Means-means-end tool choice in cotton-top tamarins (Saguinus oedipus): finding the limits on primates’ knowledge of tools

Most studies of animal tool use require subjects to use one object to gain access to a food reward. In many real world situations, however, animals perform more than one action in sequence to achieve their goals. Of theoretical interest is whether animals have the cognitive capacity to recognize the relationship between consecutive action sequences in which there may be one overall goal and several subgoals. Here we ask if cotton-top tamarins, a species that in captivity uses tools to solve means-end problems, can go one step further and use a sequence of tools (means) to obtain food (end). We first trained subjects to use a pulling tool to obtain a food reward. After this initial training, subjects were presented with problems in which one tool had to be used in combination with a second in order to obtain food. Subjects showed great difficulty when two tools were required to obtain the food reward. Although subjects attended to the connection between the tool and food reward, they ignored the physical connection between the two tools. After training on a two-tool problem, we presented subjects with a series of transfer tests to explore if they would generalize to new types of connections between the tools. Subjects readily transferred to new connections. Our results therefore provide the first evidence to date that tamarins can learn to solve problems involving two tools, but that they do so only with sufficient training.     

Representing tools: how two non-human primate species distinguish between the functionally relevant and irrelevant features of a tool

Few studies have examined whether non-human tool-users understand the properties that are relevant for a tool's function. We tested cotton-top tamarins (Saguinus oedipus) and rhesus macaques (Macaca mulatta) on an expectancy violation procedure designed to assess whether these species make distinctions between the functionally relevant and irrelevant features of a tool. Subjects watched an experimenter use a tool to push a grape down a ramp, and then were presented with different displays in which the features of the original tool (shape, color, orientation) were selectively varied. Results indicated that both species looked longer when a newly shaped stick acted on the grape than when a newly colored stick performed the same action, suggesting that both species perceive shape as a more salient transformation than color. In contrast, tamarins, but not rhesus, attended to changes in the tool's orientation. We propose that some non-human primates begin with a predisposition to attend to a tool's shape and, with sufficient experience, develop a more sophisticated understanding of the features that are functionally relevant to tools.     

Donor payoffs and other-regarding preferences in cotton-top tamarins (Saguinus oedipus)

Helping others at no cost to oneself is a simple way to demonstrate other-regarding preferences. Yet, primates exhibit mixed results for other-regarding preferences: chimpanzees and tamarins do not show these effects, whereas capuchin monkeys and marmosets preferentially give food to others. One factor of relevance to this no-cost food donation is the payoff to the donor. Though donors always receive the same payoffs regardless of their choice, previous work varies in whether they receive either a food reward or no food reward. Here, I tested cotton-top tamarins in a preferential giving task. Subjects could choose from two tools, one of which delivered food to a partner in an adjacent cage and the other of which delivered food to an empty cage. Thus, subjects could preferentially give or withhold food from a partner. I varied whether subjects received food payoffs, whether a partner was present or absent, and whether the partner was a non-cagemate or the subject’s mate. Results showed that the subjects’ overall motivation to pull either tool declined when they did not receive any food. Additionally, they did not preferentially donate or withhold food, regardless of their own payoff or their relationship with the partner. Thus, cotton-top tamarins do not take advantage of cost-free food giving, either when they might gain in the future (mates) or when they have no opportunity for future interactions (non-cagemates).     

Means to an end: Neotropical parrots manage to pull strings to meet their goals

Although parrots share with corvids and primates many of the traits believed to be associated with advanced cognitive processing, knowledge of parrot cognition is still limited to a few species, none of which are Neotropical. Here we examine the ability of three Neotropical parrot species (Blue-Fronted Amazons, Hyacinth and Lear’s macaws) to spontaneously solve a novel physical problem: the string-pulling test. The ability to pull up a string to obtain out-of-reach food has been often considered a cognitively complex task, as it requires the use of a sequence of actions never previously assembled, along with the ability to continuously monitor string, food and certain body movements. We presented subjects with pulling tasks where we varied the spatial relationship between the strings, the presence of a reward and the physical contact between the string and reward to determine whether (1) string-pulling is goal-oriented in these parrots, (2) whether the string is recognized as a means to obtain the reward and (3) whether subjects can visually determine the continuity between the string and the reward, selecting only those strings for which no physical gaps between string and reward were present. Our results show that some individuals of all species were able to use the string as a means to reach a specific goal, in this case, the retrieval of the food treat. Also, subjects from both macaw species were able to visually determine the presence of physical continuity between the string and reward, making their choices consistently with the recognition that no gaps should be present between the string and the reward. Our findings highlight the potential of this taxonomic group for the understanding of the underpinnings of cognition in evolutionarily distant groups such as birds and primates.     

How prosimian primates represent tools: experiments with two lemur species (Eulemur fulvus and Lemur catta)

The authors examined how 2 lemur species (Eulemur fulvus and Lemur catta) reason about tools. Experiment 1 allowed subjects to use 1 of 2 canes to retrieve an inaccessible food reward. Lemurs learned to solve this problem as quickly as other primates. Experiment 2 then presented subjects with novel tools differing from the originals along one featural dimension. Subjects attended more to tools' sizes than to their colors and made no distinction between tools' shapes and textures. Experiments 3 and 4 presented problems in which some of the tools' orientations had to be modified relative to the food. Subjects performed well on these problems, sometimes modifying the position of the tool. These results are discussed in light of the performance of other primates on this task.     

The role of landmarks in cotton-top tamarin spatial foraging: evidence for geometric and non-geometric features

We report experiments on captive cotton-top tamarins (Saguinus oedipus) designed to explore two components of spatial foraging. First, do tamarins have the capacity to extract geometric information concerning the spatial relationship between a landmark and a piece of food located above or below it? Second, when tamarins use a landmark to find a target location, what non-geometric features of the landmark do they encode? To explore these problems, we created an artificial jungle environment and trained subjects to find food either above or below a target object (i.e., landmark). Once subjects successfully located the food, we transformed various features associated with the landmark, including its color, orientation, and shape; we also manipulated the landmark-food reward distance, the overall shape of the jungle, and the number and position of landmarks. Results showed that the tamarins' success in finding the food reward was not affected by landmark color, orientation, number, or overall shape of the jungle, suggesting that with respect to the particular test conditions, these features are not relevant to the representation of a landmark. Subjects also generalized to novel landmark-food distances, suggesting that they had integrated geometric (i.e., above/below) with non-geometric (i.e., color/shape) features. Performance was negatively affected by changes to the shape of the landmark, indicating that this feature is critical to the representation of a landmark. Accepted after revision: 7 August 2001 Electronic Publication     

Does effort influence inequity aversion in cotton-top tamarins (Saguinus oedipus)?

The human sense of fairness entails sensitivity not just to equality, the equal division of resources, but also to merit, the relationship between an individual’s share of resources and how hard they worked for their share. Recent evidence suggests that our sensitivity to equality has deep phylogenetic roots: several nonhuman animal species show an aversion to unequal reward distributions. However, the extent to which nonhuman animals share sensitivity to merit remains poorly understood, largely because previous studies have failed to properly manipulate work effort in inequity aversion tasks. Here, we tested whether cotton-top tamarins (Saguinus oedipus) would exhibit a differential response to inequity when acquiring rewards was either (1) effortful or (2) effortless. Subjects engaged in a pulling task in which they had an opportunity to deliver a disadvantageously unequal distribution of food to themselves and a partner (one piece for self, four pieces for partner). We made delivery effortful by adding a weight to the pulling handle. Critically, effort was calibrated to each individual. Results show that individuals varied markedly in their response to effort, highlighting the importance of manipulating work effort at the individual level. Overall, subjects showed little aversion to inequity. However, subjects were slightly less likely to accept inequity when doing so was effortful, although this effect was pronounced in only one individual. Our findings suggest a new method for capturing individual variation in effort and for studying the roots of the concept of merit in nonhuman animals.     

The relationship between problem solving and inhibitory control: cotton-top tamarin (Saguinus oedipus) performance on a reversed contingency task

To explore the relationship between problem solving and inhibitory control, the authors present 4 experiments on cotton-top tamarins (Saguinus oedipus) using a reverse-reward contingency task. In Experiment 1, 1 group of tamarins was given a choice between a small and a large quantity of food. Whichever quantity the tamarins reached for first, they received the alternative. The tamarins consistently picked the larger quantity, thereby receiving the smaller. A 2nd group of tamarins was given the same task, except that if they reached for the larger quantity of food, they received nothing. The tamarins continued to pick the larger quantity, even though this resulted in no food. In addition, most of the tamarins continued to pick the larger quantity even when the food payoff for choosing the smaller quantity was increased (Experiment 2) or when the visual salience of the food was reduced (Experiment 3). Experiment 4 was based on the finding that chimpanzees (Pan troglodytes) that have been trained on the concept of number can solve the reversed contingency task if the food is replaced by Arabic numerals. With the help of a color association, and a higher cost incurred by picking the color associated with 3 food items, the tamarins learned to pick the color associated with 1 food item. These results are compared with those obtained from studies of other primate species, highlighting the importance of comparative studies of problem solving that use comparable methods.     

Gorillas (Gorilla gorilla) and orangutans (Pongo pygmaeus) encode relevant problem features in a tool-using task

Two important elements in problem solving are the abilities to encode relevant task features and to combine multiple actions to achieve the goal. The authors investigated these 2 elements in a task in which gorillas (Gorilla gorilla) and orangutans (Pongo pygmaeus) had to use a tool to retrieve an out-of-reach reward. Subjects were able to select tools of an appropriate length to reach the reward even when the position of the reward and tools were not simultaneously visible. When presented with tools that were too short to retrieve the reward, subjects were more likely to refuse to use them than when tools were the appropriate length. Subjects were proficient at using tools in sequence to retrieve the reward.     

Transport of tools to food sites in tufted capuchin monkeys (<Emphasis Type="Italic">Cebus apella</Emphasis>)

Tool use and transport represent cognitively important aspects of early hominid evolution, and nonhuman primates are often used as models to examine the cognitive, ecological, morphological and social correlates of these behaviors in order to gain insights into the behavior of our early human ancestors. In 2001, Jalles-Filho et al. found that free-ranging capuchin monkeys failed to transport tools (stones) to food sites (nuts), but transported the foods to the tool sites. This result cast doubt on the usefulness of Cebus to model early human tool-using behavior. In this study, we examined the performance of six captive tufted capuchin monkeys (Cebus apella) in a tool transport task. Subjects were provided with the opportunity to transport two different tools to fixed food reward sites when the food reward was visible from the tool site and when the food reward was not visible from the tool site. We found that the subjects quickly and readily transported probing tools to an apparatus baited with syrup, but rarely transported stones to a nut-cracking apparatus. We suggest that the performance of the capuchins here reflects an efficient foraging strategy, in terms of energy return, among wild Cebus monkeys.     

Experimental field study of hand preference in wild black-horned (<Emphasis Type="Italic">Cebus nigritus</Emphasis>) and white-faced (<Emphasis Type="Italic">Cebus capucinus</Emphasis>) capuchins: evidence for individual and species differences

In this experimental field investigation, we compare the degree to which wild capuchins in Brazil (Cebus nigritus) and Costa Rica (Cebus capucinus) exhibit individual- and population-level handedness during three visually-guided tasks. These tasks required reaching to remove a large leaf covering a hidden food reward, seizing the food reward, and manipulating a tool (pulling a wooden dowel) in order to obtain access to an embedded food reward. Studies in some populations of captive capuchins indicate evidence for both individual hand preferences and population-level handedness. In this study, six of eight wild C. capucinus and six of seven wild C. nigritus exhibited a significant hand preference during individual tasks, but no individual exhibited a consistent preference across all three tasks. Task-specialization, or the tendency for most individuals in the same group or population to use the same hand to accomplish a particular task, also was evaluated. Cebus nigritus showed a significant bias toward the use of the right hand in removing the leaf. Although the number of individual capuchins in both species that manipulated the dowels was limited (N = 7), each individual that manipulated the dowels in eight or more instances had a positive handedness index, suggesting a greater use of the right hand to accomplish this task. Overall, our results provide preliminary support for individual- and population-level handedness in wild capuchin monkeys.     

Inhibitory control and response selection in problem solving: how cotton-top tamarins (Saguinus oedipus) overcome a bias for selecting the larger quantity of food

When presented with a choice between 1 and 3 pieces of food in a type of reversed contingency task, 4 cotton-top tamarins (Saguinus oedipus) consistently chose the 3 pieces of food and received nothing, even though the choice of 1 piece would have yielded 3. However, in a task in which the tamarins received the 1 piece of food when they chose it, all subjects learned to select 1 over 3. Thus, the tamarins' prior failure on the reversed contingency task did not result entirely from an inherent inability to suppress the prepotent response of reaching to the larger of 2 quantities of food. After the experience of selecting the smaller quantity and receiving it, all of the tamarins solved the version of the reversed contingency task that they failed initially. These results suggest that the tamarins' initial failure may have reflected a difficulty with selecting an alternative response option.     

Tubes,tables and traps: great apes solve two functionally equivalent trap tasks but show no evidence of transfer across tasks

Previous studies on tool using have shown that presenting subjects with certain modifications in the experimental setup can substantially improve their performance. However, procedural modifications (e.g. trap table task) may not only remove task constraints but also simplify the problem conceptually. The goal of this study was to design a variation of the trap-table that was functionally equivalent to the trap-tube task. In this new task, the subjects had to decide where to insert the tool and in which direction the reward should be pushed. We also administered a trap-tube task that allowed animals to push or rake the reward with the tool to compare the subjects' performance on both tasks. We used a larger sample of subjects than in previous studies and from all the four species of great apes (Gorilla gorilla, Pan troglodytes, Pan paniscus, and Pongo pygmaeus). The results showed that apes performed better in the trap-platform task than in the trap-tube task. Subjects solved the tube task faster than in previous studies and they also preferred to rake in rather than to push the reward out. There was no correlation in the level of performance between both tasks, and no indication of interspecies differences. These data are consistent with the idea that apes may possess some specific causal knowledge of traps but may lack the ability to establish analogical relations between functional equivalent tasks.     

Tool choice on the basis of rigidity in capuchin monkeys

Wild capuchin monkeys select stone tools to crack open nuts on the basis of their weight and friability, two non-visual functional properties. Here, we investigated whether they would select new stick-like tools on the basis of their rigidity. In Experiment 1, subjects faced an out-of-reach reward and a choice of three unfamiliar tools differing in color, diameter, material, and rigidity. In order to retrieve the reward, capuchins needed to select the rigid tool exemplar. Capuchins gathered information regarding tools’ pliability either by (1) manipulating the tools themselves (manipulation condition), (2) observing a human demonstrator repeatedly bending the tools (observation condition), or (3) seeing the tools placed on a platform without any manipulation taking place (visual static condition). Subjects selected the rigid tool above chance levels in both the manipulation and observation conditions, but not in the visual static condition. In Experiment 2, subjects needed to select and use a flexible tool to access a liquid reward (as opposed to the rigid tool, as in previous experiment). Again, capuchins selected above chance levels the appropriate tool (i.e., flexible), thus demonstrating a good appreciation of the relation between the tool properties and the task requirements.     

The win-stay rule in foraging decisions by free-ranging titi monkeys (Callicebus cupreus cupreus) and Tamarins (Saguinus imperator imperator and Saguinus fuscicollis weddelli)

Rainforest primates need to apply distinct foraging rules for efficiently using the spatial knowledge of the distribution of resources showing different temporal patterns of renewal. A win-stay rule is very important for exploiting abundant, long-lasting resources. Here, the author tests the use of this rule in wild groups of emperor tamarins (Saguinus imperator imperator), saddle-back tamarins (Saguinus fuscicollis weddelli), and titi monkeys (Callicebus cupreus cupreus) during a series of foraging tasks. Four feeding stations composed of 8 visually similar feeding platforms (2 containing a food reward and 6 containing a sham reward) were constructed. The location of food rewards was reliable during some experiments and unreliable during others. All 3 species consistently adopted a win-stay rule for returning to reward platforms when their location was predictable over time but stopped using it when their spatial distribution changed randomly across experimental trials.     

Raking it in: the impact of enculturation on chimpanzee tool use

Recent evidence for different tool kits, proposed to be based upon culture-like transmission, have been observed across different chimpanzee communities across Western Africa. In light of these findings, the reported failures by seven captive juvenile chimpanzees tested with 27 tool use tasks (Povinelli 2000) seem enigmatic. Here we report successful performance by a group of nine captive, enculturated chimpanzees, and limited success by a group of six semi-enculturated chimpanzees, on two of the Povinelli tasks, the Flimsy Tool task, and the Hybrid Tool task. All chimpanzees were presented with a rake with a flimsy head and a second rake with a rigid head, either of which could be used to attempt to retrieve a food reward that was out of reach. The rigid rake was constructed such that it had the necessary functional features to permit successful retrieval, while the flimsy rake did not. Both chimpanzee groups in the present experiment selected the functional rigid tool correctly to use during the Flimsy Tool task. All animals were then presented with two “hybrid rakes” A and B, with one half of each rake head constructed from flimsy, non-functional fabric, and the other half of the head was made of wood. Food rewards were placed in front of the rigid side of Rake A and the flimsy side of Rake B. To be successful, the chimps needed to choose the rake that had the reward in front of the rigid side of the rake head. The fully enculturated animals were successful in selecting the functional rake, while the semi-enculturated subjects chose randomly between the two hybrid tools. Compared with findings from Povinelli, whose non-enculturated animals failed both tasks, our results demonstrate that chimpanzees reared under conditions of semi-enculturation could learn to discriminate correctly the necessary tool through trial-and-error during the Flimsy Tool task, but were unable to recognize the functional relationship necessary for retrieving the reward with the “hybrid” rake. In contrast, the enculturated chimpanzees were correct in their choices during both the Flimsy Tool and the Hybrid Tool tasks. These results provide the first empirical evidence for the differential effects of enculturation on subsequent tool use capacities in captive chimpanzees.     

Tool selectivity in a non-primate,the New Caledonian crow (Corvus moneduloides)

We present an experiment showing that New Caledonian crows are able to choose tools of the appropriate size for a novel task, without trial-and-error learning. This species is almost unique amongst all animal species (together with a few primates) in the degree of use and manufacture of polymorphic tools in the wild. However, until now, the flexibility of their tool use has not been tested. Flexibility, including the ability to select an appropriate tool for a task, is considered to be a hallmark of complex cognitive adaptations for tool use. In experiment 1, we tested the ability of two captive birds (one male, one female), to select a stick (from a range of lengths provided) matching the distance to food placed in a horizontal transparent pipe. Both birds chose tools matching the distance to their target significantly more often than would be expected by chance. In experiment 2, we used a similar task, but with the tools placed out of sight of the food pipe, such that the birds had to remember the distance of the food before selecting a tool. The task was completed only by the male, who chose a tool of sufficient length significantly more often than chance but did not show a preference for a matching length. Electronic Publication     

Task constraints mask great apes' ability to solve the trap-table task

Researchers have investigated animals' causal knowledge with a task requiring subjects to use a tool to bring a reward within reach whilst avoiding a trap. Previous studies have suggested limitations in the ability of several species to avoid traps in tubes or tables. However, certain features may have inflated task difficulty. We tested 20 chimpanzees (Pan troglodytes), 7 orangutans (Pongo pygmaeus), 5 bonobos (Pan paniscus), and 5 gorillas (Gorilla gorilla) in the trap-table--a task in which subjects have to pull one of two rakes prepositioned behind two rewards on a flat surface. One of the rewards is in front of a trap into which it will fall. We investigated the effect of trap type, tool type, the number of available tools, and reinforcement regime on performance. We replicated previous findings showing that apes failed to choose the correct rake above chance. However, when they could instead choose where to insert a single tool, around 80% of the apes solved the trap-table task in the first trial, revealing an important effect of task constraints on their performance.     

Use of experimenter-given cues in visual co-orienting and in an object-choice task by a new world monkey species, cotton top tamarins (Saguinus oedipus)

Two methods assessed the use of experimenter-given directional cues by a New World monkey species, cotton top tamarins (Saguinus oedipus). Experiment 1 used cues to elicit visual co-orienting toward distal objects. Experiment 2 used cues to generate responses in an object-choice task. Although there were strong positive correlations between monkey pairs to co-orient, visual co-orienting with a human experimenter occurred at a low frequency to distal objects. Human hand pointing cues generated more visual co-orienting than did eye gaze to distal objects. Significant accurate choices of baited cups occurred with human point and tap cues and human look cues. Results highlight the importance of head and body orientation to induce shared attention in cotton top tamarins, both in a task that involved food getting and a task that did not.     

Tool-use learning in Tonkean macaques (Macaca tonkeana)

The transmission of tool use is a rare event in monkeys. Such an event arose in a group of semi-free-ranging Tonkean macaques (Macaca tonkeana) in which leaning a pole against the parks fence (branch leaning) appeared and spread to several males. This prompted us to test individual and social learning of this behavior in seven young males. In the first experiment, three males learned individually to obtain a food reward using a wooden pole as a climbing tool. They began using the pole to retrieve the reward only when they could alternatively experience acting on the object and reaching the target. In a second experiment, we first tested whether four other subjects could learn branch leaning after having observed a group-mate performing the task. Despite repeated opportunities to observe the demonstrator, they did not learn to use the pole as a tool. Hence we exposed the latter subjects to individual learning trials and they succeeded in the task. Tool use was not transmitted in the experimental situation, which contrasts with observations in the park. We can conclude that the subjects were not able to recognize the target as such. It is possible that they recognized it and learned the task individually when we alternated the opportunity to act upon the object and to reach the reward. This suggests that these macaques could then have associated the action they exercised upon the pole and the use of the pole as a means to reach the reward.