Assessing African grey parrots' prosocial tendencies in a token choice paradigm

Prosociality is defined as a voluntary, typically low-cost behaviour that benefits another individual. Social tolerance has been proposed as a potential driver for its evolution, both on the proximate and on the ultimate level. Parrots are an interesting species to study such other-regarding behaviours, given that they are highly social and stand out in terms of relative brain size and cognitive capacity. We tested eight African grey parrots in a dyadic prosocial choice test. They faced a choice between two different tokens, a prosocial (actor and partner rewarded) and a selfish (only actor rewarded) one. We found that the birds did not behave prosocially when one subject remained in the actor role; however, when roles were alternated, the birds’ prosocial choices increased. The birds also seemed to reciprocate their partner's choices, given that a contingency between choices was observed. If the food provisioned to the partner was of higher quality than that the actor obtained, actors increased their willingness to provide food to their partner. Nonetheless, the control conditions suggest that the parrots did not fully understand the task's contingencies. In sum, African grey parrots show the potential for prosociality and reciprocity; however, considering their lack of understanding of the contingencies of the particular tasks used in this study, the underlying motivation for the observed behaviour remains to be addressed by future studies, in order to elucidate the phylogenetic distribution of prosociality further.


Introduction
Line 58: This seems like a bit of an exaggeration, given that there are ample studies that show cooperative behavior among non-kin in primates in the absence of reciprocation. For example, in studies [20,25] there was no opportunity for reciprocity within the experiment and the primates still were cooperative. I suspect the authors are suggesting that nonkin cooperate only when they are familiar (which has been difficult to study since it is challenging to find species and settings where strangers can be tested together). But this line seems like a bit of an overstatement as it is written, especially since in [20] both kin and non-kin pairs behaved similarly.
Line 108: While the parrots succeeded at the string-pulling task in [49], they did not show awareness that a partner was needed in the delay condition. By most accounts, this would be considered a failure at cooperation, since they didn't appear to notice the role of the partner.

Methods
General comment: I am sure this was unintentional, but at times the language follows Suchak & de Waal [ref 21] extremely closely, even verbatim. I recognize that when doing a replication it may be hard to think of alternative ways to phrase things, but the authors need to extensively reword large portions of the methods section to avoid violating copyright. For this reason I am suggesting a major revision. L. 546: Remove comma…sentence unclear…. L. 605: Maybe the problem is not about the birds' cognitive abilities-the experimenters did not examine social dominance issues or the extent of the 'bonding' within the pairs. Dyad 2 here seemed particularly uncooperative; I wonder if they were the least strongly bonded pair or the one with the most difference in dominance…I simply think that the tasks were confusing, and that in some instances other issues such as dominance and the order in which a bird in a pair began a task may have affected the outcomes.
We also do not know how many times the birds simply refused to work…Such behaviour is common in parrots and is not a reflection on the study, but knowing the numbers of failed trials may provide a handle on how confusing the tasks may have been.
I'm very conflicted about this paper-the experimenters tried to design appropriate controls but mainly seemed to have confused their subjects. We know nothing about the number of failed trials and nothing about the actual relationships between the birds in the dyads. The authors are honest in their overall evaluation that additional work is necessary to determine prosociality in parrots (at least Grey parrots), but leave us without much more information than we had from the earlier studies (i.e., that parrots like to play tit-for-tat to some extent, although the previous studies demonstrated that the parrots didn't simply copy the behavior of their partner-at least when the partner was human).

03-Jul-2019
Dear Dr Krasheninnikova, The editors assigned to your paper ("Assessing African grey parrots' prosocial tendencies in a token choice paradigm") have now received comments from reviewers. We would like you to revise your paper in accordance with the referee and Associate Editor suggestions which can be found below (not including confidential reports to the Editor). Please note this decision does not guarantee eventual acceptance.
Please submit a copy of your revised paper before 26-Jul-2019. Please note that the revision deadline will expire at 00.00am on this date. If we do not hear from you within this time then it will be assumed that the paper has been withdrawn. In exceptional circumstances, extensions may be possible if agreed with the Editorial Office in advance. We do not allow multiple rounds of revision so we urge you to make every effort to fully address all of the comments at this stage. If deemed necessary by the Editors, your manuscript will be sent back to one or more of the original reviewers for assessment. If the original reviewers are not available, we may invite new reviewers.
To revise your manuscript, log into http://mc.manuscriptcentral.com/rsos and enter your Author Centre, where you will find your manuscript title listed under "Manuscripts with Decisions." Under "Actions," click on "Create a Revision." Your manuscript number has been appended to denote a revision. Revise your manuscript and upload a new version through your Author Centre.
When submitting your revised manuscript, you must respond to the comments made by the referees and upload a file "Response to Referees" in "Section 6 -File Upload". Please use this to document how you have responded to the comments, and the adjustments you have made. In order to expedite the processing of the revised manuscript, please be as specific as possible in your response.
In addition to addressing all of the reviewers' and editor's comments please also ensure that your revised manuscript contains the following sections as appropriate before the reference list: • Ethics statement (if applicable) If your study uses humans or animals please include details of the ethical approval received, including the name of the committee that granted approval. For human studies please also detail whether informed consent was obtained. For field studies on animals please include details of all permissions, licences and/or approvals granted to carry out the fieldwork.
• Data accessibility It is a condition of publication that all supporting data are made available either as supplementary information or preferably in a suitable permanent repository. The data accessibility section should state where the article's supporting data can be accessed. This section should also include details, where possible of where to access other relevant research materials such as statistical tools, protocols, software etc can be accessed. If the data have been deposited in an external repository this section should list the database, accession number and link to the DOI for all data from the article that have been made publicly available. Data sets that have been deposited in an external repository and have a DOI should also be appropriately cited in the manuscript and included in the reference list.
If you wish to submit your supporting data or code to Dryad (http://datadryad.org/), or modify your current submission to dryad, please use the following link: http://datadryad.org/submit? journalID=RSOS&manu=RSOS-190696 • Competing interests Please declare any financial or non-financial competing interests, or state that you have no competing interests.
• Authors' contributions All submissions, other than those with a single author, must include an Authors' Contributions section which individually lists the specific contribution of each author. The list of Authors should meet all of the following criteria; 1) substantial contributions to conception and design, or acquisition of data, or analysis and interpretation of data; 2) drafting the article or revising it critically for important intellectual content; and 3) final approval of the version to be published.
All contributors who do not meet all of these criteria should be included in the acknowledgements.
We suggest the following format: AB carried out the molecular lab work, participated in data analysis, carried out sequence alignments, participated in the design of the study and drafted the manuscript; CD carried out the statistical analyses; EF collected field data; GH conceived of the study, designed the study, coordinated the study and helped draft the manuscript. All authors gave final approval for publication.
• Acknowledgements Please acknowledge anyone who contributed to the study but did not meet the authorship criteria.
• Funding statement Please list the source of funding for each author.
Once again, thank you for submitting your manuscript to Royal Society Open Science and I look forward to receiving your revision. If you have any questions at all, please do not hesitate to get in touch.
Kind regards, Alice Power Editorial Coordinator Royal Society Open Science openscience@royalsociety.org on behalf of Kevin Padian (Subject Editor) openscience@royalsociety.org Associate Editor's comments: Two reviewers have commented on your report. Both raise a number of concerns regarding, for instance, the methodology employed, the literature review, and how closely your text matches previously published work. To comment on the latter point, the journal's editors recognise a degree of overlap in methods may be inevitable, but would nevertheless strongly encourage the authors to seek alternative phrasing and to ensure they fully reference methodologies borrowed from elsewhere to ensure that A) your work is as original a contribution as practical and B) earlier work receives the credit it is due. A more general comment on the paper is that it is clear a substantial effort is required to get the paper to publication. You will need to work hard to persuade both the editors and reviewers that the revised manuscript has improved sufficiently to consider publication: you will need to include a point-by-point response detailing changes or any rebuttals, and also -ideally -include a tracked changes version of the manuscript to make clear how you've effectively responded to the critiques. Good luck, and we'll look forward to receiving the revision -the original reviewers will be invited to look at your responses, so do make sure they are clear and effective.

Subject Editor Comments to Authors:
Please take into account all the comments of the AE and reviewers. If you need more time than allotted for the revision, do let the office know and it will be no problem. Best wishes. major concern with the methods, but all other suggestions or questions are minor, denoted by line numbers below.

Introduction
Line 58: This seems like a bit of an exaggeration, given that there are ample studies that show cooperative behavior among non-kin in primates in the absence of reciprocation. For example, in studies [20,25] there was no opportunity for reciprocity within the experiment and the primates still were cooperative. I suspect the authors are suggesting that nonkin cooperate only when they are familiar (which has been difficult to study since it is challenging to find species and settings where strangers can be tested together). But this line seems like a bit of an overstatement as it is written, especially since in [20] both kin and non-kin pairs behaved similarly.
Line 108: While the parrots succeeded at the string-pulling task in [49], they did not show awareness that a partner was needed in the delay condition. By most accounts, this would be considered a failure at cooperation, since they didn't appear to notice the role of the partner.

Methods
General comment: I am sure this was unintentional, but at times the language follows Suchak & de Waal [ref 21] extremely closely, even verbatim. I recognize that when doing a replication it may be hard to think of alternative ways to phrase things, but the authors need to extensively reword large portions of the methods section to avoid violating copyright. For this reason I am suggesting a major revision.
Line 213: Why were the compartment sides fixed, rather than counterbalanced across days/sessions? Could this possibly have inhibited them from moving from side to side in the accessible condition?
Line 220: I think this should say the partner could witness the actor's behavior Results: Line 421: One possibility for this, which hasn't really been addressed in the discussion is that during the unequal ALT condition, the birds might experience a frustration effect, since in one trial they might get a high value reward as the partner, but in the next get a low value reward as the actor, even though the token chosen in both trials is the same.
Line 443: Are the descriptives the M±SD? If so, can you specify that?
Line 450: Have you tried a logistic regression using previous choice as a predictor, as opposed to the contingency table? Given that there were differences among individuals/pairs (as discussed in lines 613-616), it seems possible that certain individuals might be driving this effect, which would be masked in pooled data.
Line 457: I think the interpretation "indicating that the presence of a partner…" does not match the statistic presented here, since in both UNI and SFC the partner was present. It seems to me that this suggests that they weren't making their choice based on whether or not it was actually helping the partner.
Line 464: It seems odd to me that the INACC condition is the only one in which the birds changed over the 10-trial blocks. I note that half of the birds started with ACC and the other started with INACC. Is it possible this effect (and the overall high number of prosocial choices) is driven by the birds that started with ACC, since in the previous condition choosing the prosocial token was advantageous? Perhaps they needed to learn that they could not access both rewards? L. 337: I'm confused....why wouldn't the actor then treat the experimenter as the partner, responding to whatever the partner did? Wouldn't you get roughly the same results? The task is testing whether the bird responds to another bird versus the experimenter, not another bird versus the reward. I understand that the experimenter actually provides the rewards in both cases, but if bird A doesn't see bird B making a choice and only sees the experimenter's actions, then there is an obvious difference. Or maybe bird A thinks that its choice makes the experimenter act twice in some way? [That is...the experimenter distributes treats once based on bird A's choice and then once again without anything else happening...] I would think that this is an extremely confusing condition. Maybe the task would work if it was automated in some way, so that there was no visible actor, but in this case…it is confusing. L. 362ff: Again, how many times did the birds refuse to enter the empty compartment? Makes a difference in their understanding if they chose prosocially but didn't access the additional treat… And it might make a difference as to which birds started with the inaccessible versus which started with accessible conditions… If the birds started with inaccessible, and saw the experimenter remove the food after 2-3 sec, maybe they assumed it was forbidden and figured that they should not try to access it? L. 386: What did the birds understand about the blockage? Was this an alternating condition or not? If not alternating, maybe only the bird that didn't get the treat understood the condition and would respond appropriately? Maybe bird A keeps being prosocial because it wants bird B to see that it is trying to be generous no matter what the experimenter does? L. 407ff: The birds that exhibited preferences-were they first actors or first recipients? L. 411: I think the authors mean S2… L. 437: Again…did the first actor understand the issues? Only when the first recipient became the actor would the task make sense…and if the first actor wasn't very prosocial, the recipient might not be so after it became the actor when there were 30 trials of one type involved…. L. 457: I'm confused…a partner was present in both UNI and SFC, so why would an equal response suggest that the presence of a partner did not affect their choices? Would have to contrast data between the presence and absence of a partner to make that point… L. 507: Yes, they were having trouble understanding the task contingencies, but not issues of whether or not a partner was present, absent, or able to access the reward…putting oneself in the situation of the bird, a human might have trouble figuring out the various task contingencies.
L. 511ff: But by the time the unequal reward task was administered, the birds had a pretty good idea that turns would be taken, whether immediately or eventually...so they could reason that if they ever wanted the walnut, they'd better give it to their partner…and the effect seems driven mostly by one dyad… L. 530: Theoretically the option of tit-for-tat was eliminated, but one would have to show that the synergy was different to make such a claim. It was still a social system in which the birds could have been trying to influence the experimenter… L. 546: Remove comma…sentence unclear….
L. 605: Maybe the problem is not about the birds' cognitive abilities-the experimenters did not examine social dominance issues or the extent of the 'bonding' within the pairs. Dyad 2 here seemed particularly uncooperative; I wonder if they were the least strongly bonded pair or the one with the most difference in dominance…I simply think that the tasks were confusing, and that in some instances other issues such as dominance and the order in which a bird in a pair began a task may have affected the outcomes.
We also do not know how many times the birds simply refused to work…Such behaviour is common in parrots and is not a reflection on the study, but knowing the numbers of failed trials may provide a handle on how confusing the tasks may have been.
I'm very conflicted about this paper-the experimenters tried to design appropriate controls but mainly seemed to have confused their subjects. We know nothing about the number of failed trials and nothing about the actual relationships between the birds in the dyads. The authors are honest in their overall evaluation that additional work is necessary to determine prosociality in parrots (at least Grey parrots), but leave us without much more information than we had from the earlier studies (i.e., that parrots like to play tit-for-tat to some extent, although the previous studies demonstrated that the parrots didn't simply copy the behavior of their partner-at least when the partner was human).

RSOS-190696.R1 (Revision)
Review form: Reviewer 1 Is the manuscript scientifically sound in its present form? Yes

Are the interpretations and conclusions justified by the results? Yes
Is the language acceptable? Yes

Recommendation? Accept as is
Comments to the Author(s) I sincerely thank the authors for their thoughtful responses to my earlier concerns. At this point I have reviewed the revised manuscripts and have no further questions or concerns to raise.

Review form: Reviewer 2
Is the manuscript scientifically sound in its present form? No

Do you have any ethical concerns with this paper? No
Have you any concerns about statistical analyses in this paper? No

Recommendation?
Major revision is needed (please make suggestions in comments)

Comments to the Author(s)
Krasheninnikova and colleagues have revised their extremely interesting paper on prosocial behaviour in Grey parrots. They examine multiple possible situations and test the birds extensively. I nevertheless still have several comments -some involving minor clarification, others a bit more substantive-which I present in order of appearance. L. 29: I'd be more comfortable with something like "..their lack of understanding of the contingencies of the particular tasks used in this study, the underlying…" [that is the most that the authors can claim] L. 39: This paper is on pro-sociality, which the authors just described as being different from altruism. So where does this sentence with respect to altruism fit? Do they mean pro-sociality? Or at least reciprocal altruism?
L. 58-59: The edit is odd…leave out comma after "familiar"…and maybe "and in situations without a possibility…" L. 118: Remove comma after "While"…Based on the material in the paper, maybe the description should read "did not show clear reciprocity…" Again, read my comments based on the material in the paper cited: When the dominant bird led, he STARTED OUT prosocial but switched to selfish when his partner consistently acted in a selfish manner"…what is written in this manuscript is not correct.
This description of the material is not correct. Citing from the paper itself: "However, whether they shared with their partner at no cost to themselves seemed, at least for our subjects, to depend upon their dominance status and upon who led and who followed. Griffin has always been dominant to Arthur, and Figs. 3 and S1 show that when Griffin led, he became selfish, but when placed in the position of follower, he acted in a more sharing manner." That is, when Griffin led he started out sharing but when Arthur became more selfish, Griffin became more selfish. When Griffin followed, he retained his more sharing manner. Arthur, the subordinate, "..started with a slight giving tendency as both leader and follower, then became statistically significantly more selfish both as leader and follower." L. 120ff: THEREFORE, these lines should read: "If the dominant bird started the session choosing first, he initially chose the prosocial token, but became more selfish after he experienced selfish behavior from his partner. When the dominant bird was the follower, he retained his prosociality. The subordinate bird, however, became more selfish over time in both leader and follower roles. Therefore, both birds developed…" L. 125: This sentence is not correct-and the sentence has not been altered….I repeat: the authors tested for side biases and the current authors cannot disregard these tests. I will change my decision to "reject" rather than revise if this revision is not made in a subsequent version.
L. 126: This sentence is also not correct. Control was inherent in the task: Note that the birds did not copy the human, as the birds were not giving treats to the generous human but sharing treats, and they quickly learned to avoid the null cup (no treats to anyone), demonstrating that they clearly understood the task. Note, too, the follow-up study that further tested the task's assumptions. The authors of the current paper cannot blithely disregard data and criticize another paper without full justification. Avoiding the null cup if it gives no reward can be described any way you choose, but it still involves understanding the contingencies of the task. And explaining the choice of the option that provides two rewards rather than one as reinforcement learning is meaningless unless the chooser GETS more rewards. If the chooser gets to share, that means the chooser understands the contingencies of the choice. As for the control without a partner-the subject is simply likely going to be confused about what is going on-see more on that later.
The argument about color bias is justified, however.
L. 134: I think the authors mean "employed" rather than "applied" L. 201: "without shells" L. 199: This sentence is not clearly written. Try dividing the sentence into two. "…in all equal reward conditions. When the actor chose prosocially in the unequal condition, in contrast, the partner received a piece of walnut (preference for walnut was assessed prior to the experiment) while the actor received merely the same sunflower seed as in the equal condition." L. 217: I think it is important to note here that these birds did NOT express inequity aversion previously. Such finding suggest that the authors already know that the birds are not that likely to be affected by their partner getting a better reward, and that this is a further test of that finding.
L. 216: In the response to reviewers, the authors mention that they did not find any position preference in token choice…this information should be provided somewhere in the results section.
L. 226: Is this a new paragraph? L. 244: Sentence a bit confusing. Divide sentences. "The rewards inside the cups were not at that time visible to the birds. Once a choice was made, the experimenter lifted the lid(s) off the small food container(s), tilted the cup(s) forward so that both birds could see the reward(s) inside, and then distributed the reward(s)." L. 259ff: If birds don't care about inequality, why would they care if one got fed and one didn't? Isn't that finding a problem for this entire paper? The receiver could eventually associate the ring color with whether it gets food, but not necessarily with the actions of its partner if it doesn't care that its partner gets fed and it doesn't, especially if it was going to get at least half the standard reward just for hanging out near the reward site when the experimenter wanted it to move back. If a whole seed is a standard reward, half that reward is hardly "tiny" [and, yes, I do understand that sunflower seeds cannot be divided into much smaller pieces and still easily be handled, but maybe the authors should have used some other, less favored food?] l. 270: 12 sessions needed to be repeated? I'm confused…there are only 9 sessions listed in Table  1… If this statement is not a typo, then there are a LOT of repetitions. Could the authors please let the reader know how many sessions were repeated for each condition? That would help the reader know which conditions were more problematic. The authors gave some of the information in their response but not in the paper…it seems as though one dyad really had issues, and it would be helpful to know which one. Maybe that set of data needs to be dealt with separately? L. 273: What happened after the birds chose a token in the color preference test? Did it just get taken away? Did they get something or nothing for making the choice? If they had any experience in trading tokens for treats, and they chose a token and got nothing, they would-if they had any sense-alternate choice of tokens to see if choosing a different token would give a reward, even if they liked their first choice. It seems that most birds started with a color preference and those colors were eliminated, but it also seems possible that the birds just LEARNED at this point that they should choose randomly because nothing would happen almost whatever they did BUT if they chose their favorite it would disappear. I don't disagree that one has to check for color preferences, but it's a rather tricky procedure as they learn to ignore any contingencies connected with the color. L. 286ff: So now a bird was given "x" to choose and got a reward half the time and was given "y" to choose the other half and still got a reward; sometimes their partner got a reward and sometimes the partner did not. Given that they didn't care about reward inequality, why would they pay much attention to what was happening, particularly if they were actor but even if they were the powerless recipient? If the recipient couldn't do anything about what happens, and can't tell the actor what to do, why pay too much attention? L. 298: And I'm still confused…The authors mention equal and unequal rewards earlier, but Table 1 does not, so I'm confused. Does each bird have one session as actor with equal rewards then one with unequal reward before the roles are reversed (i.e., that's what is meant by 2 sessions in Table 1)? Or do reversals occur after equal reward and then again before unequal reward? And do all these trials occur before the alternating sessions? I need to understand what is happening at this point in the paper. Or do the birds cycle through UNI, ALT, YOK in the equal reward before doing unequal rewards? [It seems that way, given that different tokens were used, but I really had to work to figure this out and I'm still not sure.] If I were trying to replicate this experiment, I'd never be able to figure out exactly what was done. The material now taken out helped me before… L. 305: Two sessions EACH? Including equal or unequal rewards? L. 343: Do the authors mean 30 trials per bird acting as actor and 30 as recipient? And where do the equal/unequal fit in here…one session with equal rewards, one session with unequal rewards? Not clear… L. 355ff: I absolutely do not understand the authors' response to my query. Why would switching to a selfish strategy be the expected response? The experimenter is NOT always being selfish, so why would the actor be selfish? The authors are making AN ASSUMPTION that the birds are treating the experimenter as an automaton, and ignoring the possibility that they were playing with the experimenter. In previous studies, didn't the experimenter played partner-like roles? So why ignore that now? Given that there were lots of prosocial choices in the alternating condition, the experimenter was being prosocial a lot, so why wouldn't the actor reciprocate to that? I definitely appreciate the new material added around L. 580. L. 384: Accessible vs inaccessible: So, I'm a bird and you train me to go back and forth between compartments. I learn that I can get food in both places, but not with respect to anything to do with token choice; I just learn that it's possible. And there's no bird next door. If I'm Jelo or Bella or Sensi or Nikki (half the birds), I don't notice whether or not the passage is open but I choose to put food on the other side because I remember that it's possible to get food. If I'm Kizzi or Nina or Jack, whether or not the passage is open I'm not prosocial maybe because there's no bird there. If I'm Kimmi, I'm just confused. Actually, if the birds differentiated between the conditions, they should NOT be random but choose selfish in inaccessible because they'd otherwise be wasting food, and they should choose prosocial in the accessible condition to get more food. But none of the birds (except a tendency in Jack, in the opposite to what would be expected) seemed to understand the different conditions at all. They seemed to hit on some strategy simply with respect to the possibility of more food-hope for the best or accept the worst-and stick to it. I don't see the task as showing much of anything about understanding cooperation. I think the authors' comments in the Discussion about what they did and didn't learn and spatial confusion make more sense than the assumption that the task would be useful, particularly given that the birds' behaviours altered over time.
Overall: Individual differences seem to be extremely important in this study, from looking at the supplemental data. Maybe some Greys are good buddies and others simply are not, or maybe some birds are dim and others not…the authors really need to look at and discuss the individuals, not just the group behaviour. L. 451ff: I'm confused…if birds DID develop preferences, and there were so many individual differences in the results, shouldn't the experimenters have checked specifically to see if the birds' actions were affected by these preferences? Just looking at group behaviour is not sufficient, as the individuals varied so strongly.
L. 474: The authors must mean dashed red line, not dotted…..I'm not sure what the dotted line means.
L. 560: Correct English is "fewer prosocial", not "less prosocial"… We still do not know the relationships between the individuals in the pairs, although the authors do admit to the artificiality of the situation. Note: A 'good' relationship doesn't necessarily mean equal; if the birds have worked out a dominant-subordinate relationship and are happy with it, that makes it "good"…but if there is conflict, that is a different situation, and could easily affect the results.
We also know that parrots engage in courtship feeding, where one bird gives up food to the other with no reciprocity, and in Greys both parents feed the young directly from what they obtain when they forage (unlike mammals, who suckle, which is quite different in terms of sharing resources)…so the whole idea of unequal reward as an issue is quite different from what one would see in primates.
I'm still very conflicted about this paper-given how critical the experimenters are of the previous studies, their failure to examine the material of the previous studies carefully, and their claims that they performed their experiments in order to correct all the problems they perceived in the previous studies, they really have to be exceptionally careful about what they are saying and claiming. I re-read the paper even more carefully this time, and found other inconsistencies that need to be addressed. I'm not trying to be obnoxious-I don't want someone else to take them apart after the paper is published. I think that looking at the behaviour both of individuals and individual pairs would be very helpful.
Again, the good part of this paper is that the authors are honest in their overall evaluation that additional work is necessary to determine prosociality in parrots (at least Grey parrots). And, yes they did test more birds than in previous studies, but looking at the individual data it seems that some birds are good at being prosocial, others are not, and some learn more about the situation than do others. And I repeat that their controls were confusing…if I had been in the birds' place, I would have been confused. Just because someone else used these controls doesn't mean they are correct. Thus, again, the authors have to be extremely careful about their analyses and claims as well as criticisms.

22-Aug-2019
Dear Dr Krasheninnikova: Manuscript ID RSOS-190696.R1 entitled "Assessing African grey parrots' prosocial tendencies in a token choice paradigm" which you submitted to Royal Society Open Science, has been reviewed. The comments of the reviewer(s) are included at the bottom of this letter.
Please submit a copy of your revised paper before 14-Sep-2019. Please note that the revision deadline will expire at 00.00am on this date. If we do not hear from you within this time then it will be assumed that the paper has been withdrawn. In exceptional circumstances, extensions may be possible if agreed with the Editorial Office in advance. We do not allow multiple rounds of revision so we urge you to make every effort to fully address all of the comments at this stage. If deemed necessary by the Editors, your manuscript will be sent back to one or more of the original reviewers for assessment. If the original reviewers are not available we may invite new reviewers.
To revise your manuscript, log into http://mc.manuscriptcentral.com/rsos and enter your Author Centre, where you will find your manuscript title listed under "Manuscripts with Decisions." Under "Actions," click on "Create a Revision." Your manuscript number has been appended to denote a revision. Revise your manuscript and upload a new version through your Author Centre.
When submitting your revised manuscript, you must respond to the comments made by the referees and upload a file "Response to Referees" in "Section 6 -File Upload". Please use this to document how you have responded to the comments, and the adjustments you have made. In order to expedite the processing of the revised manuscript, please be as specific as possible in your response.
In addition to addressing all of the reviewers' and editor's comments please also ensure that your revised manuscript contains the following sections before the reference list: • Ethics statement If your study uses humans or animals please include details of the ethical approval received, including the name of the committee that granted approval. For human studies please also detail whether informed consent was obtained. For field studies on animals please include details of all permissions, licences and/or approvals granted to carry out the fieldwork.
• Data accessibility It is a condition of publication that all supporting data are made available either as supplementary information or preferably in a suitable permanent repository. The data accessibility section should state where the article's supporting data can be accessed. This section should also include details, where possible of where to access other relevant research materials such as statistical tools, protocols, software etc can be accessed. If the data have been deposited in an external repository this section should list the database, accession number and link to the DOI for all data from the article that have been made publicly available. Data sets that have been deposited in an external repository and have a DOI should also be appropriately cited in the manuscript and included in the reference list.
• Competing interests Please declare any financial or non-financial competing interests, or state that you have no competing interests.
• Authors' contributions All submissions, other than those with a single author, must include an Authors' Contributions section which individually lists the specific contribution of each author. The list of Authors should meet all of the following criteria; 1) substantial contributions to conception and design, or acquisition of data, or analysis and interpretation of data; 2) drafting the article or revising it critically for important intellectual content; and 3) final approval of the version to be published.
All contributors who do not meet all of these criteria should be included in the acknowledgements.
We suggest the following format: AB carried out the molecular lab work, participated in data analysis, carried out sequence alignments, participated in the design of the study and drafted the manuscript; CD carried out the statistical analyses; EF collected field data; GH conceived of the study, designed the study, coordinated the study and helped draft the manuscript. All authors gave final approval for publication.
• Acknowledgements Please acknowledge anyone who contributed to the study but did not meet the authorship criteria.
• Funding statement Please list the source of funding for each author.
Once again, thank you for submitting your manuscript to Royal Society Open Science and I look forward to receiving your revision. If you have any questions at all, please do not hesitate to get in touch. Ordinarily, the journal does not grant authors multiple opportunities to revise their paper(s); however, where 'good faith' attempts at improvement seem to be made by the authors in response to extensive feedback, a degree of flexibility is sometimes permitted. As the Editors consider you to have tried to respond sincerely (if not satisfactorily) to the reviewers, a further opportunity to revise is being made available -you should work hard to tackle constructively the remaining concerns. If you do not do so, and the more critical of the reviewers remains unsatisfied by your efforts, we will not be able to consider further revisions. Good luck.
Subject Editor Comments to Author: I support the AE's comments and wish you the best in your revisions.
Krasheninnikova and colleagues have revised their extremely interesting paper on prosocial behaviour in Grey parrots. They examine multiple possible situations and test the birds extensively. I nevertheless still have several comments -some involving minor clarification, others a bit more substantive-which I present in order of appearance. L. 29: I'd be more comfortable with something like "..their lack of understanding of the contingencies of the particular tasks used in this study, the underlying…" [that is the most that the authors can claim] L. 39: This paper is on pro-sociality, which the authors just described as being different from altruism. So where does this sentence with respect to altruism fit? Do they mean pro-sociality? Or at least reciprocal altruism?
L. 58-59: The edit is odd…leave out comma after "familiar"…and maybe "and in situations without a possibility…" L. 118: Remove comma after "While"…Based on the material in the paper, maybe the description should read "did not show clear reciprocity…" Again, read my comments based on the material in the paper cited: When the dominant bird led, he STARTED OUT prosocial but switched to selfish when his partner consistently acted in a selfish manner"…what is written in this manuscript is not correct.
This description of the material is not correct. Citing from the paper itself: "However, whether they shared with their partner at no cost to themselves seemed, at least for our subjects, to depend upon their dominance status and upon who led and who followed. Griffin has always been dominant to Arthur, and Figs. 3 and S1 show that when Griffin led, he became selfish, but when placed in the position of follower, he acted in a more sharing manner." That is, when Griffin led he started out sharing but when Arthur became more selfish, Griffin became more selfish. When Griffin followed, he retained his more sharing manner. Arthur, the subordinate, "..started with a slight giving tendency as both leader and follower, then became statistically significantly more selfish both as leader and follower." L. 120ff: THEREFORE, these lines should read: "If the dominant bird started the session choosing first, he initially chose the prosocial token, but became more selfish after he experienced selfish behavior from his partner. When the dominant bird was the follower, he retained his prosociality. The subordinate bird, however, became more selfish over time in both leader and follower roles. Therefore, both birds developed…" L. 125: This sentence is not correct-and the sentence has not been altered….I repeat: the authors tested for side biases and the current authors cannot disregard these tests. I will change my decision to "reject" rather than revise if this revision is not made in a subsequent version.
L. 126: This sentence is also not correct. Control was inherent in the task: Note that the birds did not copy the human, as the birds were not giving treats to the generous human but sharing treats, and they quickly learned to avoid the null cup (no treats to anyone), demonstrating that they clearly understood the task. Note, too, the follow-up study that further tested the task's assumptions. The authors of the current paper cannot blithely disregard data and criticize another paper without full justification. Avoiding the null cup if it gives no reward can be described any way you choose, but it still involves understanding the contingencies of the task. And explaining the choice of the option that provides two rewards rather than one as reinforcement learning is meaningless unless the chooser GETS more rewards. If the chooser gets to share, that means the chooser understands the contingencies of the choice. As for the control without a partner-the subject is simply likely going to be confused about what is going on-see more on that later.
The argument about color bias is justified, however.
L. 134: I think the authors mean "employed" rather than "applied" L. 201: "without shells" L. 199: This sentence is not clearly written. Try dividing the sentence into two. "…in all equal reward conditions. When the actor chose prosocially in the unequal condition, in contrast, the partner received a piece of walnut (preference for walnut was assessed prior to the experiment) while the actor received merely the same sunflower seed as in the equal condition." L. 217: I think it is important to note here that these birds did NOT express inequity aversion previously. Such finding suggest that the authors already know that the birds are not that likely to be affected by their partner getting a better reward, and that this is a further test of that finding.
L. 216: In the response to reviewers, the authors mention that they did not find any position preference in token choice…this information should be provided somewhere in the results section.
L. 226: Is this a new paragraph? L. 244: Sentence a bit confusing. Divide sentences. "The rewards inside the cups were not at that time visible to the birds. Once a choice was made, the experimenter lifted the lid(s) off the small food container(s), tilted the cup(s) forward so that both birds could see the reward(s) inside, and then distributed the reward(s)." L. 259ff: If birds don't care about inequality, why would they care if one got fed and one didn't? Isn't that finding a problem for this entire paper? The receiver could eventually associate the ring color with whether it gets food, but not necessarily with the actions of its partner if it doesn't care that its partner gets fed and it doesn't, especially if it was going to get at least half the standard reward just for hanging out near the reward site when the experimenter wanted it to move back. If a whole seed is a standard reward, half that reward is hardly "tiny" [and, yes, I do understand that sunflower seeds cannot be divided into much smaller pieces and still easily be handled, but maybe the authors should have used some other, less favored food?] l. 270: 12 sessions needed to be repeated? I'm confused…there are only 9 sessions listed in Table  1… If this statement is not a typo, then there are a LOT of repetitions. Could the authors please let the reader know how many sessions were repeated for each condition? That would help the reader know which conditions were more problematic. The authors gave some of the information in their response but not in the paper…it seems as though one dyad really had issues, and it would be helpful to know which one. Maybe that set of data needs to be dealt with separately? L. 273: What happened after the birds chose a token in the color preference test? Did it just get taken away? Did they get something or nothing for making the choice? If they had any experience in trading tokens for treats, and they chose a token and got nothing, they would-if they had any sense-alternate choice of tokens to see if choosing a different token would give a reward, even if they liked their first choice. It seems that most birds started with a color preference and those colors were eliminated, but it also seems possible that the birds just LEARNED at this point that they should choose randomly because nothing would happen almost whatever they did BUT if they chose their favorite it would disappear. I don't disagree that one has to check for color preferences, but it's a rather tricky procedure as they learn to ignore any contingencies connected with the color. L. 286ff: So now a bird was given "x" to choose and got a reward half the time and was given "y" to choose the other half and still got a reward; sometimes their partner got a reward and sometimes the partner did not. Given that they didn't care about reward inequality, why would they pay much attention to what was happening, particularly if they were actor but even if they were the powerless recipient? If the recipient couldn't do anything about what happens, and can't tell the actor what to do, why pay too much attention?
L. 298: And I'm still confused…The authors mention equal and unequal rewards earlier, but Table 1 does not, so I'm confused. Does each bird have one session as actor with equal rewards then one with unequal reward before the roles are reversed (i.e., that's what is meant by 2 sessions in Table 1)? Or do reversals occur after equal reward and then again before unequal reward? And do all these trials occur before the alternating sessions? I need to understand what is happening at this point in the paper. Or do the birds cycle through UNI, ALT, YOK in the equal reward before doing unequal rewards? [It seems that way, given that different tokens were used, but I really had to work to figure this out and I'm still not sure.] If I were trying to replicate this experiment, I'd never be able to figure out exactly what was done. The material now taken out helped me before… L. 305: Two sessions EACH? Including equal or unequal rewards? L. 343: Do the authors mean 30 trials per bird acting as actor and 30 as recipient? And where do the equal/unequal fit in here…one session with equal rewards, one session with unequal rewards? Not clear… L. 355ff: I absolutely do not understand the authors' response to my query. Why would switching to a selfish strategy be the expected response? The experimenter is NOT always being selfish, so why would the actor be selfish? The authors are making AN ASSUMPTION that the birds are treating the experimenter as an automaton, and ignoring the possibility that they were playing with the experimenter. In previous studies, didn't the experimenter played partner-like roles? So why ignore that now? Given that there were lots of prosocial choices in the alternating condition, the experimenter was being prosocial a lot, so why wouldn't the actor reciprocate to that? I definitely appreciate the new material added around L. 580. L. 384: Accessible vs inaccessible: So, I'm a bird and you train me to go back and forth between compartments. I learn that I can get food in both places, but not with respect to anything to do with token choice; I just learn that it's possible. And there's no bird next door. If I'm Jelo or Bella or Sensi or Nikki (half the birds), I don't notice whether or not the passage is open but I choose to put food on the other side because I remember that it's possible to get food. If I'm Kizzi or Nina or Jack, whether or not the passage is open I'm not prosocial maybe because there's no bird there. If I'm Kimmi, I'm just confused. Actually, if the birds differentiated between the conditions, they should NOT be random but choose selfish in inaccessible because they'd otherwise be wasting food, and they should choose prosocial in the accessible condition to get more food. But none of the birds (except a tendency in Jack, in the opposite to what would be expected) seemed to understand the different conditions at all. They seemed to hit on some strategy simply with respect to the possibility of more food-hope for the best or accept the worst-and stick to it. I don't see the task as showing much of anything about understanding cooperation. I think the authors' comments in the Discussion about what they did and didn't learn and spatial confusion make more sense than the assumption that the task would be useful, particularly given that the birds' behaviours altered over time.
Re blockage…about half the birds seemed to understand the issue and the other half did not.
Overall: Individual differences seem to be extremely important in this study, from looking at the supplemental data. Maybe some Greys are good buddies and others simply are not, or maybe some birds are dim and others not…the authors really need to look at and discuss the individuals, not just the group behaviour. L. 451ff: I'm confused…if birds DID develop preferences, and there were so many individual differences in the results, shouldn't the experimenters have checked specifically to see if the birds' actions were affected by these preferences? Just looking at group behaviour is not sufficient, as the individuals varied so strongly. L. 560: Correct English is "fewer prosocial", not "less prosocial"… We still do not know the relationships between the individuals in the pairs, although the authors do admit to the artificiality of the situation. Note: A 'good' relationship doesn't necessarily mean equal; if the birds have worked out a dominant-subordinate relationship and are happy with it, that makes it "good"…but if there is conflict, that is a different situation, and could easily affect the results.
We also know that parrots engage in courtship feeding, where one bird gives up food to the other with no reciprocity, and in Greys both parents feed the young directly from what they obtain when they forage (unlike mammals, who suckle, which is quite different in terms of sharing resources)…so the whole idea of unequal reward as an issue is quite different from what one would see in primates.
I'm still very conflicted about this paper-given how critical the experimenters are of the previous studies, their failure to examine the material of the previous studies carefully, and their claims that they performed their experiments in order to correct all the problems they perceived in the previous studies, they really have to be exceptionally careful about what they are saying and claiming. I re-read the paper even more carefully this time, and found other inconsistencies that need to be addressed. I'm not trying to be obnoxious-I don't want someone else to take them apart after the paper is published. I think that looking at the behaviour both of individuals and individual pairs would be very helpful.
Again, the good part of this paper is that the authors are honest in their overall evaluation that additional work is necessary to determine prosociality in parrots (at least Grey parrots). And, yes they did test more birds than in previous studies, but looking at the individual data it seems that some birds are good at being prosocial, others are not, and some learn more about the situation than do others. And I repeat that their controls were confusing…if I had been in the birds' place, I would have been confused. Just because someone else used these controls doesn't mean they are correct. Thus, again, the authors have to be extremely careful about their analyses and claims as well as criticisms.

Comments to the Author(s)
Krasheninnikova and colleagues have re-revised their extremely interesting paper on prosocial behaviour in Grey parrots. They examine multiple possible situations and test the birds extensively. I nevertheless still have several comments -some involving minor clarification, others a bit more substantive-which I present in order of appearance.
L. 75: Authors might be interested in a new paper: PLoS One: Female rats release a trapped cagemate following shaping of the door opening response: Opening latency when the restrainer was baited with food, was empty, or contained a cagemate Magnus H. Blystad , Danielle Andersen, Espen B. Johansen Also: Should read "methodologies" partner got a reward and sometimes the partner did not. I repeat….why would they pay much attention to what was happening, particularly if they were actor but even if they were the powerless recipient? If the recipient couldn't do anything about what happens, and can't tell the actor what to do, why pay too much attention? That's what happened in the UNI condition…The authors' response makes it clear that the birds care in the ALT condition, in which they are NOT powerless:In reply to the reviewer's last 2 sentences…we have also evidence that even recipients pays attention to the reward contingencies in both compartments as revealed by the increase in prosocial choices in our ALT (alternating) condition. In the ALT condition, the recipients were not "powerless" because they had the opportunity to reciprocate/retaliate in an alternating manner.….which is exactly my point…there is no reason to pay attention unless they can take an active role. The authors have to clarify WHY the UNI condition is useful earlier on (they do so in L. 307, but until then a reader likely thinks it's a totally useless task)… In the Discussion, they should emphasize the point that in the UNI condition, birds' failures to respond as expected could have been because they did not receive feedback from the recipient, who was powerless to respond, and because the recipient might not even have attended to the behaviour of the actor in the UNI condition for that reason.
L. 327ff: The authors seem not to understand my criticism. My point: The authors are making AN ASSUMPTION that the birds are treating the experimenter as an automaton, and ignoring the possibility (at least until the Discussion) that they were playing the game with the experimenter. The issue is not only whether the actor understood that the recipient could not choose, but also UNDERSTOOD that the experimenter was NOT deliberately choosing…How could the authors be sure of the parrot's interpretation? They suggest this possibility in the Discussion, but it needs to be made clearer.
L. ~360ff: Again, the authors seemed not to understand my criticism….I'm basically questioning the usefulness of Accessible vs Inaccessible: So, I'm a bird and you train me to go back and forth between compartments. I learn that I can get food in both places, but NOT with respect to anything to do with TOKEN choice; I just learn that it's possible when, for some odd reason, there's no bird next door. If I'm Jelo or Bella or Sensi or Nikki (half the birds), I don't notice whether or not the passage is open but I choose to put food on the other side because I remember that it's possible to get food. If I'm Kizzi or Nina or Jack, whether or not the passage is open I'm not prosocial maybe because there's no bird there. If I'm Kimmi, I'm just confused. None of the birds (except a tendency in Jack, in the opposite to what would be expected) seemed to understand the different conditions at all. They seemed to hit on some strategy simply with respect to the possibility of more food-hope for the best or accept the worst-and stick to it. I don't see the task as showing much of anything about understanding cooperation. Or…Maybe the birds kept hoping their actions would eventually help their partner, whom they expected to return because the partner had been there for all the other types of trials?
What about a different control scenario that would not be confusing to the bird?…What if the recipient had plenty of food and didn't NEED anything from the actor, would the actor have continued to be prosocial? L. 395: Again, I see this task as simply confusing to the birds, not an actual test of their understanding… L. 511: "…although such tendencies…" L. 515ff, 586ff, 642: The issue to me is that the task contingencies are inherently confusing…which is the point of my criticisms above. I understand that the authors wanted to replicate work with nonhuman primates, but I simply don't think that some of the tasks actually tested what they were purportedly designed to test. Parrots may be inherently less competitive than most nonhuman primates; tasks that might make sense to nonhuman primates might not make sense to parrots…Again, I understand that the authors wanted direct comparisons with nonhuman primates, but then they need to spend more space in the Discussion (not just a brief mention) explaining why their data differed, based on the ecology and ethology of parrots compared to nonhuman primates.
And, as some dyads were just more prosocial than others…one of the in-depth issues for the Discussion is the possibility that individual relationships may be more important to parrots than nonhuman primates and that trying to examine the issue on a group level may be more difficult than it might be in nonhuman primates…That is, maybe if these individual relationships are so important, the birds will respond in ways that differ from how nonhuman primates respond? L. 550: Yes… L. 557: Maybe the UNI prosocial walnut choice was some attempt to get the receiver to attend to what was happening? Given that the UNI unequal followed the ALT equal, maybe they thought that somehow if the recipient got a special treat, s/he would respond, breaking the UNI condition?
L. 668: I don't think the issue was the lack of cognitive resources in the parrots as much as the confusing conditions of the task.
Again, the good part of this paper is that the authors are honest in their overall evaluation that additional work is necessary to determine prosociality in parrots (at least Grey parrots). And, yes they did test more birds than in previous studies, but looking at the individual data it seems that some birds are good at being prosocial, others are not, and some learn more about the situation than do others. And I repeat that their controls were confusing…if I had been in the birds' place, I would have been confused.
What I'm trying to get at here is that the authors made an excellent, good-faith attempt at using a protocol that was designed for nonhuman primates, with very limited revision, for Grey parrots. They found that when the birds could actually interact with one another, some pairs acted prosocially, other pairs (like the pair in ref. 53), did not. They also found that the controls may not have been appropriate for the parrots…yes, the tasks might have been beyond the parrots' processing abilities, but given how many other tasks these birds solve appropriately, it is more likely that the birds "overthought" the tasks and responded in ways more appropriate to their ecology/ethology. In sum….A lot of what the authors wrote in response to the reviewer comments would be a terrific part of the Discussion, and I hope that the editors allow the authors another round so that they can insert that material. My only disagreement with the following is "unintended variables"-the issue is more likely different types and degrees of species-specific social interactions in parrots and nonhuman primates…the authors almost get to it in the current version, but what they write in their rebuttal would be a lovely end to their paper: it is the overarching dilemma of animal cognition studies, that we do not know how the test animals under investigation perceive a test situation. Most studies aim at designing their studies as ethologically valid and salient for the species under investigation as possible, but often it turns out retrospectively that unintended variables affected the animals in the test situation. This is a common pitfall of comparative cognitive studies in particular where one strives for maximum comparability by keeping the experimental protocols as similar as possible.

04-Nov-2019
Dear Dr Krasheninnikova: On behalf of the Editors, I am pleased to inform you that your Manuscript RSOS-190696.R2 entitled "Assessing African grey parrots' prosocial tendencies in a token choice paradigm" has been accepted for publication in Royal Society Open Science subject to minor revision in accordance with the referee suggestions. Please find the referees' comments at the end of this email.
The reviewers and Subject Editor have recommended publication, but also suggest some minor revisions to your manuscript. Therefore, I invite you to respond to the comments and revise your manuscript.
• Ethics statement If your study uses humans or animals please include details of the ethical approval received, including the name of the committee that granted approval. For human studies please also detail whether informed consent was obtained. For field studies on animals please include details of all permissions, licences and/or approvals granted to carry out the fieldwork.
• Data accessibility It is a condition of publication that all supporting data are made available either as supplementary information or preferably in a suitable permanent repository. The data accessibility section should state where the article's supporting data can be accessed. This section should also include details, where possible of where to access other relevant research materials such as statistical tools, protocols, software etc can be accessed. If the data has been deposited in an external repository this section should list the database, accession number and link to the DOI for all data from the article that has been made publicly available. Data sets that have been deposited in an external repository and have a DOI should also be appropriately cited in the manuscript and included in the reference list.
If you wish to submit your supporting data or code to Dryad (http://datadryad.org/), or modify your current submission to dryad, please use the following link: http://datadryad.org/submit?journalID=RSOS&manu=RSOS-190696.R2 • Competing interests Please declare any financial or non-financial competing interests, or state that you have no competing interests.
• Authors' contributions All submissions, other than those with a single author, must include an Authors' Contributions section which individually lists the specific contribution of each author. The list of Authors should meet all of the following criteria; 1) substantial contributions to conception and design, or acquisition of data, or analysis and interpretation of data; 2) drafting the article or revising it critically for important intellectual content; and 3) final approval of the version to be published.
All contributors who do not meet all of these criteria should be included in the acknowledgements.
We suggest the following format: AB carried out the molecular lab work, participated in data analysis, carried out sequence alignments, participated in the design of the study and drafted the manuscript; CD carried out the statistical analyses; EF collected field data; GH conceived of the study, designed the study, coordinated the study and helped draft the manuscript. All authors gave final approval for publication.
• Acknowledgements Please acknowledge anyone who contributed to the study but did not meet the authorship criteria.
• Funding statement Please list the source of funding for each author.
Please note that we cannot publish your manuscript without these end statements included. We have included a screenshot example of the end statements for reference. If you feel that a given heading is not relevant to your paper, please nevertheless include the heading and explicitly state that it is not relevant to your work.
Because the schedule for publication is very tight, it is a condition of publication that you submit the revised version of your manuscript before 13-Nov-2019. Please note that the revision deadline will expire at 00.00am on this date. If you do not think you will be able to meet this date please let me know immediately.
To revise your manuscript, log into https://mc.manuscriptcentral.com/rsos and enter your Author Centre, where you will find your manuscript title listed under "Manuscripts with Decisions". Under "Actions," click on "Create a Revision." You will be unable to make your revisions on the originally submitted version of the manuscript. Instead, revise your manuscript and upload a new version through your Author Centre.
When submitting your revised manuscript, you will be able to respond to the comments made by the referees and upload a file "Response to Referees" in "Section 6 -File Upload". You can use this to document any changes you make to the original manuscript. In order to expedite the processing of the revised manuscript, please be as specific as possible in your response to the referees.
When uploading your revised files please make sure that you have: 1) A text file of the manuscript (tex, txt, rtf, docx or doc), references, tables (including captions) and figure captions. Do not upload a PDF as your "Main Document". 2) A separate electronic file of each figure (EPS or print-quality PDF preferred (either format should be produced directly from original creation package), or original software format) 3) Included a 100 word media summary of your paper when requested at submission. Please ensure you have entered correct contact details (email, institution and telephone) in your user account 4) Included the raw data to support the claims made in your paper. You can either include your data as electronic supplementary material or upload to a repository and include the relevant doi within your manuscript 5) All supplementary materials accompanying an accepted article will be treated as in their final form. Note that the Royal Society will neither edit nor typeset supplementary material and it will be hosted as provided. Please ensure that the supplementary material includes the paper details where possible (authors, article title, journal name).
Supplementary files will be published alongside the paper on the journal website and posted on the online figshare repository (https://figshare.com). The heading and legend provided for each supplementary file during the submission process will be used to create the figshare page, so please ensure these are accurate and informative so that your files can be found in searches. Files on figshare will be made available approximately one week before the accompanying article so that the supplementary material can be attributed a unique DOI.
Once again, thank you for submitting your manuscript to Royal Society Open Science and I look forward to receiving your revision. If you have any questions at all, please do not hesitate to get in touch.
Kind regards, Andrew Dunn Senior Publishing Editor Royal Society Open Science openscience@royalsociety.org on behalf of Prof Kevin Padian (Subject Editor) openscience@royalsociety.org Associate Editor Comments to Author: The reviewer has presented a number of critiques of this paper through its iterations, and the Editors would like to thank them for the guidance and also encourage the authors to constructively engage with these critiques. Please do ensure that you take the time to make the changes requested, and present these in a tracked changes version of the paper, along with a full point-by-point response. This will help the Editors determine whether the reviewer needs to be consulted further.
Reviewer comments to Author: Reviewer: 2 Comments to the Author(s) Krasheninnikova and colleagues have re-revised their extremely interesting paper on prosocial behaviour in Grey parrots. They examine multiple possible situations and test the birds extensively. I nevertheless still have several comments -some involving minor clarification, others a bit more substantive-which I present in order of appearance.
L. 75: Authors might be interested in a new paper: PLoS One: Female rats release a trapped cagemate following shaping of the door opening response: Opening latency when the restrainer was baited with food, was empty, or contained a cagemate Magnus H. Blystad , Danielle Andersen, Espen B. Johansen Also: Should read "methodologies" L. 104: Odd grammar: Probably should read something like "Although few studies have focused on…in the wild, a large number of studies…in the laboratory." L. 124: The authors again do not attend to the studies. They are confusing ref. 53 and ref. 54. In a second experiment in Iref. 53, the human was either generous or selfish and the parrots modulated their responses accordingly. In the follow-up study in ref. 54, the human exactly mirrored the parrot's behavior, and the parrot learned the "tit-for-tat" response quickly. So…if the authors want to ignore the human-parrot study (ref. 54) to some extent because it is not parrot-parrot, that is a separate issue, but they should not confound the two studies….
L. 129: This description of the material is STILL not correct. Again, I am truly annoyed at the authors' inability to separate the second experiment in ref. 53 from the study in ref. 54. In ref. 54, citing from the paper itself: "Cup positions were now changed after every human choice, with positions determined by use of the program random.org." The point is that the authors did check for side biases in their subsequent study…and there were NO effects... If the same bird did not have position preferences in ref. 54, it did not have them in ref. 53.
According to the authors in ref. 54, that subsequent study was also designed to examine the parrot's understanding of the task…would the parrot learn that the recipient was playing tit-fortat and respond in kind? So, again the statement with respect to understanding is not valid, and the citation of ref. L. 146: Grammar-"We varied the reward distribution, using "equal" or "unequal" rewards, to test….distribution (i.e., making….)" [I believe a parenthesis is missing] L. 154: "…chose the prosocial…" L. 198, 199: "When the actor…" L. 226: "…overlap but were still…" L. 253: How many is "a few"? L. 268: The authors' response to the reviewer comments was a lot clearer than what they wrote in the manuscript…. In the colour preference test, any choice was rewarded in the same way. Two different tokens were presented in one session in a shallow box. After the bird chose one token, it was rewarded with a sunflower seed and the trial ended. The next trial started after a 10 sec interval. Each subject was offered 12 choices, and if any token was chosen 10 times or more by one of the two birds of one pair, the token would be replaced by another colour in the next session. This additional information about reward would be extremely helpful, so that readers understand what the bird actually experienced. L. ~276ff: In the familiarization condition, a bird was given "x" to choose and got a reward half the time and was given "y" to choose the other half and still got a reward; sometimes their partner got a reward and sometimes the partner did not. I repeat….why would they pay much attention to what was happening, particularly if they were actor but even if they were the powerless recipient? If the recipient couldn't do anything about what happens, and can't tell the actor what to do, why pay too much attention? That's what happened in the UNI condition…The authors' response makes it clear that the birds care in the ALT condition, in which they are NOT powerless:In reply to the reviewer's last 2 sentences…we have also evidence that even recipients pays attention to the reward contingencies in both compartments as revealed by the increase in prosocial choices in our ALT (alternating) condition. In the ALT condition, the recipients were not "powerless" because they had the opportunity to reciprocate/retaliate in an alternating manner.….which is exactly my point…there is no reason to pay attention unless they can take an active role. The authors have to clarify WHY the UNI condition is useful earlier on (they do so in L. 307, but until then a reader likely thinks it's a totally useless task)… In the Discussion, they should emphasize the point that in the UNI condition, birds' failures to respond as expected could have been because they did not receive feedback from the recipient, who was powerless to respond, and because the recipient might not even have attended to the behaviour of the actor in the UNI condition for that reason.
L. 327ff: The authors seem not to understand my criticism. My point: The authors are making AN ASSUMPTION that the birds are treating the experimenter as an automaton, and ignoring the possibility (at least until the Discussion) that they were playing the game with the experimenter. The issue is not only whether the actor understood that the recipient could not choose, but also UNDERSTOOD that the experimenter was NOT deliberately choosing…How could the authors be sure of the parrot's interpretation? They suggest this possibility in the Discussion, but it needs to be made clearer.
L. ~360ff: Again, the authors seemed not to understand my criticism….I'm basically questioning the usefulness of Accessible vs Inaccessible: So, I'm a bird and you train me to go back and forth between compartments. I learn that I can get food in both places, but NOT with respect to anything to do with TOKEN choice; I just learn that it's possible when, for some odd reason, there's no bird next door. If I'm Jelo or Bella or Sensi or Nikki (half the birds), I don't notice whether or not the passage is open but I choose to put food on the other side because I remember that it's possible to get food. If I'm Kizzi or Nina or Jack, whether or not the passage is open I'm not prosocial maybe because there's no bird there. If I'm Kimmi, I'm just confused. None of the birds (except a tendency in Jack, in the opposite to what would be expected) seemed to understand the different conditions at all. They seemed to hit on some strategy simply with respect to the possibility of more food-hope for the best or accept the worst-and stick to it. I don't see the task as showing much of anything about understanding cooperation. Or…Maybe the birds kept hoping their actions would eventually help their partner, whom they expected to return because the partner had been there for all the other types of trials?
What about a different control scenario that would not be confusing to the bird?…What if the recipient had plenty of food and didn't NEED anything from the actor, would the actor have continued to be prosocial? L. 395: Again, I see this task as simply confusing to the birds, not an actual test of their understanding… L. 511: "…although such tendencies…" L. 515ff, 586ff, 642: The issue to me is that the task contingencies are inherently confusing…which is the point of my criticisms above. I understand that the authors wanted to replicate work with nonhuman primates, but I simply don't think that some of the tasks actually tested what they were purportedly designed to test. Parrots may be inherently less competitive than most nonhuman primates; tasks that might make sense to nonhuman primates might not make sense to parrots…Again, I understand that the authors wanted direct comparisons with nonhuman primates, but then they need to spend more space in the Discussion (not just a brief mention) explaining why their data differed, based on the ecology and ethology of parrots compared to nonhuman primates.
And, as some dyads were just more prosocial than others…one of the in-depth issues for the Discussion is the possibility that individual relationships may be more important to parrots than nonhuman primates and that trying to examine the issue on a group level may be more difficult than it might be in nonhuman primates…That is, maybe if these individual relationships are so important, the birds will respond in ways that differ from how nonhuman primates respond? L. 550: Yes… L. 557: Maybe the UNI prosocial walnut choice was some attempt to get the receiver to attend to what was happening? Given that the UNI unequal followed the ALT equal, maybe they thought that somehow if the recipient got a special treat, s/he would respond, breaking the UNI condition?
L. 668: I don't think the issue was the lack of cognitive resources in the parrots as much as the confusing conditions of the task.
Again, the good part of this paper is that the authors are honest in their overall evaluation that additional work is necessary to determine prosociality in parrots (at least Grey parrots). And, yes they did test more birds than in previous studies, but looking at the individual data it seems that some birds are good at being prosocial, others are not, and some learn more about the situation than do others. And I repeat that their controls were confusing…if I had been in the birds' place, I would have been confused.
What I'm trying to get at here is that the authors made an excellent, good-faith attempt at using a protocol that was designed for nonhuman primates, with very limited revision, for Grey parrots. They found that when the birds could actually interact with one another, some pairs acted prosocially, other pairs (like the pair in ref. 53), did not. They also found that the controls may not have been appropriate for the parrots…yes, the tasks might have been beyond the parrots' processing abilities, but given how many other tasks these birds solve appropriately, it is more likely that the birds "overthought" the tasks and responded in ways more appropriate to their ecology/ethology.
In sum….A lot of what the authors wrote in response to the reviewer comments would be a terrific part of the Discussion, and I hope that the editors allow the authors another round so that they can insert that material. My only disagreement with the following is "unintended variables"-the issue is more likely different types and degrees of species-specific social interactions in parrots and nonhuman primates…the authors almost get to it in the current version, but what they write in their rebuttal would be a lovely end to their paper: it is the overarching dilemma of animal cognition studies, that we do not know how the test animals under investigation perceive a test situation. Most studies aim at designing their studies as ethologically valid and salient for the species under investigation as possible, but often it turns out retrospectively that unintended variables affected the animals in the test situation. This is a common pitfall of comparative cognitive studies in particular where one strives for maximum comparability by keeping the experimental protocols as similar as possible.

17-Nov-2019
Dear Dr Krasheninnikova, It is a pleasure to accept your manuscript entitled "Assessing African grey parrots' prosocial tendencies in a token choice paradigm" in its current form for publication in Royal Society Open Science. The comments of the reviewer(s) who reviewed your manuscript are included at the foot of this letter.
Please ensure that you send to the editorial office an editable version of your accepted manuscript, and individual files for each figure and table included in your manuscript. You can send these in a zip folder if more convenient. Failure to provide these files may delay the processing of your proof. You may disregard this request if you have already provided these files to the editorial office.
You can expect to receive a proof of your article in the near future. Please contact the editorial office (openscience_proofs@royalsociety.org) and the production office (openscience@royalsociety.org) to let us know if you are likely to be away from e-mail contact --if you are going to be away, please nominate a co-author (if available) to manage the proofing process, and ensure they are copied into your email to the journal.
Due to rapid publication and an extremely tight schedule, if comments are not received, your paper may experience a delay in publication. Royal Society Open Science operates under a continuous publication model. Your article will be published straight into the next open issue and this will be the final version of the paper. As such, it can be cited immediately by other researchers. As the issue version of your paper will be the only version to be published I would advise you to check your proofs thoroughly as changes cannot be made once the paper is published.

Introduction
Line 58: This seems like a bit of an exaggeration, given that there are ample studies that show cooperative behavior among non-kin in primates in the absence of reciprocation. For example, in studies [20,25] there was no opportunity for reciprocity within the experiment and the primates still were cooperative. I suspect the authors are suggesting that nonkin cooperate only when they are familiar (which has been difficult to study since it is challenging to find species and settings where strangers can be tested together). But this line seems like a bit of an overstatement as it is written, especially since in [20] both kin and non-kin pairs behaved similarly. Response: We agree and have changed the wording accordingly.
Line 57ff: "Indeed, cooperative behaviours in non-human primates seem to be directed more frequently to kin and to familiar, hence reciprocating, partners [17,18], but have also been reported between non-kin partners and without a possibility for reciprocity [20,25].".
Line 108: While the parrots succeeded at the string-pulling task in [49], they did not show awareness that a partner was needed in the delay condition. By most accounts, this would be considered a failure at cooperation, since they didn't appear to notice the role of the partner. Response: In the Péron et al. (2011) study, the parrots were able to coordinate their actions in order to pull a platform with food into reach. You are correct in stating that they failed in the delayed control (in which they needed to inhibit pulling the string until the partner arrived); however, when given the choice between two apparatuses -one which could be solved alone and one which required a partner -they correctly chose the appropriate apparatus. Consequently, indicating that the parrots were able to take the necessity for a partner into account, but potentially due to a lack of inhibitory control, they could not refrain from pulling the string when no partner was available (as in the delayed control). Nonetheless, we have changed the wording in the manuscript and no longer talk about cooperative abilities but rather the ability to coordinate actions (line 109).

Methods
General comment: I am sure this was unintentional, but at times the language follows Suchak & de Waal [ref 21] extremely closely, even verbatim. I recognize that when doing a replication it may be hard to think of alternative ways to phrase things, but the authors need to extensively re-word large portions of the methods section to avoid violating copyright. For this reason I am suggesting a major revision. Response: Thank you for noticing this overlap in the descriptions between studies. We did so unintentionally and have changed the wording of the methods.
Line 213: Why were the compartment sides fixed, rather than counterbalanced across days/sessions? Could this possibly have inhibited them from moving from side to side in the accessible condition? Response: The reason for fixed compartments sides was to ensure that the subjects "knew" their role across the conditions, i.e., who is the choosing individual. Since we provide them with a training session before starting the accessible condition during which they had to move from side to side, it is unlikely that the fixed compartments might have inhibited them from moving from walking towards the other side: only once after they had changed the sides 12 times in a row they proceeded with the test condition. We added this information in the revised version of the manuscript (line 224ff).
Line 220: I think this should say the partner could witness the actor's behavior Response: Thank you for noticing this mistake, we have changed it accordingly.

Results:
Line 421: One possibility for this, which hasn't really been addressed in the discussion is that during the unequal ALT condition, the birds might experience a frustration effect, since in one trial they might get a high value reward as the partner, but in the next get a low value reward as the actor, even though the token chosen in both trials is the same. Response: This is an interesting interpretation that we have not thought of so far. We have added this to the discussion.
Line 590ff: "The decrease of prosocial choices in the unequal ALT compared to the equal ALT condition could be explained by a frustration effect. The switch from receiving seeds (when in actor role) after having received walnut in the previous trial (if partner made a prosocial choice) might be perceived as frustrating as the reward quality decreased. Potentially, the induced frustration affected the parrots' other-regarding preferences." Line 443: Are the descriptives the M±SD? If so, can you specify that?
Response: Yes, these results depict the mean ± SD. We have added this information to the manuscript (l.528).
Line 450: Have you tried a logistic regression using previous choice as a predictor, as opposed to the contingency table? Given that there were differences among individuals/pairs (as discussed in lines 613-616), it seems possible that certain individuals might be driving this effect, which would be masked in pooled data. Response: We were not able to run a model on the data, as the model assumptions could not be fulfilled (i.e. uncorrectable overdispersion). Consequently, we looked at the contingency on a dyadic level rather than on the population level and found that none of the dyads exhibit significant contingency between choices (see Table S4 for details).
Line 457: I think the interpretation "indicating that the presence of a partner…" does not match the statistic presented here, since in both UNI and SFC the partner was present. It seems to me that this suggests that they weren't making their choice based on whether or not it was actually helping the partner. Response: You are right, the indifference between UNI and SFC actually indicates that prosocial choices might have been caused by social facilitation instead of an understanding of whether help can be provided or not. We changed the sentence accordingly in the manuscript.
Line 501ff: "The birds' prosocial choices did not differ between the UNI condition and the SFC (49.5 ± 39.2%;GLMM: β = 0.01,SE = 0.15,z = 0.08,p = 0.940); consequently, we cannot rule out that the presence of the partner facilitated prosocial choices irrespective of whether they actually helped the partner or not.".
Line 464: It seems odd to me that the INACC condition is the only one in which the birds changed over the 10-trial blocks. I note that half of the birds started with ACC and the other started with INACC. Is it possible this effect (and the overall high number of prosocial choices) is driven by the birds that started with ACC, since in the previous condition choosing the prosocial token was advantageous? Perhaps they needed to learn that they could not access both rewards? Response: We controlled for an order effect (i.e. whether birds starting with the INACC choose fewer prosocial tokens than birds starting with the ACC) and found no difference between birds. We have added the statistics to the manuscript (line 505ff).

Discussion:
Line 546: I think there is a sentence structure issue in the sentence starting "While their behaviour…"

Response:
We have clarified the sentence: Line 600ff: "The increase in prosocial choices across sessions in the unequal alternating condition could additionally be explained by reinforcement learning (i.e. pick same token as partner) or attitudinal reciprocity.".
Line 577: I am not necessarily recommending the authors do this (as I understand how difficult it is to run more conditions), but retesting the birds in the partner absent conditions with fresh tokens could identify whether it was the tokens or the set up causing this issue. If they still stayed to one side with fresh tokens, you would know it is the latter explanation, but if they showed more flexibility with fresh tokens, it would likely indicate that they formed fixed preferences for the tokens used in previous conditions. Response : We certainly agree with your suggestion and ideally, we would re-test the birds with a new set of tokens, however, considering that 1) the actual test was conducted more than one year ago and 2) a different study using other tokens was conducted in the meantime (thus potentially providing them with additional experiences), we unfortunately believe that re-testing the birds would not give us unbiased information about effects of token preferences for this study.
General comment: it seems to me that the two PCT studies [the current study and ref. 53] combined with the string pulling task [ref 49], all suggest that the parrots may not be monitoring and responding to their partner's behavior in the same way that we see in other species. This is not to suggest that they cannot truly cooperate by taking into account the partner's behavior [c.f. Ref 64], but perhaps tasks originally designed for primates are not tapping into these abilities in a parrotfriendly way. This is not meant as a criticism of this particular study (how else would we know except to try tests that have already been designed?) but rather something to consider adding to the discussion as it provides a suggestion for people in the field to think outside the primate box, so to speak. Response: Thank you for this remark, we definitely agree. We have added this aspect to the discussion in the manuscript.
Line 673ff: "While established paradigms (such as the PCT, which was originally developed for testing primates) certainly pose great potential for comparative research, future studies on avian cognition should also employ paradigms that are specifically designed for testing parrots, considering their specific behaviours and limitations and maximizing their ecological validity. This is important in order to investigate the possibility that their performance may be affected by the choice of paradigm and experimental context."

Reviewer: 2
Comments to the Author(s) Krasheninnikova and colleagues have written an extremely interesting paper on prosocial behaviour in Grey parrots. They examine multiple possible situations and test the birds extensively. I nevertheless have several comments -some involving minor clarification, others a bit more substantive-which I present in order of appearance.
whether they shared with their partner at no cost to themselves seemed, at least for our subjects, to depend upon their dominance status and upon who led and who followed. Griffin has always been dominant to Arthur, and Figs. 3 and S1 show that when Griffin led, he became selfish, but when placed in the position of follower, he acted in a more sharing manner." That is, when Griffin led he started out sharing but as Arthur became more selfish over time, Griffin also became more selfish. When Griffin was the follower, he retained his more sharing manner. Arthur, the subordinate, "..started with a slight giving tendency as both leader and follower, then became statistically significantly more selfish both as leader and follower." Response: Thank you for pointing this out. We have incorporated this into the manuscript.
Line 118ff: "While, the parrots did not show reciprocity, they changed their behaviour based on who started as the actor in each session. If the dominant bird started the session choosing first, he selected the selfish option more often, while the behaviour was reversed when roles were changed. Nonetheless, both birds developed a selfish tendency over time.". L. 122: Should read "…by decreasing or increasing prosocial choices appropriately". [Must maintain parallel construction-selfish/decrease; generous/increase] Response: Thank you for the suggestion, we have changed the sentence accordingly.
L. 123: This sentence is not correct-the authors tested for side biases. Response: We have toned the sentence down saying that the authors did not control sufficiently for side biases in our view. Although the authors state that they controlled for the spatial arrangement of cups on the tray, we are not convinced that this was sufficient to rule out side biases. For example, the spatial arrangement of cups was not consistently and randomly changed (i.e. changed only twice across 24/26 sessions). Furthermore, the authors provided statistics that assessed whether the parrots' choice differed from chance but not whether the birds actually preferred one cup over another. In order to truly rule out side preferences, as well as colour preferences, it would have been necessary to conduct preference tests before assigning cups to specific locations or reward types.
L. 124: This sentence is also not correct. Control was inherent in the task: Note that the birds did not copy the human, as the birds were not giving treats to the generous human but sharing treats, and they quickly learned to avoid the null cup (no treats to anyone), demonstrating that they clearly understood the task. Note, too, the follow-up study that further tested the task's assumptions....

Response:
We thank the reviewer for helping us to get this right, however we disagree in this point. In our view, this behaviour could easily be explained by reinforcement learning instead of an understanding of the task. The parrots quickly learn to avoid something that is not rewarding for themselves (null cup) and they preferred to choose an option that provides more rewards (i.e., two pieces under sharing cup). In order to demonstrate an understanding of the task, the bird(s) would have needed to be tested in a non-social condition, in which no partner would be present to accept food rewards. If they would have switched their preference from sharing to selfish, this would serve as an indication for the birds' understanding of the task.
Concerning the suggestive comment about preventing embarrassment later, we are actually fairly confident that given the published papers on primates (de Waal et al. 2008, Suchak et al. 2012) which we wanted to compare our data against, our manuscript would not be torn apart if published in the current form and we are not aware of any further inconsistencies that reviewer 2 mentions but fails to specify.
Concerning a stronger focus on the individual data, we counter that any reader who is interested in the individual data which we provide in the online supplementary information, is free to look at them and draw their own conclusions (or conduct meta-analyses together with other larger datasets), yet we do not feel that it is productive to dwell on individual data in order to draw speculative conclusions.
Concerning the comment about the supposed confusion of the bird; it is the overarching dilemma of animal cognition studies, that we do not know how the test animals under investigation perceive a test situation. Most studies aim at designing their studies as ethologically valid and salient for the species under investigation as possible, but often it turns out retrospectively that unintended variables affected the animals in the test situation. This is a common pitfall of comparative cognitive studies in particular where one strives for maximum comparability by keeping the experimental protocols as similar as possible. In our case, we aimed at comparing the performance of parrots to that of monkeys, and therefore implemented as few adaptations to the avian model as possible. We acknowledge and discuss that our birds seemed not to have fully grasped the contingencies of the controls. Yet, it is not the case, as reviewer 2 seems to suggests, that one could have predicted that the control would have been too confusing for the birds. On the contrary, we are still convinced that the rationale behind the non-social control was a reasonable one and might work in other parrot species or with more familiarity with the physical setup.

Reviewer: 1
Comments to the Author(s) I sincerely thank the authors for their thoughtful responses to my earlier concerns. At this point I have reviewed the revised manuscripts and have no further questions or concerns to raise.

Response:
We thank the reviewer for this feedback.