Penn et al. demonstrate the many ways in which humans and nonhuman animals are behaviorally similar. As they point out, despite these similarities, there are significant differences between human and nonhuman animals outside the lab. Penn et al. argue convincingly that nonhuman animals rely largely on situational cues (e.g., perceptual features and associations), whereas humans are able to reason using complex relational representations. We expand on this point, arguing that the ability to understand and reason with relational information magnifies the cognitive differences between humans and nonhuman animals. In particular, our ability to form role-governed categories expands our conceptual and linguistic repertoire, allowing us to transcend mere situational and perceptual cues and represent concepts as related to functions and goals (Markman & Stilwell Reference Markman and Stilwell2001). In addition, our ability to construct analogies based on relational mappings between domains that are dissimilar on the surface makes representational change and conceptual innovation possible. Finally, these differences may help to explain the uniquely human phenomenon of cumulative culture (Tomasello et al. Reference Tomasello, Carpenter, Call, Behne and Moll2005). Role-governed categories allow humans to posit the existence of objects that fill a particular relational role regardless of the perceptual properties of that object.

Human concepts can be loosely divided into three types: feature-based categories, which are represented as collections of features; relational categories, which represent a particular relational structure; and role-governed categories, which refer to items that play a particular role within a relational structure (Markman & Stilwell Reference Markman and Stilwell2001). Many nonhuman animals have feature-based categories. Feature-based categories require only representations of co-occurrences among features. Humans, however, are able to construct relational categories. Some relational categories refer to particular important relationships in the world (e.g., kinship terms, which specify relationships between people). In addition, verbs specify relationships among a set of items that are part of a sentence (Gentner & Kurtz Reference Gentner, Kurtz, Ahn, Goldstone, Love, Markman and Wolff2005; McRae et al. Reference McRae, Ferretti and Amyote1997). Verbs are particularly interesting, because any given verb must be completed with a set of objects that play different roles within the relational structure named by the verb (Ferretti et al. Reference Ferretti, McRae and Hatherell2001).

For example, in the sentence “The EMT treated the accident victim,” the EMT and accident victim play particular roles within these scenes (Ferretti et al. Reference Ferretti, McRae and Hatherell2001). These roles themselves can in turn be named by role-governed categories, of which the typical agents associated with a relation are members. So a (medical) patient is someone who is the object of the relation X treats Y. Having relational and role-based concepts allows humans to categorize entities based on goals and functions, and to associate entities that, based on surface features alone, would be considered very different. The role-governed category patient allows humans and hamsters to be part of the same category, provided they are both undergoing medical treatment.

Analogical mappings – relational mappings between two domains – are ubiquitous in human reasoning. They allow the detection of subtle relational similarities between domains. More importantly, perhaps, they also allow extension of the representation of one concept by virtue of its similarity to another. Information about one domain that is connected to the relational match can be carried over to the other domain as an analogical inference (Falkenhainer et al. Reference Falkenhainer, Forbus and Gentner1989; Markman Reference Markman1997; Spellman & Holyoak Reference Spellman and Holyoak1996). Thus, analogies provide a powerful mechanism of representational change. The importance of analogy in scientific innovation, for example, has been well demonstrated, both in historical and naturalistic settings. Gentner et al. (Reference Gentner, Brem, Ferguson, Markman, Levidow, Wolff and Forbus1997) provide a detailed analysis of the importance of analogy in Kepler's discovery of the elliptical motion of the planets, showing it was his knowledge of light and magnetism that suggested solutions to the problems that planetary motion produced for seventeenth-century scientists. Similarly, Dunbar (Reference Dunbar, Ward, Smit and Vaid1997) found that molecular biologists frequently rely on analogical mappings to problems with familiar solutions to produce novel solutions. Finally, Christensen and Schunn (Reference Christensen and Schunn2007) find that analogies are crucial for the development of innovative ideas in engineering design teams.

As argued by the target article, role-governed categories and analogical reasoning are a result of straightforward differences in representational capacity between human and nonhuman animals. We suggest that these abilities serve to magnify the apparent cognitive differences between human and nonhuman animals, because they are crucial for the development of cultural systems that increase in complexity across generations. Animals who have only feature-based concepts have no way of escaping the attributes of existing objects to suggest similarities across items that are based on relational similarities or on the possibility that two objects or individuals play the same relational role. Once animals develop a representational capacity that allows them to represent that two items play the same relational role within a relational system, it becomes possible to envision additional objects that might also fulfill that relational role. Thus, if a rock is used to break a nut, this rock becomes just one kind of breaker. Finding other objects that could fill this same role (perhaps more effectively) is crucial to the development of a technology. Role-governed technologies are the ones that are central to cultures that increase in their complexity (Tomasello Reference Tomasello1999). Thus, role-governed categories are crucial prerequisites to the development of human-like tool cultures.

This view helps to explain how the cognitive abilities of human and nonhuman animals could simultaneously appear to be very similar and very different. Small differences in representation ability support large differences in the available knowledge base that humans and nonhuman animals have to reason with. What this work does not explain is how the leap from feature-based representations to relational representations is made. Future work in cognitive science must examine the important influence of language on the development of relational representations for insight into the development of relational systems (Gentner Reference Gentner, Gentner and Goldin-Meadow2003).