Much of my research program has focused on precursors of language cognition in neonates and very young infants. In the course of pursuing this topic, I have been involved in numerous international collaborations across countries, continents, and cultures. These experiences were invaluable in shaping my understanding of the development of language and cognition and my understanding of methods and approaches in psychological science around the world.
Why should a cognitive scientist study neonate behavior? This question produces answers that tend to reflect a host of different views rather than a research program. For instance, some could reply that “the younger the age of the studied infants, the more grounded the hypothesis about how the mind grows will become.” Others would say that “the mind is mostly the outcome of the interactions of a brain with the environment; it is from the interaction that structure arises,” suggesting that the mind is almost non-existent at birth but will grow because of interactions. I could evoke multiple other answers, but instead will offer some reasons that determined my own involvement in this area of research.
During the early 1960s, I was a graduate student in the psychology department of Harvard University. Most of my teachers were prominent behaviorists, including E. G. Boring, the famous historian of psychology; R. J. Herrnstein, a famous student of animal learning who some years later wrote about the Bell curve with Charles Murray; and B. F. Skinner, one of the most influential behaviorists. Other Harvard professors had a different outlook. Professors George A. Miller and Jerome Bruner were assembling the first Center for Cognitive Studies to explore higher mental processes and ask basic questions about the mind and its biological foundations. The Center was inaugurated the year after I began my graduate studies. The ambiance was very motivating and also very creative.
As a first year student, I had the chance to participate in a seminar with Miller and Noam Chomsky. The seminar’s objective was to review and discuss their chapters in the Handbook of Mathematical Psychology (Vol II, 1963, edited by Luce, Bush, & Galanter), which discussed the formal properties of grammatical systems. Both Chomsky and Miller were exploring the specific mechanisms that might mediate language learning, a property that is only available to the human brain/mind. Concomitantly, Eric Lenneberg, also working at Harvard, made his Biological Foundations of Language available to us in manuscript form, years before publication of his famous book. Such an atmosphere may explain why a graduate student could be hooked for the rest of his life trying to answer the question “why do only humans acquire natural language”?
After my extended stay in the Boston area — interrupted only by a one year visit to Piaget’s Centre Internationale d’Epistemologie Genetique in Geneva — I moved to France where I met Claudine Amiel-Tison, a well known neuro-pediatrician working at one of the largest maternities in Paris. This encounter was propitious and helped me pursue developmental research with neonates and young infants to understand the precursors that make the acquisition of grammar possible. At the beginning of the 1970s it was still very difficult to establish a laboratory in a maternity ward. My luck or my chance encounter with Amiel-Tison made this possible.
As soon as the laboratory was installed, I started exploring whether newborns have dispositions that are essential to favor language learning. Other colleagues in the United States and Scotland were already exploring language acquisition in very young infants (see Eimas et al.,1971, in the United States among others). Moreover, other investigators were also studying non-linguistic abilities in older infants (see Gibson, 1969, Held et al., 1985, etc.). With those early investigations, the climate began to change. Investigators began to suspect that the Aristotelian/Lockian tabula rasa theory might not be correct after all. Rather, it seemed to us that many dispositions are present in the mind of the neonate in harmony with discoveries coming from brain studies of development. In fact, the neonate’s brain is not like a Lashleian jello. It is already structured and has many aptitudes that go far beyond the so called “atavic reflexes.”
Initially in the Paris laboratory we had to rely on the habituation of behavioral responses, in particular, the non-nutritive sucking method, which is cumbersome even if it allowed us to show that newborns are able to discriminate simple syllables like [pat] from [tap] while they fail to discriminate [pst] from [tsp], syllables that are highly atypical. Even more, adding vowels to the edges of the non-discriminated syllables was sufficient to re-establish discrimination behavior between the [upstu] and [utspu] items.
In the mid-1980s, we discovered that newborns habituated to one language and discriminate a switch to utterances drawn from another language if and only if the novel language belongs to a different rhythmic class. This finding became one of the central themes the laboratory explored in the years to come as it became clear to us that speech-rhythm is one of the first properties to which neonates and very young infants attend. Indeed, it is only when they turn 4 or 5 months old that infants begin to react to a language switch involving languages that belong to the same rhythmic class.
Around the same time, we realized the importance of studying functions without ignoring the underlying cortical structures that support them. In a joint effort of the Paris laboratory and the psychology department of the Universite Libre de Bruxelles in Belgium, we were able to show that lateralization in very young infants’ brains is organized in ways reminiscent of the lateralization in adults. Indeed, we habituated infants with two different simultaneously presented syllables, one to the right ear and the other to the left ear. We found the left hemisphere more responsive when using syllables, whereas the right hemisphere was more responsive to musical stimuli.
Eight years ago, I moved to Italy and pursued my work at the International School for Advanced Studies (SISSA-ISAS) located in Trieste. I continue investigating the precursors of language acquisition, using brain-imaging methods, namely NIRS (Near Infrared Spectroscopy) or OT (Optical Imaging). In one study we showed that the left perisylvian areas of the neonate’s brain is activated more when the infants listen to natural utterances than when they listen to the same utterances after they were time-reversed.
Recently, using the same brain-imaging method, we were able to show that neonates react differently to a series of structured tri-syllabic items, (e.g. megogo, fisusu, etc.) as compared to similar but unstructured bi-syllabic items (e.g. medisu, gofifi, etc.). These findings suggest that newborns are sensitive to certain input configurations in the auditory domain, a perceptual ability that might facilitate later language development.
Although our brain-imaging work is being very actively pursued, we are not about to abandon the use of behavioral methods. In fact, some of our most significant advances come from the exploration of behavior. For instance, learning word order is one of the earliest language properties that infants achieve. Our Trieste laboratory, in collaboration with Mazuka’s laboratory at Riken, Japan, presented young infants with a speech stream composed of an alternation of frequent (F) and non-frequent (nF) syllables. After familiarization, infants were then tested with six-syllable long items, (e.g., F,nF, F,nF,F,nF or nF,F,nF,F,nF,F). We showed that eight-month-old Japanese and Italian infants have opposite order preferences. Indeed, Italian babies prefer the F,nF test words whereas the Japanese prefer the opposite word, thus mirroring the orders of their respective native languages. In Italian, frequent items appear at the beginning of utterances but not at the end, while the reverse tends to be true in Japanese. This suggests that prelexically, eight-month old infants possess some representation of word order. We propose a frequency-based bootstrapping mechanism to account for our results, arguing that infants might build this representation by tracking the order of functors and content words, identified through their different frequency distributions.
With the perspective of time, I have come to the conclusion that the study of precursors of language is essential to ground the models of language acquisition. We often experience difficulty in obtaining data in maternity wards or even with families that bring their infants to the laboratory. However, such data is essential to clarify the many opposing theoretical views. Without the required data, which can inform the debate, the discussions tend to become futile. Therefore, I am grateful for the opportunities that the intellectual and the linguistic environment have provided me; they have been tremendously useful for the exploration of the cognitive processes of both the very young infant and adults. I am a firm believer that theoretical progress of our disciplines would greatly benefit from a wider range of explorations of a large variety of languages and cultures.
I hope this convinces the reader that being a citizen of the world is not a bad experience, as some people believe. I have studied and worked in Argentina, the United Kingdom, the United States, Switzerland, France, and Italy, which made me aware of the importance of observing different cultures. Although I learned how different mores can be in various cultures, I also became aware of the similarities between individuals from contrasting cultures. Moreover, having had to use several different languages in my daily activities has enlightened me of the marvelous device humans have been endowed with: the ability to acquire language.