Eurasian jays, Garrulus glandarius, are members of the crow family. These large-brained birds hide food when it is abundant, and eat it later, when it is scarce. Previous studies have found that jays avoid theft by other jays by carefully deciding what food to hide, and where. In one study, they preferred to hide their food behind an opaque barrier, rather than a transparent one, when another jay was watching. In a second study, they preferred to hide food that the watching jay had already eaten enough of, and thus did not want.

These studies suggest that jays have flexible cognitive skills when it comes to protecting their food. They respond to whether a potential thief can see their hiding place and to how much a thief might want the food they are stashing. The next question is, can Eurasian jays combine these two pieces of information? For example, if a jay has two types of food they could hide when another jay is present, but only has one place to hide them (either in view or out-of-view of the other jay), does the first jay prefer to stash the food that the second jay has already eaten, and therefore does not want anymore, only when the hiding place is visible to second jay?

To find out, Amodio et al. watched Eurasian jays hiding macadamia nuts or peanuts in the presence of another jay. In the first setup, jays were given one food to hide and two possible hiding places, one opaque and one transparent, while being watched by a jay that had either had its fill of the food, or not tried it. In the second setup, jays were given both foods to hide, but only had one place to hide them (either transparent or opaque); while being watched by a jay that had eaten enough of one of the foods.

Contrary to expectations, the jays did not seem to be able to combine the information about what the other jay could see and what it had been eating. In fact, they seemed unable to respond to either piece of information. When Amodio et al. repeated the original experiments, the jays did not seem to prefer to hide food out of sight, or to hide food that the watcher had already eaten.

These results raise questions about the repeatability of experiments on food hiding strategies in birds of the crow family. It suggests that previous findings should be further investigated, potentially to identify important factors that might affect the repeatability of food-hiding tactics. Repeating the experiments may show how best to investigate behavioural patterns in jays in the future.

Theory of mind is thought to provide a causal and flexible cognitive framework that allows humans to navigate complex social interactions (FeldmanHall and Shenhav, 2019; Tamir and Thornton, 2018). Through this framework, different observable social cues can be used to infer otherwise imperceptible mental states (e.g., perspectives, desires, knowledge, beliefs), such that the behaviour of other individuals can be interpreted, predicted, and manipulated based on an interplay of different mental states (Baker et al., 2017; Bartsch and Wellman, 1989; Bennett and Galpert, 1992). In humans, theory of mind is thought to emerge as a stepwise process, with a meta-representational framework in place by at least 5 years of age (Wellman, 2018; Wellman and Liu, 2004). However, even before they develop this meta-representational theory of mind, infants can already respond to multiple pieces of social information in an integrated manner (e.g., Moses et al., 2001; Repacholi et al., 2014). Multiple accounts exist, arguing that they may do so with or without representing others’ mental states as such (Apperly and Butterfill, 2009; Carruthers, 2020; Scott and Baillargeon, 2017; Southgate, 2019). At a minimum, though, infants seem to implicitly register social cues and exhibit adaptive responses accordingly (Apperly and Butterfill, 2009; Butterfill and Apperly, 2013). Consequently, different mechanisms may allow individuals to integrate information from multiple social cues.

Inspired by Premack and Woodruff, 1978, the past few decades have seen growing efforts to investigate whether non-human animals also possess something akin to theory of mind (e.g., primates: Buttelmann et al., 2007; Drayton and Santos, 2014; Flombaum and Santos, 2005; Hare et al., 2000; Hare et al., 2001; Kano et al., 2019; Krupenye and Call, 2019; dogs: Horowitz, 2011; Maginnity and Grace, 2014; corvids: Bugnyar et al., 2016; Dally et al., 2006; Emery and Clayton, 2001; Ostojić et al., 2013; Shaw and Clayton, 2012). This issue is significant, given foundational debates in cognitive science about whether minds can represent mental state concepts in the absence of language as well as a long-standing interest in the evolution of the key cognitive traits that make us human (e.g., Penn et al., 2008). However, within this line of research, studies most often focus on testing whether a given species/group has the ability to attribute one specific type of mental state, that is, exclusively focusing on belief, or exclusively focusing on desire, within a single study. As a result, very little is known about whether other species can – like humans (e.g., Baker et al., 2017) – integrate multiple social cues that correlate with others’ mental states and exhibit appropriate responses accordingly. This question is far from trivial: in real life scenarios, an individual’s behaviour is likely to result from the interplay of multiple factors (e.g., their perspective, desires, and previous knowledge) that can be indirectly perceived simultaneously by another agent during social interactions. Therefore, by focusing on non-human animals’ ability to respond to a single social cue at a time, comparative psychologists may overlook a crucial aspect of social cognitive complexity. Additionally, investigating whether and how other species may integrate different social cues can help shift the focus of social cognition studies in animals away from binary questions – for example does species X understand false beliefs? – towards more process-oriented and nuanced research questions, such as: what are the relative contributions of mechanisms A and B to how species X perform behaviour Y? (Buckner, 2013; Heyes, 2015). In doing so, this work may also identify mechanisms of varying complexity that operate in the absence of language and may feed into uniquely human social cognition.

Corvids are a group of large-brained birds that are hypothesised to have evolved sophisticated cognitive abilities independently from primates (Clayton and Emery, 2015; Emery and Clayton, 2004; Güntürkün and Bugnyar, 2016; Osvath et al., 2014; Seed et al., 2009). This group represents a good model for this line of research because they might be capable of responding – independently – to social cues correlating with different types of mental states (perspective: Bugnyar et al., 2016; Dally et al., 2004; Dally et al., 2005; Legg et al., 2016; Legg and Clayton, 2014; Shaw and Clayton, 2013; Stulp et al., 2009; desires: Ostojić et al., 2013, Ostojić et al., 2016, Ostojić et al., 2017Apperly and Butterfill, 2009; knowledge: Bugnyar and Heinrich, 2005; Dally et al., 2006; Emery and Clayton, 2001). In particular, previous research has reported that Eurasian jays (Garrulus glandarius) may be able to adjust their behaviour according to cues that correlate with the perspective and current desire of a conspecific. In the presence of a conspecific competitor, these jays have been reported to preferentially cache food in less visible locations (e.g., behind barriers, at distance), or in non-noisy substrates, which has been interpreted as a potential response to the visual (Legg et al., 2016; Legg and Clayton, 2014) or acoustic perspective (Shaw and Clayton, 2013) of a potential pilferer. In parallel, research investigating food-sharing behaviour found that male Eurasian jays change the type of food shared with their female partner, depending on which food the female has been sated on, and therefore which she desires (Ostojić et al., 2013; Ostojić et al., 2014; Ostojić et al., 2016). In a recent study, a similar response to another’s satiety has also been reported in the context of caching, whereby Eurasian jays and Western scrub-jays preferentially cached food that an observer, and thus potential pilferer, was sated on Ostojić et al., 2017. Notably, the effect reported in this study was unlikely to be based on theory of mind because the caching jays showed this effect also when they did not know what food the observer was pre-fed on and when the only cue available was the observer’s behaviour during the caching event itself (Ostojić et al., 2017). Thus, taken together, these studies seem to indicate that Eurasian jays employ a variety of cache protection strategies to limit the risk of cache loss, by responding to cues correlating with the perspective or current desire of a potential pilferer. Although this evidence is not sufficient to demonstrate mental state attribution, nor to pinpoint the exact underlying cognitive mechanism, it does suggest that corvids are capable of behaving flexibly on the basis of different types of social cues.

The aim of the current study was to investigate whether Eurasian jays can integrate multiple cues that correlate with different types of mental states to solve social problems. Previous work has to some extent already, at least implicitly, tested whether animals can integrate perspective cues from different sensory modalities (corvids: Shaw and Clayton, 2013; Stulp et al., 2009; primates: Bray et al., 2014; Santos et al., 2006). While these studies investigate the ability to integrate different cues regarding the same type of mental state, namely another individual’s perspective in assessing the significance of visual and acoustic perceptual cues, our study concerns the integration of cues correlating with different types of mental states, namely perspectives and desires. Building on previous studies in the caching context, we tested whether jays can integrate information about a conspecific’s perspective and current desire to selectively protect those caches that are at most risk of being pilfered. To this end, we first conducted two experiments in which we manipulated the observer’s visual access to caching locations and its current desire for different foods. We measured the caching pattern of the birds when viewed by this observer.

Our manipulation of the observer’s perspective followed the procedure by Legg and Clayton, 2014. The authors gave Eurasian jay cachers access to two locations, one that could and one that could not be seen from an adjacent compartment and thus by an observer bird (when present). One caching tray was positioned behind an opaque barrier (out-of-view tray), and the other behind a transparent barrier (in-view tray). Legg and Clayton’s experiment encompassed three conditions: jays could cache when an observer bird was housed in the adjacent compartment – the observer was either a higher ranked individual (Observed by a dominant condition) or a lower ranked individual (Observed by a subordinate condition) – or when no conspecific was present (Private condition). The authors compared the jays’ caching pattern in the Private condition with that in the Observed conditions (the two observed conditions were merged together), and found that jays cached a higher proportion of food items in the out-of-view tray in the Observed than in the Private conditions.

Our manipulation of the observer’s desire for different foods followed the procedure by Ostojić et al., 2017. The authors investigated whether Eurasian jays and California scrub-jays can protect their caches by preferentially caching the type of food that the observer was not currently motivated to pilfer. A cacher and an observer jay were housed in adjacent testing compartments. In the pre-feeding phase, the observer could feed to satiety on a specific type of food: maintenance diet in the baseline trial, and either food A or B in the two test trials. This procedure subsequently reduces the individual’s motivation for eating and caching that specific food (but not different kinds of food), a phenomenon known as ‘specific satiety’ (Balleine and Dickinson, 1998; Clayton and Dickinson, 1999; Dickinson and Balleine, 1994). Ostojić et al., 2017 found that the jays’ preference for caching food A over food B was larger after the observer was sated on food A than after the observer was sated on food B. Interestingly, this pattern was exhibited not only in the Seen Condition – when the cacher had witnessed which particular food had been provided to the observer – but also in the Unseen Condition – when the cacher had not seen the pre-feeding of the observer – thus indicating that the observer’s behaviour at the time of caching may have played a key role in the decision-making process of the cacher. Note that, when data were re-analysed for Eurasian jays only, the effect was still statistically significant in the Seen condition – although it was not statistically significant in the Unseen condition (Crosby, 2019).

In the current study we combined these two protocols. In Experiment 1, cacher jays were provided with only one type of food – which on one trial was the same food on which the observer sated, and on another trial was a different food from that on which the observer was sated – and two caching trays, one that the observer could see and one that the observer could not see. Thus, here, the jays could choose between two caching locations, one that was in-view and one that was out-of-view of the conspecific, allowing them to selectively cache food out-of-view when it was desired by the observer. In Experiment 2, cacher jays were provided with a single caching tray – which on one trial could be seen by the observer and on another trial could not be seen by the observer – and two types of food, one of which had previously been pre-fed to the observer. Thus, here, the jays could choose between two food types, one on which the observer was sated and one on which the observer was not sated, allowing them to selectively cache the less desired food when the observer could see them. Consequently, the designs of the two experiments were complementary, such that jays could most effectively protect their caches by deciding where to cache in Experiment 1, and what to cache in Experiment 2.

In Experiments 1 and 2, we investigated whether Eurasian jays can take into account two types of social cues simultaneously and perform the most advantageous behavioural output accordingly. Specifically, we tested whether caching birds can integrate information from cues correlating with a conspecific observer’s desire and perspective to most effectively protect their caches. Consistently across these two experiments, we did not detect effects that would support such integration of information from different cues. In Experiment 1, jays did not show a higher preference for caching in the out-of-view tray when they could cache a food that was highly desired by the observer relative to when they could cache a food that was less desired by the observer. In addition, in the former case (i.e., Different Food condition), jays did not cache more in the out-of-view tray than expected by chance. In Experiment 2, jays did not show a higher preference for caching the food for which the observer had a decreased desire when the observer could see them relative to when the observer could not see them. In addition, in the former case (i.e., In-view condition), jays did not cache the food for which the observer had a decreased desire more than the food for which the observer had a higher desire.

The negative results we obtained in both experiments appear inconsistent with previous effects in the literature, despite the use of set-ups that were very similar to those used in the original studies. Specifically, the negative results in the Different Food condition in Experiment 1 appear incompatible with the effect reported in Legg and Clayton, 2014, where jays were found to preferentially cache in an out-of-view tray specifically when they were observed by a conspecific. Similarly, the negative results in the In-view condition in Experiment 2 appear incompatible with the effect reported in Ostojić et al., 2017, where jays were found to preferentially cache a specific food when the observer was pre-fed on that food relative to when the observer was pre-fed on a different food.

Thus, we conducted three follow-up experiments to explore the reliability of the two previous findings (Legg and Clayton, 2014; Ostojić et al., 2017) that our first two experiments were built on. In Experiments 3 and 4, we attempted to replicate the effect reported by Legg and Clayton, 2014, but – in contrast to the original study – no statistically significant difference between the experimental conditions was detected. Similarly, in Experiment 5 no statistically significant difference was detected between the experimental conditions, a result that contrasts with the effect reported by Ostojić et al., 2017. Thus, Experiments 3–5 also yielded negative results. However, evaluating the ‘success’ of a replication study from the statistical significance of a finding alone is overly simplistic, particularly for comparative cognition research, where – like in our experiment – the sample size of replication studies are often as equally small as that of the original studies (Farrar et al., 2020). Nevertheless, the finding that we could not detect any significant effect in line with the original experiments of Legg and Clayton, 2014 and Ostojić et al., 2017 across all five of our experiments was surprising, especially given that they were conducted in the same lab, with many of the same birds and experimenters. Specifically, in four of the seven tests of the hypothesis that the jays could use social cues to protect their caches, the results were not in the direction of the prediction: Experiment 1, prediction 1; Experiment 2, predictions 1 and 2; Experiment 5, prediction 2, No-barrier condition. In the remaining three tests in which we had a directional prediction – Experiments 3 and 4, and Experiment 5 prediction 2, Barrier condition – the effects were in the correct direction but were non-significant and much smaller than the previously reported effects.

It is not possible to provide a single, clear answer for why our study was unable to detect effects that are consistent with the previous literature. We propose two explanations that could have played a role, namely low power and our re-use of a unique sample of birds whose behaviour may change across time. First, each of our experiments likely had low power to detect theoretically meaningful, and perhaps the most theoretically plausible, effect sizes. This means that even if the jays were responding to the perspective and/or desires of the observer during our experiments, the effect size might have just been too small for us to detect. However, that we failed to find any significant results across five experiments with similar designs to the original studies shows that no extremely large effect sizes were present in our jays and suggests that the original effect sizes might be overestimated – which would be expected if the original studies came from a series of underpowered research combined with a publication bias (Farrar et al., 2020; Fiedler and Prager, 2018; Hedges, 1984).

Second, our five studies used the same populations of birds as tested in the previous studies (Legg and Clayton, 2014; Ostojić et al., 2017; Table 1), but they were around 5 years older. It is possible that the behaviour of these birds has changed over time, for example as a result of aging and/or learning effects. That is, the original studies might have reported effects without much overestimation, but when we tested the jays for these experiments, they may no longer have had the motivation to protect their caches. The overall number of cached items (as a proxy for cache motivation) does not appear too dissimilar in the current experiments and the previous studies (see Figure 5 for a comparison of the number of items cached per bird per trial across all of the experiments). Thus, any differences relating to motivation of the birds would have to be specific to cache protection, rather than just caching.

Figure 5

Download asset Open asset

With regard to the potential motivational changes due to ageing, it is possible that ontogenetic variations in sociality could influence the performance in caching tasks. Although data on corvids specifically are limited, research in primates has shown that motivations potentially relevant to cache protection, such as those driving socialisation and social influence, change dramatically throughout life, declining substantially in some taxa (Machanda and Rosati, 2020). A notable example of this process is provided by rhesus macaques, which, similarly to humans, show a reduced propensity to follow gaze in older age (Rosati et al., 2016). When it comes to corvids, the only data at the moment concern age-related shifts in non-social predisposition, such as neophobia (Greggor et al., 2020). Thus, age-related changes in the jays’ motivation to attend to others’ social cues are speculative, but in light of the negative results reported here, may present a line of enquiry for future studies.

With regard to potential changes due to prior experience and learning, previous studies with Western scrub-jays suggested that the caching behaviour itself is underpinned by a motivational system that acts as a compulsion to cache and that is not sensitive to reinforcement or other external conditions and that a second motivational system influences decision making during caching through sensitivity to conditions at cache retrieval (Clayton et al., 2005; de Kort et al., 2007). In our laboratory, caching experiments are conducted such that an observer is located in an adjacent compartment and cannot access the conspecific’s caches, and in retrieval sessions the jays also receive feedback that their caches are unaffected by conditions at caching. If the jays’ cognitive system tracks conditions at retrieval, then all experiments in which caches are unaffected, i.e., when jays regularly find them intact in the retrieval phase of a study, may favour a decrease in the birds’ motivation to employ cache-protection strategies in the next similar situation. Thus, it cannot be excluded that the very first time when jays are presented with a novel set-up, their motivation to protect caches is higher than in subsequent presentations, but at the moment this must remain a speculation until studies directly test this possibility. As such, already Experiments 1 and 2 but especially the replication attempts in Experiments 3–5 may have low validity, despite them using natural caching behaviours. This issue is likely exacerbated by the fact that our Eurasian jays regularly cache outside of testing time, such that laboratory conditions do not impair the natural intensity of caching behaviour exhibited by this species (Goodwin, 1951; Goodwin, 1986). In particular, our jays frequently save several food items (e.g., nuts, insect larvae) to take with them (e.g., store items in the crop) when released from the testing compartments to be cached in the aviary, as well as cache maintenance food (e.g., seeds, vegetables, eggs) during their normal daily activities given that their aviaries offer plenty of opportunities. This behaviour is in line with one of the experiments reported by de Kort et al., 2007 in which the scrub-jays learned to inhibit caching at one location from which caches were pilfered if another location was available immediately afterwards in which caches were left intact.

To investigate the possible influences on the jays’ motivational systems described above, one would ideally use a new sample of birds lacking an extensive experimental history with caching experiments. Unfortunately, it is unlikely that a study like the one above could be conducted in a near future on the same species of corvids because Eurasian jays are not currently housed in any other laboratory in the world, and there is no guarantee that a new colony will be acquired by our laboratory. There seem to be few other ways to test the motivation of our jays to display cache-protection behaviours because to validate their motivation to protect caches would require them to demonstrate the cache-protection behaviours that we set out to test. From negative results alone, it is unlikely that we can dissociate the possibility that the jays cannot display cache-protection behaviours from the possibility that the jays are not motivated to display them. Ultimately, we believe that these issues around the interplay between cognitive and motivational systems are not trivial and that they will need to be considered – alongside issues of low power – by researchers working in the field when addressing issues of reliability and validity of results, including replication attempts, especially given that many studies in the field of comparative cognition also rely on re-using the same samples of animals.

Our results conflict not just with the two studies that our research was built on Legg and Clayton, 2014; Ostojić et al., 2017, but also, more generally, with a larger body of literature on cache-protection strategies in corvids (Bugnyar et al., 2016; Dally et al., 2004; Dally et al., 2005; Emery and Clayton, 2001; Heinrich and Pepper, 1998) and Eurasian jays in particular (Legg et al., 2016; Shaw and Clayton, 2013). We were unable to elicit the cache-protection strategies that this literature implies are consistently observable across corvid species, including in our Eurasian jays. In addition to the two possible explanations raised above, it is important to discuss a third possibility. While it is possible our findings were only local failures to find these effects, it is also possible that the general research practices and methods that have produced the corvid social cognition literature are liable to producing unreliable findings or overestimated effect sizes. If the findings of the present study were shown to be indicative of a broader pattern in the field, it could be necessary to substantially revise our current understanding of corvid social cognition and their exhibition of particular cognitive phenomena. Crucially however, while our data clearly show a local failure to elicit previously reported effects, these data are not sufficient to draw strong conclusions beyond these local failures. This means that, despite posing a challenge to the reliability of the effects described by Legg and Clayton, 2014 and Ostojić et al., 2017, in the absence of broader negative results, our findings cannot confute the notion that Eurasian jays are capable of employing cache-protection strategies by responding to cues about the visual perspective and current desire of a conspecific. Furthermore, in the lack of additional negative findings in other corvids or about different cognitive phenomena in the Eurasian jays, it is unclear to what extent our findings allow us to make inferences beyond the species and focus of this study. For instance, the fact that we could not find support that Eurasian jays respond to cues about others’ current mental states in the caching context does not necessarily challenge the reliability of the evidence about similar behaviours in other contexts. For example, in the cooperative context of food sharing, the effect whereby males adjust the pattern of food shared with the female to her specific satiety (Ostojić et al., 2013) has been reliably shown by the same males in subsequent studies (Legg, 2015; Ostojić et al., 2014; Ostojić et al., 2016). One difference between the cache-protection and the food-sharing context may be the motivation of the jays to reliably exhibit the behaviour in question. Food sharing is an integral part of jays’ courtship behaviour and is important not only in establishing but also maintaining the pair bond (Goodwin, 1951). Thus, males are likely to be motivated to respond to what the female likes or wants to eat across different time points. In contrast, as discussed earlier, after jays’ experience that their caches are not affected by any perceived risks, the motivation to employ cache-protection strategies in a very similar context may decrease.

We currently do not know how many other studies have produced negative cache-protection results but have not been published, and understanding the magnitude of the publication bias (Fanelli, 2012; Scheel et al., 2020) in this literature is therefore a necessary step to evaluating the evidential strength within the field. Concerning our failed replication of Ostojić et al., 2017, a slightly different reasoning applies as, to our knowledge, no similar study has been conducted in another laboratory. As such, we believe we have access to all the data on this topic. These are the current study, the Ostojić et al., 2017 study, and a further, unpublished, replication attempt that also did not detect the originally reported effects (Crosby, 2019). Overall, the data on these effects seem too uncertain to draw any firm conclusions about Eurasian jay cognition.

Our difficulty with replicating previous research, even in the same laboratory as the original findings and with many of the same birds and experimenters, highlights two ways in which research on corvid social cognition – and likely, on comparative cognition research more broadly – could make progress. First, understanding the extent of publication bias in our literature is key to understanding their evidential value. Retrospectively, this may be achieved through meta-analysis techniques, and prospectively through effective pre-registering of hypothesis-testing research. Second, before building on findings, researchers can, where appropriate, include reliability tests into their research programmes. For example, replication and extension studies, conducted in a two-step process, such as have been traditionally employed in comparative cognition (Beran, 2018), present a useful procedure. First, the replication stage of the study tests whether an effect previously reported in the literature can be elicited under the circumstances (e.g., specific sample of individuals, testing facility or methodological procedure) in which the novel experimental question will be investigated. Only as a second step, the main study, the extension part, is performed. Such reliability tests may be especially important for previous findings where the effects of publication bias are unknown, as well as when the same animals are tested in follow-up tests (as is the case in this study) to probe the reliability of the behavioural patterns over time. To this end, particular efforts should be put into the designing of experiments, for instance by favouring within-subject designs as a way to account for individual-level variation, and by increasing the power of analyses through designs encompassing a larger number of test trials. We believe that these approaches are necessary to strengthen or reshape our understanding of corvid cognition. For instance, if subsequent evidence in Eurasian jays backs up the results of the current study, then belief that these corvids take into account information about the perspective or desire of a conspecific to protect their caches would diminish. However, because of the low likelihood of independent replication of many studies in comparative cognition, it is questionable whether it is possible to produce data that convincingly refute previous claims (Boyle, 2021). This situation in which some important and influential claims in the literature cannot be directly assessed presents a substantial challenge for comparative cognition research and warrants special and careful consideration and discussion within the field.

In conclusion, the current study presents five experiments that are inconsistent with the previous literature on caching in Eurasian jays. Across all experiments, the effects were non-significant, and often in the opposite direction to predictions derived from the published literature. This suggests that previous effect sizes are likely overestimated, or at the very least, that the effects cannot be consistently elicited in the same or similar samples of birds. Therefore, future studies should consider this new evidence when referring to the original studies for subsequent inferences about corvid social cognition. In Experiments 1 and 2, we investigated a follow-up question that assumed the reliability of previously reported statistical effects, which we later could not replicate in Experiments 3, 4, or 5. The current series of experiments demonstrate the necessity to investigate the uncertainty of such effects and to adjust the claims – including those in previously published literature – accordingly. In regard to the behavioural effects investigated in this study, the caching patterns interpreted as cache-protection strategies in Legg and Clayton, 2014 and Ostojić et al., 2017 do not seem to be reliable enough to form the basis for follow-up studies such as the ones reported in Experiments 1 and 2, at least in our sample of jays. It would be informative, but unfortunately not currently possible, to replicate these studies at other laboratories across the world.

Data and analyses of all experiments are available at http://doi.org/10.5281/zenodo.4636561.

1. 1. Amodio P
   2. Farrar BG
   3. Krupenye C
   4. Ostojic L
   5. Clayton NS

   (2021) Zenodo

   Little evidence that Eurasian jays (Garrulus glandarius) protect their caches by responding to cues about a conspecific's desire and visual perspective.

   https://doi.org/10.5281/zenodo.4636561

1. 1. Apperly IA
   2. Butterfill SA

   (2009) Do humans have two systems to track beliefs and belief-like states

   Psychological Review 116:953–970.

   https://doi.org/10.1037/a0016923

   • PubMed
   • Google Scholar
2. 1. Baker CL
   2. Jara-Ettinger J
   3. Saxe R
   4. Tenenbaum JB

   (2017) Rational quantitative attribution of beliefs, desires and percepts in human mentalizing

   Nature Human Behaviour 1:0064.

   https://doi.org/10.1038/s41562-017-0064

   • Google Scholar
3. 1. Balleine BW
   2. Dickinson A

   (1998) The role of incentive learning in instrumental outcome revaluation by sensory-specific satiety

   Animal Learning & Behavior 26:46–59.

   https://doi.org/10.3758/BF03199161

   • Google Scholar
4. 1. Bartsch K
   2. Wellman H

   (1989) Young Children’s Attribution of Action to Beliefs and Desires

   Child Development 60:946–964.

   https://doi.org/10.2307/1131035

   • PubMed
   • Google Scholar
5. 1. Bennett M
   2. Galpert L

   (1992) Complex belief-desire reasoning in children

   Social Development 1:201–210.

   https://doi.org/10.1111/j.1467-9507.1992.tb00124.x

   • Google Scholar
6. 1. Beran MJ

   (2018) Replication and Pre-Registration in Comparative Psychology

   International Journal of Comparative Psychology 31:8654.

   https://doi.org/10.5070/P4311038654

   • Google Scholar
7. 1. Boyle A

   (2021) Replication, Uncertainty and Progress in Comparative Cognition

   Animal Behavior and Cognition 8:296–304.

   https://doi.org/10.26451/abc.08.02.15.2021

   • Google Scholar
8. 1. Bray J
   2. Krupenye C
   3. Hare B

   (2014) Ring-tailed lemurs (Lemur catta) exploit information about what others can see but not what they can hear

   Animal Cognition 17:735–744.

   https://doi.org/10.1007/s10071-013-0705-0

   • PubMed
   • Google Scholar
9. 1. Buckner C

   (2013) Morgan’s Canon, meet Hume’s Dictum: Avoiding anthropofabulation in cross-species comparisons

   Biology and Philosophy 28:853–871.

   https://doi.org/10.1007/s10539-013-9376-0

   • Google Scholar
10. 1. Bugnyar T
    2. Heinrich B

    (2005) Ravens, Corvus corax, differentiate between knowledgeable and ignorant competitors

    Proceedings. Biological Sciences 272:1641–1646.

    https://doi.org/10.1098/rspb.2005.3144

    • PubMed
    • Google Scholar
11. 1. Bugnyar T
    2. Reber SA
    3. Buckner C

    (2016) Ravens attribute visual access to unseen competitors

    Nature Communications 7:10506.

    https://doi.org/10.1038/ncomms10506

    • PubMed
    • Google Scholar
12. 1. Buttelmann D
    2. Carpenter M
    3. Call J
    4. Tomasello M

    (2007) Enculturated chimpanzees imitate rationally

    Developmental Science 10:F31–F38.

    https://doi.org/10.1111/j.1467-7687.2007.00630.x

    • Google Scholar
13. 1. Butterfill SA
    2. Apperly IA

    (2013) How to Construct a Minimal Theory of Mind

    Mind & Language 28:606–637.

    https://doi.org/10.1111/mila.12036

    • Google Scholar
14. 1. Carruthers P

    (2020) Representing the Mind as Such in Infancy

    Review of Philosophy and Psychology 11:765–781.

    https://doi.org/10.1007/s13164-020-00491-9

    • Google Scholar
15. 1. Clayton NS
    2. Dickinson A

    (1999) Motivational control of caching behaviour in the scrub jay, Aphelocoma coerulescens

    Animal Behaviour 57:435–444.

    https://doi.org/10.1006/anbe.1998.0989

    • PubMed
    • Google Scholar
16. 1. Clayton NS
    2. Dally J
    3. Gilbert J
    4. Dickinson A

    (2005) Food caching by Western scrub-jays (Aphelocoma californica) is sensitive to the conditions at recovery

    Journal of Experimental Psychology 31:115–124.

    https://doi.org/10.1037/0097-7403.31.2.115

    • Google Scholar
17. 1. Clayton NS
    2. Emery NJ

    (2015) Avian Models for Human Cognitive Neuroscience: A Proposal

    Neuron 86:1330–1342.

    https://doi.org/10.1016/J.NEURON.2015.04.024

    • PubMed
    • Google Scholar
18. Thesis

    1. Crosby RM

    (2019)

    A Comparative Investigation of the Attribution of Desires and Preferences

    PhD Thesis. University of Cambridge, UK.

    • Google Scholar
19. 1. Dally JM
    2. Emery NJ
    3. Clayton NS

    (2004) Cache protection strategies by western scrub-jays (Aphelocoma californica): hiding food in the shade

    Proceedings. Biological Sciences 271:S387–S390.

    https://doi.org/10.1098/rsbl.2004.0190

    • PubMed
    • Google Scholar
20. 1. Dally JM
    2. Emery NJ
    3. Clayton NS

    (2005) Cache protection strategies by western scrub-jays, Aphelocoma californica: implications for social cognition

    Animal Behaviour 70:1251–1263.

    https://doi.org/10.1016/J.ANBEHAV.2005.02.009

    • Google Scholar
21. 1. Dally JM
    2. Emery NJ
    3. Clayton NS

    (2006) Food-caching western scrub-jays keep track of who was watching when

    Science 312:1662–1665.

    https://doi.org/10.1126/science.1126539

    • PubMed
    • Google Scholar
22. 1. de Kort SR
    2. Correia SPC
    3. Alexis DM
    4. Dickinson A
    5. Clayton NS

    (2007) The control of food-caching behavior by Western scrub-jays (Aphelocoma californica

    Journal of Experimental Psychology 33:361–370.

    https://doi.org/10.1037/0097-7403.33.4.361

    • Google Scholar
23. 1. Dickinson A
    2. Balleine B

    (1994) Motivational control of goal-directed action

    Animal Learning & Behavior 22:1–18.

    https://doi.org/10.3758/BF03199951

    • Google Scholar
24. 1. Drayton LA
    2. Santos LR

    (2014) Capuchins’ (Cebus apella) sensitivity to others’ goal-directed actions in a helping context

    Animal Cognition 17:689–700.

    https://doi.org/10.1007/s10071-013-0700-5

    • PubMed
    • Google Scholar
25. 1. Emery NJ
    2. Clayton NS

    (2001) Effects of experience and social context on prospective caching strategies by scrub jays

    Nature 414:443–446.

    https://doi.org/10.1038/35106560

    • PubMed
    • Google Scholar
26. 1. Emery NJ
    2. Clayton NS

    (2004) The mentality of crows: convergent evolution of intelligence in corvids and apes

    Science 306:1903–1907.

    https://doi.org/10.1126/science.1098410

    • PubMed
    • Google Scholar
27. 1. Fanelli D

    (2012) Negative results are disappearing from most disciplines and countries

    Scientometrics 90:891–904.

    https://doi.org/10.1007/s11192-011-0494-7

    • Google Scholar
28. Preprint

    1. Farrar B
    2. Ostojic L

    (2019) The Illusion of Science in Comparative Cognition

    PsyArXiv.

    https://doi.org/10.31234/osf.io/hduyx

    • Google Scholar
29. 1. Farrar B
    2. Boeckle M
    3. Clayton N

    (2020) Replications in Comparative Cognition: What Should We Expect and How Can We Improve?

    Animal Behavior and Cognition 7:1–22.

    https://doi.org/10.26451/abc.07.01.02.2020

    • PubMed
    • Google Scholar
30. 1. FeldmanHall O
    2. Shenhav A

    (2019) Resolving uncertainty in a social world

    Nature Human Behaviour 3:426–435.

    https://doi.org/10.1038/s41562-019-0590-x

    • PubMed
    • Google Scholar
31. 1. Fiedler K
    2. Prager J

    (2018) The Regression Trap and Other Pitfalls of Replication Science—Illustrated by the Report of the Open Science Collaboration

    Basic and Applied Social Psychology 40:115–124.

    https://doi.org/10.1080/01973533.2017.1421953

    • Google Scholar
32. 1. Flombaum J
    2. Santos LR

    (2005) Rhesus Monkeys Attribute Perceptions to Others

    Current Biology 15:447–452.

    https://doi.org/10.1016/J.CUB.2004.12.076

    • PubMed
    • Google Scholar
33. 1. Goodwin D

    (1951)

    Some aspects of the behaviour of the jay Garrulus glandarius

    Ibis 93:414–442.

    • Google Scholar
34. Book

    1. Goodwin D

    (1986)

    Crows of the World

    British Museum.

    • Google Scholar
35. 1. Greggor AL
    2. Masuda B
    3. Flanagan AM
    4. Swaisgood RR

    (2020) Age-related patterns of neophobia in an endangered island crow: implications for conservation and natural history

    Animal Behaviour 160:61–68.

    https://doi.org/10.1016/j.anbehav.2019.12.002

    • Google Scholar
36. 1. Güntürkün O
    2. Bugnyar T

    (2016) Cognition without Cortex

    Trends in Cognitive Sciences 20:291–303.

    https://doi.org/10.1016/J.TICS.2016.02.001

    • PubMed
    • Google Scholar
37. 1. Hare B
    2. Call J
    3. Agnetta B
    4. Tomasello M

    (2000) Chimpanzees know what conspecifics do and do not see

    Animal Behaviour 59:771–785.

    https://doi.org/10.1006/ANBE.1999.1377

    • PubMed
    • Google Scholar
38. 1. Hare B
    2. Call J
    3. Tomasello M

    (2001) Do chimpanzees know what conspecifics know?

    Animal Behaviour 61:139–151.

    https://doi.org/10.1006/ANBE.2000.1518

    • PubMed
    • Google Scholar
39. 1. Hedges L

    (1984) Estimation of Effect Size under Nonrandom Sampling: The Effects of Censoring Studies Yielding Statistically Insignificant Mean Differences

    Journal of Educational Statistics 9:61–85.

    https://doi.org/10.3102/10769986009001061

    • Google Scholar
40. 1. Heinrich B
    2. Pepper JW

    (1998) Influence of competitors on caching behaviour in the common raven, Corvus corax

    Animal Behaviour 56:1083–1090.

    https://doi.org/10.1006/anbe.1998.0906

    • PubMed
    • Google Scholar
41. 1. Heyes C

    (2015) Animal mindreading: what’s the problem

    Psychonomic Bulletin & Review 22:313–327.

    https://doi.org/10.3758/s13423-014-0704-4

    • Google Scholar
42. 1. Horowitz A

    (2011) Theory of mind in dogs? Examining method and concept

    Learning & Behavior 39:314–317.

    https://doi.org/10.3758/s13420-011-0041-7

    • Google Scholar
43. 1. Hothorn T
    2. Hornik K
    3. van de Wiel MA
    4. Zeileis A

    (2006) A Lego System for Conditional Inference

    The American Statistician 60:257–263.

    https://doi.org/10.1198/000313006X118430

    • Google Scholar
44. 1. Kano F
    2. Krupenye C
    3. Hirata S
    4. Tomonaga M
    5. Call J

    (2019) Great apes use self-experience to anticipate an agent’s action in a false-belief test

    PNAS 116:20904–20909.

    https://doi.org/10.1073/pnas.1910095116

    • PubMed
    • Google Scholar
45. 1. Krupenye C
    2. Call J

    (2019) Theory of mind in animals: Current and future directions

    Wiley Interdisciplinary Reviews 10:e1503.

    https://doi.org/10.1002/wcs.1503

    • Google Scholar
46. 1. Legg EW
    2. Clayton NS

    (2014) Eurasian jays (Garrulus glandarius) conceal caches from onlookers

    Animal Cognition 17:1223–1226.

    https://doi.org/10.1007/s10071-014-0743-2

    • PubMed
    • Google Scholar
47. Book

    1. Legg EW

    (2015)

    A Comparative Approach to Testing Mental State Attribution

    PhD Thesis. University of Cambridge, UK.

    • Google Scholar
48. 1. Legg EW
    2. Ostojić L
    3. Clayton NS

    (2016) Caching at a distance: a cache protection strategy in Eurasian jays

    Animal Cognition 19:753–758.

    https://doi.org/10.1007/s10071-016-0972-7

    • PubMed
    • Google Scholar
49. 1. Machanda ZP
    2. Rosati AG

    (2020) Shifting sociality during primate ageing

    Philosophical Transactions of the Royal Society B 375:20190620.

    https://doi.org/10.1098/rstb.2019.0620

    • PubMed
    • Google Scholar
50. 1. Maginnity ME
    2. Grace RC

    (2014) Visual perspective taking by dogs (Canis familiaris) in a Guesser–Knower task: evidence for a canine theory of mind?

    Animal Cognition 17:1375–1392.

    https://doi.org/10.1007/s10071-014-0773-9

    • PubMed
    • Google Scholar
51. 1. Moses LJ
    2. Baldwin DA
    3. Rosicky JG
    4. Tidball G

    (2001) Evidence for Referential Understanding in the Emotions Domain at Twelve and Eighteen Months

    Child Development 72:718–735.

    https://doi.org/10.1111/1467-8624.00311

    • PubMed
    • Google Scholar
52. 1. Ostojić L
    2. Shaw RC
    3. Cheke LG
    4. Clayton NS

    (2013) Evidence suggesting that desire-state attribution may govern food sharing in Eurasian jays

    PNAS 110:4123–4128.

    https://doi.org/10.1073/pnas.1209926110

    • PubMed
    • Google Scholar
53. 1. Ostojić L
    2. Legg EW
    3. Shaw RC
    4. Cheke LG
    5. Mendl M
    6. Clayton NS

    (2014) Can male Eurasian jays disengage from their own current desire to feed the female what she wants

    Biology Letters 10:20140042.

    https://doi.org/10.1098/rsbl.2014.0042

    • PubMed
    • Google Scholar
54. 1. Ostojić L
    2. Legg EW
    3. Dits A
    4. Williams N
    5. Brecht KF
    6. Mendl M
    7. Clayton NS

    (2016) Experimenter expectancy bias does not explain Eurasian jays’ (Garrulus glandarius) performance in a desire-state attribution task

    Journal of Comparative Psychology 130:407–410.

    https://doi.org/10.1037/com0000043

    • PubMed
    • Google Scholar
55. 1. Ostojić L
    2. Legg EW
    3. Brecht KF
    4. Lange F
    5. Deininger C
    6. Mendl M
    7. Clayton NS

    (2017) Current desires of conspecific observers affect cache-protection strategies in California scrub-jays and Eurasian jays

    Current Biology 27:R51–R53.

    https://doi.org/10.1016/J.CUB.2016.11.020

    • PubMed
    • Google Scholar
56. 1. Osvath M
    2. Kabadayi C
    3. Jacobs I

    (2014) Independent evolution of similar complex cognitive skills: the importance of embodied degrees of freedom

    Animal Behavior and Cognition 1:249–264.

    https://doi.org/10.12966/abc.08.03.2014

    • Google Scholar
57. 1. Penn DC
    2. Holyoak KJ
    3. Povinelli DJ

    (2008) Darwin’s mistake: Explaining the discontinuity between human and nonhuman minds

    Behavioral and Brain Sciences 31:109–130.

    https://doi.org/10.1017/S0140525X08003543

    • PubMed
    • Google Scholar
58. 1. Premack D
    2. Woodruff G

    (1978) Does the chimpanzee have a theory of mind

    Behavioral and Brain Sciences 1:515.

    https://doi.org/10.1017/S0140525X00076512

    • Google Scholar
59. Software

    1. R Studio Team

    (2018) RStudio: Integrated development environment for R

    Version 1.1.447.

    http://www.rstudio.org
60. 1. Repacholi BM
    2. Meltzoff AN
    3. Rowe H
    4. Toub TS

    (2014) Infant, control thyself: Infants’ integration of multiple social cues to regulate their imitative behavior

    Cognitive Development 32:46–57.

    https://doi.org/10.1016/J.COGDEV.2014.04.004

    • PubMed
    • Google Scholar
61. 1. Rosati AG
    2. Arre AM
    3. Platt ML
    4. Santos LR

    (2016) Rhesus monkeys show human-like changes in gaze following across the lifespan

    Proceedings of the Royal Society B 283:1830.

    https://doi.org/10.1098/rspb.2016.0376

    • PubMed
    • Google Scholar
62. 1. Santos LR
    2. Nissen AG
    3. Ferrugia JA

    (2006) Rhesus monkeys, Macaca mulatta, know what others can and cannot hear

    Animal Behaviour 71:1175–1181.

    https://doi.org/10.1016/J.ANBEHAV.2005.10.007

    • Google Scholar
63. Preprint

    1. Scheel AM
    2. Schijen M
    3. Lakens D

    (2020) An Excess of Positive Results: Comparing the Standard Psychology Literature with Registered Reports

    PsyArXiv.

    https://doi.org/10.31234/osf.io/p6e9c

    • Google Scholar
64. 1. Scott RM
    2. Baillargeon R

    (2017) Early False-Belief Understanding

    Trends in Cognitive Sciences 21:237–249.

    https://doi.org/10.1016/J.TICS.2017.01.012

    • Google Scholar
65. 1. Seed A
    2. Emery N
    3. Clayton N

    (2009) Intelligence in corvids and apes: A case of convergent evolution?

    Ethology 115:401–420.

    https://doi.org/10.1111/j.1439-0310.2009.01644.x

    • Google Scholar
66. 1. Shaw RC
    2. Clayton NS

    (2012) Eurasian jays, Garrulus glandarius, flexibly switch caching and pilfering tactics in response to social context

    Animal Behaviour 84:191–1200.

    https://doi.org/10.1016/j.anbehav.2012.08.023

    • Google Scholar
67. 1. Shaw RC
    2. Clayton NS

    (2013) Careful cachers and prying pilferers: Eurasian jays (Garrulus glandarius) limit auditory information available to competitors

    Proceedings. Biological Sciences 280:20122238.

    https://doi.org/10.1098/rspb.2012.2238

    • PubMed
    • Google Scholar
68. 1. Southgate V

    (2019) Are Infants Altercentric? The Other and the Self in Early Social Cognition

    Psychological Review 127:505–523.

    https://doi.org/10.1037/rev0000182

    • PubMed
    • Google Scholar
69. 1. Stulp G
    2. Emery NJ
    3. Verhulst S
    4. Clayton NS

    (2009) Western scrub-jays conceal auditory information when competitors can hear but cannot see

    Biology Letters 5:583–585.

    https://doi.org/10.1098/rsbl.2009.0330

    • PubMed
    • Google Scholar
70. 1. Tamir D
    2. Thornton MA

    (2018) Modeling the Predictive Social Mind

    In Trends in Cognitive Sciences 22:201–212.

    https://doi.org/10.1016/j.tics.2017.12.005

    • PubMed
    • Google Scholar
71. 1. Wellman HM
    2. Liu D

    (2004) Scaling of Theory-of-Mind Tasks

    Child Development 75:523–541.

    https://doi.org/10.1111/j.1467-8624.2004.00691.x

    • PubMed
    • Google Scholar
72. 1. Wellman HM

    (2018) Theory of mind: The state of the art

    European Journal of Developmental Psychology 15:728–755.

    https://doi.org/10.1080/17405629.2018.1435413

    • Google Scholar

Author details

1. Piero Amodio

   1. Department of Psychology, University of Cambridge, Cambridge, United Kingdom
   2. Department of Biology and Evolution of Marine Organisms, Stazione Zoologica Anton Dohrn, Napoli, Italy

   Contribution

   Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing - original draft, Writing – review and editing

   For correspondence

   piero.amodio@cantab.net

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-9408-2902

2. Benjamin G Farrar

   1. Department of Psychology, University of Cambridge, Cambridge, United Kingdom
   2. Institute for Globally Distributed Open Research and Education (IGDORE), Sweden, Sweden

   Contribution

   Conceptualization, Investigation, Methodology, Writing – review and editing

   Competing interests

   No competing interests declared

3. Christopher Krupenye

   1. Department of Psychological & Brain Sciences, Johns Hopkins University, Baltimore, United States
   2. Department of Psychology, Durham University, Durham, United Kingdom

   Contribution

   Conceptualization, Methodology, Writing – review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-2029-1872

4. Ljerka Ostojić

   1. Department of Psychology, University of Cambridge, Cambridge, United Kingdom
   2. Department of Psychology, Faculty of Humanities and Social Sciences, University of Rijeka, Rijeka, Croatia

   Contribution

   Conceptualization, Investigation, Methodology, Supervision, Writing – review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-8008-1773

5. Nicola S Clayton

   Department of Psychology, University of Cambridge, Cambridge, United Kingdom

   Contribution

   Funding acquisition, Supervision, Writing – review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-1835-423X

• 1,221

  views

• 155

  downloads

• 7

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.