Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This is a good article. You have clearly displayed your argument that stylistic conventions serve as activating markers that allow abstract social categories to function as correlation devices throughout the text as an addition to David Lewis' theory of conventions. The only problem for me is in the conclusion part. I see that you once again mention the structures you insist that must be brought to life through stylistic conventions in the conclusion; however,  you do not identify them as "abstract categories" (clan- or tribe-membership, social or cultural roles) as you did in various sections (Abstract, 4th and 5th) within the stylistic conventions. Same problem with "activating markers" (in section 4), which you "name" as a good point and contribution but is lacking in the conclusion. This is a very basic addition. 

Author Response

Dear reviewer 1, I am glad you like the paper. I have made the requested changes in the conclusion.

Reviewer 2 Report

Comments and Suggestions for Authors

General comments

This manuscript argues that stylistic conventions (such as dress codes and etiquette) function as "activating markers" that make abstract social categories cognitively tractable, thereby facilitating social coordination. It draws from multiple disciplines to develop this argument, integrating insights from anthropology, sociology, psychology, and philosophy. A genuine gap in the literature is identified—specifically, how stylistic conventions fit within theoretical frameworks based on the coordinative function of convention—and sketches out a plausible solution through the concept of "correlation devices." However, the writing is difficult to follow in places, and the argument suffers from two conceptual weaknesses that I’d like to see addressed before publication.

First, the core category of "stylistic convention" requires more rigorous development and delineation at the outset. The verbal definition offered is quite broad (”culture-specific, collective, contingent customs, habits, rules, and regularities pertaining to the design of our physical and social environment”), making it difficult to draw the line between what should count as a stylistic convention as opposed to some other kind of conventions (and thus difficult to understand the nature of the problem it poses for existing accounts). Section 1 would benefit from a rewrite to set up this distinction and the puzzle that needs to be solved rather than jumping straight into intuitions about coordinating functions. What are the necessary and sufficient conditions for identifying a stylistic convention? If they are defined in part by their function, does this lead to a circularity in the proposed analysis of their function?

Second, the paper doesn't sufficiently differentiate the author's “activating markers” account from existing models of stylistic conventions arising as "group” or “identity” markers (e.g. by Paul Smaldino, Heather Burnett, Richard McElreath, etc). Especially since these existing accounts have been more carefully developed as precise formal models (e.g. stylistic variation across groups as a Lewisian signaling game), it would be helpful to articulate the distinction between the proposed coordinative function and the function of stylistic group markers in these earlier accounts. Is the "activating markers" theory meant to complement or replace the "group markers" theory? Are there specific phenomena where existing "group marker" theories and the proposed "activating markers" theory would predict different outcomes or offer different explanations? I understand this may not be the venue for it, but it would be very helpful to at least sketch out a formal model and how it differs from others.

Despite these limitations, the paper offers a promising theoretical framework that could significantly enhance our understanding of social coordination. With revisions to strengthen conceptual clarity and argumentative force, this work could make a valuable contribution to the field.

Specific Comments

• The paper would benefit from stronger empirical support for claims about cognitive intractability. People make inferences about latent, unobservable structure all the time, and it’s not clear how tractability arguments extends beyond simple group identification in ways that are theoretically significant; that is, it’s not clear how intractable inferences about latent group structure really are, and how, exactly, stylistic markers simplify the inference (e.g. Gershman & Cikara, 2020).
• On line 451, the claim that stylistic conventions have "a much more important coordinative function than is usually assigned to them" needs substantiation - what specific aspects of coordination cannot be explained by the group-marker function?
• The section on "virtual categories" (lines 297-308) introduces an important concept but I didn’t fully understand what makes these categories "virtual" rather than simply abstract or unobservable? Similarity, the metaphor or comparison to "augmented reality" (lines 442-447) is intriguing but underdeveloped.
• The claim that "without stylistic conventions, there can be no social structure" (lines 416-417) requires stronger empirical or theoretical support? it seems like there’s plenty of social structure in, e.g., preferences or beliefs, that don’t require stylistic markers?
• The "correlation devices" concept (Section 3) needs clearer explanation for readers unfamiliar with game theory.
• Some paragraphs are overly long and could be broken up for clarity (e.g., in Section 4).
• Line 26: The reference to "mindreading and/or mindshaping" could be clarified.
• Lines 262-263: The transition from Section 3 to Section 4 feels a bit abrupt and would benefit from some connective tissue.
• The Hutchins quotation is lengthy and could be trimmed to highlight the most relevant aspects.
• The footnotes (particularly footnote 2) contain pretty substantive arguments that might be incorporated into the main text.

References

Burnett, H. (2019). Signalling games, sociolinguistic variation and the construction of style. Linguistics and Philosophy, 42, 419-450.

Efferson, C., Lalive, R., & Fehr, E. (2008). The coevolution of cultural groups and ingroup favoritism. Science, 321(5897), 1844-1849.

Gershman, S. J., & Cikara, M. (2020). Social-structure learning. Current Directions in Psychological Science, 29(5), 460-466.

McElreath, R., Boyd, R., & Richerson, P. (2003). Shared norms and the evolution of ethnic markers. Current anthropology, 44(1), 122-130.

Ozaita, J., Baronchelli, A., & Sánchez, A. (2022). Ethnic markers and the emergence of group-specific norms: an experiment. Scientific Reports, 12(1), 5068.

Smaldino, P. E. (2014). The cultural evolution of emergent group-level traits. Behavioral and Brain Sciences, 37(3), 243-254.

Smaldino, P. E. (2019). Social identity and cooperation in cultural evolution. Behavioural processes, 161, 108-116.

< !-- notionvc: d26841a9-baba-42ed-b683-0d9033a80500 -->

Comments on the Quality of English Language

N/A

Author Response

Dear reviewer 2,

I would like to thank you very much for your comments, critical questions and literature suggestions. I think I have addressed all your worries and the paper has benefitted a lot from it.

As to the two main points of critique, I have added two sections to address them. In Section 1, I introduce the notion of ‘stylistic conventions’ more precisely. It is still a relatively brief section (because the paper is already very long) but I believe that the notion is defined precisely enough. In Section 8 I compare what I now call the ‘activating marker hypothesis’ with the literature on ethnic marking. This is a longer section, but it is still rather condensed (there is much more to write about this, and I probably will do that —so: thanks very much for the question!— but in this already packed paper, I believe the section is of appropriate length).

 

As to the other comments:

• I have added some text and changed other bits to address the issue of the empirical support for the claim that stylistic markers make the use of social categories more cognitively tractable. Most (but not all) of this is in Section 4. The easiest way to deal with this issue would have been to cite experimental evidence, but as far as I can see there is no relevant research on this. Instead, what I have done is (i) to make the claim more precise and realistic by comparing it with the Gershman and Cikara paper. I now argue that although social categories can also be inferred, markers allow us to skip such inferences. This makes it possible to wield various social categories more or less simultaneously as is required in many social interactions in complex cultural organizations (think e.g. of Bourdieu’s habitus). (ii) I also added a paragraph on the cognitive costs of role-coordination, citing two studies that imply that complexity of division of labour is a function of the available resources for role-coordination. I argue that stylistic marking is a resource that has been overlooked in the literature.
• The issue of line 451 is taken care of in Section 8.
• I have changed ‘virtual’ into ‘abstract’ in what is now Section 4, and I have added a paragraph and a half to say a bit more about the metaphor of augmented reality.
• The claim that there can be no social structure without stylistic conventions was, indeed, too unprecise. What I mean—and now say—is that complex role-divisions are substantially aided by stylistic marking. This is connected with the new bit in Section 4.
• I have replaced the traffic light example as an introduction to the notion of correlation devices with an example taken from Guala’s book on institutions. I think that example works better and is accessible to readers who do not know or like game theory.
• I have tried to break up longer paragraphs and sentences.
• I have added a brief footnote on the mindreading/mindshaping difference.
• I have re-written the transition from section 3 to section 4 (now section 4 to section 5).
• I have trimmed the Hutchins quote.
• The former footnote 2 did not survive because I deleted some parapraphs to make room for all the new text. So there is no need to put it in the main text.

 

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for the improvements to the paper. I look forward to future work.