Dan Sperber

Dan Sperber (born 1942) is a French cognitive and social scientist whose work has redefined how we understand cultural transmission, human communication, and the architecture of inference. Trained in anthropology and philosophy, Sperber moved from the study of symbolism and ritual toward a rigorous naturalistic program: explaining culture not through interpretive hermeneutics but through the mechanisms of cognition and the dynamics of information spread. His intellectual trajectory makes him one of the few thinkers to have produced foundational contributions in both cognitive science and the theory of cultural evolution — and to have insisted, at every step, that the two domains are inseparable.

Relevance Theory

With linguist Deirdre Wilson, Sperber developed Relevance Theory, a cognitive-pragmatic framework that reconceptualizes human communication as a process of inference rather than coding-decoding. The classical model of communication — sender encodes a message, channel transmits it, receiver decodes it — treats meaning as a package to be delivered intact. Sperber and Wilson reject this. On their account, an utterance is not a container of meaning but a piece of evidence: it provides the hearer with clues that, when processed through a relevance-oriented cognitive architecture, yield an interpretation.

Relevance is defined in terms of cognitive effort and cognitive effect: the best interpretation is the one that yields the greatest effect for the least effort. This is not a heuristic but a putative law of cognition — a claim about the fundamental operating principle of the human mind. If true, it has sweeping implications. It means that comprehension is not a special-purpose linguistic module but the natural output of a general inferential engine tuned for relevance. It also means that the literal meaning of an utterance is merely a starting point for inference, not its terminus. Irony, metaphor, implicature, and even standard literal communication are all products of the same relevance-seeking process.

The Epidemiology of Representations

Sperber's second major contribution is the Epidemiology of Representations, articulated most fully in his 1996 book of that title. The central claim is that cultural transmission is not a process of copying — as Richard Dawkins' meme model assumes — but a process of reconstruction. When one person tells another a story, demonstrates a technique, or expresses a belief, what is transmitted is not a discrete packet of information but a stimulus that prompts the receiver to construct a representation of their own. The receiver's representation will differ from the sender's in ways determined by the receiver's cognitive architecture, prior knowledge, and interpretive context.

What makes cultural stability possible, then, is not high-fidelity replication but cognitive attractors: patterns of representation toward which human minds tend to converge. A cognitive attractor is not a fixed idea but a basin in cognitive space. Representations that fall within the basin are reconstructed in similar ways by different minds; representations outside the basin are transformed toward the attractor or forgotten. This explains why some cultural variants are stable across populations despite low copying fidelity, and why others — despite being actively promoted — fail to spread.

The epidemiological model is explicitly systemic. It treats culture as a population-level phenomenon emergent from the micro-dynamics of individual inference, just as epidemiology treats disease spread as emergent from the micro-dynamics of individual infection and transmission. The parallel is not merely metaphorical: both fields study the aggregate consequences of local interaction rules, and both require statistical population thinking rather than single-case narrative explanation.

The Critique of Memetics

Sperber's epidemiological framework is the deepest theoretical challenge to memetics — the project of treating culture as Darwinian evolution among discrete replicating units. Sperber does not deny that cultural items spread, vary, and are selected. He denies that they spread by copying. The difference is decisive. If cultural transmission is reconstruction-toward-attractor, then the meme-as-gene-analogue is the wrong basic unit. There is no "meme pool" in the sense Dawkins intended, because there is no reliable mechanism of high-fidelity replication to maintain a pool.

This is not merely a quibble about mechanism. It transforms the explanatory questions. Memetics asks: what makes a meme fit — why does it replicate more than rivals? The epidemiology of representations asks: what makes a representation cognitively attractive — why do minds tend to reconstruct it in a stable form? The first question assumes a copying dynamics; the second assumes an inferential dynamics. The answers are not intertranslatable, because the underlying processes are structurally different.

Daniel Dennett's defense of memetics — that virtual patterns can be genuine replicators without physical token identity — is, on Sperber's account, a retreat to abstraction that loses the empirical bite the concept needs. If memes are whatever patterns spread, then memetics becomes true by definition and empty by consequence. Sperber's attractor model retains empirical content because it makes testable predictions about which representations will be stable and which will transform, based on what we know about human cognition.

The debate between memetics and the epidemiology of representations is not a disagreement about whether culture evolves. It is a disagreement about what kind of system culture is — a replicator dynamics or an attractor dynamics. The evidence, Sperber insists, favors the attractor model, and the failure of memetics to produce predictive success across decades is precisely what one would expect if the underlying system were inferential rather than replicative. To persist with the meme framework is not scientific conservatism; it is commitment to a formalism that misdescribes the phenomenon it purports to explain.