Summary
In playback studies, wild dolphins respond distinctly to group members regardless of individual relationships, implying a mediating associative concept of the group. However, confirming associative concept formation in wild animals is out of our current technological reach, and theories of associative concepts do not account for reciprocal, cooperative relationships which may cause the individual to be a member of the associative concept.
Comparative psychology has shifted from finding the function that separates humans from nonhuman animals toward understanding the situational and evolutionary conditions around cognitive behavior. In these ventures, the abilities to use tools, solve problems, and generate insights are often highlighted, but equally complex are the cognitive abilities that enable us to achieve cooperative goals—communication, empathy, and social intelligence. While most animals seem to lack these abilities to a meaningful degree, dolphins process social identity from auditory cues (i.e., using signature whistle) and have evolved natural cooperative behaviors that utilize this communication.
King et al. (2021) studied dolphins’ social behavior using methods made possible by modern technology and extensive prior data collection. These authors produced a signature whistle from a known member of a target dolphin’s cooperative social network, and a drone video recorded the target dolphin’s behavioral change while an array of specialized microphones recorded any communicative responses. Critically, the authors’ extensive knowledge of the dolphins’ social behavior enabled them to produce whistles from a known closely allied individual (i.e., a frequent first-order ally), from an individual within a cooperative group (i.e., a second-order ally), or from an individual within a larger social network (i.e., a third-order ally). The authors report that across these three levels of social connection, the dolphins’ responses to the playback was most influenced by the mid-level relationship; they conclude that this is evidence for a “cooperation-based concept.”
Concept formation in animals is contentious. A concept is an abstraction that allows an agent to generalize learned knowledge to something novel. Studies of animals are unable to verify that abstraction, however, so the hallmark of concept formation in animals is accurate categorization behavior to novel exemplars. Associative concepts are therefore especially problematic: How can one generalize to a novel or unexperienced member of the concept, when the concept is fundamentally based on experience? This philosophical problem drove investigations of associative concept learning toward demonstrating that new learning with a subset of the items in the associative concept changes behavior to other items in the concept that had not received the new training (Zentall et al., 2014).
Applying these tenets of concept formation to the dolphin study raises obvious difficulties. Foremost is the typical challenge with studies in the animals’ natural ecology: controlling for experiences and providing the right class of stimulus to verify the hypothesis. These dolphins had previous experiences with the dolphins in both second-order and third-order categories (though maybe not all third-order category members). If all of the animals have been experienced, how can the transfer of learned knowledge to something novel be evaluated? Furthermore, within the dolphins’ ecology and the framework of the study, is there a place for novelty? Even in those situations where episodes of common mating were not witnessed between two individuals, bond strength between third-order allies was assessed based on observations of being in proximity with each other. The operational definition of allyship is fundamentally based on the researchers observing the dolphins experiencing each other, and therefore “truly novel” relations could not be evaluated (especially considering the possibility of unobserved interactions).
A more insidious challenge regards how to treat associative concepts during affiliative interactions. The majority of well-explored theories and experiments with associative concepts use stimuli that have limited perceptual similarity and are initially neutral for the animal. The benefits of these designs are that changes in behavior cannot be mediated by low-level factors and the learned behavior can only be explained by way of the training and learning within the experiment. In considering the design of King et al. (2021), using this lens, other dolphins within the alliance arguably have little to distinguish each other, except experiential history, and without knowledge of the experiences, they would not be recognized as second-order allies. The authors affirm that in dolphin societies, few, if any, physiological markers of individuality affect the social relationship. However, the putative social situations generated by the playback study are potential reciprocal affiliative interactions with individuals who may have once been instrumental in successful mate acquisition. During playback, the target dolphin must “decide” to orient and approach, leaving open the nature of the behavioral response involved—perhaps affiliative to second-order allies while aggressive to third-order allies? There is considerable variation and potential “purpose” in the dolphin’s response, while a pigeon’s peck or a rat’s lever press in comparable experiments is relatively unambiguous. The situation may also mediate the dolphin’s behavior. Prior to playback, the dolphin was not waiting for the next trial in the operant chamber but instead processing a larger world; how does the target dolphin weigh the signature whistle’s potential meaning against current behavior? Additionally, during a natural interaction, what communication or other noncommunicative auditory signals might follow the signature whistle?
The nature of this problem potentially extends beyond conflating a single species-specific natural response as indicative of responding to a single specific underlying construct. Fundamentally, associative concepts have not been studied in social contexts in which the critical behaviors include reciprocity. Social animals’ concepts of individuals may include a representation of the “self,” and associative concepts about individuals might generate a representation of the allyship or team. If there were an associative concept cohering the allyship together, and something changes that affects the individuals in the allyship, how “should” the individual who is a part of that concept or category change? If several dolphin allies start producing a particular sound during foraging, does an individual within the group represent their allies as producing that sound in a purely external fashion? Does the animal lack individuality and associate themself with the novel feature because their meaningful social group is engaging in it? Does the definition of the social group change in a meaningful way, now excluding the self from the representation of the group?
These questions of self-membership in social groups by nonhuman animals verge on anthropomorphism. Eliminating the human-like “self” from the investigation is possible—simply refer to the allies as the individuals with whom operant behavior can yield probabilistic reinforcement on a long delay. Secondary reinforcers of social cohesion or game-theoretic explanations may help explain the momentary investment. Discussions of social behavior in animals often avoid this discussion (Seyfarth & Cheney, 2015), though as King et al. (2021) note, those investigations focus on external relationships. Human social networks benefit from concepts of the group that is both independent of and intertwined with each individual (e.g., as in Hogg & Reid, 2006). Ignoring (and in doing so, scientifically denying) the possibility of the dolphin “self” in effect deanimates the individual, removing that which is animal and in its stead leaving a biological robot.
King et al. (2021) provide insight into dolphins’ natural behavior toward individuals in their social network; there is a potentially meaningful difference in the minds of dolphins between second-order allies and others. Perhaps these social concepts are critical to the evolution of the complex social systems, like those in modern human society. Technology and significant investment enabled the authors to investigate this question with greater accuracy than possible when the dolphins were first being studied in this area decades ago. Modern experiments can use drones to less-obtrusively monitor dolphins closely, with high-fidelity technical equipment to produce and record vocalizations. As in many areas of animal cognition, the ability to investigate more complex questions will become possible as technology evolves. For example, GPS tags with high-accuracy, low-drift accelerometers for communities of dolphins may be used to learn about complex interanimal motion-based communication (like birds in flight). As the precision of our ability to predict these animal behaviors evolves, the utility of a group-concept (and a self-concept) may become more apparent.
Toward this end, in studying animal social structures, some have questioned whether the study of social groups in animals (animal ethnography) could crucially benefit from the participant-observer methodology seen in anthropology. By generating behavior and meaningfully engaging in the animal society, theories of animal social cognition may be more effectively tested. Complex artificial social interactions may be the next frontier of studies into nonhuman animal social cognition. At present, artificial robotic dolphins have the potential to substitute for real dolphins in some entertainment venues. If made popular and accessible to the scientific community, perhaps an artificial dolphin mind can be developed, enabling the artificial dolphin to integrate as a participant-observer into wild dolphin communities—and a dolphin “Turing test” will reveal the accuracy of our knowledge of dolphin social behavior.
References
Hogg, M. A., & Reid, S. A. (2006). Social identity, self-categorization, and the communication of group norms. Communication Theory, 16(1), 7–30. https://doi.org/10.1111/j.1468-2885.2006.00003.x
King, S. L., Connor, R. C., Krützen, M., & Allen, S. J. (2021). Cooperation-based concept formation in male bottlenose dolphins. Nature Communications, 12(1), 2373. https://doi.org/10.1038/s41467-021-22668-1
Seyfarth, R. M., & Cheney, D. L. (2015). Social cognition. Animal Behaviour, 103, 191–202. https://doi.org/10.1016/j.anbehav.2015.01.030
Zentall, T. R., Wasserman, E. A., & Urcuioli, P. J. (2014). Associative concept learning in animals. Journal of the Experimental Analysis of Behavior, 101(1), 130–151. https://doi.org/10.1002/jeab.55
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Qadri, M.A.J. A putative social concept in dolphins. Learn Behav 50, 1–2 (2022). https://doi.org/10.3758/s13420-021-00502-4
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.3758/s13420-021-00502-4