the experiments proposed in the article by steels & belpaeme (s&b) can be considered as a starting point toward a general methodology for the automatic learning of conceptual spaces.

consideration of color alone can give a misleading impression of the three approaches to category coordination: the nativist, empiricist and culturalist models. empiricist models can benefit from a wider range of correlational information in the environment. also, all three approaches may explain a set of perceptual categories within the human repertoire. finally, a suggestion is offered for supplementing the naming game by varying the social status of agents.

we propose a way to achieve across-population sharing within the authors' model in a way that is plausibly in accordance with human evolution, and also a simple way to capture ecological structure. finally, we briefly reflect on the model's scope and limits in modeling linguistic communication.

to vallortigara & rogers's (v&r's) evidence of everyday directional asymmetries in the natural environment of a variety of species, we offer one more example for human beings. it is the bias for holding an infant on the left side, and it illustrates several themes in the target article.

the debate of nativisim versus empiricism is over the relative importance of evolutionary versus ontogenetic mechanisms. this is mostly seen today as a false dichotomy. the synthesis of these positions provides a modern viewpoint of grounded category formation. this combined view places equal importance on feedback between these levels in guiding development, and is more appropriately compared to culturalist positions.

extrapolation of steels & belpaeme's (s&b) results show that colour is a culturalist category. populations will only share the category of colour if it is built into the system. if “left to themselves” different populations may or may not stumble on the colour category. populations that do not share a colour category may still be able to communicate in a wide variety of environments.

an evolutionarily stable strategy (ess) may apply to characters expressed across species for predation and feeding, because these characters are conservative. however, the evolution of complex, polymorphic behaviors is more difficult to define as an ess. lateralization may be selective for certain simple traits, but lateralization of complex traits is likely the result of coadaptation of otherwise non-lateralized features.

i suggest that a communicative context that has the potential to establish and maintain a shared advantage of behavioral lateralization should be identified in the domain of sexual selection, specifically in the interactions that individuals exploit to assess the fitness of potential mates.

the general structure of steels & belpaeme's (s&b's) central premise is appealing. theoretical stances that focus on one type of mechanism miss the fact that multiple mechanisms acting in concert can provide convergent constraints for a more robust capacity than any individual mechanism might achieve acting in isolation. however, highlighting the significance of complex constraint interactions raises the possibility that some of the relevant constraints may have been left out of s&b's own models. although abstract modeling can help clarify issues, it also runs the risk of oversimplification and misframing. a more subtle implication of the significance of interacting constraints is that it calls for a close relationship between theoretical and empirical research.

sunstein argues that heuristics misguide moral judgments. principles that are normally sound falter in unusual cases. in particular, heuristics generate erroneous judgments about regulation of risks. sunstein's map of moral reasoning omits some prominent contours. the simple heuristics he suggests neglect a reasoner's attempt to balance the pros and cons of regulating a risk.

steels & belpaeme's (s&b's) simulations contain all the right components, but they are put together wrongly. color categories are unrepresentative of categories in general and language is not merely naming. language evolved because it provided a powerful new way to acquire categories (through instruction, rather than just the old way of other species, through trial-and-error experience). it did not evolve so that multiple agents looking at the same objects could let one another know which of the objects they had in mind, co-coining names for them on the fly.

this article proposes a number of models to examine through which mechanisms a population of autonomous agents could arrive at a repertoire of perceptually grounded categories that is sufficiently shared to allow successful communication. the models are inspired by the main approaches to human categorisation being discussed in the literature: nativism, empiricism, and culturalism. colour is taken as a case study. although we take no stance on which position is to be accepted as final truth with respect to human categorisation and naming, we do point to theoretical constraints that make each position more or less likely and we make clear suggestions on what the best engineering solution would be. specifically, we argue that the collective choice of a shared repertoire must integrate multiple constraints, including constraints coming from communication.

this commentary explores the use of interaction between moral heuristics and emotional intelligence (ei). the main insight presented is that the quality of moral decisions is very sensitive to emotions, and hence this may lead us to a better understanding of the role of emotional abilities in moral choices. in doing so, we consider how individual differences (specifically, ei) are related to moral decisions. we summarize evidence bearing on some of the ways in which ei might moderate framing effects in different moral tasks such as “the asian disease problem” and other more real-life problems like “a divorce decision.”

we present behavioral lateralizations of spiders and ants and their probable survival value. they clearly conform to the vertebrate lateralizations reviewed by vallortigara & rogers (v&r) and to earlier arthropod studies. we suggest two complementary reviews: (1) differences in lesion susceptibility and muscle strength between left and right body side, and (2) perceptual biases and predator inspection in invertebrates.

if language is crucial to the development of shared colour categories, how might cultural constraints influence the development of divergent category sets? we propose that communities arrive at different sets of categories because the tendency to group by perceptual similarity interacts with environmental factors (differential access to dying and printing technologies), to make different systems optimal for communication in different situations.

a common objection to utilitarianism is that it clashes with our common moral intuitions. understanding the role that heuristics play in moral judgments undermines this objection. it also indicates why we should not use john rawls' model of reflective equilibrium as the basis for testing normative moral theories.

although there are some common underlying mechanisms for many nonhuman behavioural asymmetries, the evidence at present is not compelling for commonalities in cerebral organisation across vertebrates. phylogenetic analysis of detour behaviour in fish suggests that more closely related species are not particularly similar in the direction of turning; contingency and demands of ecological niches may better explain such asymmetries.

by emphasizing that color categories are the collective achievement of a language community, the methodology of steels & belpaeme (s&b) suggests a number of corollaries. it focuses attention on whether a system of categories is optimized to match color experience. if a hypothesis can be operationalized about the nature of the optimality – about how color language becomes standardized – it becomes testable.

crossing the border into the moral domain changes moral thinking in two ways: (1) the facts at hand become “anthropocentric” facts not easily open to revision, and (2) moral reasoning is often the servant of moral intuitions, making it difficult for people to challenge their own intuitions. sunstein's argument is sound, but policy makers are likely to resist.

preferences for options that do not secure optimal outcomes, like the ones catalogued by sunstein, derive from two sources: cognitive heuristics and deontological rules. although rules may stem from automatic affective reactions, they are deliberately maintained. because strongly held convictions have important behavioral implications, it may be useful to regard cognitive heuristics and deontological rules as separate sources of nonconsequential judgment in the moral domain.