2 have the same forms as the original tones 1 and 2, respectively, is it possible to argue that they are compatible underlyingly, but we cannot hear the difference between with and without the lexical suprafixes in today’s Cantonese?"
Response 4: This is a very good point. I have added a few more sentences at the relevant locations towards the end of section 2.1, expressing my deeply-considered agnosticism towards this issue.  
 
Minor comment 5: "In endnote 5, the tone marks of the Chinese “four-corner” system are not displayed properly in the file."
Response 5: I will try to coordinate with the publisher to make sure that a good typeface is used, so that these symbols are correctly displayed in the article.

Reviewer 3 Report

Comments and Suggestions for Authors

Please see attached document.

Comments for author File: Comments.pdf

Author Response

I would like to thank Reviewer 3 for their detailed comments. I have learnt immensely from them.

General comment 1: "My first general comment is that the whole discussion on the origin of the suprafix is presented in a quite confusing and redundant fashion. The discussion is scattered here and there in the paper. I read about it in §3.1, where the A is moving to the analysis of non-Cantonese Yue and Pinghua dialects. Why this long digression here? I suggest to do this earlier or later, and separate it from the treatment of individual varieties. Or, at least, to connect better the different parts. And then the topic is taken up again in §3.3. It's really confusing. Maybe the A could hint at the issue earlier on in the paper, and then defer the extended discussion to when all the data have already been presented, in §3.3."

Response: Thank you very much for the input. I now see the problem that you have raised. Nonetheless, due to how cumbersome the publisher's template is (the publisher mandates endnotes, and the numbering of endnotes will get majorly messed up), I do not dare to rearrange parts of my manuscript in anyway. What I can offer instead is a paragraph in §1, outlining the rationale of the ordering of the discussions (which, thanks to you, makes me realise is greatly needed). General

General comment 2: "My second point is more of a suggestion/question than a comment. Usually, when discussing suprafixes (see e.g. Chappell 2023, quoted in this article), the basic assumption is that the tone of an earlier suffix is 'transferred' or 'merges' with the tone of the root noun or verb(/adjective). This is particularly straightforward in the case of perfective tone change. Here, however, the hypothesized starting point seems to be that the suprafix is marked on the suffix, most likely *ɲiH 兒. What would be the source for that suprafix? Again. I have always assumed that the suffix itself was the source of the changed tone. Maybe I am missing something here, though."

Comment: That is a very good point. I have now added a few more lines talking about (my agnosticism towards) that in §3.2. While it is true that suprafixes often come from the tone of a segmental affix, evidently that need not be the case. The -ɲiA suffix gained a high tone suprafix to become -ɲiH due to the "iconicity" that you objected / cautioned about.

(Response to the following two comments are given together after the second comment:)

Comment: "L. 25-26: it's a bit unusual to read "modern Standard Cantonese"; the definition of a 'standard' Cantonese language is tricky, since it is not taught in school or used as a real official language (Hong Kong's language policy is ambiguous, and it has no official status in Mainland China). Maybe the A could elaborate a bit on what she means by 'standard' here."

Comment: "L. 36-37: "Yuè is a dialect group within the Sinitic language family. Standard Yuè is Standard Cantonese" – I am not sure I agree with this. Cantonese is the variety which is commonly considered as representative of Yue, like e.g. Meixian for Hakka, but it is not a 'standard' in the usual sense of the word, I think. See the comment above: the A needs to clarify what she means by 'standard' here.

Response: I changed this line to "Standard Cantonese is the representative of the Yue dialect group". A comment on "Standard" Cantonese is provided in the following paragraph, highlighting the fact that only a very small portion of Cantonese is codified in the Hong Kong education system, and the suprafixes as used in colloquial Cantonese are not codified.

Comment: "L. 33-34: "Kwok Bit-chee 郭必之 2016" – I think the Chinese characters are unnecessary here, as well as the name (Kwok 2016 should do). The same applies to the other Chinese authors cited in this article; the use of the name can be limited to case when two different authors with the same surname have a publication in the same year (e.g. Li Rong 2015 and Li Jie 2015)."

Response: I would be happy to comply with what the editors tell me if they prefer me to follow a different standard. Before that, I would refrain from making changes. (Lest I do all these manual changes which are later deemed not necessary; I have never had this request from a reviewer.)

In general, I resist the Chinese-unfriendly practises that are common in Western publications on Chinese. There are reasons why articles in Chinese usually quote entire names rather than simply the surname: Chinese surnames are less distinctive. Amongst my bibliography, more than half of the Chinese surnames are shared by at least one author. For a manuscript on Sinitic languages, even (or especially) one that is aimed at an Anglophone audience, in general I do not follow the subpar Western practise of only quoting the surname of Chinese authors. Also, the names in Chinese characters are there for papers that are written in Chinese. It is a reader-oriented device for readers who can read Chinese.  

Comment: "L. 60-61: "and the Hong Kong data are checked against the online dictionary words.hk 粤典" – I wonder whether this is an authoritative source for a scientific paper. I'd switch to some other 'traditional' dictionary as the main reference, maybe using online dictionaries as additional, complementary sources.

Response: I have now also consulted Bauer (2020)'s dictionary, and added lines in the manuscript to reflect this. I have now also mentioned that I specifically mean HK Cantonese as spoken by people under 50yo or so. (The initial decision to use words.hk was because the data therein best represent modern spoken Cantonese in Hong Kong, by people under 50yo or so. All traditional dictionaries of "Hong Kong Cantonese" are either too old and/or dominated by people from Mainland China.)

Comment: "L. 74: what's an "additive suprafix"? Does the A mean those cases (discussed later) when the tone contour of the root is lengthened? If this is the case, the A should clarify this term when it is first used."

Response: Thank you for the comment. I have now added an explanation of "additive suprafix" at this point of the manuscript. 

Comment: "L. 92-94: "The glossing in this article mostly follows the Leipzig glossing rules (see also Chén Yùjié 陈玉洁 et al. 2014). The two most frequently used abbreviations are SUPR for suprafix, and DIM for diminutive." – It would be clearer for the reader if the A simply added a footnote with the list of glosses she uses (except for those included in the Leipzig list)."

Response: The publisher mandates the use of endnotes; putting this important information in an endnote would not make it clearer for the reader. (In addition, the template provided by the publisher makes it extraordinarily cumbersome if I want to add an extra endnote. I fear for my life if I am made to add another endnote this early in the article.)

Comment: "L. 96: "All morphology in Cantonese is derivational" – not sure I agree. Tonal morphology is also used to mark perfective aspect (the A mentions it later), which is definitely not derivational. Also, compounding and reduplication are common non-derivational morphological phenomena in Cantonese."

Response: I see that we use the term "derivation" very differently. The way that I use it is simply in the sense of morphology being on a cline between inflection and derivation. Compounding and reduplication are strategies of word building, and can be either inflectional or derivation. 
Nonetheless, to avoid confusion for people from different schools of linguistics, I have now deleted that line.

Comment: "L. 108: "Modern Standard Cantonese has no tone sandhi" – what about redulplication with the 哋 suffix? Or cases like bohng yat bohng > bóng-bohng 'weigh'

Response: These involve suprafixes, more specifically, non-lexical suprafixes. (For me at least, as discussed in the manuscript, tone sandhi is the change of tone that is triggered by another tone in the vicinity.) Thank you for the query; I have now included examples of these in the following paragraph on non-lexical suprafixes.   

Comment: "L. 122: "ji4mɐn4 移民 'migrate' --> ji4-2mɐn4 'migrated'" – a genuine question: if this is the product of the contraction of the perfective suffix, shouldn't tone change occur on the second verb syllable, since it's the closest to the suffix itself? Unless this is a separable verb (I don't think so)."

Response: 移 is the verb in Cantonese. The norm in Cantonese is to say 移咗民 'migrated'. Googling (with quotation marks) "移咗民" gives 22100 results, and "移民咗" gives 9560 results. I have added a note at that location in the manuscript to talk about "移咗民". 

Comment : "L. 131: "not uncommonly sounds a bit awkard. Maybe 'quite often'?"
Response: Thank you for the suggestion. Changed as suggested.

Comment: "L. 136: "but is neutral towards whether they should be considered diminutives" – I guess a reasonable compromise is to consider them as deriving from diminutives, independently from the meaning they express now (as is generally the case for suffixes deriving from 'child' in Sinitic and beyond)."

Response: Agreed. This is indeed a good way of putting it. I have made slight changes in the manuscript partially reflecting your suggestion.

Comment: "L. 151, Table 1: the notation used in this table (A, B, C...) is opaque to anyone who has no background in historical Chinese phonology. I strongly suggest to use different terminology, or provide explanations (there is an endnote, but it's not very clear for non-specialists). Maybe there is no need to use those traditional categories? Also, why the stars?"

Response: 1. I have now deleted the stars, you made a good point; 2. I have added lines saying that the Middle Chinese categories are there only for the convenience of specialists, and are basically not essential for the understanding of the article.

Comment: "L. 155: ""Like many other derivational affixes" – maybe too strong? "Like some other derivational affixes"?"

Response: My statement was indeed too strong. Thank you for the suggestion. I have now made that change.

Comment: "L. 166: "With a tone 1 suprafix, -mui6-1 妹 is a suffix meaning '-girl'" – I really don't understand why calling this 'suffix'. To me, it's a regular compound constituent. I see no reason to treat it differently. Same goes for -pʰɔ4 婆 '-woman'. 

Reponse: I disagree; the suffix -mui6-1 is morphologically more bound than the bound mui6 (the one used in compounds), which in turn is more bound than the morphologically free mui6-2. You can still kinda use mui6 in isolation as an independent utterance, but this is certainly not the case for the suffix -mui6-1. 

Comment: "L. 206. "One harder-to-explain example is the 206 name of the suburb of joeŋ4-2 kei1 楊箕 (Yángjī) in Canton"

Response: The Reviewer's comment looks incomplete. My guess is that the reviewer finds the expression "harder-to-explain example" awkward, which I now realise. I have now changed that to "one example which is harder to explain". 

Comment: "L. 228: "I suppose the diminutive form of this would be ap8-2 -tsɐi2 鴨仔." – I suggest deleting this. It sounds irrelevant and speculative."

Response: Thank you for the suggestion. I have now deleted this.

Comment: "L. 230: "One examples is tʰɔŋ4 糖 'sufar' vs. tʰɔŋ4-2 糖 'lolly / candy'" – There is a paper by Jurafsky (1996), "Universal Tendencies in the Semantics of the Diminutive", where this is discussed: one common function of diminutives is the so-called 'individuation/partitive' (delineated part of a mass). The A cites this paper in §3.1."

Response: Thank you for the suggestion. I have now added one line mentioning this at this point of the manuscript.

Comment: "L. 330: "Lǐ Xīnkuí 李新魁 et al. (1995) has" > have

Response: Thank you. I have now changed this.

Comment: "L. 368-369: "the high tone suprafix was an additive suprafix: the citation tone was pronounced, and then the pitch rose" > this suggests an intermediate stage in which the tone merger wasn't complete. It's very interesting evidence for the historical development of tonal affixation. The A mentions this in §3.1."

Response: I have now added one line to highlight that this shows an intermediate stage in the development in Cantonese. A signpost is already there towards §3.2. (The mentioning is not at §3.1.)

Comment: "L. 426: "still has" > still have"

Response: Thank you. I have now made this change.

Comment: "L. 498, (n): "-文 '-script'" > '-script/language'?"

Response: Thank you for the comment. I have now incorporated this.

Comment: "L. 559-560: "for instance ou3mun4-2 jɐn4 澳門人 is more common than ou3mun4jɐn4 澳門人 in Macau these days" – I suggest that the A avoid the use of statements which are not backed by evidence. It's better to phrase this as "ou3mun4-2 jɐn4 澳門人 is quite common in Macao these days". 

Response: I understand, and share the review's concern. Nonetheless, the suggestion of saying "ou3mun4-2 jɐn4 澳門人 is quite common in Macao these days" is also not backed by evidence. I would also like to be able to quote a study on this. Despite its deficiencies, personal observations have a place in scientific publications. What I have now done is adding hedges to that statement in the manuscript, saying that that is my impressionistic observation as a native speaker.

Comment: "L. 571: "Intrim" > Interim

Response: Thank you. I have now corrected this. 

Comment: "L. 608: "made up of a host syllables" > syllable

Response: Thank you very much. I have now corrected this.

Comment: "L. 664: where does this Middle Chinese reading come from? Besides, this looks like a reconstruction, as it is starred: this is not usually done for Middle Chinese."

Response: Thank you for this comment. I have given this some thoughts, and I have now put in an endnote talking about this. (It is OK putting in an endnote this late in the manuscript.) For forms which I explicitly say "Middle Chinese", I have now deleted the star. For others, I have kept the star. For instance, there is *-tsVj, which is my reconstruction. For *-ɲiH, this is a hypothetical earlier stage in the local speech varieties. Even for *ɲiA, this is a reconstruction, and most probably not the same as Middle Chinese 兒 ɲeA. In addition, without a star, it would be difficult to tell whether I am talking about the -ɲi form in a particular modern Yue or Pinghua dialect, or whether I am talking about this type of suffixes in general.

Comment: "L. 674: "is that all their "Lower" tones (i.e. the onset was voiced in Middle Chinese)" – see comment above. Here the A is referring to the yin/yang distinction in tone registers, and mentions the reason behind the distinction. To a reader who is not familiar with Middle Chinese phonology and traditional Chinese terminology, this can be confusing."

Response: This information on the Middle Chinese categories is important to the arguments. I have added some extra explanations for readers who are not familiar with these terminologies. 

Comment: "L. 689-693: I strongly suggest to delete this. Sound symbolism and iconicity in this connection are very controversial topics, which are best avoided (and not really relevant to the discussion here)."

Response: Thank you for the input, which I partially accept. I need something to link my proposal of low tone diminutives being a preceding step to the high tone diminutives. People would ask why would high tone diminutives develop from my proposed step of low tone diminutives. It is true that I cannot give an answer to this, and all I can do is to bring up suggestions by others, especially that of Zhū Xiǎonóng (2004). I have changed the relevant section in my manuscript to create distance between myself and these suggestions.  

Comment: "L. 772: "There are" > "There are some scholars"

Response: Thank you. I have changed this as recommended. 

Comment: "L. 984-991: "I suggest to drop the brackets () and just leave the text between them. It's very awkward to conclude an article with a long digression between brackets. Acutally, it would be even better if this could be turned into an endnote.

Response: Thank you for this input. I have turned this into an endnote.

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

I appreciate the reply, revision, and clarification. Although I am not fully satisfied with the originality of this paper, I appreciate the expansion in Section 1, which clarifies which parts contain original content while acknowledging that certain sections are not intended to present new findings. This adjustment helps set clearer expectations for readers. I would suggest applying a similar approach to the Conclusions section to make the contributions of this article more explicit. (That said, I would not “force” the author(s) to do so, and so my overall recommendation is “Accept in present form.”)

 

The strength of this article lies in its in-depth and rigorous description on the topic. It translates a subject previously studied mainly within traditional Chinese linguistics into a contemporary linguistic framework, and makes excellent research originally published in Chinese accessible to an Anglophone audience. Although I still have some reservations about its originality, I believe the paper fulfills the promise of its title, “Some Observations on the Cantonese Lexical Suprafixes.”