how do we know that insects “don't know anything”?

Why 'insects know nothing' feels obvious—and why it's misleading

Insects are small, have tiny brains, and often behave in ways that look like fixed reflexes. That makes it tempting to assume there is no knowledge inside. But 'knowledge' in this context means information an animal can learn, store, and use flexibly—not human-style self-awareness. The question is whether insects can do that, and the evidence says some can, at least in narrow ways [2].

A useful comparison is a calculator versus a person doing arithmetic. A calculator 'knows' nothing in the human sense, yet it can produce correct answers. The interesting scientific question is whether insects are more like calculators or more like animals that form internal representations they can apply to new situations. Some bee experiments suggest the latter, at least for simple spatial concepts [2].

Bees can learn 'above' and 'below'—a small but real concept

In one set of experiments, honeybees were trained to process spatial relations between visual stimuli—specifically, whether one thing was above or below another. The bees could indeed apply these relations, which is a step beyond simply reacting to a single feature like color or shape [2]. This matters because 'above' and 'below' are relational concepts: they depend on how two things are arranged, not on either thing alone.

That does not mean bees think about space the way we do. The finding is limited to a specific visual task under controlled conditions. But it does undercut the blanket claim that insects 'don't know anything,' because applying a relational rule to new stimuli is a form of learned, flexible knowledge [2].

Building blocks of cognition—and knowledge stored outside the brain

A broader review of social insects argues that invertebrates offer a window into the 'building blocks of cognition' and possible emotional expression [1]. The phrase 'building blocks' is important: it suggests components of cognition rather than full human-like thought. The review synthesizes current knowledge rather than proving rich inner lives, so it should be read as evidence that the simple 'no knowledge' view is too crude [1].

There is also a collective angle. Insect nests and architecture have been discussed as structures that can preserve isolated pieces of knowledge, with social insect nests viewed as part of collective cognition [4]. This is a proposal about how information might be held in a colony and its structures, not a demonstration that individual insects are knowledgeable. It does, however, widen what 'insect knowledge' could mean beyond a single tiny brain [4].

What we still can't say—and why the question stays open

None of this shows that insects have conscious knowledge, beliefs, or a human-like understanding of the world. The bee result is a specific laboratory demonstration of relational learning [2], and the review and architecture papers are syntheses or proposals about mechanisms and collective information, not proof of inner experience [1][4]. A historical study of Merian's insect observations touches on perceptual knowledge in a different, interpretive register and does not settle the cognitive question [3].

So the honest answer is that the claim 'insects don't know anything' is not established. What the evidence supports is narrower: some insects can learn and apply abstract spatial relations [2], social insects show cognitive building blocks worth studying [1], and colonies may hold information in collective structures [4]. Whether that adds up to 'knowing' depends on what we mean by the word—and that is a definitional question as much as an empirical one.

Sources used in this answer

Related research questions