Burling's (1999) target article raises relevant questions but it seems unsatisfactory in several respects: (1) the restricted framework, (2) the apparently arbitrary and uncongenial selection of the prerequisites, (3) the avoidance of the question of how the emergence of those prerequisites could lead to human language, and especially (4) the strategy for dealing with phylogeny in terms of ontogeny. What is special about language evolution is that, while child language acquisition takes place in an already linguistic environment, the same does not hold for the origin of language (that is why we speak of its origin). Finally, pointing is mentioned as an overlooked, but relevant, prerequisite to language.
2. One has no qualms about the selectionist approach to language that this target article assumes, and the various questions Burling raises are all relevant to understanding language origin (and evolution). Less satisfactory is the way of setting them all together; in particular, the restricted and linear way in which the question of the origin of language is formulated - like the origin of words, it is seen as a progressive, step by step, process. I will try to elaborate this dissatisfaction in a more detailed way.
3. In the first place, it is highly contentious to reduce the question of the origin of language - and its cognitive prerequisites - to the question of the origin of words, or, for that matter, to any other single aspect of language. (As a matter of fact, Burling's focus of attention is just names, a subset of words). Ever since Condillac (1746), who was probably the first to attempt a naturalistic account of the origin of human language, it has been clear that in order to understand the origin of language one must pay attention to the feedback effects of each novelty, and especially to the cognitive effects of language itself (as a powerful means of representation). But attention must also be paid to the social as well as the neuroanatomical transformations that may accompany this, not with regard to the selective pressures for the prerequisites for language (as Burling briefly does in paragraph 14), but the different aspects of the single process of language origin.
4. It may seem mysterious that any of these changes took place at some point in the past, because they may only make sense when viewed from a complex stance that considers the possible interactions of these different factors (Dunbar, 1993). From this perspective, it is also surprising that Burling's questions concern the selective pressures that could favour the prerequisites he considers, instead of the ones that could favour language itself, for it is conceivable that the appearance of these different prerequisites (if we accept them for the sake of argument) did not necessitate the origin of language - not to speak of its functionality, and the reasons for its retention, evolution, and proliferation. This is not to say, however, that Burling's questions are irrelevant ones.
5. There are also some problems with Burling's preferred set of cognitive prerequisites, or to be more precise, his strategy of modifying the work of Bloom & Markson (1998) - which addresses the ontogenetic issue of word learning - by eliminating one of their prerequisites and including another two. The difficulty here is not the possibility that the ability to infer referential intentions and the imitation and use of motivated signs "can be subsumed under a broader concept of a theory of mind", as Burling concedes in passing (paragraph 5), but the fact that they do not seem to be homogeneous concepts. Take imitation. Initially, it is defined as limited to behavioural reproduction (paragraph 10), and in this sense its connection with recognition of iconic similarity, which is supposed to be a different prerequisite, is acknowledged. Later on, however, imitation is taken to imply recognition of intentions (paragraph 12), a capacity associated with the prerequisite of inferring referential intentions. Either way, it is not clear how imitation can stand on the same level as the other prerequisites. Moreover, if Burling were to restrict his definition of imitation to mimicry, he would be doubling the mechanisms of sign understanding/use, to include intentional and imitative use ("mimicry") of motivated signs. That's because what makes signs motivated is that their use/understanding does not necessarily involve a referential intention: its meaning is contextually transparent.
6. The four prerequisites distinguished by Burling, then, are not individuated by the same classificatory principles of cognitive competences. One might say that this list of prerequisites lacks explanatory power: because it adopts a mixture of approaches, it contributes little to understanding what is involved in word use.
7. Part of the problem is Burling's understanding of Bloom and Markson's prerequisites. (I have to put it this way because I haven't been able to get their paper yet.) The claimed "ability to infer the referential intentions of others" (paragraph 5) is insufficient unless linked to the ability to have competent referential intentions oneself, and it is not explained what this involves, or whether or not it is the same competence. It is risky to extrapolate competences involved in the use of language as prerequisites to language acquisition. First, the initial stages of language may take place without having these competences in full, and second, they may "co-develop" in ontogeny (Pinker, 1987)(where "co-development" in ontogeny parallels "co-evolution" in phylogeny).
8. One can fully agree that "only by knowing what a speaker is attending to can a child correctly determine the referent of a previously unknown word" (paragraph 7), but this is not enough to show that this intentional understanding is logically independent and ontogenetically precedes getting the referent right. For it could well be that its development occurs in trying to make sense of the referential behaviour of others. The ability to infer referential intentions (as something more complex than simply understanding the referential use of signs) is fully in place a couple of years after the period of massive word acquisition at the age of two (Gleitman & Landau, 1994). Much the same holds for Bloom & Markson's first prerequisite, which is "a rich understanding of the external world", and an enormous amount of our understanding of the world is gained through the acquisition of language in infancy (Gelman & Coley, 1991). Bloom & Markson's third prerequisite, "an appreciation of syntactic cues to word meaning" suggests that syntactic competence may have an innate basis (see below).
9. This approach could be improved by extending it to try and bridge the gap between "motivated signs" and "signs of institution": symbolic, conventional, linguistic signs. In other words, to try to illustrate how having these different available prerequisites gave rise to the appearance of new types of signs, whose most distinctive property, Burling insists, was that they could be used "in an unlimited variety of circumstances" (paragraph 2). In my view, this is the most difficult part of the question of the origin of language, and Burling's approach would be strengthened if it could shed some light on this crucial matter. In particular, the point about chimps' pokes (Tomasello, Gust and Frost, 1989) needs more discussion: this appears to be exactly the right place to look, since they exhibit the change from motivated to conventional signs. Traditionally this has been a major setback for any naturalistic account of language origin, since it was assumed that conventions require language to get established (Lewis, 1969), therefore the very constitution of language cannot be explained as a convention, on pain of regress. The chimps' pokes may provide an alternative way of understanding the creation of conventions, as well as insight into what else is involved in making these conventions community-wide.
10. This is not to say that we should recover a progressive view of evolution, from which to view our species as "improved" chimps. It is just that had our ancestors had similar competences in this respect as chimps have now, it would be easier to understand why and how they evolved. The reason these pokes are not "words", according to Burling, is that first, "they are not learned by imitation" (paragraph 4), hence they are unfit to be shared by a community or transmitted to the next generation; and, second, they cannot be used in a variety of contexts, but function only as imperatives.
11. This second aspect has to do with the issue of syntax, to which I'll turn next. The first aspect strikes me as a deep claim, and one which deserves more support, given that it is far from obvious; moreover, it is well established that chimps share community-wide "technical" practices, which are transmitted to subsequent generations (Sabater-Pi, 1974). So, even if the claim were well-founded, an account of why chimps do not include signs in their rudimentary "cultural" practices would be needed. At the very least, this makes it interesting to compare this species and our ancestors; at most, a whole theory of symbolic development in terms of the regulation of social interaction could be made out of it.
12. It could be thought that sign languages could be equally interesting in this respect, in that they also exhibit this evolution from iconic and indexical to symbolic signing (Klima & Bellugi, 1 979). But there is a deep asymmetry between the phylogenetic stance and the ontogenetic one which makes it particularly misleading in this case to read phylogenesis from ontegenesis, quite apart from the reasons that make such inference problematic in general. It is not simply that language development in infancy takes place in a linguistic environment, but also that language development seems to be endogenuosly driven. In other words, contemporary human children do not come to the task of language learning as the first linguistic humans did; they are already equipped with some neuronal "pre-wiring" or a "language acquisition device" (Osherson & Lasnik, 1990), that facilitates and guides this task. It is to this developmental background that the third prerequisite of Bloom and Markson alludes.
13. Burling's disregard for this demonstrates the simplicity of his approach. It is not just the question of language origin that must be answered, but also the question of its proliferation, as a species-specific feature, and this second question must necessarily account for the functional changes in the brain that became genetically coded (as long as this theory maintains its privileged status). This is perhaps most clearly exemplified by the specificity of the auditive processing of linguistic sound: if the motor theory of speech perception is on the right track (Mattingly & Studdert-Kennedy, 1991), it cannot be taken for granted that when the first words were proffered, the specific mechanisms involved in language perception were already in place. This disregard for syntax also gives a simplistic flavour to Burling's project. He insists that names can be used in an unlimited variety of circumstances, but the examples he gives are not of names in isolation, but names within propositions ("The snake has gone, Did you see the snake?, Where is the snake?, I don't like the snakes", paragraph 2 ). It is not easy to figure out how, with just "snake", such diverse uses might be possible. From this perspective, it is clear that understanding the evolution of human language cannot be reduced to understanding the use of names. This is not to say that human language is just animal signals plus syntax, but rather that the determinate character of linguistic meanings is linked to their structured character. This also suggests that it is contentious to interpret animals' signals as imperatives: they are much more ambiguous than this, involving all those various dimensions (Seyfarth & Cheney, 1993; Millikan, 1984). The error lies in trying to grasp those signals' meanings in terms of our own conceptual capacities, which are clearly much finer-grained.
14. One strategy for addressing this complex question of human language (of what came first, words or syntax) is to try to think of a possible intermediate stage, a proto-language, which could soften the transition from the most complex signals used by the most intelligent animals, and human language (Bickerton, 1990). But this necessitates taking the phylogenetic approach seriously, without substituting it for the ontogenetic approach - which is not to say that ontogenesis has nothing relevant to contribute.
15. There has been an attempt here to arrange these comments in terms of increasing distance from Burling's framework. I would like to close "by pointing to pointing", as a cognitive ability I missed in the paper. Pointing is involved in the most basic forms of linguistic - and protolinguistic - communication, long before the child is able to refer with names (Franco & Butterworth, 1996). It is another example of a transition from motivated signs to symbols, in that the understanding of the latter is facilitated by contextual clues. Pointing cannot work just by imitation, because its understanding involves perspective-shifting, and intentional attribution. From an evolutionary point of view, pointing seems to be plurifunctional, and hence it constitutes a good candidate for having been selected. Finally, pointing is also present in other primates (Gomez, Sarria & Tamarit, 1993). It is no wonder that, ever since Condillac, it has been thought that gestures, such as pointing, constituted the first human proto-languages: community-wide, discrete, conventional, and intentional but contextually transparent symbols, from which a more complex stage could develop.
This work has received the support of the Spanish Department of Education, through the project PB95-585.
Bickerton, D., (1990), Language and Species. Chicago: University of Chicago Press
Bloom, P., & Markson, L., (1998), "Capacities underlying word learning." Trends in Cognitive Science, Vol. 2. No 2, pp.67-73.
Burling, R., (1999), The cognitive prerequisites for language: Target article on language prerequisites. PSYCOLOQUY 10(32). ftp://ftp.princeton.edu/pub/harnad/Psycoloquy/1999.volume.10/ psyc.99.10.032.language-prerequisites.1.burling http://www.cogsci.soton.ac.uk/psyc-bin/newspy?10.032
Condillac, E. B., (1746), Essai sur l'origine des connaissances humaines. Paris.
Dunbar, R., (1993), "Coevolution of neocortical size, group size and language in humans". Behavioural and Brain Sciences 16, pp.681-735. ftp://ftp.princeton.edu/pub/harnad/BBS/.WWW/bbs.dunbar.html
Franco, F., & Butterworth, G., (1996), "Pointing and social awareness: declaring and requesting in the second year". Journal of Child Language 23, pp.307-336.
Gelman, S., & Coley, J. D., (1991), "Language and Categorization: the Acquisition of natural kind terms". In S. Gelman & J. P. Byrnes (eds.), Perspectives on Language and Thought. Cambridge: Cambridge University Press
Gleitman, L., & Landau, B., (1994), The acquisition of the lexicon. Cambridge, Mass.: MIT Press.
Gomez, J.C., Sarria, E., & Tamarit, J., (1993), "The comparative study of early communication and theories of mind: ontogeny, phylogeny and pathology". In S. Baron-Cohen et. al., (eds.), Understanding other minds: perspectives from autism. Oxford: Oxford University Press, pp.397-426.
Klima, E. S., & Bellugi, U., (1979), The signs of language. Cambridge, Mass.: Harvard University Press
Lewis, D., (1969), Convention. Cambridge, Mass.: Harvard University Press
Mattingly, I.G., & Studdert-Kennedy, M., (eds.), (1991), Modularity and the Motor Theory of Speech Perception. Hillsdale: Lawrence Erlbaum.
Millikan, R.G., (1984), Language, thought and other biological categories. Cambridge, Mass.: MIT Press. ftp://ftp.princeton.edu/pub/harnad/BBS/.WWW/bbs.millikan.html
Osherson, D., & Lasnik, H., (1990), Language: an invitation to cognitive science. Vol. I. Cambridge, Mass.: MIT Press.
Pinker, S., (1987), "The bootstraping problem in language acquisition". In B. MacWhinney (ed.), Mechanisms of language acquisition. Hillsdale: Lawrence Erlbaum.
Sabater-Pi, J., (1974), "An elementary industry of the chimpanzees in the Okorobik mountains, Rio Muni (Republic of Equatorial Guinea), West Africa", Primates 15 (4), pp.351-364.
Seyfarth, R.M., & Cheney, D.L., (1993), "Meaning, Reference and Intentionality in the Natural Vocalizations of Monkeys". In H. L. Roitblat et. al., (eds.), Language and Communication. Comparative Perspectives. Hillsdale, NJ: Lawrence Erlbaum.
Tomasello, M., Gust, D., & Frost, G.T., (1989), "A longitudinal investigation of gestural communication in young primates", Primates 30, pp.35-50.