Abstract and Keywords
The contemporary field of animal cognition began over 150 years ago when Charles Darwin posed questions regarding the abilities of the animal mind. Animal cognition is a science dedicated to understanding the processes and mechanisms that allow nonhumans to think and behave. The techniques that are used and the species that are studied are diverse. The historical questions originally proposed by ethologist Nikolas Tinbergen as a framework for studying animal behavior remain at the core of the field. These questions are reviewed along with the domains and methods of animal cognition with a focus on concept learning, memory, and canine cognition. Finally, ideas on how a field rich in tradition and methodological strength should proceed in the future are presented.
The Field of Animal Cognition
The field of animal cognition draws from concepts of cognitive psychology to study the mental abilities of nonhuman animals. As depicted in Figure 1, several disciplines have influenced the modern study of animal cognition spanning the biological and social sciences (Shettleworth, 2009). Animal cognition researchers are often concerned with how a single (or a few) species process and utilize information from the environment, but research questions and approaches may vary across different theoretical perspectives. Some researchers use animals as models of human psychological processes in order to study general cognitive mechanisms; similarly, the related field of comparative cognition is concerned with identifying those processes that are common across diverse species in order to understand the evolutionary origins of human and nonhuman cognition (Shettleworth, 2009). On the other hand, ethological approaches to animal cognition explore how animals solve ecologically relevant problems in the wild. Beyond answering basic empirical questions of animals’ cognitive abilities, understanding cognitive capacities also has important implications for animal welfare. Whereas the ability to feel pain has historically formed the basis of welfare considerations for nonhumans, addressing the cognitive “needs” of animals as well as considering how cognitive processes may influence welfare are now major concerns for animals kept in captivity (Rogers, 2010). Similarly, a better appreciation of cognition can inform conservation efforts for wild species (Vonk, 2016).
The study of animal cognition can be traced to Darwin’s anecdotal contemplations of animal intelligence, but the experimental study of animal intelligence began with Thorndike’s systematic observations of trial-and-error learning (Darwin, 1871; Thorndike, 1898; Thorndike, 1911). In Thorndike’s classic studies, animals (including cats and chickens) were placed inside puzzle boxes that required the subject to learn a behavioral response in order to escape the box. Over repeated trials the subjects gradually learned the correct response and eventually managed to escape faster. Thorndike argued that such behavior could be explained on the basis of simple, associative learning rather than complex and anthropomorphic constructs such as insight and reasoning. Thorndike’s premise was based on Morgan’s Canon, which asserts that behavior should be interpreted according to the most parsimonious account rather than attributing behavior to overly complex constructs. To this day Morgan’s Canon remains a strong tenet in the field of animal cognition. Thorndike’s work initiated the study of behaviorism, which dominated the study of animal learning and behavior for several decades. Behaviorism operates on the premise that behavior is the result of conditioned associations between behavior and the environment. Early work by Pavlov demonstrated the powerful effect of stimulus-response associations through classical conditioning experiments in which dogs learned to salivate in response to a signal that was previously associated with food (Pavlov, 1927). Much later, B. F. Skinner demonstrated the powerful effect of consequences on animal (and human) behavior through operant conditioning (Skinner, 1938). However, central to the field of behaviorism was the focus on only observable behavior, with no assumption of mental processes. In this view, animal behavior was seen as reflexive responses due to mechanistic stimulus-response associations.
Traces of animal cognitive research were apparent in the early 20th century (Dewsbury, 2000), but it was not until the 1970s that animal cognition became a recognized subfield. The field’s emergence followed on the heels of the cognitive revolution in human psychology as a response to the strict reductionist views of behaviorism. Instead of rejecting concepts requiring speculation about internal, mental processes, researchers began accepting that the behavior of animals is a representation of their cognitive processes (Wasserman & Zentall, 2006). Contemporary animal cognition researchers measure behavior in order to infer cognitive processes, using careful controls to rule out simple associative explanations for behavior in a wide range of domains.
The 21st-century research in animal cognition has a wider global reach than that of the 20th century and is uniformly influenced by Niko Tinbergen. Tinbergen, founder of the ethologist school of animal psychology, proposed a framework consisting of four broad questions by which to study animal behavior that has since been applied to the study of cognition (Tinbergen, 1963). Animal cognition researchers typically invoke one or more of Tinbergen’s questions when attempting to interpret some behavioral phenomenon. The first question is concerned with the function of the behavior in order to identify its adaptive purpose and is used to understand why a species or individual may exhibit a particular behavior. The second question considers the evolutionary history of a behavior and how the behavior varies across diverse phylogenetic lineages in order to understand how the behavior evolved. The third question concerns the ontogenetic development of a behavior across an individual’s lifetime and includes both age-related changes as well as effects of learning and experience. The fourth question asks what proximate mechanisms, such as brain activity or hormones, may cause a behavior to be exhibited. Of note, Tinbergen’s four questions have been implored in neuroscience too (Krakauer, Ghazanfar, Gomez-Marin, & Poeppel, 2017).
Animal cognition researchers study a wide range of cognitive mechanisms that can be categorized according to different domains: physical, social, and general. The physical cognition domain includes cognitive processes that enable animals to understand and utilize the physical properties of objects and surroundings. Physical cognition is involved in spatial navigation, numerosity, timing, causal inference, object permanence, and tool use, and typically underlies foraging-related problem solving (Nawroth et al., 2019). The domain of social cognition includes cognitive processes involved in relating to others and includes communication, cooperation, learning from others, and understanding others’ emotions, attentional states, and intentions. Like physical cognition, social cognition can be important for foraging-related tasks such as tracking the movement and locations of predators, prey, or conspecifics. Social cognition is also particularly important for navigating group dynamics in social species, as well as for interacting with humans in the case of domesticated animals.
The domestication hypothesis posits that domestication-related changes have led to an enhanced ability to communicate with and understand humans compared to their wild counterparts (Hare et al., 2010). Selection for enhanced socio-cognitive skills for adaptation to an anthropocentric world may have also led to a social-physical cognition tradeoff, due to relaxed ecological pressures for physical problem-solving during domestication. For example, dogs are considered to possess exceptional socio-cognitive skills for solving problems requiring communicating or cooperating with humans but fail to solve problems requiring advanced object tracking or causal inference skills. The opposite is true for great apes, reported to excel at foraging-related tasks requiring solving a problem using information about the physical properties of objects, while failing to solve similar problems requiring the use of social information conveyed by a human (Bräuer et al., 2006).
Other cognitive processes not pertaining to the social or physical domain are general fundamental processes that underlie most behavior and allow for adaptive responding in varying contexts, such as attention, memory, and learning. Domain-general mechanisms also include executive functions, which consist of effortful, goal-directed behavior mediated by the prefrontal cortex, such as working memory and inhibitory control. In humans, executive functions are related to life outcomes including academic performance, interpersonal skills, and risky behavior; however, executive functions in animals may be important for adaptive and optimal behavioral patterns (Olsen, 2018).
The diversity of experimental methods used by animal cognition researchers reflects the range of disciplines integrated in the field. Laboratory experiments confer rigorous experimental control and commonly employ traditional apparatuses such as operant chambers with touch screens or levers and radial-arm and water mazes. Using these methods, researchers can precisely present stimuli or problems and measure animals’ reactions and solutions. They can also require animals to learn more complex tasks and measure the learning acquisition and performance across various conditions. Traditionally, pigeons, rats, and primates have largely represented the animal subjects in which these methods are employed in order to study topics such as discrimination and reversal learning, categories and concepts, and mechanisms related to attention and memory. While laboratory studies may be viewed as artificial environments lacking ecological validity, advantages of such controlled settings include the ability to probe cognitive faculties by manipulating and controlling variables not possible in naturalistic observations (Wasserman & Zentall, 2006). Such standardization of methods also allows for cross-species comparisons critical to the field of comparative cognition.
Alternatively, field experiments study animals in their natural environment, which may include introducing some kind of manipulation and then observing responses. For example, playback experiments play recordings of sounds such as animal calls, and the behavioral responses to the calls are observed. In a classic example of the playback method, Seyfarth et al. (1980) recorded alarm calls produced by vervet monkeys in response to the sighting of different ground and aerial predators, each of which elicited a different sounding call from the monkeys. Results from the playback experiments showed that different behavioral responses were evoked by monkeys hearing the call depending on the type of call heard. Monkeys hearing the snake call reared up on their hind legs and looked at the ground, whereas monkeys hearing the leopard call ran up into trees. Conversely, aerial predator calls (e.g., eagles) caused monkeys in trees to run down and hide at ground level. These findings were significant in the field of animal cognition as they demonstrated a form of communication, as well as showing that monkeys were capable of functionally categorizing predators and discriminating between them. Another common naturalistic experiment takes advantage of some bird species’ food-caching behavior to study spatial memory by tracking visits to different caching sites across periods of time. Experimental settings for the study of animal cognition continue to diversify, expanding beyond the laboratory or the wild to include zoos, aquariums, farms, sanctuaries, and even pet owners’ homes (Vonk, 2016).
The domains and their underlying topics are too vast to cover extensively, and the reader can find empirical reports of animal cognition research published in journals such as Animal Cognition, Behavioural Processes, Comparative Cognition & Behavior Reviews, Current Biology, International Journal of Comparative Psychology, Journal of Comparative Psychology, Journal of Experimental Psychology: Animal Learning and Cognition, Learning & Behavior, and others. Comprehensive overviews of animal cognition can also be found in several other resources (Wasserman & Zentall, 2006; Zentall & Wasserman, 2012; Shettleworth, 2010; Vonk & Shackelford, 2012). In the remaining sections three prominent topics in the field of animal cognition are discussed. “Concept Learning,” “Memory,” and “Canine Cognition” were selected due to their history and contemporary relevance.
Animals live in a world that is constantly changing; failure to adapt efficiently to an environment in flux can reduce overall fitness. One way in which animals adapt to their environment is to learn concepts that transfer from one object or context to another. Concept learning can involve either categorization or abstract relational rules. Categorization (sorting) occurs when animals generalize responses to stimuli based on common specific features that have been learned from a trained category (stimulus set). Evidence of such concept learning has been found in many species including pigeons (e.g., Herrnstein, Loveland, & Cable, 1976; Wasserman, Kiedinger, & Bhatt, 1988), gorillas (Vonk & MacDonald, 2002), chimpanzees (Oden, Thompson, & Premack, 1988), goldfish (Goldman & Shapiro, 1979), and cichlids (Schluessel, Fricke, & Bleckman, 2012), suggesting widespread evolutionary convergence. Categorization can also involve second-order conditioning based on a common response outcome regardless of the perceptual similarity of the objects within a category (e.g., Urcuioli, 2006; for reviews see Huber, 2001; Lazareva, Shimizu, & Wasserman, 2012; Zentall et al., 2008). In contrast, abstract concepts are based on judging relationships among two or more stimuli. Abstract-concept learning cannot be accounted for by generalization processes bound to specific stimuli. Such concepts are based on relational rules (e.g., sameness, differentness, oddity, addition, subtraction). Conceptual behavior is considered abstract when these rules can be successfully applied to novel stimuli.
While evidence of concept learning via categorization in nonhuman animals has been generally accepted for some time, abstract concepts are traditionally more difficult to both learn and evince (e.g., Katz, Wright, & Bodily, 2007). Abstract-concept learning has been at the heart of experimental studies with animals since Thorndike and is central to theories of animal intelligence (Thorndike, 1911). As noted, abstract concepts require comparisons between stimuli based on the relationship between items (e.g., relative brightness, size, saturation, sameness). One example is the same/different concept, which involves learning to discriminate between pairs of objects based on the relative sameness and differentness. The ability to represent this same/different concept has been called “the very keel and backbone of our thinking” (James, 1890) and is the basis for higher cognitive functions such as mathematics and language. As such, animal cognition researchers have sought to understand the extent to which the ability to learn the same/different concept, and abstract concepts generally, is part of a process that is found in a variety of taxa or is exclusively found in humans.
To answer this question, experiments on abstract-concept learning frequently consist of two general phases: acquisition and transfer testing. Animals first learn to discriminate a training set of stimuli that form examples of the abstract concept to be learned. After they have acquired the relevant discrimination with the training stimuli (e.g., same/different), they are then presented with completely novel stimuli in the transfer test. If the subject responds to the novel stimuli in a manner consistent with the previous training set (i.e., transfer performance is equivalent to training performance), the subject is considered to have learned the abstract concept (Wright, Cook, Rivera, Sands, & Delius, 1988; Katz, Wright, & Bodily, 2007). A common procedure to test for abstract- concept learning is the match-to-sample (MTS) task, in which subjects are first presented with a sample stimulus (e.g., orange circle), which is then followed by two comparison stimuli (e.g., orange circle and blue circle). The subject must select the matching stimulus (orange circle) from the comparison items. The inverse of this task, the nonmatch-to-sample (NMTS) task, is also frequently used. The key difference is that subjects trained on this procedure must pick the item that does not match the sample (i.e., the correct choice would be “blue circle”). Both of these procedures have been extensively used in the study of concept learning, as well as other cognitive processes.
Early investigations into animal abstract-concept learning often failed to provide supporting evidence. For example, Zentall and Hogan (1974) trained pigeons on an MTS procedure that required subjects to match or nonmatch colors. While pigeons could learn the task over time, when presented with novel stimuli, transfer performance was poor. Premack (1978) argued that such attempts at concept learning experiments were often flawed due to the fact that transfer tests repeated stimuli, and animals could rapidly learn stimulus-specific responses across trials. This further supported Premack’s claim that only humans and language-trained chimpanzees, which showed immediate transfer (without repeating transfer stimuli), could form abstract concepts. Later, Wright et al. (1988) addressed Premack’s concerns about transfer tests by limiting analysis of transfer performance to the first trial with novel stimuli, preventing the potential effects of learning about the novel stimuli over repeated exposures. Wright and colleagues found evidence of the abstract concept of sameness in pigeons as a function of the number of stimuli used to learn the MTS task, such that pigeons trained with 152 stimuli successfully transferred to novel stimuli, while pigeons trained with two stimuli did not. Using a set-size expansion procedure where animals learn a small training set and then experience a cycle of doubling the training set followed by more training and novel stimulus testing has consistently shown increasing transfer as the training set is expanded to the point where animals fully demonstrate the abstract concept (i.e., transfer = baseline).
The functional relationship between set size and task performance is seen across tasks (NMTS, MTS, same/different) as well as across a variety of species. Figure 2 shows the functional relationship between set size and transfer in a same/different (S/D) task for capuchin monkeys, rhesus monkeys, Clark’s nutcrackers, pigeons, and black-billed magpies (Wright, Katz, & Kelly, 2018). As an added control, the same training and transfer stimuli were used in training and transfer tests for each of the species. In solving the S/D discrimination, these data show a qualitative similarity (i.e., learning the abstract S/D concept) across all species with a quantitative difference (i.e., species fully apply the abstract S/D rule at different set sizes). Hence, set-size expansion provides insight into one mechanism for abstract-concept learning. This process works by exposing subjects to a variety of exemplars of the abstract concept. Eventually, with enough experience with a sufficiently varied set of exemplars, animals will fully attend to the relationship between stimuli allowing transfer to novel items. Additional analyses have shown that stimulus generalization accounts of the data do not explain the excellent transfer performance (Bodily, Katz, & Wright, 2008; Daniel, Goodman, Thompkins, Forloines, Lazarowski, & Katz, 2016; Wright & Katz, 2006).
Focusing on how animals learn abstract concepts via revealing functional relationships will continue to reveal similarities and differences across species. There also continues to be an emphasis on what species can learn abstract concepts. Recently, Martinho and Kacelnik (2016) explored how the process of imprinting in ducklings may allow for recently hatched ducks to learn a relational, same/different concept. Imprinting is a phenomenon wherein hatchlings form an emotional bond with the first moving object they see. After imprinting, ducklings respond to the moving object by following it (Hess, 1964). Martinho and Kacelnik found that ducklings could imprint onto sets of matching or nonmatching shapes that moved around a room, and then transfer this imprinted concept onto novel pairs of objects that were presented in the same configuration. In another example, Newport, Wallis, and Siebeck (2015) attempted to train archerfish, a species of fish known for their unique hunting technique of spitting at aerial prey, to discriminate between same and different pairs of objects. While the fish that acquired the discrimination failed to transfer, this experiment demonstrates the inclusion of less common species in animal cognition, and a potential difference in the cognitive abilities of more general groups of animals (i.e., fish versus mammals). However, absence of evidence cannot be taken as evidence that a process does not exist, as one never knows if the right conditions have been met to evince a particular process.
One such condition is sensory modality, which can be critical to a species’ ability to learn and demonstrate abstract-concept learning. For example, it was previously assumed that rats were incapable of abstract-concept learning after several studies failed to find evidence of transfer in a visual MTS task (Iversen, 1997). However, when rats were trained in MTS using olfactory stimuli, they readily acquire the concept and transferred to novel stimuli (Peña, Pitts, & Galizio, 2006; Lazarowski, Goodman, Galizio, & Bruce, 2019). Similarly, pigeons show faster acquisition and better transfer when trained using procedures that mimic natural foraging behavior, such as projecting stimuli from the floor and requiring downward pecking rather than vertical (Wright, 1997), or digging through different colors of gravel stimuli (Wright & Delius, 1994). These findings highlight the importance of maintaining ecological validity when assessing cognitive capacities of animals. In terms of cognitive processes, Martinho and Kacelnick’s (2016) imprinting ducklings suggests a process that does not require a fully developed brain. Studies using corvids (Wright, Magnotti, Katz, Leonard, Vernouillet, & Kelly, 2017) and other avian species suggest that a neocortex is also not necessary for learning concepts. However, the nature of the neural representations of concepts themselves are unknown. The future of animal cognition should strive to answer this question in order to better understand the processes that underlie concept learning.
Memory is undoubtedly necessary for animals to both survive and thrive: from magpies that remember the location of their long-hidden caches to seals that recognize the vocalizations of their mother after a year of separation (Insley, 2000). The questions that drive research on animal memory are similar to all subfields of animal cognition: how memory affects behavior and how it relates to memory in humans (Shettleworth, 2010). As in humans, memory is discussed in terms of limitations, such as the rate of forgetting (Eichenbaum, 2008). Researchers answer these questions through the careful development of methods that rely on animals’ behavior rather than verbal ability. This section outlines some of the main findings in animal memory research and provides insight on the current direction of the field. Specifically, working and episodic memory will be discussed due to their important translational value to human memory.
Working memory in nonhuman animals is defined as short-term memory for stimuli within a given testing session (Honig, 1978; Olton & Samuelson, 1976). It is an important aspect of executive functioning that plays a role in intelligence, decision making, and future planning (Baddeley, 2017; Engle & Kane, 2004). Although working memory has been thoroughly investigated in many species (e.g., pigeons, primates, and rats) the construct remains a subject of theoretical debate (Dudchenko Talpos, Young, & Baxter, 2013). In the majority of these studies, working memory is measured as the rate of forgetting. Specifically, the duration in which a species can remember a given stimulus within an experimental session. Its strength can be tested by the effects of interference.
A modification to the MTS task is the most common way to assess the temporal duration of working memory in animals. In a typical delayed match-to-sample (dMTS) task, an animal is required to choose a comparison stimulus that matches a previously presented sample following a delay interval (e.g., Wright, 1991). If the animal responds to the to-be-remembered stimulus, the trial is marked correct and the animal’s working memory duration is recorded as the length of the delay. An incorrect response may indicate that the animal is unable to retain the to-be-remembered information after such an interval. A common forgetting function depicts a decrease in performance as the delay interval between sample and comparisons increases. Figure 3 shows multispecies performance on the dMTS (Lind, Enquist, & Ghirlanda, 2015). A number of factors can influence performance on the dMTS. For example, proactive interference occurs when a stimulus from a previous trial appears on the current trial and disrupts performance. Larger stimulus set sizes and the implementation of novel stimuli during testing alleviate these effects by reducing confusion and account for the ability of many species to remember stimuli at longer delays (Wright, Kelly, & Katz, 2018).
Working memory can also be measured as a function of the number of stimuli to remember (i.e., storage capacity). For example, the odor span task (OST) was developed as a nonhuman analogy to the digit-span task used to measure working memory capacity in humans. The two-choice incrementing NMTS task requires the animal to choose the novel odor (within the current session) on each trial as the number of odors to remember is increased across trials. The comparison odor is randomly selected from the odors previously encountered in the session. Using the OST, it has been argued that rats have a working memory capacity for up to 72 odors (April, Bruce, & Galizio, 2013; Dudchenko, Wood, & Eichenbaum, 2000). These results challenge the notion that the OST could be used as a model of human limited memory capacity, which is often accepted to be memory for 7 +/- 2 items (April et al., 2013; Miller, 1957). The validity of the OST as a test of working memory capacity and the definition of working memory in animals has since become a topic of debate. To assess the cross-species reliability of the OST, the April et al. 2013 study was replicated in a group of dogs (Krichbaum, Rogers, Cox, Waggoner, & Katz, in press). Although the experiment yielded similar results in terms of accuracy across trial blocks, the researchers additionally assessed memory in terms of the number of intervening odors or the number of discriminations since the previously encountered odor was last presented. The findings suggested that dogs are successful at choosing the novel odor following up to eight discriminations since the previously encountered odor was last presented.
An alternative explanation for these results could be that the rats and dogs are making their choices based on relative familiarity (April et al., 2013). Familiarity refers to a process that works parallel to working memory but only encodes whether a stimulus has been encountered previously (Yonelinas, 2002; Basile & Hampton, 2013). For example, subjects could have avoided the comparison odor due to a sense that they had already encountered the incorrect odor (April et al., 2013), rather than actively holding all of the previously encountered odors in working memory for comparison on each trial. Research suggests that animals can perform well on memory tests by relying solely on familiarity and that it may have an infinite capacity unlike its working memory counterpart (Basile & Hampton, 2013).
Recent studies have attempted to tease apart these constructs in an effort to better understand the mechanism of working memory (Basille & Hampton, 2013; Brady & Hampton, 2018). In a variation on the dMTS procedure, monkeys underwent either an image familiarization study phase or a regular study phase (i.e., traditional dNMTS) before the introduction of a five-second delay. The image familiarization phase was implemented so the monkeys were unable to use familiarity to solve the task. On every trial in this phase, the sample (marked by a circle) and all comparisons were presented simultaneously. Therefore, following the delay, the images presented at test had the same level of familiarization. The results indicated that image familiarization strongly decreased accuracy when using a large stimulus set size suggesting that animals rely heavily on familiarization when the stimulus set is large (Figure 4; Brady & Hampton, 2018).
In contrast to working memory, studies on long-term remembering focus on animals’ ability to retain information across extended periods of time (i.e., longer than within a single experimental session). Researchers who study long-term animal memory are currently involved in their own debate, one centered on a decades-old question: Is episodic memory uniquely human? Defined as a memory for personal happenings, it is distinguished from semantic memory, which is not limited by an individual’s personal experience (Tulving, 1972; Tulving & Markowitsch, 1998). For example, if one is asked what they did on their birthday last year, it is likely they will remember what they did, where it took place, and who was there. Tulving went on to explain three main tenets of episodic memory: recognition of self, autonoetic consciousness, and mental time travel. “Self” most simply refers to a person’s ability to recognize his or herself as an individual separate from the environment and other individuals. Autonoetic consciousness is the ability to mentally place oneself in the past or future and analyze one’s own thoughts. Finally, mental time travel, or chronesthesia, is the ability of the “rememberer” to travel back in his or her mind to an earlier situation and relive the experience (Tulving, Terrace, & Metcalfe, 2005).
Unfortunately, concepts such as “self” and “autonoetic consciousness” require verbal reports leaving episodic memory a difficult concept to study in animals. Despite this, scientists have endeavored to find a model of non-verbal episodic memory in animals. Instead of verbal reports, these studies rely on an objective criterion for episodic-like memory. This criterion is called What-Where-When and requires evidence that the animal’s observable behavior is due to what event occurred, when the event occurred, and when the event occurred (Clayton & Dickinson, 1998). One of the earliest of these studies focused on caching behaviors in scrub jays (Clayton & Dickinson, 1998) that showed birds were able to differentially recover caches based on what food they cached, where they cached it, and when they were permitted to recover their caches. In response to these results, researchers attempted to find evidence for episodic-like memory in a number of species including dogs and rats (Lo & Roberts, 2019; Pañoz-Brown et al., 2016).
An important part of episodic memory is that it involves a recollection of events in sequential order or in-context memory (Pañoz-Brown et al., 2016). In-context memory is defined as the recollection of the context in which multiple unique events were encountered. In a recent study by Pañoz-Brown et al. (2016) rats were trained on the previously described OST in two different arenas. They were trained to always choose the new odor (S+; had never been presented in that context) and avoid choosing the old odor (S-; had been presented in that context) when being paired across different contexts. In order to rule out the confound of familiarity or the ability of the rat to use the semantic rule of how long ago they encountered a specific scent, the researchers presented sequences of odors that put familiarity and in-context memory in conflict (see Figure 5). It was hypothesized that if rats were using the semantic rule to respond to the scent encountered the longest time ago (familiarity), they would respond to scent A and display below-chance accuracy on the task. It was also hypothesized that rats were using in-context memory and would respond to scent B and have above-chance accuracy on the task. Rats displayed above-chance accuracy on the in-context memory task indicating that they relied on in-context memory rather than cues of familiarity to respond to the odors.
The study of memory in animal cognition is currently being driven by theory. As discussed, many species have displayed powerful abilities to remember stimuli in both the long and short term, but interpretation of these findings remains a topic of debate. Future work should focus on pinpointing the cognitive processes being assessed. Due to the relevance of both working and episodic memory to the treatment of human aging, learning more about these constructs could influence future translational work. It is of high importance that researchers ground their questions in theory and strive to expand the number of species studied.
Rats, monkeys, and pigeons have long dominated animal cognition research (Shettleworth, 2009). However, dogs have become increasingly popular subjects over the last two decades. While dogs are no stranger to animal psychology research, as they were central to Pavlov’s classic respondent conditioning experiments and often serve as biomedical models for translational research, interest in canine cognition in its own right has rapidly evolved into a thriving subfield of animal cognition. Initial interest in canine cognition was initially sparked by the discovery that dogs outperformed nonhuman primates in tasks requiring the use of social information conveyed by a human, leading to the aforementioned domestication hypothesis as the basis for unique socio-cognitive skills in dogs (Hare et al., 2002). This hypothesis led to a surge of studies examining the relative roles of domestication status and life history on social cognition in dogs as well as their wild counterparts, resulting in data both in support of and against accounts invoking domestication. For example, evidence that dogs also outperformed wolves on human-based social tasks further implicated the role of domestication in the evolution of canine socio-cognitive skills (Hare et al., 2002), whereas other experiments with human-reared wolves and dogs with varying life histories have demonstrated the powerful effects of ontogenetic factors such as socialization, rearing history, and testing conditions (D’Aniello, Alterisio, Scandurra, Petremolo, Iommelli, & Aria, 2017; Udell, Dorey, & Wynne, 2008; Udell, Dorey, & Wynne, 2010). Thus, the two-stage hypothesis suggests that ostensibly superior socio-cognitive skills may be the results of socialization to humans during critical periods of development, which are optimized for human interaction in domestic dogs, combined with salient experience with human-relevant social stimuli (Udell, Dorey, & Wynne, 2010). This account does not negate the contribution of domestication but rather emphasizes the important modulating effects that experience can have on the development of such skills. Taken together, the data suggest that selection for a sensitivity and responsiveness to humans that occurred during domestication likely predisposes dogs to readily learn and acquire sophisticated socio-cognitive abilities.
While investigations into the socio-cognitive abilities of dogs have dominated the modern study of canine cognition, researchers have also become interested in dogs’ non-social cognitive abilities and how they compare to humans and other species. When placed in the context of dogs’ evolutionary history and social ecology, it appears that dogs’ cognitive abilities—outside of those requiring communication and cooperation with humans—may not be exceptional relative to other species (Lea & Osthaus, 2018). Nonetheless, impressive feats of cognitive abilities have been demonstrated in dogs such as a border collie capable of learning and recalling the names of over 1,000 objects (Pilley & Reid, 2011). Demonstrations of such remarkable capabilities are likely partly due to the high trainability required to reveal evidence of such skills, as well as ease of access to canine participants. Likewise, dogs’ inherent trainability has given rise to a new method for comparative neuroimaging by training dogs to remain still for awake functional Magnetic Resonance Imaging (fMRI). A recently emerging field, awake canine fMRI will provide greater understanding of the neurocognitive processes of the canine brain. For example, discovering commonalities and differences in the neural circuitry of dog and human brains will shed light on the evolutionary history of cognitive processes and their underlying neural mechanisms (e.g., Berns & Cook, 2016; Kyathanahally et al. 2015; Robinson et al., 2016; Thompkins Deshpande, Waggoner, & Katz, 2016).
Animal cognition has evolved since Darwin built the foundation in 1871 in his classic The Descent of Man, and Selection in Relation to Sex. Darwin emphasized the mental differences in species were in “degree” and not “kind.” In modern animal cognition, exploring the abilities of a variety of species is a standard in the field. Diversity in species is critical for a comprehensive comparative perspective (Shettleworth, 2009), but it is also important because testing a number of species can exhibit replicability and hence strengthen the field as a whole (Beran, Parrish, Perdue, & Washburn, 2014). A recent analysis of species published in 2018 in the journal Animal Cognition showed that researchers continue to study diverse species (Kelly & Leonard, personal communication). The species included amphibians, arthropods, birds, canids, fish, primates, reptiles, and ungulates. While continued extension of the animal kingdom remains critical, an emphasis should also be placed on examining the mechanisms that underlie cognition. Specifically, the future of animal cognition, in every sub-area, should continue to focus on revealing functional relationships within and across species. It is also essential to combine what animals can do with the more delicate questions of why animals behave in specific ways and what mechanisms are driving those behaviors. As such, Tinbergen’s four questions will continue to influence what questions are asked in modern animal cognition research.
April, L. B., Bruce, K., & Galizio, M. (2013). The magic number 70 (plus or minus 20): Variables determining performance in the rodent odor span task. Learning and Motivation, 44(3), 143–158.Find this resource:
Baddeley, A. D. (2017). The concept of working memory: A view of its current state and probable future development. In Exploring working memory (pp. 99–106). London, U.K.: Routledge.Find this resource:
Basile, B. M., & Hampton, R. R. (2013). Dissociation of active working memory and passive recognition in rhesus monkeys. Cognition, 126(3), 391–396.Find this resource:
Beran, M. J., Parrish, A. E., Perdue, B. M., & Washburn, D. A. (2014). Comparative cognition: Past, present, and future. International Journal of Comparative Psychology, 27(1), 3.Find this resource:
Berns, G. S., & Cook, P. F. (2016). Why did the dog walk into the MRI? Current Directions in Psychological Science, 25(5), 363–369.Find this resource:
Bodily, K. D., Katz, J. S., & Wright, A. A. (2008). Matching-to-sample abstract-concept learning by pigeons. Journal of Experimental Psychology: Animal Behavior Processes, 34, 178–184.Find this resource:
Brady, R. J., & Hampton, R. R. (2018). Nonverbal working memory for novel images in rhesus monkeys. Current Biology, 28(24), 3903–3910.Find this resource:
Bräuer, J., Kaminski, J., Riedel, J., Call, J., & Tomasello, M. (2006). Making inferences about the location of hidden food: social dog, causal ape. Journal of Comparative Psychology, 120(1), 38–47.Find this resource:
Clayton, N. S., & Dickinson, A. (1998). Episodic-like memory during cache recovery by scrub jays. Nature, 395, 272–274.Find this resource:
Daniel, T. A., Goodman, A. M., Thompkins, A. M., Forloines, M. R., Lazarowski, L., & Katz, J. S. (2016). Generalization cannot predict abstract-concept learning. In M. C. Olmstead (Ed.), Animal Cognition: Principles, Evolution and Development (pp. 131–145).Find this resource:
Darwin, C. (1871). The descent of man and selection in relation to sex. New York, NY: D. Appleton.Find this resource:
D’Aniello, B., Alterisio, A., Scandurra, A., Petremolo, E., Iommelli, M. R., & Aria, M. (2017). What’s the point? Golden and Labrador retrievers living in kennels do not understand human pointing gestures. Animal Cognition, 20(4), 777–787.Find this resource:
Dewsbury, D. A. (2000). Comparative cognition in the 1930s. Psychonomic Bulletin and Review, 7(2), 267–283.Find this resource:
Dudchenko, P. A., Talpos, J., Young, J., & Baxter, M. G. (2013). Animal models of working memory: A review of tasks that might be used in screening drug treatments for the memory impairments found in schizophrenia. Neuroscience & Biobehavioral Reviews, 37(9), 2111–2124.Find this resource:
Dudchenko, P. A., Wood, E. R., & Eichenbaum, H. (2000). Neurotoxic hippocampal lesions have no effect on odor span and little effect on odor recognition memory but produce significant impairments on spatial span, recognition, and alternation. Journal of Neuroscience, 20(8), 2964–2977.Find this resource:
Eichenbaum, H. (2008). Learning & memory. New York, NY: W.W. Norton.Find this resource:
Engle, R. W., & Kane, M. J. (2004). Executive attention, working memory capacity, and a two-factor theory of cognitive control. Psychology of Learning and Motivation, 44, 145–200.Find this resource:
Goldman, M., & Shapiro, S. (1979). Matching-to-sample and oddity-from-sample in goldfish. Journal of the Experimental Analysis of Behavior, 31, 259–266.Find this resource:
Hare, B., Brown, M., Williamson, C., & Tomasello, M. (2002). The domestication of social cognition in dogs. Science, 298(5598), 1634–1636.Find this resource:
Hare, B., Rosati, A., Kaminski, J., Bräuer, J., Call, J., & Tomasello, M. (2010). The domestication hypothesis for dogs’ skills with human communication: A response to Udell et al. (2008) and Wynne et al. (2008). Animal Behaviour, 79(2), 1–6.Find this resource:
Herrnstein, R. J., Loveland, D. H., & Cable, C. (1976). Natural concepts in pigeons. Journal of Experimental Psychology: Animal Behavior Processes, 2, 285–302.Find this resource:
Hess, E. H. (1964). Imprinting in birds. Science, 146(3648), 1128–1139.Find this resource:
Honig, W. K. (1978). Studies of working memory in the pigeon. In Cognitive processes in animal behavior (pp. 211–248). London, U.K.: Routledge.Find this resource:
Huber, L. (2001). Visual categorization in pigeons. In R. G. Cook (Ed.), Avian visual cognition.Find this resource:
Insley, S. J. (2000). Long-term vocal recognition in the northern fur seal. Nature, 406(6794), 404.Find this resource:
Iversen, I. H. (1997). Matching-to-sample performance in rats: A case of mistaken identity? Journal of the Experimental Analysis of Behavior, 68(1), 27–45.Find this resource:
James, W. (1890). The principles of psychology (Vol. 1). New York, NY: Dover Publications.Find this resource:
Katz, J. S., Wright, A. A., & Bodily, K. D. (2007). Issues in the comparative cognition of abstract-concept learning. Comparative Cognition & Behavior Reviews, 2, 79–92.Find this resource:
Krakauer, J. W., Ghazanfar, A. A., Gomez-Marin, A., MacIver, M. A., Poeppel, D. (2017). Neuroscience needs behavior: Correcting a reductionist bias. Neuron, 93, 480–493.Find this resource:
Krichbaum, S., Rogers, B., Cox, E., Waggoner, P., Katz, J. S. (in press). Odor span task in dogs (Canis familiaris). Animal Cognition.Find this resource:
Kyathanahally, S. P., Jia, H., Pustovyy, O. M., Waggoner, P., Beyers, R., Schumacher, J., … Deshpande, G. (2015). Anterior–posterior dissociation of the default mode network in dogs. Brain Structure and Function, 220(2), 1063–1076.Find this resource:
Lazareva, O. F., Shimizu, T., & Wasserman, E. A. (2012). How animals see the world: Comparative behavior, biology, and evolution of vision. Oxford, U.K.: Oxford University Press.Find this resource:
Lazarowski, L., Goodman, A., Galizio, M., & Bruce, K. (2019). Effects of set size on identity and oddity abstract-concept learning in rats. Animal Cognition, 22, 7331–74210.Find this resource:
Lea, S. E., & Osthaus, B. (2018). In what sense are dogs special? Canine cognition in comparative context. Learning & Behavior, 46(4), 335–363.Find this resource:
Lo, K. H., & Roberts, W. A. (2019). Dogs (Canis familiaris) use odor cues to show episodic-like memory for what, where, and when. Journal of Comparative Psychology, 133(4), 428–441.Find this resource:
Lind, J., Enquist, M., & Ghirlanda, S. (2015). Animal memory: A review of delayed matching-to-sample data. Behavioural Processes, 117, 52–58.Find this resource:
Martinho, A., & Kacelnik, A. (2016). Ducklings imprint on the relational concept of ‘same or different.’ Science, 353(6296), 286–288.Find this resource:
Miller, G. A. (1957). The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review, 101(2), 343–352.Find this resource:
Nawroth, C., Langbein, J., Coulon, M., Gabor, V., Oesterwind, S., Benz-Schwarzburg, J., … von Borell, E. (2019). Farm animal cognition-linking behavior, welfare and ethics. Frontiers in Veterinary Science, 6, 1–16.Find this resource:
Newport, C., Wallis, G., & Siebeck, U. E. (2015). Same/different abstract concept learning by archerfish (Toxotes chatareus). PloS One, 10, 1–14.Find this resource:
Oden, D. L., Thompson, R. K., & Premack, D. (1988). Spontaneous transfer of matching by infant chimpanzees (Pan troglodytes). Journal of Experimental Psychology: Animal Behavior Processes, 14, 140–145.Find this resource:
Olsen, M. R. (2018). A case for methodological overhaul and increased study of executive function in the domestic dog (Canis lupus familiaris). Animal Cognition, 21(2), 175–195.Find this resource:
Olton, D. S., & Samuelson, R. J. (1976). Remembrance of places passed: Spatial memory in rats. Journal of Experimental Psychology: Animal Behavior Processes, 2(2), 97.Find this resource:
Pañoz-Brown, D., Corbin, H., Dalecki, S., Gentry, M., Brotheridge, S., Sluka, C., … Crystal, J. (2016). Rats remember items in context using episodic memory. Current Biology, 26(20), 2821–2826.Find this resource:
Pavlov, P. I. (1927). Conditioned reflexes: An investigation of the physiological activity of the cerebral cortex. Annals of Neurosciences, 17(3), 136.Find this resource:
Peña, T., Pitts, R. C., & Galizio, M. (2006). Identity matching-to-sample with olfactory stimuli in rats. Journal of the Experimental Analysis of Behavior, 85(2), 203–221.Find this resource:
Pilley, J. W., & Reid, A. K. (2011). Border collie comprehends object names as verbal referents. Behavioural Processes, 86(2), 184–195.Find this resource:
Premack, D. (1978). On the abstractness of human concepts: Why it would be difficult to talk to a pigeon. In Cognitive processes in animal behavior (pp. 423–451). London, U.K.: Routledge.Find this resource:
Robinson, J. L., Bazi, M., Katz, J. S., Waggoner, P., Beyers, R., Morrison, E., … Deshpande, G. (2016). Characterization of structural connectivity of the default mode network in dogs using diffusion tensor imaging. Scientific Reports, 6, 36851.Find this resource:
Rogers, L. J. (2010). Cognition and animal welfare. Wiley Interdisciplinary Reviews: Cognitive Science, 1(3), 439–445.Find this resource:
Schluessel, V., Fricke, G., & Bleckmann, H. (2012). Visual discrimination and object categorization in the cichlid Pseudotropheus sp. Animal Cognition, 15, 525–537.Find this resource:
Shettleworth, S. J. (2009). The evolution of comparative cognition: Is the snark still a boojum? Behavioural Processes, 80(3), 210–217.Find this resource:
Shettleworth, S. J. (2010). Cognition, evolution, and behavior. Oxford, U.K.: Oxford University Press.Find this resource:
Seyfarth, R. M., Cheney, D. L., & Marler, P. (1980). Monkey responses to three different alarm calls: evidence of predator classification and semantic communication. Science, 210(4471), 801–803.Find this resource:
Skinner, B. F. (1938). The behavior of organisms: An experimental analysis. Cambridge, MA: B.F. Skinner Foundation.Find this resource:
Thompkins, A. M., Deshpande, G., Waggoner, P., & Katz, J. S. (2016). Functional magnetic resonance imaging of the domestic dog: Research, methodology, and conceptual issues. Comparative Cognition & Behavior Reviews, 11, 63–82.Find this resource:
Thorndike, E. L. (1898). Animal intelligence: An experimental study of the associative processes in animals. Psychological Review, Monograph Supplements, No. 8. New York, NY: Macmillan.Find this resource:
Thorndike, E. L. (1911). Animal intelligence. New York, NY: Macmillan.Find this resource:
Tinbergen, N. (1963). On aims and methods of ethology. Zeitschrift für tierpsychologie, 20(4), 410–433.Find this resource:
Tulving, E. (1972). Episodic and semantic memory. Organization of Memory, 1, 381–403.Find this resource:
Tulving, E., & Markowitsch, H. J. (1998). Episodic and declarative memory: Role of the hippocampus. Hippocampus, 8(3), 198–204.Find this resource:
Tulving, E., Terrace, H. S., & Metcalfe, J. (2005). The missing link in cognition: Origins of self-reflective consciousness. Oxford, U.K.: Oxford University Press.Find this resource:
Udell, M. A., Dorey, N. R., & Wynne, C. D. (2008). Wolves outperform dogs in following human social cues. Animal Behaviour, 76(6), 1767–1773.Find this resource:
Udell, M. A., Dorey, N. R., & Wynne, C. D. (2010). What did domestication do to dogs? A new account of dogs’ sensitivity to human actions. Biological Reviews of the Cambridge Philosophical Society, 85(2), 327–345.Find this resource:
Udell, M. R., Dorey, N. R., & Wynne, C. D. (2010). The performance of stray dogs (Canis familiaris) living in a shelter on human-guided object-choice tasks. Animal Behaviour, 79(3), 717–725.Find this resource:
Urcuioli, P. J. (2006). Responses and acquired equivalence classes. In E. A. Wasserman & T. R. Zentall (Eds.), Comparative cognition (pp. 405–421). New York, NY: Oxford University Press.Find this resource:
Vonk, J. (2016). Advances in animal cognition. Behavioral Sciences, 6(4), 27.Find this resource:
Vonk, J., & MacDonald, S. E. (2002). Natural concepts in a juvenile gorilla (Gorilla gorilla) at three levels of abstraction. Journal of the Experimental Analysis of Behavior, 78, 315–332.Find this resource:
Vonk, J., & Shackelford, T. K. (Eds.). (2012). The Oxford handbook of comparative evolutionary psychology. New York, NY: Oxford University Press.Find this resource:
Wasserman, E. A., Kiedinger, R. E., & Bhatt, R. S. (1988). Conceptual behavior in pigeons: Categories, subcategories, and pseudocategories. Journal of Experimental Psychology: Animal Behavior Processes, 14, 235–246.Find this resource:
Wasserman, E. A., & Zentall, T. R. (2006). Comparative cognition: Experimental analysis of animal intelligence. Oxford, U.K.: Oxford University Press.Find this resource:
Wright, A. A. (1997). Concept learning and learning strategies. Psychological Science, 8(2), 119–123.Find this resource:
Wright, A. A. (1991). Concept learning by monkeys and pigeons. Memory mechanisms: A tribute to GV Goddard, 247–273.Find this resource:
Wright, A. A., Cook, R. G., Rivera, J. J., Sands, S. F., & Delius, J. D. (1988). Concept learning by pigeons: Matching-to-sample with trial-unique video picture stimuli. Animal Learning & Behavior, 16, 436–444.Find this resource:
Wright, A. A., & Delius, J. D. (1994). Scratch and match: Pigeons learn matching and oddity with gravel stimuli. Journal of Experimental Psychology: Animal Behavior Processes, 20(1), 108.Find this resource:
Wright, A. A., & Katz, J. S. (2006). Mechanisms of same/different concept learning in primates and avians. Behavioural Processes, 72, 234–254.Find this resource:
Wright A. A., Katz J. S., Kelly, D. M. (2018). Same/different learning. In J. Vonk & T. Shackelford (Eds.), Encyclopedia of animal cognition and behavior. Cham, Switzerland: Springer.Find this resource:
Wright, A. A., Kelly, D. M., & Katz, J. S. (2018). Comparing cognition by integrating concept learning, proactive interference, and list memory. Learning & Behavior, 46(2), 107–123.Find this resource:
Wright, A. A., Magnotti, J. F., Katz, J. S., Leonard, K., Vernouillet, A., & Kelly, D. M. (2017). Corvids outperform pigeons and primates in learning a basic concept. Psychological Science, 28(4), 437–444.Find this resource:
Yonelinas, A. P. (2002). The nature of recollection and familiarity: A review of years of 30 years of research. Journal of Memory and Language, 46(3), 441–517.Find this resource:
Zentall, T., & Hogan, D. (1974). Abstract concept learning in the pigeon. Journal of Experimental Psychology, 102, 393–398.Find this resource:
Zentall, T. R., & Wasserman, E. A. (Eds.). (2012). The Oxford handbook of comparative cognition. Oxford, U.K.: Oxford University Press.Find this resource:
Zentall, T. R., Wasserman, E. A., Lazareva, O. F., Thompson, R. K. R., & Rattermann, M. J. (2008). Concept learning in animals. Comparative Cognition & Behavior Reviews, 3, 13–45.Find this resource: