Atran reifies Fodor's metaphor of modularity to create a truth-producing apparatus to generate a priori taxonomies or natural kinds that lock a tautology in place.

Deliberate practice and experience may suffice as predictors of expertise, but they cannot account for spectacular achievements. Highly variable environmental and biological factors provide facilitating as well as constraining conditions for development, generating relative plasticity rather than absolute plasticity. The skills of virtuosos and idiots savants are more consistent with the talent account than with the deliberate-practice account.

Study of “theory of mind” in nonhuman primates is hampered both by the lack of rigorous methodology that Heyes stresses and by our lack of knowledge of the cognitive neuroscience of nonhuman primate conceptual structure. Recent advances in this field indicate that progress can be made by first asking simpler research questions.

There are problems with Atran's argument for an innate cognitive module for folk biology. He has been too quick to assume innate origins for what might plausibly be learned. Furthermore, in his characterization he includes aspects – essentialist reasoning and inductions from classes – that are not domain-specific. Finally, his characterization compromises his argument that the module is pretheoretical.

My criticisms of parapsychology are neither based on its subject matter per se, nor simply on a charge of sloppy research, but rather on the whole pattern of theory and research in this domain. The lack of a positive definition of psi, the use of ad hoc principles such as psi-missing and the experimenter psi effect to account for failures to confirm hypotheses, and the failure to produce a single phenomenon that can be replicated by neutral investigators are among the major problems that keep parapsychology outside regular science. Glicksohn and I agree that anomalous experiences should be investigated.

This commentary stresses the importance of Atran's work for the development of a new cognitive anthropology, but questions both his particular use of Dawkins's “meme” model and the general usefulness of the meme model for understanding folk-taxonomies as cultural phenomena.

Heyes argues that we need alternative experiments to study those animal abilities generally considered to involve “theory of mind.” The studies she proposes, however, have as many problems as those that she criticizes. Further interactions should exist among researchers examining these capacities before additional experiments are undertaken.

Wilson and Sober's (1994t) revival of group selection theory may have failed with some readers because its simple arithmetic foundation was obscured under the complexities of its presentation. When that uncontrovertible principle is uncovered, it broadens dramatically the fundamental motives that social scientists may impute to human nature and still be consistent with Darwinian evolutionary theory.

Current evidence shows that “talent” is an unreliable concept, but the competing concept of “expertise” suffers from inadequacies too: its research designs are structurally incomplete (talent and training are confounded) and individual factors are largely ignored. The deliberate practice account should be advanced by investigating in more detail the motivational and individual determinants of successful practice.

The target article examines the role of practice in the development of excellence in several domains that vary in the degree of innovation involved. The authors do not differentiate domains, but they discuss the development of specifically creative skills only in passing. This commentary presents evidence that practice plays a role in the development of musical composition, through an examination of the careers of Mozart and the Beatles.

We reinforce Thompson's points by providing a second example of the paradox that makes group selection appear counterintuitive and by discussing the wider implications of multilevel selection theory.

The empirical advances the target article makes over Atran (1990) tend not so much to enrich our knowledge of the “folk taxonomic” hierarchy as to militate against the idea of one. Folk-biological domain-specific universals are to be found not in “taxonomic” kind-kind subordination relations, but in the relation of individual organisms to low ranking kinds and in the peculiarities of those kinds.

Progress in the “theory of mind” debate would be better served at this point by abandoning the search for a perfect “critical experiment” and developing an incremental research program based on a systematic theory of “theory of mind.” Studies using the goggle procedure advocated by Heyes should dissociate the ability to see from possible behavioral artifacts of “blind” trainers.

Howe et al. fail to provide evidence that practice is sufficient and ignore evidence of high ability before instruction. They unsuccessfully discount savants; provide weak evidence against heritability of music, criticize retrospective evidence selectively, using it when it supports their position; and ultimately both accept and deny talent. Finally, they conflate a scientific question with one of policy.

Clark & Thornton take issue with my claim that parity is not a generalisation problem, and that nothing can be inferred about back-propagation in particular, or learning in general, from failures of parity generalisation. They advance arguments to support their contention that generalisation is a relevant issue. In this continuing commentary, I examine generalisation more closely in order to refute these arguments. Different learning algorithms will have different patterns of failure: back-propagation has no special status in this respect. This is not to deny that a particular algorithm might fortuitously happen to produce the “intended” function in an (oxymoronic) parity-generalisation task.

The hierarchical reach of Atran's cognitive universals is unclear, and some of the key concepts used to discuss them are notorious for their imprecision. Although ideas of class hierarchy pervade Atran's discussion, other ways of thinking are also allowed. The history and practice of systematic biology suggests that a nonclass hierarchical and continuity-based way of thinking has been common there until recently.

We doubt that theory of mind can be sufficiently demonstrated without reliance on verbal tests. Where language is the major tool of social manipulation, an effective theory of mind must use language as an input. We suspect, therefore, that in this context, prelinguistic human and nonhuman minds are more alike than are human pre- and postlinguistic minds.

Atran conjectures that a triggering algorithm for a living- kind module could involve inputs from other modules that detect animacy and intentionality. Here we further speculate about how algorithms for detecting specific intentions could be used to trigger between- or within-species categorization. Such categorization may be adaptively important in Eldredge's energy and information realms.

Great apes show behavioral evidence for secondary representation similar to that of children of about two years of age. However, there is no convincing evidence for metarepresentation in apes. A good evolutionary interpretation should be parsimonious and must bring developmental and comparative data in accord. I propose a model based on the work of Perner (1991) and close by pointing out a logical flaw in Heyes's second proposed experiment.

There is a problem with the definition of talent as presented by Howe et al. that makes it dependent on experts' ability to detect it in the untrained. In addition, the choice of musical performance as the example for innate talent is inappropriate, and musical board results are selective and biased tests of it. Outstanding mathematical reasoning ability offers much better evidence of genetic influence.