Analogical reasoning in a capuchin monkey (Cebus apella)

Previous evidence has suggested that analogical reasoning (recognizing similarities among object relations when the objects themselves are dissimilar) is limited to humans and apes. This study investigated whether capuchin monkeys (Cebus apella) can use analogical reasoning to solve a 3-dimensional search task. The task involved hiding a food item under 1 of 2 or 3 plastic cups of different sizes and then allowing subjects to search for food hidden under the cup of analogous size in their own set of cups. Four monkeys were exposed to a series of relational matching tasks. If subjects reached criterion on these tasks, they were exposed to relational transfer tasks involving novel stimuli. Three of the monkeys failed to reach criterion on the basic relational matching tasks and therefore were not tested further. One monkey, however, revealed above-chance performance on a series of transfer tasks with 3 novel stimuli. This evidence suggests that contrary to previous arguments, a member of a New World monkey species can solve an analogical problem.     

Can rhesus monkeys spontaneously subtract?

Animals, including pigeons, parrots, raccoons, ferrets, rats, New and Old World monkeys, and apes are capable of numerical computations. Much of the evidence for such capacities is based on the use of techniques that require training. Recently, however, several studies conducted under both laboratory and field conditions have employed methods that tap spontaneous numerical representations in animals, including human infants. In this paper, we present the results of 11 experiments exploring the capacity of semi-free-ranging adult rhesus monkeys to spontaneously compute (i.e. single trial, no training) the outcome of subtraction events. In the basic design, we present one quantity of objects on one stage, a second quantity on a second stage, occlude both stages, and then remove one or no objects from each stage. Having watched these events, a subject is then allowed to approach one stage and eat the food objects behind the occluder. Results show that rhesus monkeys correctly compute the outcome of subtraction events involving three or less objects on each stage, even when the identity of the objects is different. Specifically, when presented with two food quantities, rhesus monkeys select the larger quantity following subtractions of one piece of food from two or three; this preference is maintained when subjects must distinguish food from non-food subtractions, and when food is subtracted from either one or both initial quantities. Furthermore, rhesus monkeys are capable of representing zero as well as equality when two identical quantities are contrasted. Results are discussed in light of recent attempts to determine how number is represented in the brains of animals lacking language.     

Spatial relational memory in 9-month-old macaque monkeys

This experiment assesses spatial and nonspatial relational memory in freely moving 9-mo-old and adult (11-13-yr-old) macaque monkeys (Macaca mulatta). We tested the use of proximal landmarks, two different objects placed at the center of an open-field arena, as conditional cues allowing monkeys to predict the location of food rewards hidden in one of two sets of three distinct locations. Monkeys were tested in two different conditions: (1) when local visual cues marked the two sets of potentially baited locations, so that monkeys could use both local and spatial information to discriminate these locations from never-baited locations; and (2) when no local visual cues marked the two sets of potentially baited locations, so that monkeys had to rely on a spatial relational representation of the environment to discriminate these locations. No 9-mo-old or adult monkey associated the presence of the proximal landmarks, at the center of the arena, with the presence of food in one set of three distinct locations. All monkeys, however, discriminated the potentially baited locations in the presence of local visual cues, thus providing evidence of visual discrimination learning. More importantly, all 9-mo-old monkeys tested discriminated the potentially baited locations in absence of the local visual cues, thus exhibiting evidence of spatial relational learning. These findings indicate that spatial memory processes characterized by a relational representation of the environment are present as early as 9 mo of age in macaque monkeys.     

Metacognitive-like information seeking in lion-tailed macaques: a generalized search response after all?

Previous research has demonstrated that Old World primates (both apes and monkeys) seek information about the location of a hidden food item, unless they are privy to the hiding process. This has been cited as evidence of metacognition. However, these results could also be interpreted using non-metacognitive accounts, including a generalized search response to uncertainty, in which subjects reach for food when it is seen, or search for food until it is spotted. In the present research, lion-tailed macaques were tested on an object-choice task. Conditions varied with respect to the visibility of the baiting process, and whether the location of the hidden food could be inferred by logical exclusion. Additionally, the hidden food could be located visually before a choice was made, by peering under the objects through a Plexiglas tray. Across conditions, monkeys consistently looked for the food when it had not been seen, even if its location could be inferred, despite the fact that these monkeys are capable of inference by exclusion. This suggests that apparently ‘metacognitive’ information seeking in monkeys may instead reflect a generalized search strategy. Alternatively, it is possible that monkeys only have metacognitive access to certain types of knowledge, including that obtained visually. Results are discussed with respect to the likelihood of metacognition in this species and the evolutionary emergence of metacognition across species.     

Categorical Perception and Conceptual Judgments by Nonhuman Primates: The Paleological Monkey and the Analogical Ape

Studies of the conceptual abilities of nonhuman primates demonstrate the substantial range of these abilities as well as their limitations. Such abilities range from categorization on the basis of shared physical attributes, associative relations and functions to abstract concepts as reflected in analogical reasoning about relations between relations. The pattern of results from these studies point to a fundamental distinction between monkeys and apes in both their implicit and explicit conceptual capacities. Monkeys, but not apes, might be best regarded as “paleo‐logicans” in the sense that they form common class concepts of identity on the basis of identical predicates (i.e., shared features). The discrimination of presumably more abstract relations commonly involves relatively simple procedural strategies mediated by associative processes likely shared by all mammals. There is no evidence that monkeys can perceive, let alone judge, relations‐between‐relations. This analogical conceptual capacity is found only in chimpanzees and humans. Interestingly, the “analogical ape,” like the child, can make its analogical knowledge explicit only if it is first provided with a symbol system by which propositional representations can be encoded and manipulated.     

Dyslexia and reasoning: The importance of visual processes

Recent research has suggested that individuals with dyslexia rely on explicit visuospatial representations for syllogistic reasoning while most non‐dyslexics opt for an abstract verbal strategy. This paper investigates the role of visual processes in relational reasoning amongst dyslexic reasoners. Expt 1 presents written and verbal protocol evidence to suggest that reasoners with dyslexia generate detailed representations of relational properties and use these to make a visual comparison of objects. Non‐dyslexics use a linear array of objects to make a simple transitive inference. Expt 2 examined evidence for the visual‐impedance effect which suggests that visual information detracts from reasoning leading to longer latencies and reduced accuracy. While non‐dyslexics showed the impedance effects predicted, dyslexics showed only reduced accuracy on problems designed specifically to elicit imagery. Expt 3 presented problems with less semantically and visually rich content. The non‐dyslexic group again showed impedance effects, but dyslexics did not. Furthermore, in both studies, visual memory predicted reasoning accuracy for dyslexic participants, but not for non‐dyslexics, particularly on problems with highly visual content. The findings are discussed in terms of the importance of visual and semantic processes in reasoning for individuals with dyslexia, and we argue that these processes play a compensatory role, offsetting phonological and verbal memory deficits.     

Assessing the concreteness of relational representation

Research has shown that people's ability to transfer abstract relational knowledge across situations can be heavily influenced by the concrete objects that fill relational roles. This article provides evidence that the concreteness of the relations themselves also affects performance. In 3 experiments, participants viewed simple relational patterns of visual objects and then identified these same patterns under a variety of physical transformations. Results show that people have difficulty generalizing to novel concrete forms of abstract relations, even when objects are unchanged. This suggests that stimuli are initially represented as concrete relations by default. In the 2nd and 3rd experiments, the number of distinct concrete relations in the training set was increased to promote more abstract representation. Transfer improved for novel concrete relations but not for other transformations such as object substitution. Results indicate that instead of automatically learning abstract relations, people's relational representations preserve all properties that appear consistently in the learning environment, including concrete objects and concrete relations.     

Core knowledge and its limits: The domain of food

Adults, preschool children, and nonhuman primates detect and categorize food objects according to substance information, conveyed primarily by color and texture. In contrast, they perceive and categorize artifacts primarily by shape and rigidity. The present experiments investigated the origins of this distinction. Using a looking time procedure, Experiment 1 extended previous findings that rhesus macaques (Macaca mulatta) generalize learning about novel food objects by color over changes in shape. Six additional experiments then investigated whether human infants show the same signature patterns of perception and generalization. Nine-month-old infants failed to detect food objects in accord with their intrinsic properties, in contrast to rhesus monkeys tested in previous research with identical displays. Eight-month-old infants did not privilege substance information over other features when categorizing foods, even though they detected and remembered this information. Moreover, infants showed the same property generalization patterns when presented with foods and tools. The category-specific patterns of perception and categorization shown by human adults, children, and adult monkeys therefore were not found in human infants, providing evidence for limits to infants’ domains of knowledge.     

Long-tailed macaques (Macaca fascicularis) understand what conspecifics can see in a competitive situation

Visual perspective taking (VPT), an understanding of what others can see, is a prerequisite for theory of mind (ToM). While VPT in apes is proven, its presence in monkeys is much-debated. Several different paradigms have been developed to test its existence, but all face interpretational problems since results can be explained by simpler cognitive mechanisms than VPT. Therefore, we adjusted one method where two individuals compete for access to food, visible or invisible for the dominant competitor, to preclude cognitively simpler mechanisms. The subordinate long-tailed macaques tested, selected significantly more often the food item invisible than the item visible to the dominant. In most trials, subjects retrieved only one food item and preferred the invisible food item. Surprisingly, they occasionally adopted an alternative strategy to obtain both food items, by first choosing the visible, most at risk food item. Faster animals adopted this strategy proportionally more often than slower ones. Contrary to previous research, our results cannot be explained by simpler cognitive mechanisms, since behavioural reading was prevented by a one-way mirror between the competitor and the food, and accessibility was equal to both food items. This is the first unequivocal evidence of VPT in a monkey species, suggesting that this precursor to ToM is an evolutionarily conserved capacity present in monkeys, apes and humans.     

An assessment of memory awareness in tufted capuchin monkeys (Cebus apella)

Humans, apes, and rhesus monkeys demonstrate memory awareness by collecting information when ignorant and acting immediately when informed. In this study, five capuchin monkeys searched for food after either watching the experimenter bait one of four opaque tubes (seen trials), or not watching (unseen trials). Monkeys with memory awareness should look into the tubes before making a selection only on unseen trials because on seen trials they already know the location of the food. In Experiment 1, one of the five capuchins looked significantly more often on unseen trials. In Experiment 2, we ensured that the monkeys attended to the baiting by interleaving training and test sessions. Three of the five monkeys looked more often on unseen trials. Because monkeys looked more often than not on both trial types, potentially creating a ceiling effect, we increased the effort required to look in Experiment 3, and predicted a larger difference in the probability of looking between seen and unseen trials. None of the five monkeys looked more often on unseen trials. These findings provide equivocal evidence for memory awareness in capuchin monkeys using tests that have yielded clear evidence in humans, apes, and rhesus monkeys.     

Intuitive optics: what great apes infer from mirrors and shadows

There is ongoing debate about the extent to which nonhuman animals, like humans, can go beyond first-order perceptual information to abstract structural information from their environment. To provide more empirical evidence regarding this question, we examined what type of information great apes (chimpanzees, bonobos, and orangutans) gain from optical effects such as shadows and mirror images. In an initial experiment, we investigated whether apes would use mirror images and shadows to locate hidden food. We found that all examined ape species used these cues to find the food. Follow-up experiments showed that apes neither confused these optical effects with the food rewards nor did they merely associate cues with food. First, naïve chimpanzees used the shadow of the hidden food to locate it but they did not learn within the same number of trials to use a perceptually similar rubber patch as indicator of the hidden food reward. Second, apes made use of the mirror images to estimate the distance of the hidden food from their own body. Depending on the distance, apes either pointed into the direction of the food or tried to access the hidden food directly. Third, apes showed some sensitivity to the geometrical relation between mirror orientation and mirrored objects when searching hidden food. Fourth, apes tended to interpret mirror images and pictures of these mirror images differently depending on their prior knowledge. Together, these findings suggest that apes are sensitive to the optical relation between mirror images and shadows and their physical referents.     

Analog number representations in mongoose lemurs (Eulemur mongoz): evidence from a search task

A wealth of data demonstrating that monkeys and apes represent number have been interpreted as suggesting that sensitivity to number emerged early in primate evolution, if not before. Here we examine the numerical capacities of the mongoose lemur (Eulemur mongoz), a member of the prosimian suborder of primates that split from the common ancestor of monkeys, apes and humans approximately 47–54 million years ago. Subjects observed as an experimenter sequentially placed grapes into an opaque bucket. On half of the trials the experimenter placed a subset of the grapes into a false bottom such that they were inaccessible to the lemur. The critical question was whether lemurs would spend more time searching the bucket when food should have remained in the bucket, compared to when they had retrieved all of the food. We found that the amount of time lemurs spent searching was indicative of whether grapes should have remained in the bucket, and furthermore that lemur search time reliably differentiated numerosities that differed by a 1:2 ratio, but not those that differed by a 2:3 or 3:4 ratio. Finally, two control conditions determined that lemurs represented the number of food items, and neither the odor of the grapes, nor the amount of grape (e.g., area) in the bucket. These results suggest that mongoose lemurs have numerical representations that are modulated by Webers Law.     

Multimodal communication by captive chimpanzees (Pan troglodytes)

Many studies have shown that apes and monkeys are adept at cross-modal matching tasks requiring the subject to identify objects in one modality when information regarding those objects has been presented in a different modality. However, much less is known about non-human primates’ production of multimodal signaling in communicative contexts. Here, we present evidence from a study of 110 chimpanzees demonstrating that they select the modality of communication in accordance with variations in the attentional focus of a human interactant, which is consistent with previous research. In each trial, we presented desirable food to one of two chimpanzees, turning mid-way through the trial from facing one chimpanzee to facing the other chimpanzee, and documented their communicative displays, as the experimenter turned towards or away from the subjects. These chimpanzees varied their signals within a context-appropriate modality, displaying a range of different visual signals when a human experimenter was facing them and a range of different auditory or tactile (attention-getting) signals when the human was facing away from them; this finding extends previous research on multimodal signaling in this species. Thus, in the impoverished circumstances characteristic of captivity, complex signaling tactics are nevertheless exhibited by chimpanzees, suggesting continuity in intersubjective psychological processes in humans and apes.     

Gaze alternation during “pointing” by squirrel monkeys (<Emphasis Type="Italic">Saimiri sciureus</Emphasis>)?

Gaze alternation (GA) is considered a hallmark of pointing in human infants, a sign of intentionality underlying the gesture. GA has occasionally been observed in great apes, and reported only anecdotally in a few monkeys. Three squirrel monkeys that had previously learned to reach toward out-of-reach food in the presence of a human partner were videotaped while the latter visually attended to the food, a distractor object, or the ceiling. Frame-by-frame video analysis revealed that, especially when reaching toward the food, the monkeys rapidly and repeatedly switched between looking at the partner’s face and the food. This type of GA suggests that the monkeys were communicating with the partner. However, the monkeys’ behavior was not influenced by changes in the partner’s focus of attention.     

Picture recognition of food by macaques (<Emphasis Type="Italic">Macaca silenus</Emphasis>)

Pictorial representations of three-dimensional objects are often used to investigate animal cognitive abilities; however, investigators rarely evaluate whether the animals conceptualize the two-dimensional image as the object it is intended to represent. We tested for picture recognition in lion-tailed macaques by presenting five monkeys with digitized images of familiar foods on a touch screen. Monkeys viewed images of two different foods and learned that they would receive a piece of the one they touched first. After demonstrating that they would reliably select images of their preferred foods on one set of foods, animals were transferred to images of a second set of familiar foods. We assumed that if the monkeys recognized the images, they would spontaneously select images of their preferred foods on the second set of foods. Three monkeys selected images of their preferred foods significantly more often than chance on their first transfer session. In an additional test of the monkeys’ picture recognition abilities, animals were presented with pairs of food images containing a medium-preference food paired with either a high-preference food or a low-preference food. The same three monkeys selected the medium-preference foods significantly more often when they were paired with low-preference foods and significantly less often when those same foods were paired with high-preference foods. Our novel design provided convincing evidence that macaques recognized the content of two-dimensional images on a touch screen. Results also suggested that the animals understood the connection between the two-dimensional images and the three-dimensional objects they represented.     

Strategies used to combine seriated cups by chimpanzees (Pan troglodytes), bonobos (Pan paniscus), and capuchins (Cebus apella).

The authors investigated strategies used to combine seriated cups by apes (Pan troglodytes and P. paniscus) and monkeys (Cebus apella) using a protocol reported in P. M. Greenfield, K. Nelson, and E. Saltzman's (1972) study with children. It was hypothesized that apes would exhibit more hierarchical combinations of cups than monkeys, given apes' language capacity, and that apes would seriate the cups more efficiently than monkeys. As predicted, apes made many structures with the cups using a variety of strategies, and monkeys rarely combined the cups. After a training phase to orient monkeys to the task, the 2 genera did not differ in the strategies used to combine the cups or in efficiency in seriating the cups. Success in this task suggests that sensorimotor versions of hierarchically organized combinatorial activity are well within apes' and monkeys' abilities.     

Structural Alignment during Similarity Comparisons

Similarity comparisons are a basic component of cognition, and there are many elegant models of this process. None of these models describe comparisons of structured representations, although mounting evidence suggests that mental representations are well characterized by structured hierarchical systems of relations. We propose that structured representations can be compared via structural alignment, a process derived from models of analogical reasoning. The general prediction of structural alignment is that similarity comparisons lead subjects to attend to the matching relational structure in a pair of items. This prediction is illustrated with a computational simulation that also suggests that the strength of the relational focus is diminished when the relational match is impoverished, or when competing interpretations lead to rich object matches. These claims are tested in four experiments using the one-shot mapping paradigm, which places object similarity and relational similarity in opposition. The results support the hypothesis that similarity involves the alignment of structured representations.     

Bayesian analogy with relational transformations

How can humans acquire relational representations that enable analogical inference and other forms of high-level reasoning? Using comparative relations as a model domain, we explore the possibility that bottom-up learning mechanisms applied to objects coded as feature vectors can yield representations of relations sufficient to solve analogy problems. We introduce Bayesian analogy with relational transformations (BART) and apply the model to the task of learning first-order comparative relations (e.g., larger, smaller, fiercer, meeker) from a set of animal pairs. Inputs are coded by vectors of continuous-valued features, based either on human magnitude ratings, normed feature ratings (De Deyne et al., 2008), or outputs of the topics model (Griffiths, Steyvers, & Tenenbaum, 2007). Bootstrapping from empirical priors, the model is able to induce first-order relations represented as probabilistic weight distributions, even when given positive examples only. These learned representations allow classification of novel instantiations of the relations and yield a symbolic distance effect of the sort obtained with both humans and other primates. BART then transforms its learned weight distributions by importance-guided mapping, thereby placing distinct dimensions into correspondence. These transformed representations allow BART to reliably solve 4-term analogies (e.g., larger:smaller::fiercer:meeker), a type of reasoning that is arguably specific to humans. Our results provide a proof-of-concept that structured analogies can be solved with representations induced from unstructured feature vectors by mechanisms that operate in a largely bottom-up fashion. We discuss potential implications for algorithmic and neural models of relational thinking, as well as for the evolution of abstract thought.     

Does the brain's ventral visual pathway compute object shape?

A rich behavioral literature has shown that human object recognition is supported by a representation of shape that is tolerant to variations in an object's appearance. Such 'global' shape representations are achieved by describing objects via the spatial arrangement of their local features, or structure, rather than by the appearance of the features themselves. However, accumulating evidence suggests that the ventral visual pathway – the primary substrate underlying object recognition – may not represent global shape. Instead, ventral representations may be better described as a basis set of local image features. We suggest that this evidence forces a reevaluation of the role of the ventral pathway in object perception and posits a broader network for shape perception that encompasses contributions from the dorsal pathway.