A history of psycholinguistics in the pre-Chomsky an era

A history of psycholinguistics in the pre-Chomsky an era

By Willem Levelt

           How do we speak and how do we understand language? It is widely believed that the scientific study of these uniquely human abilities was launched during the 1950s with the advent of Noam Chomsky’s generative linguistics. True, modern psycholinguistics received a major impulse from this “cognitive revolution,” but the empirical study of how we speak and listen and how children acquire these amazing skills has its roots in the late 18th century. By the end of the 19th century the psychology of language was an established science and the field was booming up to World War II. Empirical psycholinguistics emerged from four roots.

The Viennese engineer Wolfgang von Kempelen spent 20 years constructing a “speaking machine”. His 1791 book contains a precise construction manual. Copies have been built and indeed, the machine can articulate complex utterances such as Leopoldus secundus. It is the first serious working model of the vocal tract. During the 19th century the study of speaking became an experimental endeavor. It became possible to exactly measure the “mental durations” involved in naming pictures, colors, or numbers. Wilhelm Wundt’s psychology laboratory in Leipzig, the first of its kind, became the cradle of experimental psycholinguistics.

Franz Joseph Gall, also in Vienna, was the first to develop serious brain anatomy during the final two decades of the 18th century. His dissection classes there and later in Paris attracted some of the best medical students. Gall proposed the theory that mental faculties such as the memory for words were localized in specific regions of the brain. The stronger such an innate ability, the larger the corresponding brain region. This idea was never entirely lost in neuroscience. Paul Broca’s advanced brain anatomy made it possible in 1865 to localize an important region involved in the production of speech in the left frontal lobe. With Carl Wernicke’s localization of a second region, involved in speech understanding, the study of language in the brain had become a mature chapter of psycholinguistics.

Jean-Jacques Rousseau’s Émile of 1762 pleaded for a reform of education, a “natural” education without drill. Rousseau’s plea for the careful observation of children initiated the keeping of diaries by parents and teachers. Philosopher Dietrich Tiedemann was the first to publish a diary, in 1787. It follows his son’s development during the 30 months since his birth and includes a number of observations on Friedrich’s acquisition of speech. More diaries followed during the 19th century, but diary studies became a real boom after Darwin (1877) published his own observations on son William’s early development. Studies of language acquisition, for a variety of languages, kept appearing till the present day. They became an important database for theories of language acquisition.

Sanskrit scholar William Jones formulated the lexical affinities between Sanskrit, Greek, and Latin in his 1786 lecture for the Asian Society of Calcutta. Such affinities among Indo-European languages had been observed since medieval times, but the budding Romantic notion of evolution became the impetus of explaining these affinities from a common origin of these languages. There must have been some proto-language from which all languages in the family evolved. This raised the question of how primordial human beings began to speak such a simple proto-language. This, one realized, was a psychological issue. Ever since, the empirical study of language origins and language functions in human communication has been an important chapter of psycholinguistics. Studying the emergence of language, in particular of sign languages, is still a rich chapter of psycholinguistics.

Peace did not always reign in the community of psycholinguists. Major controversies arose around World War I. In the European tradition it had always been a matter of course that language use is a mental phenomenon. But this was anathema for emerging American behaviorism. Speech acts are mere responses to stimuli; there is no mind mediating between the two. But peace was literally and seriously disturbed during Hitler’s regime. European leaders in psycholinguistics emigrated, mostly to the United States, in two waves. First, right after Hitler came to power in 1933, almost immediately ordering the dismissal of Jewish staff at German universities. Second, after the Austrian Anschluss in 1938 and the following invasions all over the European continent. It was only after World War II that the four roots of psycholinguistics sprang to live again as an interdisciplinary theory of human communication.

The progression of modern psycholinguistics can be traced in four major periods: the Formative Period, the Linguistic Period, the Cognitive Period, and the current Cognitive Science Period.

In the Formative Period, both linguistics and psychology were committed to an operation list philosophy which derived theoretical constructs from observable data by using a set of highly explicit and verifiable operations. Structuralism was the prevailing paradigm in linguistics and defined units like the phoneme and the morpheme in terms of operational procedures. Behaviorism was predominant in psychology and also gave primacy to observable data, methodologically focusing upon rigorous experimental design and statistical analysis of data.

The Linguistic Period is characterized by the rise of transformational generative grammar in linguistics, followed by its theoretical domination of psycholinguistic research. Noam Chomsky, the originator of transformational generative grammar, successfully argued that behaviorism, structuralism, and information theory, the scientific perspectives which had oriented psycholinguistic research in the Formative Period could not provide an adequate explanation of natural language if it was based on an operation list philosophy. He further argued that a deductive approach was required, and that linguistic theory has as its proper domain the underlying competence of speakers, and not their actual performance. By extension, psycholinguistics also took as its starting point the study of competence, with the study of performance a secondary activity. The centrality of grammar was taken as a basic assumption, with the sentence emerging as the primary unit in most psycholinguistic experiments during this period. Psycholinguistic studies in the 1960s tested whether the number and complexity of mental operations performed during processing was a function of the number and complexity of formal transformations seen in the grammatical derivation of that sentence, as postulated by linguists.

Because generative grammar was inherently interested in the nature of human language, there was also considerable interest in language acquisition as well as in linguistic universals. It was commonly held that the capacity for language acquisition is species-specific and is a genetically determined attribute of humans and humans alone.
The Cognitive Period questioned what the intrinsic capacity to learn language really means, and whether what is really brought to bear is a set of general learning principles. The major premise that underlies a cognitive approach is the dependence of language upon human cognition, the notion that language is but one of several outcomes of more fundamental cognitive processes. The centrality and independence of grammar was rejected, in favor of a view that the cognitive capacity described in grammatical accounts of competence is only one manifestation of human language, and is in no way prior to or independent of other cognitive and behavioral systems involved in the acquisition and use of language. Linguistic structures are not learned independently of semantic concepts and discourse functions, and more importantly, cognitive principles must be assumed to govern the acquisition of linguistic structures. Thus, although linguistic theory continued to play a role in psycholinguistic theory and practice, the role was not as directive as before.

Psychology in the 1970s also began to be disenchanted by linguistic theory, given its limited applicability to dealing with how we actually produce or comprehend language, and psychological investigations of language processing made less reference to linguistic theory. There was robust evidence for the role of surface structure in processing, but not for transformations. The Derivational Theory of Complexity, suggesting that the number and complexity of mental operations performed during processing was a function of the number and complexity of transformations, never received sufficient support to continue as a working hypothesis. Instead, much of the information gleaned from experimental results along the way was extremely informative about natural language processing, and showed that there was much more to be learned from psycholinguistic research than just this weak hypothesis.
The last and most current period in psycholinguistics is one matched by the development of Cognitive Science as an interdisciplinary activity. Psycholinguistics is now involved in a larger field of inquiry, that is, the nature of knowledge, the structure of mental representations, and how these are used in mental processes like reasoning and decision making. Current psycholinguistic theory also reflects a considerable theoretical variety in both psychology and linguistics, but there is nonetheless a tremendous amount of truly interdisciplinary activity. More than ever before, researchers are very much aware of developments in adjacent fields, exhibiting research goals and a breadth of knowledge not confined to a single discipline. Linguists and psychologists can no longer afford to ignore scientific answers in the other field which impinge on research problems in their own field and the explanations they offer for them.

The linguistic approach which one sees increasingly is one which takes information processing constraints into account. “Correctness” of a grammatical theory is no longer as vehemently argued, for grammatical theories can all be internally “correct”; it becomes more a question of compatibility of a system of grammatical description that is attuned to the problems, as well as the results, of psycholinguistic research. Language is now increasingly seen as a symbolic process, often leading to decisions based on knowledge, and not as the ultimate and only interesting knowledge set.

In turn, psychologists are once again interested in the role of linguistic structures in language processing. Some claim that psycholinguistics now is more vibrant than it has been since the 1960s. The interaction of linguistic structure and language processing may be the focal topic of psycholinguistic research, but that interaction is now subject to the double, and equally important, rigors of linguistic analysis and psychological experimentation. Although psychology and linguistics are once again equal partners in the psycholinguistic enterprise, they do not always subscribe to same philosophy of science. One of the current questions is whether the language processing system is modular or interactive, and much recent research probes the degree of modularity or interaction in the system. Linguists generally have favored the modularity hypothesis, while psychologists have been more attracted to interactive explanations of language and learning principles. This is an exciting time for psycholinguistic research, with its potential for contributing to an exacting and realistic science of the human mind, a cognitive science.