Cognitive Science

It is difficult to define cognitive science in a way that is unambiguous and uncontroversial. In his recent overview, Keil (1998, p. 354) wrote that “Most people know what cognitive science is when they see it; they have far more difficulty providing a strict definition”. As a working definition, cognitive science may be defined as a multi-disciplinary approach to studying how mental representations enable an organism to produce adaptive behavior and cognition. Although this definition is very general, it captures three aspects of cognitive science which have always been important in the field: a faith in the necessity of multi-disciplinarity, an agreement that the object to be explained is directed behavior and cognition, and a recognition that internal knowledge representation is relevant to that explanation.

There are three main reasons why a more precise definition is difficult.

The first reason is the multidisciplinary nature of cognitive science. Cognitive science arose historically from a confluence of several academic disciplines. Because of this, it encompasses a wider range of intellectual territory than many other empirical disciplines.

The multi-disciplinary nature of the subject was motivated by- and is compounded by- the inherent complexity of the phenomenon whose study defines the field. Cognition admits of multiple levels of analysis and/or description- from the sub-cellular to the cultural- which interact in a myriad of ways. This multiplicity of description has made it possible for very different approaches to coexist under the same rubric, even within a single identified academic discipline. For example, social psychologists and neuropsychologists working in the same department may both consider themselves to be cognitive scientists without ever finding a common explanatory vocabulary.

A third difficulty in defining cognitive science arises from the dynamic nature of the discipline itself. Rapid progress in the four decades since cognitive science first appeared has rendered doubtful the definitional utility of some of the original central features of the discipline. At the same time, several features that were originally under-appreciated or simply ignored have come to play a central role in current theorizing.

In light of these complexities, the best way to expand upon the definition of cognitive science offered above is to situate the discipline in an historical framework. The decision about where to begin with a history of cognitive science must be made somewhat arbitrarily. It is possible to trace the roots of cognitive science back many centuries, by arguing that cognitive science cannot be understood without being situated in the tradition of Western thinking which goes back at least to Aristotle. Any number of other starting points within the last two millenia might be defensible. However, there is widespread agreement among those who have written on the history of cognitive science that the discipline in its modern form arose in between 1955 and 1960 (Gardner, 1985). One scientist who was a participant in the activities which gave rise to the discipline at that time has even argued that the birth of cognitive science may be plausibly pinned down to a precise day: September 11, 1956 (Miller, 1979; see also Breuer, 1993). On that day a symposium held at the Massachusetts Institute For Technology (MIT) in Boston brought together thinkers from various disciplines who were to play important roles in the emergence of cognitive science. The topic of the symposium was ‘information science’.

Information science was, along with computer science, one of the two related disciplines underlying cognitive science that had been spurred on by the practical challenges posed by the Second World War. The British and American governments wanted to improve their understanding of how communicated information could be coded, decoded, comprehended, and otherwise manipulated. They had strong practical motivation to give heavy funding to disciplines that might be able to make practical breakthroughs in these areas.

In England, one result was the development by Colin Cherry and Donald Broadbent of a new metaphor for understanding human beings, inspired (as have been so many metaphors for understanding human beings) by recent technological advances: the information processing metaphor. Cherry and Broadbent proposed that we think about human beings in the same terms as were being used to describe simpler information processing devices, as a set of input and output channels with known, limited capacities (Broadbent, 1958). This metaphor defined a research program dedicated to exploring the constraints and functioning of human sensory channels, allowing for an unambiguous means of stating and exploring an old idea which scientific progress was making increasingly thinkable: that the human being might be a kind of complex machine.

In the United States, the demands of war led to the development of ENIAC, the first electronic computer, which had been designed for computing weapons trajectories. ENIAC was switched on in 1946, a little too late to help the war effort. Its design was directly based on theoretical principles outlined in a famous paper published in 1936 by the British mathematician Alan Turing, and in a MIT master’s thesis published in 1938 by the mathematician, Claude Shannon. The publication of either of these papers might be taken as an earlier plausible starting point for cognitive science. Turing’s paper announced a startling proof, showing that there was a simple, general form underlying all possible algorithmic manipulations of information: a universal language for describing computation. Shannon built on this work by demonstrating that the functions of existent electronic relays and switches were sufficient to implement a machine that could simulate the universal (but theoretically infinitely large) computing device that Turing’s paper had described.

The American mathematician John Von Neumann was closely following the developments in the emerging field of electronic calculation. He contributed so many important conceptual breakthroughs to the organization of the computer hardware that the modern computer architecture is often referred to as ‘the Von Neumann architecture’. Von Neumann was also among the first to speculate on the connection between the new technology and the human brain. He died of cancer before seeing the publication of his own contribution to the field that would become cognitive science: his book, The Computer And The Brain (Von Neumann, 1958).

Another talk given at the 1956 MIT symposium, by Alan Newell and Herbert Simon, built on Von Neumann’s dream of trying to understand human thinking in terms of the electronic computer. Newell and Simon presented the first machine-generated proof of a theorem in symbolic logic, using a computer named ‘Johnniac’ in honor of Von Neumann. In some ways the presentation was a mere formality, since Simon had proven by hand some months earlier that such a demonstration was in principle possible, and since the theorem the computer proved had already been proven true decades earlier by the British mathematicians Alfred North Whitehead and Bertrand Russell. However, Newell and Simon made strong claims about what they had shown. They insisted that they were not merely demonstrating machine intelligence, but actually demonstrating the general laws underlying all thinking (Simon, 1969). In their insistence on this strong interpretation of their work, they stated a major claim of early cognitive science: that one could study the general organization of cognitive processing independently of how those processes were implemented in the human brain. There was a dissociation between the form of the algorithm that was used to solve a problem, and its implementation. The former could be studied without paying attention to the latter.

The developments in information processing and artificial computation provided researchers from numerous fields with new intellectual resources for thinking in a formal and rigorous way about cognitive processes. They also provided a new terminology for psychological theorists, enabling them to think in terms such as algorithms, information buffers, flow bottlenecks, and recursion loops.

The growing popularity of such computational ideas both contributed to and was driven by a parallel change occurring simultaneously in the fields of philosophy and psychology. Practitioners in these fields had begun to realize that methodologies which were defined by their insistence upon limiting themselves only to empirically-accessible sense data (behaviorism in psychology, and positivism and verificationism in philosophy) were too limited to explain all the phenomena we might wish to explain (Hebb, 1949). The old dominant metaphor of the brain as a telephone switchboard which connected stimuli directly to response gave way to a more complex metaphor which saw the brain as “a map room where stimuli were sorted out and arranged before ever response occurred” (Bruner, Goodknow, and Austin, 1956, p. vii).

A third talk given at the 1956 MIT symposium on information processing directly emphasized the need for postulating such cognitive pre-processing. The speaker was a young linguist in the process of single-handedly re-inventing his field: Noam Chomsky. Chomsky proved that the simple computational principles that had been outlined by Shannon were not sufficient to account for human language. He presented a new way of thinking about language which postulated computational transformations across the morphological units of language (Chomsky, 1957). Language, Chomsky claimed, could not be explained unless the black box of the brain was opened up – unless theorists allowed themselves to speak of inferred computational events defined with operations which had no obvious external manifestation and which went beyond those defined in logic. Chomsky’s work contributed to the acceptance of the key idea that understanding the mind’s functionality was going to require the development of new ideas about how information could be represented and manipulated. As well as having implications for the kinds of information processing the brain must be doing, Chomsky’s work underscored the need to consider that some kinds of knowledge might be innately specified, hard-wired into the human nervous system at birth.

In the years following the 1956 MIT conference, interest in the promise of the new computational sciences grew rapidly. In 1960 two psychologists at Harvard University – Jerome Bruner and George Miller – gave us another reasonable date for the beginnings of cognitive science, when they founded the Center For Cognitive Studies. The center was explicitly devoted to furthering the new ways of thinking about thinking, and played a central role in the early years of cognitive science (Posner & Shulman, 1979; Bruner, 1988; Norman & Levelt, 1988). In 1969, the first influential general textbook devoted to the exposition (and criticism) of the ‘new science’ was released: Ulric Neisser’s Cognitive Psychology.

Three years later Newell and Simon published their own massive book, Human Problem Solving, in which they described a general computational approach to problem solving. Their approach – means-end analysis – can be seen as an extension to cognition of the earlier cybernetic principle of feedback (Wiener, 1948). In feedback systems, information about the current state of the system is compared to the desired state of the system, and the current state is nudged in the desired direction. This simple principle is widely used in both machine and natural control systems. The application of the idea to cognition is very simple. It involves the comparison (often on several dimensions) of the current state of a problem to the desired goal state, and the application of an operator that reduces the distance between those states. Newell and Simon’s book demonstrated that this simple approach could be used in an algorithmic fashion to solve a range of problems that had previously been considered to require human intelligence. Although its applications have become increasingly sophisticated and its limitations increasingly apparent, variations of the general approach described by Newell and Simon are still widely used in artificial intelligence and cognitive science today.

Cognitive science received a boost in the mid 1970s, when the privately-administered Sloan Foundation chose to give major funding to cognitive science initiatives. Soon thereafter the discipline’s first dedicated journal, Cognitive Science, began publishing. The new approach to the mind was coming into its own.

In the twenty years since cognitive science became sufficiently defined to have its first journal, the discipline has undergone a number of changes. Four in particular have drastically altered (or are currently altering) the field’s basic conceptions about cognitive processing:

The first is the rise of distributed models of mind, and the resulting decline of top-down methods and architectures. Early cognitive science labored under a view of computation too strongly based on Von Neumann architecture. It was wedded to a single central processing unit in control of a single stream of computation. Today’s cognitive science is increasingly emphasizing forms of parallel, distributed, interrupt-driven, stochastic computation.

The second is the integration of neuroscience with cognitive science. Advances in technology (such as functional brain-imaging and connectionist models) and the growing body of traditional neuropsychological work are together making it increasingly tenable to try to constrain cognitive theories of function with neurological evidence. Moreover, the recognition that many purely cognitive theories are woefully under-specified is making the evidence of neuroscience increasingly important in adjudicating between cognitive theories.

The third trend is the emergence of evolutionary psychology (Dennett, 1995; Deacon, 1997; Bogdan, 1997; Hendriks-Jansen, 1996; Pinker, 1997). Much evidence suggests that human beings are far from optimal computational devices. Many puzzles about why this is so can be addressed from an evolutionary point of view, since evolutionary fitness has no need of perfection as an explanatory principle.

The fourth trend is a re-consideration of the role of context in cognitive function. In his history of cognitive science, Gardner (1985, p. 41) wrote that “Though mainstream cognitive scientists do not necessarily bear any animus against the affective realm, against the context that surrounds any action or thought, or against historical or cultural analyses, in practice they attempt to factor out these elements to the maximum extent possible”. A small number of anthropologists and philosophers outside of the field argued in the formative years of cognitive science that context must be taken into account in understanding human beings (see Wittgenstein, 1958; Winch, 1958; Austin, 1962; Garfinkel, 1967; Geertz, 1973). However, context tended to be treated by early cognitive scientists as noise, to be controlled for but not allowed into theory, since “For most psychologists, the idea that context can differentiate cognitive processing is akin to acknowledging the fragility of our theories” (Ceci & Roazzi, 1994, p. 74). However, there is now incontrovertible evidence that many functions of interest to cognitive psychologists – including reasoning (Ceci & Roazzi, 1994; Ferrari & Sternberg, 1998), memory (Anderson, 1982; Chase and Simon, 1973; Ceci & Leichtman, 1992) and even low-level motor control (Mowrey & MacKay, 1990; Kelso, 1995) – are more sensitive to context than many early cognitive scientists had suspected. Today an increasing number of cognitive scientists have accepted that accounting for the effects of context is important, or even crucial, to understanding human cognition. Not all cognitive scientists had ever believed otherwise. Jerome Bruner, recognized as one of the founders of cognitive science for his role in setting up the Harvard Center For Cognitive Study, published a book in 1990 in which he not only approvingly declared that “the contextual revolution (at least in psychology) is occurring today” (p. 105-106), but also insisted that the current growing interest in context constituted a return to one of the original, long-neglected goals of cognitive science. Bruner’s vision for cognitive psychology, which he calls ‘transactional contextualism’, emphasizes action in context as being constitutive of experienced meaning (for discussion of the same theme in from a cognitive science perspective, see Varela, Thompson & Rosch, 1991; Franklin, 1995; Hendriks-Jansen, 1996; Shore, 1996).

These changes have made cognitive science more complex than its original vision, which was in part an attempt to idealize the problem of cognition to render it simple enough to be tractable. Cognition has to date resisted such idealization. Although the future may yet hold an insight as profound in its simplifying implications for cognition as the Copernican insight was for understanding astronomy, few cognitive scientists today cling to such a hope. It now seems more likely (as Steve Pinker, 1997, has suggested in his recent summary of the field) that cognitive science will continue to deconstruct cognition into an increasing number of specialized computational tricks, which compete together for control of the organism and somehow manage to give the impression of something coherent enough to warrant the single label ‘cognition’. The science of cognition may forever be a record of increasing diversity and multiplicity, resisting our every attempt to fit the components neatly into a simple, unified- and easily-definable- package.

Acknowledgements: This article draws from an earlier document entitled ‘Knowledge representation in cognitive science: Implications for education’, by Chris Westbury and Uri Wilensky, which was internally commissioned by the Ministry of Education of the Government of Peru.