Jones claims that his extended framework of Optimality Theory (OT) overcomes the limitations of previous accounts of kin terminology. His application of the study of conceptual structure to kin terminology and his finding that the latter “borrows much of its organization from the conceptual structure of space … [and] the tracing of close parallels in conceptual structure in both cases” (sect. 5.1) promises to integrate kin terminology within human cognition.

In spite of these promising aspects of the paper, several problems remain. First, Jones has given us no new reason to believe that kinship terminology has to have a universal grammar (UG). Taking kinship UG for granted, he never addresses standard criticisms of UG in the linguistic domain from developmental (e.g., Behme & Deacon Reference Behme and Deacon2008; Edelman & Waterfall Reference Edelman and Waterfall2007; Elman et al. Reference Elman, Bates, Johnson, Karmiloff-Smith, Parisi and Plunkett1996; Monahagan & Christiansen Reference Monahagan, Christiansen and Behrens2008; Redington et al. Reference Redington, Chater and Finch1998; Sagae et al. Reference Sagae, MacWhinney and Lavie2004; Solan et al. Reference Solan, Horn, Ruppin and Edelman2005; Tomasello Reference Tomasello2003; Reference Tomasello, Kuhn and Siegler2006) or evolutionary (e.g., Arbib et al. Reference Arbib, Liebal and Pika2008; Botha Reference Botha1999; Christiansen & Chater Reference Christiansen and Chater2008; Deacon Reference Deacon2007; MacWhinney Reference MacWhinney, Bjorklund and Pellegrini2005; Studdert-Kennedy Reference Studdert-Kennedy, Hurford, Studdert-Kennedy and Knight1998; Tomasello Reference Tomasello2008) perspectives. While it seems possible that the universal rules Jones discusses exist, he has not shown that all humans rely on the same innate “template” to assign kinship terms. Given the complexity of constraints and rankings that become already evident in the two languages Jones discusses in detail, it appears questionable that the suggested version of OT could apply to all existing human cultures without spiraling into a computationally unmanageable system. Nor is it evident how such an intricate “machinery of kin classification” (sect. 5.2) might have evolved though natural selection. The constraints and trade-offs suggested just for the small segment of existing kinship terms Jones discusses give a flavour of the complexity required by a system that covers universally all human societies. And Jones does not even consider complicating factors such as within language variation of kinship assignments, polygamous or same sex marriages, divorce, and remarriage. It would appear that kinship terminology needs to remain flexible to account for the fluidity of the input, and Jones does not make it clear how an innate system could offer such flexibility.

Jones observes that kin terminology is highly constrained: “Languages around the world have independently hit on similar patterns, while largely steering clear of other imaginable possibilities. Kin terminologies are not perfectly regular, but overwhelmingly so” (sect. 2.3). He takes this as support for his theory. Yet, it is also possible that kin terminologies are similar because what they describe is similar: There are limits on how human beings can be related to one another, how property can be passed on to the next generation, how sexual relationships (e.g., marriage) can be executed, and more. When learning kin terminology, individuals face a very limited specific subset of the potentially possible kin arrangements. Thus, it is imaginable that they learn the relevant kin terms from the input they receive without having to rely on an innate system that has to accommodate all possible arrangements. Furthermore, some of Jones' examples appear to be post hoc. Seemingly, we have to know how English and Seneca assign kinship terms before we can design the system that accounts for these assignments. Potential counterexamples are then accounted for by invoking additional “rules.” Take the Seneca derivation of “Older Mother's Mother's Brother's Son,” which results in ‘cross cousin.’ Jones admits that this violates constraints invoked earlier because “in the case of first cousins, aunts, and uncles, Seneca is meticulous about keeping adjacent matrikin separate” (sect. 4.1). Yet, given that ‘cross cousin’ is the term used by the Seneca, Jones holds that “apparently when Mother's Brother is buried deeply enough inside a larger expression, the contrast between adjacent matrikin is neutralized. This is a markedness effect.” (sect. 4.1). It is not clear how Jones' system could have predicted when to invoke the markedness effect had we not already known what the output is. This holds true even if we accept that Jones uses a revised form of OT that incorporates iterated processes which restrict the size of moves and allow the output to be fed back into the input repeatedly. It may be true that this revision “can generate the full array of Seneca cousin terms out to indefinite distances” (sect. 4.1). But, given that “cousin terms to indefinite distance” have no practical application in actual language use and that individuals often struggle to assign kin terms to distant relatives, it is not clear that such a system exists and how it could have evolved. Overall, it would have been helpful had Jones shown how his system can predict unfamiliar kinship terms after receiving only partial input from an arbitrarily chosen language. If those predictions do not match with the observed terms, it stands to reason that the proposed system has only limited applicability.

Finally, Jones may be correct to suggest that even though the content of individual kin terms varies widely across cultures “the conceptual structure of kinship manifest in the grammar of kin terms is more universal – and shows every sign of being adapted for tracking genealogical connections” (sect. 5.2). However, it does not follow without further argument that a universal grammar of kin terms exist. Just as the existence of language universals has been challenged recently (e.g., Evans & Levinson Reference Evans and Levinson2009) it remains to be seem whether or not the suggested kinship universals exist. Given that individuals grow up in particular communities and thus encounter there one specific subset of kin-terminology the coordination problem Jones discusses (sect. 5.3) could be resolved by data-driven learning. Learning kin terminology does not occur in isolation but is embedded in a broader context of language learning. Here like elsewhere it is plausible to assume that learners rely on multiple cues to arrive at the correct solution. And, considering that kin terms only cover a very small segment of human cognition, the complexity of Jones' proposed machinery needed to handle this segment casts doubt on the existence of innate structures that cover the complete linguistic domain.