Studies of copular forms are extremely relevant to issues in philosophy, psychology, and linguistics. Psychologists have recently argued that the most distinctive aspect of human language is its combinatorial nature (e.g., Gentner, Reference Gentner2003; Spelke, Reference Spelke2003). They argue that this linguistic component might be what separates human from animal cognition. As combinatorial elements, copular forms have no equal. However, as the quote from Bertrand Russell implies, copular forms are multi-dimensional and complex. A full understanding of their use, linguistic variation, acquisition, and relation to cognition is a formidable task. While other papers in this special-issue volume of Bilingualism: Language and Cognition will address the different syntactic and semantic functions carried by these forms, in this commentary, I will focus on the relation between the acquisition of copular forms and cognitive development, or more specifically on the development of categorization.

Categories play a central role in components of cognition such as memory, language, and perception. We use categories to organize the stimulus world, to capture similarities, and as a basis from which to draw inferences. There continues to be considerable debate and controversy in the literature on categorization within adults as to what the structure of categories is (see e.g., Murphy, Reference Murphy2002; and Rogers and McClelland, Reference Rogers and McClelland2004 for recent overviews). However, one thing that is not debated is that categories change with development. Ever since developmentalists have studied cognition, conceptual differences between children and adults have been identified by the most prominent developmental psychologists of the century, and figure heavily in nearly all theories of cognitive change. Vygotsky (1934/1962), for example, viewed the child's concepts as undergoing a fundamental change from poorly organized in terms of familiar instances to rule-governed, and cites an example of a child who first understood the term “aunt” to refer to females of a particular age and later to refer to a kinship relation. Since Vygotsky, there have been several views of perceptual and cognitive development that suggest analogous shifts.

One of the functions of the copular forms is to assign properties to referents, and through its dual-form copular system Spanish assigns different kinds of properties to referents. Many studies have shown that language use and its development can reflect aspects of underlying cognition. Thus, by studying the development of copular uses we may gain insights into the underlying changes in categorization that accompany them. More specifically, by understanding how a child begins to learn the different meanings of ser and estar, we may gain insights into how a child comes to understand that there are different kinds of categories, and that they have different kinds of properties. This is important because even though there have been many studies documenting conceptual change in childhood, there are surprisingly few explanations of how those changes might take place.

One clear example of how studying ser/estar can offer information about cognitive development involves the distinction between different kinds of properties. Considerable evidence in cognitive development indicates that young children (at least monolingual English-speaking children) start the world as phenomenologists (e.g., Flavell, Green and Flavell, Reference Flavell, Green and Flavell1987). That is, they take reality to be what they apprehend through their senses. In one such demonstration, three-year-olds might be given a white toy lamb, and they acknowledge that the lamb is white. Then, a red-colored filter is put between the child's line of sight and the lamb, and the child is asked, “What is the lamb's real color, is it white or red?” Children often state that the lamb's real color is red even though they agreed seconds earlier that the lamb was white. Additional evidence suggests that English-speaking children tend to view “temporary” properties as permanent instead of vice versa (e.g., Bales and Sera, Reference Bales and Sera1995). Thus, children are more likely to believe that if one is happy, one has to stay happy for ever and ever, rather than mistakenly believe that people can change from being a person to being a dog. Why might children be confused about the relation between properties and their bearers?

The way children acquire ser and estar may shed light on more general processes that underlie the development of property differentiation. In documenting the path children take to distinguish ser from estar, Silva-Corvalán and Montanari suggest that initially children use ser with nominal predicates to identify entities in sentences such as “It is a man” and “It is a lion”. Even though this may be driven by discourse and pragmatic factors, these uses may initially establish the way that properties are viewed by young children. In these frames, properties identify referents in unchangeable ways and they may dominate the way young children view the status of properties. In contrast, the child's earliest uses of estar involve describing the location of entities in sentences such as “Here it is”. Thus, the initial partitioning of properties by ser and estar involves identity versus spatial locations. According to Silva-Corvalán and Montanari, from this initial partitioning of predicates, children begin to extract the different meanings of ser and estar. Ser is taken to assign identities to objects (which are typically permanent), and estar assigns locations (which are typically temporary). Thus, the properties that initially have estar status are more limited in kind (to spatial locations) than the properties that have ser status. Consequently, children have to gradually expand the kinds of properties that they use with estar and that are changeable. This may be why children have difficulty in conceptual development tasks that require attributing “temporary” status to properties such as the color of a lamb. The dominance and almost exclusive use of estar with spatial locations may both reflect and compound children's difficulty with property differentiation. For children, and especially for Spanish-speaking children, only spatial locations may hold a temporary status.

Another example of how the acquisition of ser and estar sheds light on category differentiation (or lack thereof) involves the distinction between objects and events. In a classic study on the differentiation of categories by children, Keil (Reference Keil1979) proposed that children initially do not make a distinction between objects and events, and that they treat events as if they were objects. Thus, for example, five-year-old English-speaking children say that events such as parties and recesses can have colors and shapes just as solid objects such as tables and trees do. In the acquisition of ser and estar, there is evidence of the same difficulty. For example, at 2;7 Nico uses the incorrect copulas when he says Mi cumpleaños está en mayo “My birthday is in May”. So Nico appears to be overgeneralizing the form that is used for objects to describe events. Other studies have reported similar patterns (e.g., Sera, Reference Sera1992; Sera, Gathje and Castillo, Reference Sera, Gathje and Castillo1999). In this case again, the path of acquisition of ser and estar may offer insight. The dominance of estar with spatial locations appears to be driven by almost exclusive devotion to discussing the spatial locations of physical objects (versus events). So, just as sheer frequency dictates that only spatial locations have estar (and temporary) status, sheer frequency may also dictate that all categories have object status. In short, frequency may dictate which kinds of categories (and properties) form a sort of psychological “magnet” from which other categories and properties have to be extracted or pulled out.

Although others have noted these patterns of use, the thoroughness of the Silva-Corvalán and Montanari analysis rules out other (less interesting) reasons for the restricted uses by children. For example, one (uninteresting) possibility for the protracted use of ser may be the fact that ser is an irregular verb whereas estar is a regular verb. So ser might be harder to conjugate and, thus, harder to produce than estar. However, the analysis of uses as a function of verb inflection in Silva-Corvalán and Montanari's Table 6 suggests comparable ages of producing both copulas across different persons. So it is clear from the evidence reported by these authors that the potential difficulty of conjugating forms of ser is not the reason why children are failing to use it to describe the location of events.

Understanding the acquisition of ser and estar should also advance our understanding of bilingualism and cognition. Much attention has recently been placed on processing advantages by bilinguals (e.g., Bialystok, Craik, Grady, Chau, Ishii, Gunji and Pantev, Reference Bialystok, Craik, Grady, Chau, Ishii, Gunji and Pantev2005). For example, bilinguals seem to be able to inhibit information and switch their attention across different tasks better than monolinguals. Studies of specific structural differences across languages (such as the Spanish dual-form copula) point to potential structural differences in bilingual cognition. For example, bilinguals learning ser–estar must learn to accommodate different organizations of properties across different languages. These language differences could lead to differences between bilinguals and monolinguals in the organization of conceptual properties and attribution. Yet, these potential differences have received little attention in the psychological literature.

In sum, given the difficulty that cognitive psychologists have had in describing the status of a category's property, it is not too surprising that linguists who have studied Spanish have encountered similar difficulties in describing the different functions of ser and estar. However, just as developmental psychologists have gained insights into categories by studying how they change with development, similar insights should be gained about ser and estar by examining their acquisition by children. Some will read the Silva-Corvalán and Montanari paper, and immediately become skeptical because it is a case study of a single child. So in closing, I would like to end with the quote from a classic work in developmental psychology by Fraiberg (Reference Fraiberg1977), which starts with the question “What can we learn from just one baby?”. As anyone who has read that book knows, Fraiberg's case study became the cornerstone for education of the blind in his country. More generally, and as the Silva-Corvalán and Montanari paper again illustrates, we can learn a lot from the study of a single child.