SPECIAL ISSUE Editor's introduction: Education and cognition--Naturalistic explorations into the nature of mind and learning.

My editorship of the Australian Journal of Education terminates with this special, Special Issue. I began my editorship (AJE, 40(1), 1996) by noting that I, like my predecessors, would be keen to stress issues in educational research that I considered of special importance for the field. In particular, conceptions of research methodology and the justification of research were so identified. How well we have done in coveting these areas will have to be judged by the readers of this journal. In publishing research from a multiplicity of theoretical and methodological perspectives, the AJE has, I believe, maintained its character as a broad-based generalist education research journal, one of the few remaining in the English-speaking domain of educational research. It has also maintained and increased the contributions by researchers from outside Australia, as evidenced by the writers published in this issue.

Cognition and its importance for education, the theme to which this Special Issue is dedicated, deserves to become a top priority of research in Australian education. For good reason, this century has been described as the century of the brain, and I believe that it is incumbent upon educators, whose business is teaching and learning, to engage with the most current cognitive science research in order to find out what is known about human brain functioning and the acquisition of knowledge and information processing. After all, and this still seems to puzzle some, there is a causal connection between the architecture of our brains and what we are able to learn and the manner in which we learn. It is such knowledge that helps reshape our conceptions of human learning with subsequent implications for the structuring of learning-teaching situations, among other things.

More particularly, what is known as `connectionism', or the neural net account of brain functioning, has generated fundamentally challenging insights into our standard, sentential theories of teaching and learning which presume the primacy of symbol manipulation such as use of language; of how to think about the human mind; the nature of knowledge and human practice; of intelligence and rationality; in short: of what it means to be human. The consequences for education theory and philosophy of education, too, are far reaching, as are the practical implications for structuring learning in the preparation of trainee teachers. The contributions by the outstanding writers in this issue discuss the importance of connectionist research from various perspectives and also raise implications of the challenges raised by cognitive science for the most fundamental aspects of human learning, and hence for the business we are all engaged in. The contributions are original, and their importance for pushing Australian research in education into a new dimension cannot be overestimated.

The articles are organised somewhat thematically, although this is not an entirely satisfactory way of classifying the order of appearance. The common element which threads its way through all the contributions is the assumption that detailed knowledge of recent connectionism or, more broadly, cognitive science holds the key to our understanding of the human brain, and that it is the human brain which holds the key to understanding learning and teaching, and indeed all other activities we are engaged in whether these happen inside or outside classrooms.

The theoretical perspective, more or less explicitly expressed, is that of naturalism, in that it is assumed that human beings are evolved creatures, linked to the vast chain of other living creatures, and endowed with a rather large brain. Just as Dewey argued in his pragmatist philosophy (of education), there is no substantive distinction to be drawn between a person's brain and the mind, the many ongoing disputes about the relationship between these purportedly separate and distinct entities notwithstanding (examples of good discussions are found in Dreyfus & Dreyfus, 1986 and Dennett, 1996). For naturalists, there are no such hard distinctions and there is no ontologically different realm called the mind, which is superior and different from the physical brain. Rather the explanation of the architecture of our brains, the result of the history of the brain's developmental context, helps us understand why we have developed the image of ourselves as beings whose most proud and determining feature is that they are rational, that is, symbol manipulating creatures.

According to the results of modelling biological brain functioning in artificial neural nets (ANNs), we are now able to understand that the use of language and other symbol systems and their representation in symbolic form is due to the brain's primary ability to recognise patterns. Symbols of any kind are primarily neuronal patterns whose specific feature is that they are amenable to representation. Human practice, however, as expressed in playing the piano, leading a school or reading the mood of a meeting, is not the kind of activity that can be represented in symbolic form. We have a hard time trying to express in words what we know how to do well. This inability to explain what is often called tacit knowledge, the other side of the propositional/tacit dichotomy so pervasive in the social sciences, or knowing that and knowing how, has bedevilled much of education as well. Creating good practitioners, for example, is the raison d'etre of every teacher education program; hence capturing that elusive tacit knowledge became the most important task of the whole enterprise. But since the tacit/propositional distinction has shaped education theory and practice so fundamentally with its assumption that we learn primarily through the use of language, the theoretical lenses prohibited getting into focus that which could not be captured in sentential/propositional form. The tacit remained mysterious and unexplained. As long as this unhelpful dichotomy is maintained, accounts of practice are bound to fall short of causal explanation. (For further detailed discussion of this issue, see Evers and Lakomski, 2000, who have developed a naturalistic-coherentist account of practice in the context of educational administration.)

On the understanding that the primary mode of representing human knowledge is by means of neurological configurations or patterns where the strength of the connection weights is what makes the difference between `reading' an event/action/practice, there is no longer the distinction between what we can represent symbolically and what remains tacit or hidden. Since both are examples of neuronal patterns of activation that differ in the manner of their representation--at symbolic or sub-symbolic level--the traditional dichotomy dissolves and makes way for a more productive way of refashioning our account of human cognition which becomes more inclusive of all human cognitive capacity and does not define it as only that which can be captured in symbolic form.

What connectionist research allows us to do is to see that we have taken the part for the whole. We have assumed from our ability to manipulate symbols that this is the sum total of human reason and rationality, the essence of what it means to be human. The history of western philosophy is eloquent testament to this assumption. But, as we now know, symbol manipulation is only one of our special features. Once it is realised that we have privileged that part of our cognitive capacity that can be represented symbolically and once we see that all our cognitive functioning is a matter of pattern recognition and activation, it is possible to understand the whole spectrum of human cognitive functioning: from debating Descartes' cogito ergo sum to playing the violin, and running a school.

These are complex matters to which I can only allude in this brief introduction (the authors of this issue present many excellent references for further study on all these aspects), but it will be clear at least that the scope of what is being researched is vast, and the potential consequences for the business of teaching and learning quite enormous and very exciting. One of the most important tasks connectionists face is to explain just why and how language has become as important as it has, given that it is a relatively late evolutionary development. This task is intimately connected with the need to reconsider the development of human culture from a naturalistic perspective since we live and learn in communities of different cultural contexts. In other words, the task is not only to explain how knowledge is generated in individual skulls, but also to explain its socially distributed character.

This might sound contradictory at first glance. Knowledge on this account is not the property of individual skulls although generated by them, but is also socially owned and generated. The reason for this is that, although the human brain is an enormously powerful processor, it is also finite in how much it can process, a biological limitation which requires the `out-sourcing' of cognitive tasks. Modern organisation is the result of this biological necessity. No one person can comprehend and know all there is to know about modern flight, for example, hence specialisation was required. In the same manner, no teacher or principal can know all there is to know about how to educate. It is this biological feature of our brains that led to the development of culture and organisation. Our creations, such as the invention of the wheel or the mechanical loom, in turn made it possible to develop the modern transport system and mass production. So we are engaged in a perennial cognitive boot-strapping with the objects, processes, and artifacts we have created. This insight is being developed at present in the pioneering work of Strauss and Quinn (1997) within the field of cultural anthropology. In this sense, human knowledge is embedded. It is also embodied (see Clark, 1997, who examines how we can put `brain, body, and world together again'; also Damasio, 1996) in that brains are located in physical bodies that exert their own contingencies on cognition.

Finally, a word of caution. The picture I have painted above is nothing more than the merest outline of some fundamental, new developments in our understanding of human cognition. Our knowledge has not yet developed far enough to be specific about how these insights might lead directly to better educational practice. The articles that follow nevertheless provide some initial explorations into the realm of application, and in this they are truly pioneers in a terrain still vastly under-explored.

Since it is important to know at least in outline some of the scientific background and what some of the most critical developments in cognitive science research are, the article that leads off this Special Issue is Colin Evers's `Connectionist modelling and education'. Evers provides a detailed, technical introduction to where the cognitive science research is at present. In particular, he discusses the rise of the `new cognitive science', especially artificial neural net models, and explains one particular influential model. In the final sections, he provides some quite diverse applications and discusses their implications for education. The connectionist research agenda implicit in these applications is breathtaking in scope and will challenge many of our current understandings of a vast range of human cognitive activity.

In the second article, Carl Bereiter's `Keeping the brain in mind', two models of mind are discussed, which have quite different consequences for how we think the brain relates to mind and knowledge. The first, hugely influential, model is the `mind-as-container' where the image of the mind is more akin to that of the computer, and where knowledge is considered to be more like data stored in the computer's memory. The other model fundamentally challenges this notion in that the brain does not actually store or contain knowledge in the sense in which we traditionally believed it to do so. It is this second, connectionist, model that Bereiter advocates. A theory of mind for education that is based on a realistic understanding of the neural architecture and functioning of the brain will be educationally more productive. It will also provide a more realistic basis for dealing with the `knowledge age'.

Allan Yuen, in his article `Teaching computer programming: A connectionist view of pedagogical change', argues that teachers' views of knowledge especially affect their educational perspectives and their pedagogical practices. Over the years, he points out, two predominant computational theories of mind have emerged from the study of artificial intelligence: the symbolic and the connectionist models. Based on a qualitative study of 12 computer studies teachers in Hong Kong, Yuen argues that the connectionist view provides an alternative way of thinking about understanding and knowledge that leads to better insights on teaching and facilitates pedagogical change.

John St. Julien's article, `Changing conceptions of human intelligence and reasoning: Implications for the classroom', discusses an alternative view of what makes human competence possible, framed by complexity theory and drawing on connectionism as well as situated cognition. St. Julien suggests that, based on these theoretical frameworks, it is possible to develop a perspective whose implications can provide the basis for a more fully articulated theory of instruction. Such a theory will have beneficial implications for the practice of education.

The contributions by Nicholas Allix and Howard Gardner and Michael Connell which conclude this Special Issue differ in format from the preceding since the Gardner/Connell piece is a reply to Allix. It is quite fitting that this Issue should end with such a lively exchange, a de facto critical dialogue, since the issues under debate are contentious, fundamental to the business of teaching and learning, and will not admit of easy answers.

Allix's article `The theory of multiple intelligences: A case of missing cognitive matter' is an examination of Howard Gardner's influential theory of multiple intelligences. Allix argues that, although Gardner's conception of human cognition, characterised by a set of multiple and distinct cognitive capabilities, is an advance over the narrow conception of IQ, it nevertheless runs into fundamental difficulties of a methodological kind. Analysis uncovers that it is founded upon a discredited empiricist theory of knowledge. As a consequence, Allix claims, the theory is unable to muster the necessary methodological resources for adequately explicating and defending the integrity of the multiple intelligences themselves, and for underwriting associated explanatory generalisations. Allix proposes to treat matters of epistemology and methodology in theory construction naturalistically and to draw from the best and most coherent theories of mind and cognition available, for insights into the nature of intelligent cognition, and as a basis for building better and more realistic theories of intelligence.

The Gardner/Connell reply takes Allix to task over the issue of interpreting the theory of multiple intelligences in terms of the goals the theory wanted to achieve on the one hand, and the difficulties inherent in contemporary neural net research on the other. On some matters, there is agreement between the discussants but there are also substantially divergent views on the nature of what constitutes evidence as well as the role (theory of) knowledge might play in the context of justification. Specifically Gardner/Connell believe that philosophy of science and neural networks for determining `multiple intelligences' are concerns far removed from their enterprise, and thus do not approach the real `heartland' of the theory of multiple intelligences.

The arguments put forward by all parties in this exchange are complex and detailed, and readers will decide for themselves how to evaluate them, both within their own as well as the wider context of the other contributions in this issue.

The extended discussion of education and cognition presented in this Issue provides an excellent insight into the current state of affairs of what we know about human cognition. And what we have learnt so far from the attempts to unveil the brain's architecture and functioning suggests radically new ways of thinking about what it means to learn and to teach, with radical consequences for both in the offing. If nothing else, the contributions in this Issue should have made it clear that the stakes are high, but that the rewards for investigating the biological brain are immense because we will know how to structure practice on the basis of knowledge rather than folk psychology, a worthwhile endeavour by anyone's reckoning.

Keywords
cognitive processes     knowledge representation    neurology
educational practices   learning                    teaching process

References

Clark A. (1997). Being there: Putting brain, body, and world together again. Cambridge, MA: MIT Press.

Damasio, A.R. (1996). Descartes' error. London: Macmillan.

Dreyfus, H. L. & Dreyfus, S. E. (1986). Mind over machine. New York: Free Press.

Dennett, D. C. (1996). Kinds of minds. London: Weidenfeld & Nicolson.

Evers, C. W. & Lakomski, G. (2000). Doing educational administration. Oxford: Elsevier.

Strauss, C. & Quinn, N. (1997). A cognitive theory of cultural meaning. Cambridge: Cambridge University Press.

Gabriele Lakomski is Professor of Education and Director of the Centre for Organizational Learning and Leadership, Faculty of Education, The University of Melbourne, Parkville, Victoria 3010.