The proposal that language has evolved to conform to general cognitive and learning constraints inherent in the human brain calls for specification of these mechanisms. We propose that just as cognition appears to be grounded in cross-modal perceptual-motor capabilities, so too must language. Evidence for perceptual-motor grounding comes from non-arbitrary sound-to-meaning correspondences and their role in word learning.

The commentary discusses three components of the target proposal: (1) analogy as a host of phenomena, (2) relations as transformations, and (3) analogy as priming. The commentary argues that the first component is potentially productive, but it has yet to be fully developed, whereas the second and third components do not have an obvious way of accounting for multiple counterexamples.

Central to Hurley's argument is the position that imitation is “automatic” and requires inhibition. The evidence for this is poor. Imitation is intentional, involves active comparison between self and other, and involves new learning to improve self-other likeness. Abnormal imitation behaviour may result from impaired learning rather than disinhibition. Mentalizing may be similarly effortful and dependent upon learning about others.

We distinguish the question whether only human minds are equipped with a language of thought (LoT) from the question whether human minds employ a single uniquely human learning mechanism. Thus separated, our answer to both questions is negative. Even very simple minds employ a LoT. And the comparative data reviewed by Penn et al. actually suggest that there are many distinctively human learning mechanisms.

With regard to social-cognitive deficits in autism and psychosis, Crespi & Badcock's (C&B's) theory does not incorporate the developmental context of the disorders. We propose that there is significant overlap in social-cognitive impairments, but that the exact manifestation of social-cognitive deficits is highly dependent on the dynamics of cognitive development and hence different in autism as compared to psychosis.

The proposal of Rips et al. is motivated by discontinuity and input claims. The discontinuity claim is that no continuity exists between early (nonverbal) numerical representations and natural number. The input claim is that particular experiences (e.g., cardinality-related talk and object-based activities) do not aid in natural number construction. We discuss reasons to doubt both claims in their strongest forms.

After offering a brief account of how we understand the shared circuits model (SCM), we divide our response into four sections. First, in section R1, we assess to what extent SCM is committed to an account of the ontogeny and phylogeny of shared circuits. In section R2, we examine doubts raised by several commentators as to whether SCM might be expanded so as to accommodate the mirroring of emotions, sensations, and intransitive actions more generally. Section R3 responds to various criticisms that relate to the account of social-learning Hurley proposes in the target article. We conclude in section R4 by responding to a number of commentators who argued for the limitation of control theory as a framework for studying social cognition.

We suggest there is somewhat more potential than Christiansen & Chater (C&C) allow for genetic adaptations specific to language. Our uniquely cooperative social system requires sophisticated language skills. Learning and performance of some culturally transmitted elements in animals is genetically based, and we give examples of features of human language that evolve slowly enough that genetic adaptations to them may arise.

In this commentary, we outline an epistemological continuum between earlier and later number concepts, showing how empirical findings support the view that specific and general underpinnings to number develop in parallel in children; and we raise the question, based on cross-syndrome comparisons in infancy, as to whether exact or approximate number abilities underlie these later skills.

Crespi & Badcock (C&B) have presented a novel view that the influence of genomic imprinting causes diametrically opposite disorders: namely, psychoses and autism. I propose an extended hypothesis that while genomic imprinting is likely to have an influence on the pathogenesis of psychoses and autism, it might contribute to phenomenological antithesis between as well as within these disorders.

We argue that formal analogical reasoning is not a uniquely human trait but is found in chimpanzees, if not in monkeys. We also contest the claim that the relational matching-to-sample task is not exemplary of analogical behavior, and we provide evidence that symbolic-like treatment of relational information can be found in nonhuman species, a point in contention with the relational reinterpretation hypothesis.

The target article by Leech et al. presents a compelling computational theory of analogy-making. However, there is a key difficulty that persists in theoretical treatments of analogy-making, computational and otherwise: namely, the lack of a detailed account of the neurophysiological mechanisms that give rise to analogy behavior. My commentary explores this issue.

Christiansen & Chater (C&C) have taken the interactionist approach to linguistic universals to an extreme, adopting the metaphor of language as an organism. This metaphor adds no insights to five decades of analyzing language universals as the result of interaction of linguistically unique and general cognitive systems. This metaphor is also based on an outmoded view of classical Darwinian evolution and has no clear basis in biology or cognition.

Rips et al. raise important questions about the relation between infant quantification and achievement of natural number concepts. However, they may be oversimplifying the interactions that characterize actual development in real time. Though they propose a worthwhile agenda for future research, its explanatory power will be limited if it does not address developmental issues with greater sensitivity.

Crespi & Badcock (C&B) hypothesize that psychosis and autism represent opposite poles of human social cognition. I briefly outline how computational models of cognitive brain function may be used as a resource to further develop and experimentally test hypotheses concerning “autism-psychosis spectrum disorders.”1

If connectionist computational models explain the acquisition of complex cognitive skills, errors in such models would also help explain unusual brain activity such as in creativity – as well as in mental illness, including childhood onset problems with social behaviors in autism, the inability to maintain focus in attention deficit and hyperactivity disorder (ADHD), and the lack of motivation of depression disorders.

Penn et al. argue that the complexity of relational learning is beyond animals. We discuss a model that demonstrates relational learning need not involve complex processes. Novel stimuli are compared to previous experiences stored in memory. As learning shifts attention from featural to relational cues, the comparison process becomes more analogical in nature, successfully accounting for performance across species and development.

Our understanding of the sense of taste is largely based on research designed and interpreted in terms of the traditional four “basic” tastes: sweet, sour, salty, and bitter, and now a few more. This concept of basic tastes has no rational definition to test, and thus it has not been tested. As a demonstration, a preliminary attempt to test one common but arbitrary psychophysical definition of basic tastes is included in this article; that the basic tastes are unique in being able to account for other tastes. This definition was falsified in that other stimuli do about as well as the basic words and stimuli. To the extent that this finding might show analogies with other studies of receptor, neural, and psychophysical phenomena, the validity of the century-long literature of the science of taste based on a few “basics” is called into question. The possible origins, meaning, and influence of this concept are discussed. Tests of the model with control studies are suggested in all areas of taste related to basic tastes. As a stronger alternative to the basic tradition, the advantages of the across-fiber pattern model are discussed; it is based on a rational data-based hypothesis, and has survived attempts at falsification. Such “population coding” has found broad acceptance in many neural systems.

Leech et al. present a connectionist algorithm as a model of (the development) of analogizing, but they do not specify the algorithm's associated computational-level theory, nor its computational complexity. We argue that doing so may be essential for connectionist cognitive models to have full explanatory power and transparency, as well as for assessing their scalability to real-world input domains.

Augmenting and supplementing the arguments of Crespi & Badcock (C&B), I show that digit ratio (2D:4D), a putative marker of prenatal androgen action, indeed appears differentially altered in autism-spectrum disorders (lower/masculinized) versus schizophrenic-spectrum disorders (higher/feminized). Consistent with C&B's framework, some evidence (substantial heritability, assortative mating, sex-specific familial transmission) points to possible sex chromosome and imprinted genes effects on 2D:4D expression.