Tool use in animals is the employment of an external object to alter another object, another organism, or the user itself. Familiar examples include chimpanzees extracting termites with plant stems, crows probing for concealed prey, and sea otters opening shellfish with stones. Tool use occurs in several distantly related groups, although its frequency, complexity, and importance vary greatly among species and populations. It is not a single behavior or a universal measure of intelligence: a readily available stone used once and a carefully manufactured probe used throughout life pose different biological questions.
The subject connects ethology, comparative cognition, ecology, and the study of behavioral traditions. Tools can give animals access to otherwise inaccessible foods, protect vulnerable body parts, or modify social interactions. Their use depends on an interaction between bodily capabilities, environmental opportunities, individual experience, and sometimes learning from others. Research therefore examines not only successful performances but also how animals acquire techniques, select materials, and respond when a familiar problem changes.
Animal tool use is also important for understanding human technological evolution. Comparisons with living species reveal possible precursors of technological behavior without treating any contemporary animal as a direct model of a human ancestor. Field observations, controlled experiments, and archaeological traces provide complementary evidence about how object manipulation becomes an effective and persistent part of an animal’s way of life.
Definitions, Boundaries, and Research History
A widely used family of definitions describes tool use as the controlled manipulation of an external, unattached or manipulable attached object to change the form, position, or condition of a target. The user must orient or operate the object appropriately during, or immediately before, its application. Exact wording differs among researchers, especially over whether an object must be held and whether fixed environmental features can count. These disagreements matter because they determine which behaviors enter comparative datasets.
The distinction between a tool and a construction is particularly important. A bird carrying twigs to build a nest ordinarily uses the twigs as building material rather than as implements acting on a separate target. A beaver dam is generally classified as construction rather than tool use. Likewise, striking a shell against a fixed rock is sometimes distinguished as “anvil use” or “proto-tool use,” whereas picking up a stone and striking the shell usually meets stricter definitions. These categories describe different mechanical relationships; they do not establish a ranking of cognitive sophistication.
Tool manufacture involves altering an object before or during use so that it functions more effectively. Removing leaves from a twig, shaping a hooked probe, and fraying a stem into a brush can qualify. Manufacture need not involve a detailed mental blueprint: repeated actions guided by feedback may produce a functional implement. Researchers also distinguish sequential tool use, in which several implements act on the same resource, from metatool use, in which one tool is used to obtain or modify another. Using a short stick to retrieve a longer stick is a common experimental example of the latter.
Scientific attention grew through both experimentation and field observation. Wolfgang Köhler’s studies of captive chimpanzees on Tenerife, published in German in 1917, examined such actions as joining sticks and stacking boxes to reach food. They helped establish animal problem-solving as a subject for systematic investigation. In 1960, Jane Goodall observed wild chimpanzees at Gombe using stems to obtain termites and modifying vegetation for that purpose. These observations challenged claims that making and using tools uniquely distinguished humans.
Later work broadened the subject beyond apes. Long-term studies documented tool traditions in monkeys, dolphins, and birds, while experiments tested unfamiliar materials and causal problems. Benjamin Beck’s 1980 synthesis, Animal Tool Behavior, helped organize a scattered literature into a comparative field. Subsequent definitions increasingly emphasized observable relationships among user, implement, and target rather than assumptions about intention. This approach allows a behavior to be identified as tool use while leaving the animal’s underlying understanding open to separate investigation.
Primates: Extraction, Percussion, and Tool Sets
Among nonhuman primates, chimpanzees display an especially varied repertoire. They insert stems into termite galleries, dip for ants, use leaf sponges to collect water, and crack nuts with hammers and anvils. These behaviors are not distributed uniformly across Africa. Particular techniques occur in some communities but not others, and their form varies with available materials, prey characteristics, local conditions, and learning opportunities. A statement that chimpanzees “use tools” therefore compresses substantial diversity within the species.
Termite gathering illustrates the importance of tool properties. A probe must enter a narrow passage yet remain stiff enough to be controlled. At some sites, chimpanzees manufacture brush-tipped probes by separating plant fibers, increasing the surface available for termites to grip. In the Goualougo Triangle of the Republic of the Congo, some termite-extraction sequences involve a sturdy implement for opening access followed by a more flexible fishing probe. Such tool sets divide a task into mechanically distinct operations rather than requiring one implement to perform every function.
Nut cracking presents a different problem. In parts of West Africa, chimpanzees place nuts on an anvil and strike them with stone or wooden hammers. Success requires coordinating the target, support, and striking object, while adjusting force to break the shell without unnecessarily crushing the kernel. Hammer choice is affected by nut hardness, stone availability, and transport costs. Youngsters may manipulate the correct objects for years before becoming proficient, showing that participation and mastery are not equivalent.
Wild orangutans also use tools, notably at some Sumatran sites where sticks help extract insects or obtain seeds from protected fruits. Their behavior demonstrates that sophisticated manipulation need not depend on the chimpanzee pattern of social organization or terrestrial activity. Wild gorillas have been observed using detached branches in contexts such as assessing water depth or providing support in swampy terrain. Such reports establish that tool use occurs, but they do not imply that it is as routine or diverse as chimpanzee extractive foraging.
Habitual stone use is not confined to great apes. Capuchin monkeys in Brazil crack hard foods with stone hammers, sometimes transporting substantial stones to suitable anvils. At some sites they also use stones for digging or other pounding activities. Their behavior is particularly valuable for comparisons because capuchins are evolutionarily distant from apes and differ greatly in body size and hand anatomy.
Some macaques, including populations of long-tailed macaques along Southeast Asian coasts, use stones to process shellfish and other foods. Different actions suit attached oysters, mobile shellfish, and nuts. Across these examples, the common feature is not a single primate “tool instinct,” but the integration of manipulative ability with specific feeding problems and developmental experience.
Birds: Probes, Hooks, and Percussive Techniques
Bird tool use demonstrates that hands are not necessary for precise object manipulation. Bills and feet can position, carry, stabilize, and modify implements. The best-studied examples occur among corvids, but tools also appear in woodpecker finches, some parrots, Egyptian vultures, and other birds. As in primates, isolated observations must be distinguished from behaviors that recur regularly in natural populations.
The New Caledonian crow is a particularly accomplished manufacturer. In its native range, it makes probes from plant material to extract invertebrates from holes and crevices. Some tools incorporate hooks fashioned from forked twigs; others are cut or torn from pandanus leaves. Manufacture involves choices about raw material, working end, and dimensions. Hooked tools can engage prey that a straight probe might fail to extract, although performance depends on the particular target and cavity.
A captive crow named Betty attracted attention in 2002 when she bent a piece of wire into a hook and used it to retrieve a food-containing bucket from a tube. The event demonstrated flexible use of an unfamiliar material, but its interpretation requires context. Later field studies showed that bending plant material is part of the species’ natural tool-making repertoire. The experiment therefore does not require an explanation based on wholly unprecedented invention; it can also be understood as the flexible application of evolved and learned manufacturing routines.
The woodpecker finch of the Galápagos uses twigs or cactus spines to dislodge arthropods from crevices. Tool use becomes especially important under some dry conditions, when concealed prey can be a valuable resource. Birds may shorten or otherwise adjust implements, and proficiency develops through experience. This example links tool use directly to feeding ecology: an implement effectively extends the bird’s reach into spaces its bill cannot exploit alone.
Goffin’s cockatoos provide evidence from both captivity and the wild. Captive individuals can manufacture implements from materials such as wood or cardboard and solve tasks requiring tools of particular dimensions. Field observations have also documented wild individuals using different wooden implements to access seeds within certain fruits. Captive results reveal capacities under controlled conditions, whereas field observations establish when those capacities are expressed in ordinary life.
Other birds use simpler but mechanically effective techniques. Egyptian vultures throw stones at large eggs, helping break shells that are difficult to open directly. Some crows drop hard foods onto firm surfaces; whether all such cases qualify as tool use depends on whether manipulating the food and selecting a substrate meet the chosen definition. Experiments in which birds drop stones into water-filled tubes require additional caution: successful performance can reflect learned preferences or feedback from a rising reward, rather than a general understanding of water displacement.
Aquatic Tool Users: Mammals, Fishes, and Octopuses
Water changes both the mechanics and the visibility of tool use. Buoyancy reduces effective object weight, currents complicate positioning, and many behaviors occur below the surface. Nevertheless, aquatic animals provide several well-supported examples. These are especially informative because they involve different body plans and sensory environments from those of terrestrial primates and birds.
The sea otter uses stones and other hard objects to open resistant prey. An otter floating on its back may place an anvil on its chest and strike shellfish against it, or use a handheld object as a hammer. Otters can carry tools and prey in loose skin beneath the forelimbs. Tool use varies among individuals and populations and is closely related to diet: animals specializing in hard-shelled prey face different processing demands from those feeding mainly on softer foods. A persistent individual habit is therefore not necessarily a species-wide requirement.
In Shark Bay, Western Australia, some bottlenose dolphins carry marine sponges over their rostra while searching the seafloor. The sponge acts as a protective covering during foraging in substrate where abrasion or injury is possible. The behavior occurs in a subset of the population and is strongly associated with particular maternal lineages and foraging habitats. The sponge itself is an animal, but its biological origin does not prevent the detached object from functioning as a tool.
Certain ray-finned fishes, including tuskfishes and wrasses, transport hard-shelled prey to rocks and repeatedly strike the prey against them. These observations establish controlled use of a particular surface for food processing. Classification remains definition-dependent because the fish may manipulate the prey rather than the rock. Such cases are often discussed as anvil use, illustrating why behavioral descriptions are more informative than a simple yes-or-no label.
A striking invertebrate example concerns the veined octopus, reported in 2009 carrying coconut-shell halves across the seafloor and later arranging them as shelter. Transport is awkward and provides no immediate covering while the animal moves, distinguishing the behavior from merely occupying an available cavity. The shells become portable defensive resources. Researchers nevertheless distinguish this use of detached objects from the broader and more common behavior of arranging material around a den.
These examples differ in function as well as anatomy. Otter tools transmit force, dolphin sponges provide protection during searching, and octopus shells create refuge. Water jets raise another boundary issue: archerfish knock prey down by expelling water, but liquid jets are usually treated separately from conventional object tools. Including every externally directed substance would greatly expand the category. Aquatic research consequently places particular emphasis on describing what is controlled, how it acts on the target, and whether the proposed implement is an independently manipulable object.
Invertebrates and Other Terrestrial Animals
Tool-like behavior also occurs among animals whose nervous systems and manipulative structures differ markedly from those of vertebrates. These cases help separate the mechanical definition of tool use from assumptions about large brains, flexible hands, or elaborate planning. They also show why the boundary between tool use, construction, and ordinary handling must be applied consistently.
Some ants place absorbent materials in liquid food and later retrieve the soaked objects. Studies of Aphaenogaster ants have documented the use of soil particles, plant fragments, and, in experiments, novel absorbent materials. An object functions as an intermediate carrier, allowing liquid to be transported in a form that workers can handle. Material properties matter: absorbency, mass, and ease of retrieval affect usefulness. A colony’s successful performance need not imply that each worker represents the entire transport process in advance.
Certain solitary wasps use a small pebble to tamp soil over a nest entrance. The wasp holds the object in its mandibles and applies it repeatedly to the substrate. This differs from simply carrying a stone as building material because the object acts as a working implement during compaction. Such behavior can be highly specialized and linked to a particular stage of nesting. Its narrow context does not make it less real as object-mediated action, but it limits what can be inferred about flexibility in unrelated tasks.
Other insect behaviors are more difficult to classify. Larvae, silk, sticky secretions, and construction materials can function as parts of elaborate extended systems. Weaver ants, for example, manipulate larvae whose silk joins leaves during nest construction. This is sometimes discussed as the use of living tools, but it falls outside definitions restricted to inanimate objects and overlaps with cooperative construction. Explicitly identifying such alternatives prevents a broad label from obscuring biologically different processes.
Among larger terrestrial animals, elephants manipulate branches during activities such as fly switching. Captive experiments have also shown that an Asian elephant can move a support into position and stand on it to reach otherwise inaccessible food. These observations concern different forms of environmental modification: a branch is directly wielded, whereas a movable platform changes the animal’s position. Performance also depends on sensory and anatomical constraints. Holding an object in the trunk can interfere with using that same organ to investigate a target.
Evidence is uneven across taxa. A repeatedly observed, filmed sequence provides a stronger basis for analysis than an isolated anecdote, especially when an object’s function is ambiguous. Reports of animals throwing material, rubbing against objects, or carrying decorations require attention to what the behavior actually achieves. Neither rarity nor apparent simplicity is grounds for dismissal, but neither is an unusual movement sufficient to establish tool use. Comparative accounts are most reliable when they retain the original behavioral detail rather than reducing every report to a species count.
Development, Learning, and Physical Cognition
Tool use develops through interactions among inherited dispositions, motor maturation, exploration, and experience. A species may have a strong tendency to manipulate particular materials without possessing a fully formed technique at birth. Conversely, an animal may be capable of an action in an experiment but rarely encounter the relevant problem in nature. Distinguishing predisposition, competence, and everyday performance is essential to explaining how tools enter a behavioral repertoire.
Young chimpanzees and capuchins illustrate prolonged skill acquisition. They handle materials, visit processing sites, and attempt actions long before matching experienced adults’ efficiency. Nut cracking requires control of impact, alignment, and stability, while probing requires coordination between movements of the implement and events occurring partly out of sight. Errors are therefore informative. A youngster may select suitable objects but fail to position the nut correctly, or perform an appropriate movement with an unsuitable implement.
Some tool-related actions appear without a skilled demonstrator. Hand-reared New Caledonian crows have shown species-typical tool-oriented behavior, indicating that social copying is not necessary for every component of the repertoire. This does not establish that wild skills are independent of experience. A developmental predisposition can make certain discoveries more likely, while practice and social exposure refine material selection, manufacture, and deployment. The alternatives are not simply “innate” versus “learned.”
Experiments in animal cognition investigate whether subjects attend to functional properties rather than superficial cues. Researchers may vary a tool’s length, rigidity, connectivity, or orientation while controlling reward location. Transfer to an unfamiliar apparatus can provide stronger evidence of generalization than success after repeated training on one device. Even then, several mechanisms may explain performance: learned associations, perceptual rules, sensitivity to available actions, or a more explicit representation of physical relationships.
Classic tasks reveal the interpretive difficulty. In a trap-tube problem, food must be moved along a tube without falling into a trap. A subject may learn to avoid a visible feature without understanding why the trap matters. Rotating or modifying the apparatus can test this possibility, but altered tasks can introduce new perceptual demands. Similarly, an animal that chooses a sufficiently long stick may be matching visible distances rather than mentally simulating the complete action.
Planning experiments separate tool selection from immediate reward. Some apes and corvids can retain or select an object that will be useful later under particular laboratory conditions. The strongest designs control for immediate attraction, prior reinforcement, and alternative uses. Such findings concern specific capacities, not a general scale on which every tool user can be ranked. A highly practiced natural technique may depend on different cognitive processes from an unfamiliar laboratory puzzle, even when both produce an apparently ingenious result.
Social Transmission and Animal Traditions
A tool behavior becomes a tradition when it persists within a group partly through social learning. Social learning includes more than copying an action. An observer may become interested in a location, attend to a material, discover the result of another animal’s behavior, or reproduce a movement. These routes can lead to similar group patterns while imposing very different cognitive demands.
Young animals often learn in environments already structured by experienced individuals. A nut-cracking site contains anvils, abandoned hammers, shell fragments, and partially processed food. These remains direct attention toward relevant objects and preserve opportunities for practice. In this sense, a technological environment can transmit information without a demonstrator intentionally teaching. The physical durability of stones makes such effects particularly persistent.
Chimpanzee communities provide major evidence for geographically patterned repertoires. A comparative study published in 1999 identified behavioral variants across long-term field sites, including several involving tools. Such distributions helped establish the study of animal culture, but geographic differences alone do not prove social transmission. Differences in prey, vegetation, material availability, genetic history, and observation effort may also contribute. Stronger explanations combine distributional evidence with individual histories, experiments, and direct observations of acquisition.
Teaching is a narrower category than social influence. In a functional definition, an experienced individual modifies its behavior in the presence of an inexperienced learner, incurs some cost or gains no immediate benefit, and helps the learner acquire a skill more efficiently. Chimpanzee mothers in the Goualougo Triangle have been observed transferring termite-fishing tools to offspring, sometimes interrupting their own feeding. Such transfers provide access to functional implements and opportunities to practice, supporting teaching interpretations under specified criteria.
Shark Bay dolphin sponging provides a different social system. Calves spend years accompanying their mothers, and the technique is transmitted predominantly along maternal lines, especially to daughters. Analyses of association patterns, habitat use, and relatedness help distinguish social inheritance from simple co-occurrence. The resulting tradition can persist within a subset of a population without becoming universal, because individuals differ in social exposure, specialization, and foraging location.
Social transmission does not necessarily produce cumulative culture, in which successive modifications yield forms that depend on earlier innovations. Many animal traditions are stable rather than progressively elaborated, and independent rediscovery can contribute to their persistence. Claims of cumulative change require evidence that later forms build on earlier ones, not merely that tool designs differ among populations. Dispersal, social tolerance, and access to skilled partners can all influence whether a useful innovation spreads, remains local, or disappears when its practitioners die.
Ecological Drivers and Evolutionary Explanations
Tool use can be understood as an adaptation to particular opportunities and constraints rather than as an inevitable consequence of increasing intelligence. A useful implement must be available, manageable, and capable of improving the outcome of an activity. Animals must also encounter a target worth processing and have sufficient time or safety to learn the technique. Failure of any one condition may make tool use uncommon even in a cognitively capable species.
Two influential explanations emphasize necessity and opportunity. Under a necessity account, animals turn to tools when preferred foods are scarce or difficult to obtain. Under an opportunity account, tool use becomes more likely when animals frequently encounter suitable materials and targets together. These explanations are not mutually exclusive. Scarcity may increase motivation, but it cannot produce stone hammering where suitable stones and stable supports are absent. Conversely, abundant materials may support habitual tool use even when starvation is not a threat.
The costs and benefits depend on local ecological niches. A hammer opens a hard nut, but searching for it, transporting it, and processing the food take time and energy. A probe can expose hidden prey while also occupying an appendage that might otherwise handle food or support movement. Tools may reduce tooth wear or injury risk, yet their acquisition and use can create other risks. Their value is therefore conditional, not intrinsic.
Anatomy influences which solutions are feasible. Grasping hands facilitate some operations, but beaks, feet, trunks, mouths, and flexible arms can perform others. Posture also matters: a stable surface can free appendages for manipulation, while arboreal movement may constrain the use of heavy objects. Sensory systems determine what information is available during work. A probing animal may rely on touch and resistance rather than vision to detect contact with concealed prey.
The scattered distribution of tool use indicates repeated convergent evolution of object-mediated solutions. Similar tasks can favor comparable actions in distantly related species, such as striking hard food against a resistant surface. Nevertheless, an observed behavior need not be a genetically specialized adaptation in every detail. General manipulative tendencies, individual learning, and socially inherited practices can generate functional techniques whose precise form varies locally.
Tool use can also alter relationships within a food web. Access to protected seeds or concealed invertebrates changes which resources an animal can exploit and may support individual dietary specialization. Repeated use can modify local sites through accumulated debris and worn surfaces. These effects connect behavior to niche construction, although their ecological scale differs greatly among cases. Identifying an evolutionary explanation requires measurements of resource availability, costs, developmental pathways, and consequences for survival or reproduction, rather than assuming that every effective tool behavior necessarily provides a large fitness advantage.
Archaeological Evidence, Research Methods, and Conservation
Tools can leave evidence after the user has departed. Stone hammers acquire impact marks, anvils develop pits or depressions, and processing sites accumulate shell or plant residues. Primate archaeology applies excavation, spatial recording, and material analysis to such traces. Its central advantage is temporal depth: direct observation covers only a fraction of the history of a long-standing behavioral tradition.
Excavations in Côte d’Ivoire have documented chimpanzee-associated nut-cracking evidence approximately 4,300 years old. At Serra da Capivara in Brazil, capuchin stone-tool deposits extend back about 3,000 years. These findings demonstrate that some nonhuman tool practices persist on archaeological timescales. They do not establish that every detail remained unchanged, nor that present-day individuals are directly connected to every recovered episode. Dating, stratigraphy, artifact characteristics, and ecological context must be considered together.
Comparison with human archaeology is informative but requires restraint. Nonhuman percussion produces some patterns resembling those of early human activities. Capuchins can also unintentionally detach sharp stone flakes while striking stones, showing that flake morphology alone does not always establish deliberate cutting-tool manufacture. This does not erase distinctions between archaeological assemblages. Instead, it highlights the need to combine technological patterns, use wear, residues, spatial associations, and evidence of what the objects were used to do.
Modern field research integrates video, focal observation, camera traps, mapping, and measurements of tools and raw materials. Experiments can test preferences or learning mechanisms, but introducing food or unfamiliar objects may change the behavior being studied. Captivity provides control while also altering experience, motivation, material availability, and human exposure. A complete account therefore distinguishes spontaneous wild behavior, experimentally elicited field performance, and captive demonstrations rather than treating them as interchangeable evidence.
Tool traditions can be vulnerable to habitat fragmentation and population decline. Removing a plant species may eliminate suitable probes; losing access to a shoreline may disrupt shellfish processing; the death of experienced individuals may reduce opportunities for juveniles to acquire a local technique. Conservation biology increasingly recognizes that maintaining a species’ numbers does not automatically preserve all of its behavioral diversity. Nevertheless, demonstrating the loss of a tradition requires reliable earlier records and adequate observation, since failure to detect a rare behavior is not proof of disappearance.
Important research problems concern the durability and visibility of evidence. Wooden probes decay, underwater activity is difficult to observe, and heavily studied species are more likely to enter comparative databases. Apparent taxonomic absences may therefore reflect limited observation as well as genuine biological differences. Long-term records linking identified individuals, learning histories, environmental changes, and surviving artifacts are especially valuable for distinguishing independent invention, social transmission, and behavioral loss.