PONE-D-23-25119Sound Symbolism in Japanese Names: Machine Learning Approaches to Gender ClassificationPLOS ONE

Dear Dr. Ngai,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

See my comments at the end of this message. Please submit your revised manuscript by Oct 28 2023 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

• A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
• A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
• An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Søren Wichmann, PhD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at 

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and 

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please note that PLOS ONE has specific guidelines on code sharing for submissions in which author-generated code underpins the findings in the manuscript. In these cases, all author-generated code must be made available without restrictions upon publication of the work. Please review our guidelines at https://journals.plos.org/plosone/s/materials-and-software-sharing#loc-sharing-code and ensure that your code is shared in a way that follows best practice and facilitates reproducibility and reuse.

3. We note that the grant information you provided in the ‘Funding Information’ and ‘Financial Disclosure’ sections do not match. 

When you resubmit, please ensure that you provide the correct grant numbers for the awards you received for your study in the ‘Funding Information’ section.

4. We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide.

5. Please include a new copy of Table 2,4,5 in your manuscript; the current table is difficult to read. Please follow the link for more information: https://blogs.plos.org/plos/2019/06/looking-good-tips-for-creating-your-plos-figures-graphics/

6. Please include a copy of Table 7 and 8 which you refer to in your text on page 15.

7. We note you have included a table to which you do not refer in the text of your manuscript. Please ensure that you refer to Table 9 in your text; if accepted, production will need this reference to link the reader to the Table.

Additional Editor Comments:

I have not sent this out for reviewing yet because you need to provide a cleaner manuscript before bothering reviewers. On the first few pages alone I came across numerous typos and stylistic problems. I got tired of noting these when arriving at p. 8. What I noted up to then is indicated below. I also noted numerous problems in the list of references. See also below. But these are just things I happened to quickly note. It is not the case that you can just take care of these things, you need to be more thorough than that. So you need to carefully revise the manuscript taking care of these presentational issues first. Possibly you need to involve a professional copy-editor. When I get a better presented manuscript I will send it out for review.

Three typos in the abstract:

are shown to reasonably -> are shown to be reasonably

was associated -> were associated

and which -> ???

p. 4, clumsy formulation: Combined, evidence suggests a strong tendency to map certain phonemes with the imagery of size, other than speakers of the Bahnar language (13)

p. 4 could potential deceive -> could potentially deceive

p. 4 correlates with body size -> correlate with body size

p. 6 it remains an open question as to how -> it remains an open question how

p. 6 For example, the name ‘Catherine’: the way this is transcribed (which is wrong) it has four syllables; also, it is stressed on the first syllable, while the text says "non-initial stressed syllable"

p. 7 Kanji characters are logographic characters adopted from early Chinese religious texts: a bit weird statement; they are not somehow extracted from specific texts, but adopted from early Chinese writing

p. 8 Kanji characters is -> Kanji characters are

Ref. 2: what is K.A.?

Ref. 5 is garbled

Ref 11 incomplete

Ref 13 incomplete

Ref 15 incomplete

Ref 23 inconsistent use of capitalization

Ref 38 capitalization

Ref 56 capitalization

Ref 58 Forest -> forests

Ref 60 incomplete?

Ref 63 incomplete

Ref 73 capitalization

Ref 75 extra space

[Note: HTML markup is below. Please do not edit.]

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PONE-D-23-25119R1Sound Symbolism in Japanese Names: Machine Learning Approaches to Gender ClassificationPLOS ONE

Dear Dr. Ngai,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process. The reviewers provide many constructive comments, which are largely complementary, since one reviewer looks more at methods and the other is more focussed on the context of the general study of sound symbolism. I strongly advice paying close attention to the comments of both reviewers. Additionally, below my signature I offer some observations on stylistic issue and typos. Please take those into account as well.

Please submit your revised manuscript by Jan 04 2024 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

• A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
• A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
• An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Søren Wichmann, PhD

Academic Editor

PLOS ONE

Additional Editor Comments:

References are to line numbers

54-56:

feature importance is examined to determine whether associations previously reported in

English are found in Japanese which suggest that the systematic sound symbolic expressions of

gender are universal.

->

feature importance is examined to determine whether associations previously reported in

English are found in Japanese since if they are, this would suggest that the systematic sound symbolic expressions of

gender are universal.

167: the loanword London /ro.n.do.n/ would be consisted of four mora ->  the loanword London /ro.n.do.n/ would consist of four mora

278: the Altmann methods [65] method -> the Altmann method [65]

325: poison regression -> Poisson regression

332: Poison regression -> Poisson regression

347: typologically unique from -> typologically distinct from

360: femineity -> femininity

369-370: in contrary to -> contrary to

367-378: "Considering that these elements adhere to the mora structure of Japanese, and seldomly encountered in other languages, it is fitting to classify these elements as morphemes." Seems to me a logical non sequitur.

404-406

"The superior performance of the XGBoost algorithm suggests that the boosting technique, which optimizes an

objective function through sequential tree-building, allowing for a more effective capture of

complex non-linear relationships within the data." Maybe "allows for" rather than "allowing for" is meant? Something is wrong with this sentence.

414-415: sound symbolism sound-gender associations extend to Japanese names, a language distinct from Indo-European languages. -> sound symbolic sound-gender associations extend to Japanese names, a language distinct from Indo-European languages.

462-464:

Blasi DE, Hammarström H, Stadler PF, Christiansen MH. Sound-meaning association

biases evidenced across thousands of languages. Proc Natl Acad Sci.

2016;113(39):10818–23.

->

Blasi DE, Wichmann S, Stadler PF, Hammarström H, Christiansen MH. Sound-meaning association

biases evidenced across thousands of languages. Proc Natl Acad Sci.

2016;113(39):10818–23

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

Reviewer #2: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The motivation of the study needs to be explained better. In some research communities (e.g., NLP, machine learning), there was a huge discussion on the ethics of doing machine learning based gender classification in the current age (e.g., see this online discussion: https://www.reddit.com/r/MachineLearning/comments/qmm6uh/d_ethical_concerns_for_ml_to_predict_race_gender/) I am surprised the paper does not even mention such issues. Further, I found the paper to describe a very baseline approach, taking an existing dataset (whose creation process is unclear), and a vague algorithm for feature extraction (my impression is that this is the novel contribution for this paper because data and models are coming from somewhere else). Hence, I felt it needs major revisions.

Three main issues I found problematic are:

1. The dataset: it is clearly taken as is from the website. ""Gender is listed as a distribution between male and female." - in what data is that distribution calculated and how? You write that this data was inspected by a Japanese linguist. But, what was the inspection? What did you find out? Is the data 100% correct or are there errors? Is it totally obvious to distinguish male and female names in Japanese? In many languages, there are some gender neutral names and it is hard to guess the gender from name alone. What was your strategy for such cases?

2. The features: You write: "an algorithm was constructed to convert Japanese names into phone counts, the output of the algorithm was checked by the same native Japanese linguist" but no details are given. This seems to be main original contribution of this paper, considering that the dataset is external, and machine learning approaches are also standard ones. But there is no further detail on this algorithm. What did the manual checking of the output reveal? Were there any disagreements between the algorithm and human evaluation? How much? What is the accuracy of this phone counting program? They should be discussed.

3. Modeling: Why are only Random Forest and XGBoost chosen? Why not other simpler approaches like nearest neighbors or logisitic regression etc? Considering the size of your dataset, perhaps nearest neighbors would have worked too. In a paper like this, it would be good to see a comparison with more algorithms. In terms of feature selection too, there are approaches to estimate more predictive features irrespective of the algorithm used (e.g., looking for correlation of a feature with the predicted class). I think those should also be explored to understand the data better. The fold split also seems rather arbitrary. Why can't you just do a 5-fold or 10-fold stratified CV like most researchers report in their papers? Other than these, some error analysis showing where the algorithm failed and how to improve over this baseline approach would be good too.

Potential limitations of the paper needs to be discussed too, in my opinion.

Reviewer #2: I find this manuscript both interesting and valuable for iconicity research as it demonstrates how relatively new methodologies can be applied to address the challenge of handling increasingly available large datasets. The argumentation is easy to follow, and the language is clear. The statistical analysis also appears sound, although I am not an expert in machine learning, and all data underlying the findings is fully available. My recommendation is that it should be accepted with some minor but crucial revisions.

Main points:

Throughout

Although drawing universal conclusions from the occurrence of iconicity in names used in a single language can be challenging, I would like to stress that thorough language-specific iconicity studies serve the same purpose as descriptions of poorly documented languages. The compiled data and analyses deepen our knowledge of the diversity of iconicity in the linguistic system. This, in turn, helps us gain a clearer picture of the cross-linguistic situation in the field and guides future studies. With this being said, it is also important to clearly point out that the comparison between Japanese and English/Indo-European languages arises from the relatively sparse material on name iconicity research, which is mostly confined to Indo-European languages. Without such clarification, it might sound like English/Indo-European languages are a default when it comes to expected iconic patterns and naming conventions. For example:

“Our current results demonstrate that sound symbolism sound-gender association is also prevalent in Japanese names, a language that is typologically unique from Indo-European languages.”

“In conclusion, our study provides compelling evidence that sound symbolism sound gender associations extend to Japanese names, a language distinct from Indo-European languages.”

Throughout

The predictions made based on the frequency code are fitting and the authors have included an extensive list of relevant literature. The authors also draw some conclusions based on cross-linguistic patterns, such as in: “The observed association of the consonant /m/ with femininity could be attributed not only to the frequency code [15,16], but potentially to the idea of the concept of breast [12] or cuteness [67,68].” I wonder why these are not used as predictions along with the frequency code.

Iconicity in male and female names might appear binary, but the grounds for their associated sounds do not necessarily have to be of the same origin (cf. /m/ > femininity possibly via iconicity in words for ‘breast,’ etc.). Erben Johansson & Cronhamn (2022) tested the presence of iconicity in nominal classification systems, i.e., gender and classifier systems. Based on 344 languages from 212 language families, they found that morphological markings for masculine gender involved front, central, and back vowels, while markings for femininity involved nasal and stop consonants. While not directly name data, the material touches on the same type of male/female distinctions as the manuscript, and the findings align.

Furthermore, markings for human ~grammatical gender (in human/non-human classification systems) were largely associated with the same sound features as feminine. This raises the question of markedness in female and male names as well.

It would also be possible to connect the findings to the language-specific lexicon in general, and possibly to the bouba-kiki effect. References:

Sidhu, D. M., & Pexman, P. M. (2018). Lonely sensational icons: Semantic neighbourhood density, sensory experience and iconicity. Language, Cognition and Neuroscience, 33(1), 25–31. https://doi.org/10.1080/23273798.2017.1358379

Erben Johansson, N. E., & Cronhamn, S. (2022). Vocal iconicity in nominal classification. Language and Cognition, 15(2), 266-291. https://doi.org/10.1017/langcog.2022.36

Sidhu, D. M., Westbury, C., Hollis, G., & Pexman, P. M. (2021). Sound symbolism shapes the English language: The maluma/takete effect in English nouns. Psychonomic Bulletin & Review, 28(4), 1390–1398. https://doi.org/10.3758/s13423-021-01883-3

Page 8

Since this manuscript focuses on iconicity in Japanese, it is important to define what mimetics/ideophones are and their role in the Japanese language, especially in contrast to their status in other languages. This is crucial, considering the frequent comparisons with English and Indo-European languages throughout the text. Additionally, on page 17, “phonestheme” and “phonesthemic” appear without definition and should be clarified.

Page 10

The data used seems sound and the procedure for evaluation the data is well-described, but the authors do not explain why JTALK was chosen in the first place. For convenience, because JTALK is the only available database of Japanese names, or something else?

Page 19-20

The discussion brings up many relevant factors, including morphology and culture. However, while interesting, I find the connection between naming conventions and improvements in women’s rights too speculative without any data to back it up. To draw any conclusions, a comparison to older Japanese names would be needed. Additionally, a shorter paragraph at the beginning of the manuscript summarizing naming conventions from across the world would help contextualize the analyzed material, especially if the authors wish to maintain the point about potential phonological changes in Japanese names over the last decades. For instance, politically motivated names, as seen in Mandarin Chinese, could provide a useful comparison.

Minor points:

Page 3

I think it would benefit the reader if there were a clear linguistic example of sound symbolism in the first paragraph. Not necessarily an example like “bouba-kiki”, but just an association that is cross-linguistically common and can also be found in English or some other global language for familiarity.

Page 3

Since “iconicity” is used in the manuscript, it should be stated whether the authors consider sound symbolism as the same as (vocal) iconicity to avoid confusion. A distinction does not have to be drawn, but in that case, state that you use the terms interchangeably. I also want to mention that I am glad the authors highlighted that the “symbolism” in “sound symbolism” can be misleading.

Page 4

“thatthe” > “that the”

Page 5 and other places

While iconicity research in Pokémon names across languages has increased significantly in recent years, I think these findings should be contextualized. They are created in a manner that is arguably more deliberate than names and words. Therefore, their informative value about iconicity in language ought to be lesser and/or qualitatively different.

Page 8

“Typologically, Japanese differs from many Indo-European languages in many aspects. Japanese is a member of the Japonic language family. Although not limited to Japanese, one of the distinct features of Japanese are…”

The beginning of this paragraph sounds stunted. If the sentence “Japanese is a member of the Japonic language family” were expanded upon, it would likely flow better into the next sentence. For example, “Japanese is a member of the Japonic language family, together with Ryukyuan and Hachijō”, or something similar.

Page 8-9

“alphabets” > “syllabaries”

Hiragana and Katakana are not alphabets.

Page 9-10

Several technical terms are introduced here without description. While some are explained in the method section, others, such as “weak/strong learners”, are not. It would be helpful for the reader if the authors could add a sentence indicating that these terms will be described in detail later in the manuscript.

Page 12

“A partial Latin square revealed 28 possible combinations of subsets, and each combination was 237 used resulting in 28 iterations for each algorithm.”

This sentence should be explained in more detail for less statistics-savvy readers. Why is a partial Latin square used?

Discussion/conclusion

The post hoc analysis highlights the role of sound combinations, which I believe is crucial for understanding iconicity in spoken language data. I am not suggesting that the authors redo the entire analysis to include both phonemes and all possible phoneme combinations, but adding a sentence or two about the implications for future studies would be beneficial.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Niklas Erben Johansson

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Sound Symbolism in Japanese Names: Machine Learning Approaches to Gender Classification

PONE-D-23-25119R2

Dear Dr. Ngai,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Søren Wichmann, PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments: