Tool Use Behavior

Abstract

Tool use behavior is a form of animal behavior in which an animal employs an external, unattached object to act on another object, surface, or organism to reach a goal. Once thought uniquely human, it is now documented across primates, corvids, and other taxa, from chimpanzee termite-fishing and nut-cracking to the hook and compound tools that New Caledonian crows manufacture. This article develops the operational definition of tool use, the comparative evidence, the manufacture and metatool sequences that mark flexible use, and the live debate over what tool use reveals about cognition. A worked example and three interactive demonstrations accompany the text.

Keywords: tool use, animal cognition, comparative psychology, cumulative culture

In 1960, Jane Goodall watched a wild chimpanzee strip leaves from a twig and push it into a termite mound to fish out insects, a report that forced a redefinition of the boundary between human and animal (Goodall, 1964). The claim that toolmaking set humans apart could not survive a chimpanzee doing it in the field. In the decades since, tool use has been found to be neither rare nor uniquely primate: New Caledonian crows manufacture hooked foraging tools, capuchin monkeys crack nuts with stone hammers and anvils, and a whole crow species was discovered to use tools as a matter of course.

This article treats tool use as a problem in comparative cognition rather than a single trick. The central questions are definitional and mechanistic. What counts as a tool, as opposed to an incidental use of the environment? When does tool use require planning, causal understanding, or learning from others, and when is it a stereotyped foraging routine that looks cleverer than it is? Pulling these apart is the work of the field, and it is why a careful operational definition does as much as any single observation.

What Tool Use Behavior Is

Tool use behavior is the use of an external object, not attached to the substrate, to change the position, form, or condition of another object, surface, or organism, in a way that gives the user a functional advantage. In MeSH it is classified under animal behavior. The qualifiers in that definition do real work: *external* and *unattached* exclude an animal pressing prey against a fixed rock, and *functional advantage* excludes incidental contact. The most cited modern treatment, by Robert St Amant and Thomas Horton, tightens the classic Beck definition by requiring that the user hold or manipulate the tool and that the tool mediate a change in another entity, drawing a defensible line around the behavior while admitting borderline cases honestly (St Amant & Horton, 2008).

The definition matters because it determines what the comparative record contains. Amanda Seed and Richard Byrne, surveying the field, stress that a single operational definition lets observations from an ant, a crow, and an ape be compared on the same terms, and that the interesting variation is not whether an animal uses a tool but how flexibly it does so (Seed & Byrne, 2010). Flexibility is what the later sections pursue: whether an animal can make a tool, select the right one, or build a new one when the situation demands it.

The Comparative Record: Apes, Monkeys, and Birds

The strongest early evidence came from wild chimpanzees. Beyond termite-fishing, Christophe Boesch and Hedwige Boesch documented the Taï Forest chimpanzees cracking hard-shelled nuts by placing them on a stone or root anvil and striking them with a wooden or stone hammer, a two-object technique that takes years to master (Boesch & Boesch, 1990). Tool use in chimpanzees is not uniform across the species: Andrew Whiten and colleagues catalogued thirty-nine behavior patterns, many of them tool-related, that are customary in some communities and absent in others with no ecological explanation, the pattern that defines a cultural tradition (Whiten et al., 1999). Monkeys contribute an independent primate line: Dorothy Fragaszy and colleagues described wild capuchin monkeys using stone pounding tools and anvils to open nuts, showing that heavy percussive tool use is not confined to the great apes (Fragaszy et al., 2004).

Birds supply the most striking case of all, because their tool use evolved on a branch separated from primates by hundreds of millions of years. Gavin Hunt reported that New Caledonian crows in the wild manufacture hooked twig tools and cut stepped tools from the barbed edges of pandanus leaves, fashioning standardized shapes rather than seizing whatever is at hand (Hunt, 1996). Nathan Emery and Nicola Clayton framed this convergence directly, arguing that corvids and apes have arrived at comparable cognitive abilities independently, so that intelligence is better understood as a set of solutions evolution reaches more than once than as a single primate inheritance (Emery & Clayton, 2004).

Manufacture, Metatools, and Compound Tools

Making a tool is cognitively more demanding than using a found one, and sequencing tools more demanding still. In a laboratory test of manufacture, Alex Weir, Jackie Chappell, and Alex Kacelnik gave a captive New Caledonian crow, Betty, straight wire and an out-of-reach bucket of food in a vertical tube; she spontaneously bent the wire into a hook to lift the bucket, a case of an animal shaping a novel tool to fit a problem it had not faced before (Weir et al., 2002). Flexibility of this kind is what separates a fixed foraging action from a general problem-solving capacity.

Two further behaviors push the point, both forms of sequential tool use in which an ordered series of actions must be planned ahead. A *metatool* is a tool used on another tool rather than on food: Alex Taylor, Gavin Hunt, Jennifer Holzhaider, and Russell Gray showed that New Caledonian crows spontaneously use a short stick to extract a long stick, then use the long stick to reach food, solving a three-stage sequence on first encounter (Taylor et al., 2007). A *compound tool* is one assembled from parts that are useless alone: Auguste von Bayern, Kacelnik, and colleagues found that the same species can combine two or more non-functional pieces into a single elongated tool long enough to reach a reward, a form of physical construction not previously seen outside humans and great apes (von Bayern et al., 2018). Each behavior raises the ceiling on what the animal must represent: a goal, an intermediate subgoal, and the physical relation between separate objects.

What Tool Use Reveals About Cognition

That an animal uses, makes, or sequences tools does not settle what it understands. Much of the current theoretical work concerns the gap between competent behavior and the cognition behind it. One position holds that flexible tool use rests on causal or physical understanding, a grasp of how objects interact that lets an animal generalize to new problems. A competing account is more deflationary: some tool-use behaviors may be *latent solutions*, actions that lie within a species' individual behavioral repertoire and reappear without social learning or insight whenever the ecological trigger is present (Bandini & Tennie, 2017). The two accounts predict different things about whether a behavior needs a demonstrator to spread.

A third strand reframes the human case. François Osiurak and Emanuelle Reynaud argue that what matters cognitively in cumulative technological culture is not primarily imitation but *technical reasoning*, the capacity to reason about physical object properties and mechanical actions, which lets improvements accumulate across generations (Osiurak & Reynaud, 2020). Whether the ratchet of cumulative culture is driven by high-fidelity social copying or by individually reinvented technical reasoning is unresolved, and it is the question that connects animal tool use to the human technological record. These debates are revisited in the Current Directions section.

Worked Example

The ratchet of cumulative culture is the idea that a population retains and builds on improvements that individuals rarely rediscover alone, so that complexity accumulates only when gains are faithfully transmitted. A simple arithmetic model makes the dependence on transmission fidelity concrete. Let each generation start from the complexity it inherits, add a fixed increment of innovation, and then lose a fraction of the total to imperfect transmission. Write the inherited complexity as $C$, the per-generation innovation as a flat +1 unit, and the transmission fidelity as a retained fraction $f$ between 0 and 1. Each generation computes $C_{\text{next}} = (C + 1) \times f$.

Consider two populations starting from $C_0 = 0$, one with high fidelity $f = 0.9$ and one with low fidelity $f = 0.5$:

The high-fidelity population keeps climbing, because each generation retains most of what it inherited and adds to it; the low-fidelity population stalls near a ceiling, because losses cancel innovation almost as fast as it arrives. Each line converges on the fixed point $C^{*} = f / (1 - f)$: for $f = 0.9$ that is $0.9 / 0.1 = 9$ units, and for $f = 0.5$ it is $0.5 / 0.5 = 1$ unit. The ninefold gap in the ceiling comes entirely from the difference in transmission fidelity, not from any difference in how clever the innovators are. This is why the debate over cumulative culture turns on *how* gains are transmitted. The CumulativeCultureDemo lets the reader vary the fidelity and watch the ceiling move; the values are illustrative of the model, not measured from any population.

Discussion

The study of tool use has moved from a search for the behavior to an analysis of its varieties. Establishing that chimpanzees, capuchins, and crows all use tools was the work of the first decades, and it dismantled the idea that toolmaking marks a human boundary (Goodall, 1964; Hunt, 1996). The convergence of apes and corvids on similar abilities, from distant evolutionary branches, reframed the whole question: tool use is not a single inheritance but a solution that evolution finds repeatedly where ecology rewards it (Emery & Clayton, 2004).

What remains open is the interpretation. A behavior can be real, reliable, and even manufactured, and still leave unsettled whether it rests on causal understanding, individual trial-and-error, or social transmission (Seed & Byrne, 2010). The metatool and compound-tool results show animals solving multi-stage physical problems (Taylor et al., 2007; von Bayern et al., 2018), but the latent-solution and technical-reasoning accounts offer leaner explanations for part of the record (Bandini & Tennie, 2017; Osiurak & Reynaud, 2020). The field's progress is now measured less by new species added to the list than by sharper experiments that separate what an animal does from what it knows.

Current Directions

The active research front has three strands, each represented by recent work. The first is the discovery of tool use as a species-wide trait rather than a local tradition: Christian Rutz and colleagues found that the Hawaiian crow, a species not previously known as a tool user, uses stick tools pervasively and from an early age, implying that a capacity for tool use can be built into a species' biology rather than invented afresh in each population (Rutz et al., 2016). The second is the deflationary turn: Elisa Bandini and Claudio Tennie report that naïve chimpanzees with no opportunity to copy others spontaneously reinvent a wild tool-use behavior, scooping, which supports the latent-solutions account that some tool behaviors need no social transmission to appear (Bandini & Tennie, 2017).

The third strand concerns construction and the cognition behind cumulative culture. The demonstration that New Caledonian crows assemble compound tools from non-functional parts extends physical construction to a bird and sharpens the question of what representations such assembly requires (von Bayern et al., 2018). In parallel, Osiurak and Reynaud's argument that technical reasoning, rather than imitation alone, drives cumulative technological culture has reoriented the human-evolution debate toward the physical-reasoning abilities that animal tool use also taps (Osiurak & Reynaud, 2020). Together these lines are shifting the field from cataloguing tool use toward explaining its cognitive and cultural basis.

Common Misconceptions

Only humans make and use tools.

Goodall's observation of chimpanzees stripping twigs to fish for termites, and later reports of crows manufacturing hooked tools, ended this view. Toolmaking is documented across apes, monkeys, and birds (Goodall, 1964; Hunt, 1996).

Tool use proves an animal understands cause and effect.

Competent tool use can arise from trial-and-error or from species-typical latent solutions without any causal insight, so the behavior alone does not establish understanding (Bandini & Tennie, 2017).

Tool use is a single ability an animal either has or lacks.

Using a found object, manufacturing a tool, selecting between tools, and sequencing tools are different capacities with different cognitive demands; an animal may show one and not another (Seed & Byrne, 2010).

Glossary

Animal Behavior.

The scientific study of what animals do and why, the MeSH category under which tool use behavior is classified.

Anvil.

A stable surface, such as a stone or root, against which an animal strikes a resource with a hammer tool to break it open.

Causal understanding.

A grasp of how objects physically interact, invoked to explain flexible tool use but not established by competent use alone.

Compound tool.

A tool assembled from two or more parts that are non-functional on their own, such as combinable sticks joined to reach a distant goal.

Cumulative culture.

A process in which socially transmitted behaviors accumulate improvements across generations beyond what an individual could invent alone.

Latent solution.

A behavior within a species' individual repertoire that reappears without social learning whenever the ecological trigger is present.

Manufacture.

The modification of an object into a tool, such as bending wire into a hook or trimming a twig to a usable shape, before or during use.

Metatool.

A tool used to obtain or act on another tool rather than directly on a resource, as when a short stick is used to retrieve a long stick.

Operational definition.

A definition stated as the observable criteria an instance must meet, used here to decide which behaviors count as tool use.

Pandanus tool.

A stepped tool that New Caledonian crows cut from the barbed edge of a pandanus leaf, carrying a counterhook of natural barbs.

Percussive tool use.

Tool use involving striking, such as cracking a nut with a hammer stone against an anvil.

Sequential tool use.

The use of tools in an ordered series, each step enabling the next, as in retrieving a tool with a tool before reaching food.

Technical reasoning.

Reasoning about the mechanical properties of objects and actions, proposed as a core driver of cumulative technological culture.

Tool use behavior.

The employment of an unattached external object to change the state of another object, surface, or organism toward a goal.

Key Researchers

Christophe Boesch

(1951–2024). Founding director of the Department of Primatology at the Max Planck Institute for Evolutionary Anthropology; he documented nut-cracking and tool traditions in wild Taï Forest chimpanzees and established the study of chimpanzee culture in the field.
Wikipedia - Wikidata

Nicola S. Clayton

(University of Cambridge). Professor of comparative cognition whose work on corvid intelligence, with Nathan Emery, framed the convergent evolution of complex cognition in crows and apes.
ORCID - Wikipedia

Russell D. Gray

(Max Planck Institute for Evolutionary Anthropology). Director studying cultural and cognitive evolution; co-author of the studies demonstrating spontaneous metatool use and tool manufacture in New Caledonian crows.
ORCID - Wikidata

Alex Kacelnik

(University of Oxford, emeritus). Behavioral ecologist whose laboratory work on New Caledonian crows, including the wire-bending crow Betty and compound-tool construction, revealed flexible tool manufacture in birds.
Wikipedia - Wikidata

Christian Rutz

(University of St Andrews). Behavioral ecologist who discovered species-wide tool use in the Hawaiian crow and pioneered bio-logging methods to study tool use in wild crows.
ORCID

Elisabetta Visalberghi

(Institute of Cognitive Sciences and Technologies, CNR, Rome). Primatologist whose field and laboratory studies of wild capuchin monkeys established heavy percussive stone-tool use outside the great apes.
ORCID

Frequently Asked Questions

What is tool use behavior?

Tool use behavior is the employment of an external, unattached object to change the position, form, or condition of another object, surface, or organism toward a goal. In MeSH it is classified as a form of animal behavior (St Amant & Horton, 2008).

Do animals other than humans really use tools?

Yes. Wild chimpanzees fish for termites and crack nuts, capuchin monkeys pound nuts with stones, and New Caledonian crows manufacture hooked foraging tools, among many documented cases (Goodall, 1964; Fragaszy et al., 2004).

Who first showed that chimpanzees use tools?

Jane Goodall, who in 1960 observed wild chimpanzees at Gombe stripping leaves from twigs and inserting them into termite mounds to extract insects, published in 1964 (Goodall, 1964).

What is a metatool?

A metatool is a tool used on another tool rather than directly on a resource. New Caledonian crows will use a short stick to extract a longer stick and then use the long stick to reach food (Taylor et al., 2007).

Can animals make new tools, not just use found ones?

Yes. A captive New Caledonian crow bent straight wire into a hook to retrieve food, and the species can assemble compound tools from non-functional parts, both forms of manufacture (Weir et al., 2002; von Bayern et al., 2018).

Does tool use mean an animal understands cause and effect?

Not necessarily. Some tool behaviors are latent solutions that reappear without social learning or insight, so competent use alone does not establish causal understanding (Bandini & Tennie, 2017).

Why do crows and apes both use tools despite being distantly related?

Because complex cognition can evolve convergently: corvids and apes have independently arrived at similar abilities, suggesting intelligence is a solution evolution reaches more than once (Emery & Clayton, 2004).

How does tool use relate to human cumulative culture?

Human technology accumulates improvements across generations, and a central debate is whether that ratchet is driven by high-fidelity social copying or by individually reinvented technical reasoning about objects (Osiurak & Reynaud, 2020).

References