In human reasoning about the characteristic appearances and behaviors of natural kinds such as folk genera and chemical substances, the cognitive timeline, according to Cimpian & Salomon (C&S), is as follows: First comes the inherence heuristic and thus an assumption that the causes of those appearances and behaviors – the tiger's stripes, water's transparency – is something internal. Then comes essentialism and so the further suppositions that (a) this internal property is for each appearance or behavior the same, so that a single internal property explains all characteristic appearances and behaviors, and (b) the property in question is an essence, necessary and sufficient for kind membership.
Essentialism in its sparest form does not attribute to the thinker the belief that the essence is something wholly internal; it is allowed, in particular, that the thinker is agnostic about the essence's location. The hypothesis that essentialism emerges from a prior commitment to inherence can be tested, then, by asking whether the essence is, from the first moment that the child begins to think essentially, represented as something definitely inside the animal, plant, or substance.Footnote
1
An original aim of psychological essentialism was to explain the results of Keil's (Reference Keil1989) “discovery” and “transformation” experiments. In the biological version of the discovery experiment, a creature with the external characteristics of one species is discovered to have the insides and lineage of another – an animal that looks and acts like a horse, for example, is found to have the insides and ancestry of a cow. In the biological version of the transformation experiment, a creature is cosmetically transformed to have the appearance and behavior of another species – a raccoon, for example, is made up to look and smell like a skunk. In both experiments, subjects are asked to classify the animal in question: Horse or cow? Raccoon or skunk?
Keil and his collaborators found that younger children tend to classify in accordance with appearances, answering “horse” and “skunk” when presented with the scenarios described in the previous paragraph, whereas older children and adults tend to classify in accordance with deep properties (“cow” and “raccoon”). Crucially, the older children sooner and more confidently make “deep” classifications in the transformation cases than in the discovery cases. This is the datum that tells against the inherence explanation of essentialism.
How so? If essences are thought to be wholly internal properties, then an animal with cow insides must possess the cow essence and so must be a cow. An essentialist thinker, then, provided that they have enough courage in their convictions to classify in contradiction to external properties, should be confident that the animal in the discovery experiment is a cow. Suppose that the advocates of psychological essentialism are correct in thinking that the results of the transformation experiments are to be explained by essentialist thinking. Then, any subject who gives the “deep” answer in the transformation experiment is an essentialist thinker and so ought at the same time to give the deep answer with equal confidence in the discovery experiment. That is not what Keil found: On the whole, deep answers in the discovery experiments come later and with less certainty.
Let me make the same point in a different way. In the transformation experiment, subjects are told that the animal begins as a raccoon. In the discovery experiment, they are not told that the animal begins as a cow, but only that it has cow insides. But if C&S are correct in supposing that essences are from the first believed to be internal, then it should be straightforward for subjects to reason from insides to essence – from having cow innards to being a cow. The two cases ought therefore be on a par. (Indeed, the deep answer ought if anything to come more easily in the discovery cases.) Quite the contrary effect is observed; this shows, I suggest, that many young essentialist thinkers (and quite possibly many adults, too) are agnostic about the location of the essence.
More generally, where we see “essentializing” in human thought we are not always seeing internalizing. New Yorkers are expected to behave the same way wherever they go; does this show that naive reasoners believe that New Yorkers have some internal property that causes their famous characteristics? Not necessarily; it would be enough that the reasoners believe that being a New Yorker causes those characteristics and that being a New Yorker, like being a raccoon, is a persistent state of affairs. Expectations about New Yorkers are explained, on this approach, by attributing to the reasoner a belief in the persistence and causal efficacy of category membership.
Such an attribution is at the core of the psychological essentialist's explanatory strategy.Footnote
2
Essentialism does not rely on a commitment to inherence at all: What is persistent need not be internal, and what is internal need not be persistent. Thus, the inherence theorist's explanatory schema is more a rival to than a complement to the essentialist's explanatory schema: Whereas the essentialist emphasizes category membership, the inherence theorist emphasizes the physical constitution of individual category members. The two can live together, as they do in an “internal essentialism,” according to which category membership is determined by something physically internal. But the evidence suggests that, psychologically, often enough they do not.
In human reasoning about the characteristic appearances and behaviors of natural kinds such as folk genera and chemical substances, the cognitive timeline, according to Cimpian & Salomon (C&S), is as follows: First comes the inherence heuristic and thus an assumption that the causes of those appearances and behaviors – the tiger's stripes, water's transparency – is something internal. Then comes essentialism and so the further suppositions that (a) this internal property is for each appearance or behavior the same, so that a single internal property explains all characteristic appearances and behaviors, and (b) the property in question is an essence, necessary and sufficient for kind membership.
Essentialism in its sparest form does not attribute to the thinker the belief that the essence is something wholly internal; it is allowed, in particular, that the thinker is agnostic about the essence's location. The hypothesis that essentialism emerges from a prior commitment to inherence can be tested, then, by asking whether the essence is, from the first moment that the child begins to think essentially, represented as something definitely inside the animal, plant, or substance.Footnote 1
An original aim of psychological essentialism was to explain the results of Keil's (Reference Keil1989) “discovery” and “transformation” experiments. In the biological version of the discovery experiment, a creature with the external characteristics of one species is discovered to have the insides and lineage of another – an animal that looks and acts like a horse, for example, is found to have the insides and ancestry of a cow. In the biological version of the transformation experiment, a creature is cosmetically transformed to have the appearance and behavior of another species – a raccoon, for example, is made up to look and smell like a skunk. In both experiments, subjects are asked to classify the animal in question: Horse or cow? Raccoon or skunk?
Keil and his collaborators found that younger children tend to classify in accordance with appearances, answering “horse” and “skunk” when presented with the scenarios described in the previous paragraph, whereas older children and adults tend to classify in accordance with deep properties (“cow” and “raccoon”). Crucially, the older children sooner and more confidently make “deep” classifications in the transformation cases than in the discovery cases. This is the datum that tells against the inherence explanation of essentialism.
How so? If essences are thought to be wholly internal properties, then an animal with cow insides must possess the cow essence and so must be a cow. An essentialist thinker, then, provided that they have enough courage in their convictions to classify in contradiction to external properties, should be confident that the animal in the discovery experiment is a cow. Suppose that the advocates of psychological essentialism are correct in thinking that the results of the transformation experiments are to be explained by essentialist thinking. Then, any subject who gives the “deep” answer in the transformation experiment is an essentialist thinker and so ought at the same time to give the deep answer with equal confidence in the discovery experiment. That is not what Keil found: On the whole, deep answers in the discovery experiments come later and with less certainty.
Let me make the same point in a different way. In the transformation experiment, subjects are told that the animal begins as a raccoon. In the discovery experiment, they are not told that the animal begins as a cow, but only that it has cow insides. But if C&S are correct in supposing that essences are from the first believed to be internal, then it should be straightforward for subjects to reason from insides to essence – from having cow innards to being a cow. The two cases ought therefore be on a par. (Indeed, the deep answer ought if anything to come more easily in the discovery cases.) Quite the contrary effect is observed; this shows, I suggest, that many young essentialist thinkers (and quite possibly many adults, too) are agnostic about the location of the essence.
More generally, where we see “essentializing” in human thought we are not always seeing internalizing. New Yorkers are expected to behave the same way wherever they go; does this show that naive reasoners believe that New Yorkers have some internal property that causes their famous characteristics? Not necessarily; it would be enough that the reasoners believe that being a New Yorker causes those characteristics and that being a New Yorker, like being a raccoon, is a persistent state of affairs. Expectations about New Yorkers are explained, on this approach, by attributing to the reasoner a belief in the persistence and causal efficacy of category membership.
Such an attribution is at the core of the psychological essentialist's explanatory strategy.Footnote 2 Essentialism does not rely on a commitment to inherence at all: What is persistent need not be internal, and what is internal need not be persistent. Thus, the inherence theorist's explanatory schema is more a rival to than a complement to the essentialist's explanatory schema: Whereas the essentialist emphasizes category membership, the inherence theorist emphasizes the physical constitution of individual category members. The two can live together, as they do in an “internal essentialism,” according to which category membership is determined by something physically internal. But the evidence suggests that, psychologically, often enough they do not.