Problems of humanism in cognitive science

In the previous chapters, the authors take a neutral stance between two radically different perspectives – the secular and the religious. We describe some of the evidence and arguments for each perspective from a relatively dispassionate standpoint. We thus hope to indicate how developments in cognitive science can be used to illuminate both secular and religious interpretations of the human being and, in turn, be illuminated by them. In Chapters 9 and 10, we suggest that hermeneutic and symbolic approaches coming from the social sciences can be seen as potentially continuous with cognitive science. Just as scientific models of the mind are schemas for perception, learning, and intelligence, so hermeneutics, which studies the interpretation of texts from within different symbol systems, can be seen as expressing schemas for the interaction of the individual and the social. Indeed, we show how the schema theorist can use the concepts of hermeneutic circle, dialectic, and dialogue as heuristic hypotheses for extending the science from individual to social.

We have, however, touched on some issues indicating that this conciliatory view is not sufficient to solve the problems of the secular view of the human nor to do justice to the religious perspective. This chapter makes explicit some irreducible controversies between the two perspectives that we broadly describe as the “secular” and the “religious.”

To some people, Freud's psychoanalysis is a word therapy that has provided invaluable new tools for psychiatry; for others, it is a failed pseudoscience. Yet for many humanists and literary scholars, clinical or scientific criteria are simply beside the point: for them, Freud has supplied a language to chart the human mind, not just its conscious rationality but also the unconscious and repressed sources of the darkness in people's souls.

Freud was trained as a neurologist and also received an excellent nineteenth-century European education and read widely in the classics. Thus, when creating the metapsychology of psychoanalysis, he built on concepts rooted in scientific materialism yet shaped them in the light of the human, yet transcendent, dramas of Greek myth. His work took him from neurology to reaches of the human mind that, at that time, resisted neurological explanation. His studies of the individual mind were complemented by studies of society and religion, in which he saw these as expressions of the individual psyches of people coming together into groups. God was the projection of human fears and wishes, not a transcendent reality constitutive of human meaning.

We thus devote this chapter to a critique of the work of Freud.

This chapter examines artificial intelligence (AI) – the dimension of cognitive science that focuses on programming computers to exhibit aspects of intelligence, without necessary regard for the constraints of human behavior (cognitive psychology) or of brain function (brain theory). We suggest that the achievements of AI to date are limited, but we deny that these are limitations in principle. Nonetheless, we argue that much of what is human about intelligence depends on our being embodied within human bodies and being members of human societies.

In this chapter, we distinguish formal systems in which words are related only to other words within a closed system from systems in which symbols are linked to action “in the world.” We then argue that Gödel's incompleteness theorem does not prove that machines cannot be intelligent, but rather that an intelligent machine must learn, and that some element of inconsistency is an inescapable facet of intelligence, whether in human being or machine. In the next chapter, we see the view of The Construction of Reality Piaget offers in his study of assimilation and accommodation in the child. We then build on Piaget's insights and work in brain theory and artificial intelligence to outline a schema theory that provides our bridge from cognitive science to epistemology.

Myth and history

In this chapter, we turn to consider explicitly the Bible as religious text, regarding this as a complex and holistic symbol system that, roughtly speaking, constructs the worlds of Judaic and Christian religion. We have two purposes in mind here. First, we want to exhibit some workings of a symbol system in this particular case. Second, we want to use the Bible to address two problems that have dogged our discussion of schemas and symbolisms throughout: the problems of relativism and transcendence. Can Biblical religion be seen as anything but another functional or chance manifestation of particular societies at particular times, or can we give some sense to its claim to universality, to be speaking truly of a transcendent God?

In considering the Bible as an internally knit symbol system, we are, of course, following the tradition of hermeneutics in social theory (of which indeed Biblical hermeneutics was the first example) and in social anthropology. This approach first became explicit in the nineteenth century; it should be noted, however, that in the mainstream of interpretation, the Bible has never been taken only as a “naïve story” or as factual history. The texts established (with more or less minor variations) in the canon by the third century a.d. were already chosen for a variety of theological, moral, and political as well as historical reasons.

From reductionism to schema theory

The language of mental phenomena often sets the stage for our study of the brain, and many neural phenomena can best be described in terms of their role within mental acts or the behaviors of organisms. Yet, we do not see neural levels of description as superseded by the mental. Rather, our aim is to develop a schema theory that makes contact with the phenomena of everyday experience but that can, where appropriate, be supplemented at the neural level of description.

All of the many kinds of mind/brain reductionism have in common a naturalistic basis: there are no mental events not explicable by brain science. It follows from this statement that the terms of mental language are all in principle describable in terms of brain language. As we shall see, however, it is possible that mind and brain science are not yet complete enough for the descriptions to be carried out in practice. Dualism, on the other hand (as exemplified by Eccles, Section 4.2), holds that mind involves some entity or substance not identifiable with anything in the material brain.

A multilevel view of schemas

In discussing scientific knowledge in Chapter 1, we saw that scientific theories are testable and mutable and that changes yielding increasing pragmatic success may be accompanied by radical changes in the ontology posited to underlie the observable phenomena. We have based our epistemology on a schema theory with a similar quality. Our schema assemblages are not the precise formal statements of scientific explanation and observation; but they do allow us to make sense of our world, to plan courses of action, and yet to change the schemas (not necessarily consciously) when the expectations they support fail to be met.

There is a paradox at the heart of this chapter – the paradox of the individual actor discovering that much personal individuality appears to be the playing out of social forces. In this chapter, we come to a transition. The last five chapters emphasize schemas as units of cognition within the head of the individual; the next four chapters turn from psychology to the sociology of knowledge. They gather insights from philosophy of language, from social anthropology, and from hermeneutics to develop an account of social schemas that act as a reality external to the individual members of a society. These chapters culminate in the presentation in Chapter 11 of the Biblical worldview as a “Great Schema” that locates human reality in a God reality transcending space and time.

The Construction of Reality develops an integrated perspective on human knowledge, extending ideas from cognitive science and philosophy of science to address fundamental questions concerning human action in the world and whether the space–time world exhausts all there is of reality. We seek to reconcile a theory of the individual's construction of reality through a network of schemas or mental representations with an account of the social construction of language, science, ideology, and religion. Along the way, we take account of much current research and debate in philosophy, linguistics, artificial intelligence, brain theory, cognitive science, evolutionary biology, social anthropology, history of religions, theology, and biblical and literary criticism. However, the reader will find no breathless pastiche here but a cumulative marshalling of evidence for a coherent and integrated view of the individual and social dimensions of human knowledge. We hope that it will stimulate the reader to find that within this integrated perspective there remains much scope for lively debate, particularly in our discussion of free will and of the reality of God.

For many people, even in today's secular world, God is the fundamental reality that gives meaning to human existence; for others, God does not exist, and whatever meaning human existence holds is to be found in society and in the more intimate groupings of family and friends.

Lord Gifford's brief

“I wish the lecturers to treat their subject as a strictly natural science, the greatest of all possible sciences, indeed, in one sense, the only science, that of Infinite Being, without reference to or reliance upon any supposed special exceptional or so-called miraculous revelation. I wish it considered just as astronomy or chemistry is.”

This is Lord Gifford's statement in the deed of foundation of his lectures in 1885. Much water has flowed under the bridge since then, and we can no longer take for granted either the character of “natural science” or of “revelation” in the sense in which these concepts were understood in Gifford's time. We therefore make no excuse (and believe it to be within the spirit of his foundation) that this book concentrates on the problem of what it is to be a “science,” and what kind of continuity, if any, exists between the knowledge of “nature,” of “persons,” or of “society,” and the possibility of knowledge of God.

We speak of “knowledge,” but our difficulty today in addressing Lord Gifford's brief is that the theory of knowledge (epistemology) has come to mean almost exclusively the methodology of the natural sciences and, more recently and belatedly, the social sciences, to the exclusion of any possibility of knowing extraspatiotemporal reality, if such can be said to exist. Our culture leads us to believe in a natural space–time reality that is explored and increasingly discovered to us in natural science.

Language as metaphor

In the previous chapter, we made some suggestions about a schema theory of language acquisition by the child, for whom language is initially an external reality to be mastered. We also noted that the external language is itself a social construction – a “collective representation,” as Durkheim called it. In this section, and more particularly in Chapter 10, we start to build some bridges between the essentially individualist approach and a more comprehensive picture of language as also embodying the constructions and classifications of a culture. In this area there are plenty of empirical studies on which to draw, deriving from literary criticism, social anthropology, and the history of ideas and of science, as well as Wittgensteinian philosophy. However, cognitive science is as yet an infant in these worlds, and we cannot claim to have more than hints for an adequate theory of such social schemas. In this section, we concentrate not on the empirical aspects of sociolinguistics but on the philosophical implications of our theory of language so far.

Our emphasis on the dynamics of meaning change and its holistic character already brings us into conflict with a long-entrenched philosophical tradition. Speaking of the suggestion that the meaning of words changes whenever our mental state changes, when, for example, we acquire more knowledge about the subject matter, Putnam (1981, p. 22n) says this “would not allow any words to ever have the same meaning, and would thus amount to an abandonment of the very notion of the word ‘meaning’.”

Natural language is an integral part of our lives. Language serves as the primary vehicle by which people communicate and record information. It has the potential for expressing an enormous range of ideas, and for conveying complex thoughts succinctly. Because it is so integral to our lives, however, we usually take its powers and influence for granted.

The aim of computational linguistics is, in a sense, to capture this power. By understanding language processes in procedural terms, we can give computer systems the ability to generate and interpret natural language. This would make it possible for computers to perform linguistic tasks (such as translation), process textual data (books, journals, newspapers), and make it much easier for people to access computer-stored data. A well-developed ability to handle language would have a profound impact on how computers are used.

The potential for natural language processing was recognized quite early in the development of computers, and work in computational linguistics – primarily for machine translation – began in the 1950s at a number of research centers. The rapid growth in the field, however, has taken place mostly since the late 1970s. A 1983 survey by the Association for Computational Linguistics (Evens and Karttunen 1983) listed 85 universities granting degrees in computational linguistics. A 1982 survey by the ACM (Association for Computing Machinery) Special Interest Group on Artificial Intelligence (Kaplan 1982) listed 59 university and research centers with projects in computational linguistics, and the number continues to grow.

The objectives of computational linguistics

Computational linguistics is the study of computer systems for understanding and generating natural language. In this volume we shall be particularly interested in the structure of such systems, and the design of algorithms for the various components of such systems.

Why should we be interested in such systems? Although the objectives of research in computational linguistics are widely varied, a primary motivation has always been the development of specific practical systems which involve natural language. Three classes of applications which have been central in the development of computational linguistics are

Up to now, we have restricted ourselves to determining the structure and meaning of individual sentences. Although we have used limited extrasentential information (for anaphora resolution), we have not examined the structure of entire texts. Yet the information conveyed by a text is clearly more than the sum of its parts – more than the meanings of its individual sentences. If a text tells a story, describes a procedure, or offers an argument, we must understand the connections between the component sentences in order to have fully understood the story. These connections are needed both per se (to answer questions about why an event occurred, for example) and to resolve ambiguities in the meanings of individual sentences. Discourse analysis is the study of these connections. Because these connections are usually implicit in the text, identifying them may be a difficult task.

As a simple example of the problems we face, consider the following brief description of a naval encounter:

The most evident linguistic problem we face is finding an antecedent for ‘it’. There are four candidates in the first sentence: ‘dawn’, ‘Valiant’, ‘Zwiebel’, and ‘torpedoes’. Semantic classification should enable us to exclude ‘dawn’ (*‘dawn sinks’), and number agreement will exclude ‘torpedoes’, but that still leaves us with two candidates: ‘the Valiant’ and ‘the Zwiebel’ (which are presumably both ships of some sort).