More than a century ago, naturalist Henry Smith Williams observed that birds nesting in his garden differed in their colour preferences for nest-building materials. His observations challenge the assumption that individuals within the same species respond similarly to the same colour cues, but the stability and consequences of these apparent preferences remain largely unexplored. Here, we tested the colour preferences of laboratory-raised zebra finches (Taeniopygia guttata) with access to pink and orange nest-building string. Birds were tested under conditions lacking relevant environmental cues, social information, or previous experience with the experimental colours. Individuals differed in both their preferred colour and the strength of preferences, which were more strongly repeatable across short-term than long-term intervals. We next asked whether these stable individual preferences influenced how birds respond to information, reanalyzing published social-learning datasets spanning different informational contexts. Stronger initial preferences generally predicted reduced use of socially demonstrated nonpreferred colours during nest building. These findings suggest that stable individual preferences represent an underappreciated source of variation in how animals acquire and use information during decision making.
This study examined the Weibull model for individual learning curves in Pavlovian conditioning from a theoretical perspective. Although the Weibull model effectively captures variability in individual learning, it remains unclear what the parameters in this model represent. First, we identified similarities in the mathematical representations of the Weibull and error-correction models and showed that two of the three parameters in the Weibull model can be interpreted in the same manner as those in the error-correction model. Next, to explore possible psychological interpretations of the remaining parameter, α , and to examine how α is distributed across individual learning curves, we analyzed four fear-conditioning datasets obtained from both humans and animals. Although it was difficult to establish a definitive interpretation of α in this paper, we discuss possible variables that may be associated with α .
In natural environments, the outcomes of actions are often delayed and separated from the actions themselves by intervening events, posing a credit assignment problem for learning. Although both humans and nonhuman animals can associate actions with immediate subsequent outcomes, little is known about learning when action–outcome delays span multiple, unpredictably ordered events. We investigated whether chimpanzees could associate a choice with a reward delivered several trials later, using a touchscreen-based task. Six chimpanzees completed four experimental rounds involving three trial types: an immediately rewarded discrimination, a non-rewarded choice that triggered a delayed reward, and a reward-delivery trial contingent solely on prior choices. Three chimpanzees showed a significant increase in choosing the stimulus associated with delayed reward. Reversal of stimulus roles in a later round confirmed that increased selection of the delayed-reward stimulus was not attributable to fixed stimulus preferences. Although performance remained variable across individuals and sessions, the results suggest that chimpanzees possess learning mechanisms capable of bridging extended action–outcome gaps, likely operating implicitly rather than through explicit understanding of task structure. Our study contributes to comparative perspectives on the evolutionary origins of learning beyond the “here and now.”
Spatial navigation is considered essential for survival and therefore likely to have been selected for across vertebrates; however, similarities and differences in the mechanisms underlying these abilities across species remain unclear. The current experiment was designed to investigate strategies underlying spatial navigation in green frog (Rana clamitans) tadpoles. Tadpoles were trained to navigate a maze either by using external cues to navigate (place learners), or by remembering the specific egocentric actions that led to escape from the maze (response learners). There were no significant differences in acquisition following place or response learning, which suggests some flexibility in their spatial navigation strategies. However, only ten tadpoles reached the learning criterion before metamorphosis, which was fewer than that predicted by chance alone. Of these, six were place learners and four were response learners. Three place learners appeared to retain this information in the short term - pointing to the intriguing possibility that green frog tadpoles may retain this information for a period of several weeks or longer. Future studies should set a stricter learning criterion and focus on how memory is retained following metamorphosis.
Previous work has demonstrated that repeated presentations of a conditional stimulus (CS) with a weak shock following training with a strong shock can reduce later conditional responding to that CS. While this procedure shares characteristics with extinction, several converging lines of evidence suggest that weak shock exposure, with steps to minimize prediction error between acquisition and the weak shock phase, will have an advantage over extinction to persistently reduce fear behaviors. A similar experiment by Rescorla demonstrated that unconditional stimulus (US) habituation (repeated exposure to the US alone) reduced responding to a CS that previously predicted that US. Here, we apply this same logic to our weak shock procedure to determine if weak shock presentations in the absence of the CS could reduce responding to subsequent presentations of the same CS. In Experiment 1, using a within-subject procedure, we found that the weak shock-driven reductions in CS-elicited behavior did not transfer to a different CS that was not paired with a weak shock. In Experiments 2 and 3, we demonstrated that weak shock presentations in the absence of a CS (in either a novel context or the acquisition context, respectively) did not attenuate later CS-elicited freezing. Together, these results suggest that, like extinction, weak shock presentations reduce fear behavior in a stimulus-specific manner.
This study investigated how variations in the physical properties of an operant manipulandum (i.e., lever height), affect behavior under a progressive ratio (PR) schedule. Lever height was analyzed from two complementary frameworks: the ecological theory of affordances and the economic theory of utility maximization, specifically through the Modular Maximization Theory. Rats were exposed to 60 progressive ratio (PR) sessions in which lever height varied daily pseudorandomly between 1, 11, and 18 cm. Results showed that breakpoints, response rates, and parameter estimates of response capacity (w) were significantly lower at 1 cm than at 11 and 18 cm, suggesting reduced motor affordance at floor level. The parameter z, indexing the value of the reinforcer relative to leisure, was higher at 1 cm, indicating a lower cost of effort at that position. Breakpoints indicated no carryover effects from prior sessions and stabilized rapidly from the first session of each condition, suggesting that adaptation to the affordances of the lever may be a distinct process from adaptation to reinforcement contingencies. These findings extend previous work on lever affordances and support the notion that operant performance is shaped by the interaction between environmental affordances and economic trade-offs. The integration of ecological and economic perspectives provides an account of how animals allocate behavior under varying physical and motivational constraints.
The ability to categorize objects and events is a highly adaptive property common among organisms with different nervous systems. Using a bead floor learning paradigm, we systematically analyzed the parameters that enable the rapid formation of a new visual category in young domestic chicks (Gallus gallus domesticus). In one short training session, newborn chicks successfully learned to distinguish food crumbs from the unfamiliar food-like plastic beads and group the latter into a new category of inedible objects. We found that a critical prerequisite for the category acquisition was the occurrence of several non-reinforced exploratory pecks, which served as a behavioral driver of categorization. The presence of food or previous association of the learning context with food reinforcement was not required for the formation of the new "inedible object" category. Chicks were also capable of forming two categories - edible and inedible objects - simultaneously, demonstrating parallel categorization abilities. Analysis of object features showed that changes in the shape or size of beads interfered with their generalization, whereas changes in color did not disrupt assignment of beads of a new color to a previously existing category. Our results provide an ecologically valid model of rapid visual categorization learning in chicks that allows integration of classical avian categorization paradigms with modern neurobiological methods, thus establishing a powerful framework for analysis of dynamics and neural mechanisms of categorical memory formation in avian species.
Ghost crab (Ocypode quadrata) live in dynamic environments filled with numerous conspecifics, predators, and anthropogenic threats. The capacity to learn in these treacherous contexts could drastically improve the chances of survival and, perhaps, reproduction. Operant conditioning encompasses a suite of paradigms all meant to foster the acquisition of predictive relations between stimuli and behavior. Recent evidence suggests ghost crab are capable of relatively complex mental processes, and to add to the known cognitive repertoire evolved in this species, the current research exposed them to an operant conditioning regime. In a “reinforced” condition, animals were placed in a square experimental arena that adjacently positioned a green sticker and false escape hatch covered by a movable door on the same wall. Travel to the green sticker lifted the door to give brief visual access to a sand compartment (i.e. reward) visible through the false hatch. In a control condition, the environment was identical, but view of the sand compartment was yoked to sticker approaches from animals in the reinforced group. Across 3 hours of training, crabs in both conditions significantly increased their sticker approach rates. However, group differences across training blocks emerged and persisted after only 1 hour, such that crabs in the reinforced group displayed significantly higher approach rates compared to yoked controls. An exploratory analysis also revealed that learning rates were correlated positively to carapace size. Thus, this research demonstrates that ghost crabs are able to learn through positive reinforcement, and this skill is one that may develop across the lifespan.
Past work has shown that pigeons can take efficient routes during traveling or spatial optimization problems in laboratory environments. Whereas pigeons are relatively efficient travelers during these tasks, it remains unclear whether the pigeons primarily rely upon local information in the array of locations (such as a nearest-neighbor [NN] solution that minimizes the distance their current location to their next location), or if they are able to use the global geometry an array of locations to take the most optimal routes. In the current study, we examined the paths that pigeons took that included five locations: a start box, three different feeders, and an end box. We tested the pigeons with pairs of mirrored test configurations of feeders that were similar, but for which the end box was in a different position. If pigeons were attending to the global features of the array, then they should take the most efficient global route based on the end location, compared with a less efficient NN route based on minimizing the distance of the current leg. Overall, the pigeons were efficient and took routes that were more effective than a random traveler. The results also showed that the pigeons tended to take NN rather than global routes. Different testing environments or reward structures may lead to differences in how pigeons differentially weigh the use of local versus global information.
Could bumblebees have emotional states that can be transferred between individuals? A recent study provides positive evidence and illustrates the logic and methods employed to investigate these challenging questions.
Resurgence and spontaneous recovery are two extensively studied behavioral phenomena in which previously extinguished operant behavior increases or reemerges. These findings are relevant to understanding flexibility in behavior but also relapse. Resurgence is the re-return of an extinguished target response (R1) when reinforcement conditions worsen for a more recently reinforced alternative response (R2), such as with extinction. Spontaneous recovery is the return of an eliminated R1 following time off from experiencing extinction conditions, such as with a retention interval. The present experiment examined whether combining tests for resurgence (removing R2 reinforcement) and spontaneous recovery (retention interval) produced greater increases in R1 than either test in isolation. University students played a computer video game and learned to perform R1 during Phase 1. During Phase 2, R1 was placed on extinction while simultaneously performing R2 was reinforced (Ext + DRA). Finally, the four groups of participants were tested for (1) neither resurgence or spontaneous recovery, (2) only resurgence, (3) only spontaneous recovery, or (4) both resurgence and spontaneous recovery. Increases in R1 were additive, with combined testing producing significantly greater responding that either of the two tests in isolation. We discuss theoretical issues as well as the implications of our findings for the treatment and relapse of challenging behavior.
Ants exhibit remarkable navigational abilities, flexibly integrating private information such as path integration and learnt visual cues with social information in the form of trail pheromones. Far from being simple or rigid directional signals, pheromone trails serve diverse, context-dependent roles in navigation. For instance, they can act as scaffolds during early foraging trips, tethering naïve ants to the nest while they acquire spatial knowledge. Pheromones also function as reassurance signals, confirming that a forager is on the correct path, and as a fail-safe when other cues become unreliable. Additionally, they appear to support multi-vector way-pointing through gating part of the path integrator expression, enabling ants to segment complex routes and to leave and re-enter trails. The regulation of pheromone laying itself is also influenced by private information streams, including internal state, prior foraging success, and navigational memory, highlighting the nuanced interplay between individual experience, environmental cues, and social signals. Together, these findings reveal the trail pheromone is not merely a recruitment or directional signal, but is an integral component of a sophisticated, multimodal navigational system, interwoven with private memories and the individual’s internal state to support flexible navigation.
Participants were trained and tested on their ability to discriminate among geometric angles presented to them as freestanding three-dimensional objects located within a larger experimental room. Three different angle sizes were used during training and participants were grouped based on their assigned training angle (110°, 135° or 160° angle), which they learned to discriminate from the remaining two angles. Following a learning phase, participants experienced test trials in which their training angle was paired with a novel test angle, and they made a choice between the two angles. Participants in the smallest angle group (Group 110) biased their choices toward test angles that were smaller than their training angle; conversely, participants in the largest angle group (Group 160) biased their choices toward test angles that were larger than their training angle. When it came to trials involving fine discrimination between angles, both groups showed a high degree of precision when needed. Results support cognitive strategies of geometric estimation that include relational rule learning and more fine-grained judgements to maximize overall accuracy.
A growing acceptance that many nonhuman animal communities have distinct cultures – group-variable patterns of behavior and information sustained over time by social learning – is beginning to reshape thinking about animal conservation. Culture, in this sense, can significantly influence how different populations interact with their environment and respond to environmental changes, and, therefore, has important implications for conservation. The literature on animal culture and conservation has led to valuable insights about how to protect endangered cultural animals. It has also led to some challenging questions. Should protecting animal cultural diversity become a new conservation goal, along the lines of preserving biodiversity? Should culture be an important consideration in prioritizing populations for conservation? Should we be designating animal “cultural heritage sites” for special protection, analogous to heritage sites of special significance for humans? This paper explores these questions and various arguments for preserving animal culture that have been offered in the literature. These include both instrumental arguments and arguments that suggest that animal cultures are of intrinsic value in their own right. These arguments raise important considerations, but they do not address all the ways in which animal cultures matter. We argue for more sustained attention to animals’ own interests in culture: animal cultures matter first and foremost because they matter to the animals themselves. Animals’ interests with respect to culture are not only about preserving practices or geographical locations, but include more abstract interests in protecting opportunities for agency, self-determination, and changing cultural traditions.
Pavlovian instrumental transfer (PIT) is a phenomenon in which a conditioned stimulus (CS), previously paired with a specific outcome, can influence performance of an instrumental response (R). PIT can be specific (the CS selectively elevates performance of an R paired with the same outcome as that CS) or general (CSs elevate performance of an R paired with any outcome of the same motivational value). PIT has been proposed as a mechanism underlying maintenance of addictive behaviours, while exposure to stressors is regarded as a risk factor for the development, maintenance and relapse of addiction. Thus, we used a quasi-avoidance-based PIT task to explore whether stress can influence the magnitude of specific or general PIT effects. We used stress induction procedures to elevate stress levels. These were either conducted online, exposing participants to negative pictures and heavy metal music (Experiment 1), or in person, telling participants they had to give a speech which would be filmed (Experiment 2). In neither experiment could we detect any effect of higher stress levels on specific or general PIT. The results do not support the proposal that stress influences the size of PIT effects in human participants.
Evidence has shown significant sex differences in freezing and darting behaviors in a rat model of aversive learning using fear conditioning. The present study explored sex differences in a rat model of aversive learning using a fear-conditioning method via measuring freezing and darting behaviors. Fear conditioning was induced by three footshocks (0.8 mA, 3 s, 30-s interval) paired with an auditory conditioned stimulus (75 dB, 3 s). Extinction was performed by broadcasting 20 auditory conditioned stimuli (75 dB, 3 s, 30-s interval), with no shocks, in three, or four, of five sessions. Freezing and darting behaviors, locomotor activity and time spent in the center squares (anxiety-like behavior) in the open field test, and brain-derived neurotrophic factor (BDNF) in the infralimbic region of the mPFC (medial prefrontal cortex) were evaluated. The results showed both sexes showed a high rate of freezing, with males showing more freezing. Females were more responsive to extinction. Darting behavior was only observed in females and diminished following extinction. Locomotion and anxiety-like behavior were increased and decreased following extinction learning in both sexes, respectively. BDNF expression level in the infralimbic region of the mPFC was increased following extinction learning, with a greater increase in females. In conclusion, we showed that females have a diverse behavioral response to the anticipation of a threat in a rat model of fear conditioning. The important role of BDNF in the modulation of both freezing and darting behaviors was also shown.
Two online experiments investigated whether the weakening of cue-outcome temporal contiguity impacts generalization gradients. Participants experienced strong temporal contiguity (the offset of a cue coincides with the onset an outcome) or weak temporal contiguity (there is a temporal gap between the offset of a cue and the onset of an outcome) in a continuous predictive learning task. Initial training involved a double negative discrimination, in which there was one reinforced cue (aqua-color rectangle) and two non-reinforced cues (one greener and one bluer than the reinforced cue), and participants had to make a response whenever they predicted the outcome. After training, in the generalization test phase, outcome expectancy in the presence of several stimuli along the full green-blue dimension was tested. Outcome expectancy was assessed with discrete ratings in Experiment 1 (n = 252) and a behavioral dependent measure (button presses, similar to the training phase) in Experiment 2 (n = 250). Regardless of the type of measure, when participants experienced weak cue-outcome temporal contiguity they showed a broader generalization response, although differences were more prominent in the green side of the color dimension. Furthermore, no significant differences emerged between the two groups experiencing strong temporal contiguity but differing in cue duration (controlling stimulus onset asynchrony, SOA), suggesting that cue-outcome temporal contiguity plays a more critical role than cue duration in shaping generalization. Overall, this result advances our understanding of the role of temporal contiguity at a basic level, while opening the possibility to study it as a critical factor in more applied settings.
Inhibitory control is an executive function that allows humans and non-human animals to suppress a prepotent response and continue in goal-directed behavior when outcomes are delayed. In non-human animals, inhibitory control has been proposed to be an important component of behavioral flexibility, where animals must inhibit previously used behaviors to establish new ones. The cylinder task has become a widely used, detour-reaching task to test motor response inhibition in many non-human animal species. Here we directly test motor inhibition in raccoons for the first time using a modified version of the cylinder task. Raccoons have demonstrated behavioral flexibility traits such as repeated innovation and reversal learning but have not been tested for motor response inhibition specifically. Five wild-caught, captive raccoons participated in the familiarization and detour procedures of the cylinder task, and we found an overall 60% pass rate in their cylinder task performance. We used k-means clustering to compare raccoons to other species tested in the cylinder task and found that raccoons cluster in the moderate performance group. Raccoons' expression of persistence may lead to their moderate cylinder task performance while benefiting them in tasks related to behavioral flexibility, such as problem solving.
Attention to relevant information is necessary to solve a categorization task. Prior studies monitoring peck location indicated that pigeons narrowly focus their attention on the relevant features of a category exemplar, whereas computational modeling suggested that pigeons' attention can be widely distributed. To help resolve these apparently contradictory results, we deployed both strategies - peck tracking and computational modeling - in the same experiment. Pigeons' peck-tracking scores neatly aligned with their category choices and with the results of the computational modeling. In addition, pigeons displayed distinctive learning and attentional patterns during training and testing. Half of our pigeons showed preferential, even exclusive, use of the most relevant category features, but the other half were inclined to use several of the available features. These results highlight the importance of combining an individual differences approach along with the pursuit of general cognitive principles.
I am honored to have been asked to curate and introduce this special issue recognizing the vast contributions of Dr. Suzanne MacDonald, professor of psychology at York University, Toronto Ontario, Canada, to the field of comparative psychology. As can readily be determined by only a cursory glance at the contents of this special issue, Suzanne’s influence has been felt across disciplines as her work encompasses an impressive breadth of species, topics, and contexts. Suzanne is the true embodiment of a comparative psychologist. It is rare that one has impacted so many areas of study and amassed such diversity of experiences, even in a field defined by breadth and comparative analyses. This breadth is reflected in the contributions to this special issue, inspired by the special symposium organized by the Comparative Cognition Society to honor Suzanne at the annual conference in Albuquerque, New Mexico, USA in March of 2025. I am humbled not just by recognition of the quantity of Suzanne’s contributions, both in terms of research and service to the field, but also because of the astounding practical significance of her contributions.
