Grush describes a variety of different systems that illustrate his vision of representation through emulation. These individual data points are not necessarily sufficient to determine what level of detail is required for a representation to count as emulation. By examining one of his examples closely, this commentary suggest that salience of the information supplied is a critical dimension.

If alleles that predispose to schizophrenia have reduced Darwinian fitness, their persistence in modern times is puzzling. Burns identifies the evolutionary genetics of schizophrenia as a central issue, but his treatment of it is not clear. Recent advances in evolutionary genetics can help explain the persistence of alleles that predispose to debilitating disorders such as schizophrenia, and can buttress Burns' core argument.

The fish-sharing patterns on Ifaluk Atoll underscore several limitations of the explanations of food sharing offered by Gurven and suggest that non-foraging labor activities may provide insights into reciprocity and punishment relevant for understanding food-sharing patterns. I also argue that future food-sharing studies should focus on signaling rather than resource holding potential (RHP).

Strong recent selection for social cognition may well explain the persistence of genes that predispose to schizophrenia. The specific mechanism responsible may be a skewed fitness function in which selection pushes the mean for advantageous mental traits perilously close to a “fitness cliff” where the system fails catastrophically in some individuals.

Although in general agreement with Carpendale & Lewis's (C&L's) claims, I argue that (1) gradual development is better supported by within-task eye gaze/verbal comparisons; (2) gradual development and social construction do not contradict the theory-theory view; (3) there is good evidence for an early developing self-other distinction; and (4) the language–false belief link could be mediated by parental talk.

We address the issue of the normativity of representation and how Grush might address it for emulations as constituting representations. We then proceed to several more detailed issues concerning the learning of emulations, a possible empirical counterexample to Grush's model, and the choice of Kalman filters as the form of model-based control.

Each of the food-sharing models that Gurven considers demands unique cognitive capacities. Reciprocal altruism, in particular, requires a suite of complex abilities not required by alternatives such as tolerated scrounging. Integrating cognitive constraints with comparative data from other species can illuminate the adaptive benefits of food sharing in humans.

All mammals have social brains, and there is presently no evidence that humans have relatively more genetically dictated social brain circuitry than other species. The postulation that schizophrenia arises from disruption of brains systems uniquely devoted to social traits is obviated not only by the large number of anatomical and biochemical brain differences, but also by nonsocial symptoms of schizophrenic disorders.

I criticize Carpendale & Lewis's (C&L) attempt to produce a Wittgensteinian theory, as an alternative to work in the “theory of mind” tradition, not because I disagree with it as theory, but because Wittgenstein would be critical of any attempt to make such a use of his work. His concern is with descriptions, not theories.

In his article Grush proposes a potentially useful framework for explaining motor control, imagery, and perception. In our commentary we will address two issues that the model does not seem to deal with appropriately: one concerns motor control, and the other, the visual and motor imagery domains. We will consider these two aspects in turn.

Although Carpendale & Lewis (C&L) correctly emphasize the importance of conversation in children's social understanding, they neglect several complex issues. Contrary to their assertion, the focus on mental state processing has not been misplaced, and there is a need to recognize that different aspects of social understanding are liable to undergo distinctive developmental changes that vary in relation to social interaction.

The most developed and the latest-to-mature mental skills represented in the creation of mono- versus polysemantic contexts are related respectively to the left and right frontal lobe. A polysemantic way of thinking is responsible for the subject's successful integration in the polydimensional world. The functional insufficiency of this right-hemispheric way of thinking displays a predisposition toward the development of mental disorders, including schizophrenia.

Although the application of the emulation model to the control of simple positioning movements is relatively straightforward, extending the scheme to actions requiring multisegmental, interlimb coordination complicates matters a bit. Special consideration of the demands in this case, both on sensory processing and on the process model (two key elements of the Kalman filter), are discussed.

I agree with Gurven that costly signaling can explain food-sharing phenomena. However, costly signaling may also explain the role of food sharing in deterring rivals. Details of food-sharing interactions may reveal gains and losses in the social prestige of the interacting parties. The evolutionary models of kin selection and of reciprocal altruism are unstable and should be avoided.

Evolutionary theories of schizophrenia must account for the maintenance of putative alleles in past and present populations despite reduced fitness among the affected. Such models must also account for extant intersex and population-level variability in the expression of schizophrenia. We argue that genetic balanced-polymorphism hypotheses remain the most robust in terms of modeling and testing these processes in populations.

Mental representations are more than emulations. Different types of representations, including external representations, various mental models (distorted and abstract), and emulative models, can all play important roles in human cognition. To explain cognitive performance in a specific task, a systematic analysis of the underlying representational structures and their interactions is needed.

Carpendale & Lewis's (C&L's) reanalysis of Chapman's (1999) epistemic triangle dealing with the coordination of interactions with physical objects and people's communication is misleadingly incomplete. An alternative proposal is outlined combining the causality of action with the normativity of knowledge in acts of judgment. This alternative is empirical and developmental, with a focus on rich but neglected phenomena.

Empirical data on food sharing in native Dolgan, Nganasan, and Nenets communities in Siberia provide evidence for hunter control over big game and fish, as well as likely benefits of inter-household sharing. Most food sharing occurs with kin and, thus, kin-selection-based nepotism cannot be ruled out. Reciprocal interhousehold sharing at meals occurs less often. Social context is discussed.

Many of Grush's arguments should apply equally to animals with small brains, for which the capacity to internally model the body and environment must be limited. The dilemma may be solved by making only very approximate predictions, or only attempting to derive a “high-level” prediction from “high-level” output. At the extreme, in either case, the “emulation” step becomes trivial.

Four models commonly employed in sharing analyses (reciprocal altruism [RA], tolerated scrounging [TS], costly signaling [CS], and kin selection [KS]) have common features which render rigorous testing of unique predictions difficult. Relaxed versions of these models are discussed in an attempt to understand how the underlying principles of delayed returns, avoiding costs, building reputation, and aiding biological kin interact in systems of sharing. Special attention is given to the interpretation of contingency measures that critically define some form of reciprocal altruism.