Thursday, 1 December 2005

Poetry and the Science of Mind

From The Philosopher, LXXXXIII No. 2 Winter 2005

With acknowledgements to the The Association of Teachers of Mathematics


By James Aitchison

Science has long been the dominant intellectual force in Western societies. Some fields of empirical enquiry, from the cosmos to the neurone, are necessarily scientific. In other fields, though, for example, concerned with the nature of our humanity and our relationship with the natural world, scientists also seem to have achieved more than have poets or philosophers or the social scientists who practise psychology and sociology. Perhaps, one reason for the dominance of science has been the willing submission to its ascendancy by academic disciplines that were once regarded as subject areas in the arts and humanities.

I. A. Richards, whose Principles of Literary Criticism and Practical Criticism influenced a generation of critics and students of poetry, put it:
... the theory of literary analysis is at an extremely interesting point in its development, on the point of making, through experiment, those contacts with actuality that would transform it into a science, and a science from which very important practical utilities may be expected to result. 
 Scientific experiment usually leads to empirical observations and measurable results, and scientists consider their findings verified when repetitions of the experiment lead to the same results. One could argue that the meaning of a poem and the truth of that meaning are verified when critics and other readers recognize the experience in the poem, but that kind of recognition varies from one reader to another. And the practical utility that could emerge from the close reading of a poem, a greater precision in the readers' use of language, perhaps, or in their ability to understand other writings, or an enhanced memory cannot be predicted like the outcome of a scientific experiment and cannot have the practical usefulness of applied science.

A striking instance of philosophy's surrender to science was the influential paper, 'Psychology In The Language Of Physics', published in 1931 by Rudolf Carnap, the German-American logical positivist. Carnap's thesis is that every sentence in psychology can be formulated in the language of physics. He states:
Once the language of physics, due to its nature as a universal language, has become the system language of science, all science will become physics. Metaphysics will be eliminated as meaningless.
If metaphysics is taken to be the study of the human condition in terms of being and selfhood, substance and essence and orders of reality (including mind, time and eternity, knowing and not knowing), then it is impossible that these things will ever be rendered meaningless. Even if they ceased to be taught in universities they would still be the subjects of thought and of poetry because the poet is drawn irresistibly to consider these things; indeed, the poet's mind is designed to consider them. Carnap continues:
If our thesis is correct, even the general laws of psychology, i.e., the psychological laws, can be translated into the language of physics. They are, therefore, physical laws. 
Carnap writes as if the disciplines of psychology and philosophy were sciences of mind, but the fact that a subject can be explored rigorously and meticulously does not make that subject a science. Even if all metaphysical concerns were ignored, psychology and philosophy could not fully meet the criteria of the physical sciences; for example, the criterion of independence from the phenomenon under investigation and the criterion of causality cannot be met if the phenomenon is the working of the human brain or the mind. There are many theories of mind: behaviourism, centralism, connectionism, determinism, dualism in its various forms, functionalism, functional isomorphism, identity theory, interactionism, materialism, monism, anomalous monism, peripheralism, physicalism, pragmatism, reductionism and others. Some of these are contradictory, and if scholars disagree about the very nature of their subject, they cannot claim that the subject is a science.

Carnap influenced A. J. Ayer, whose book, Language, Truth And Logic makes several favourable references to Carnap. The title of Ayer's first chapter, 'The Elimination Of Metaphysics', is similar to Carnap's claim that metaphysics will be eliminated as meaningless. And, in the second chapter, Ayer again echoes Carnap:
There is no field of experience which cannot, in principle, be brought under some form of scientific law, and no type of speculative knowledge about the world which it is, in principle, beyond the power of science to give. 
 This kind of reductive argument is at least as old as Plato, who states in the Republic:
... the part [of the mind] which relies on measurement and calculation must be the best part of us, and the part which contradicts them an inferior one. 
Plato is discussing the mind of the poet and other creative artists, who do not rely on measurement and calculation. Some proponents of behaviourism and determinism, however, such as the American psychologist, B. F. Skinner, have argued that mind does not exist and that the brain is essentially an 'input-output device'. Gilbert Ryle, in The Concept Of Mind, even writes:
The phrase 'in the mind' can and should always be dispensed with. Its use habituates its employers to the view that minds are queer 'places', the occupants of which are special-status phantasms. 
 Ryle uses the words, 'queer' and 'phantasms', as ridicule to reinforce his argument. His little case study in the chapter, 'The Intellect', illustrates the extent to which reductive philosophy has been overtaken by neuroscience. Ryle dismisses the life of the mind as the impenetrable shadow-life and as the ghost in the machine; readers of poetry know that their enjoyment and understanding of the poem involve their imaginative penetration of the poet's mind, and the poet's penetration of theirs.

Nowadays psychologists and philosophers are unlikely to be so dismissive of mind. They may regard it as an emergent system from its neurophysiological substrate, the brain, and they accept that their subject areas cannot meet all the criteria of the physical sciences. Psychologists, philosophers and other social scientists now accept that, even if their findings can be formulated as laws, these laws cannot have the status of the laws of physics. But what is a law of physics?

The law is not an integral part of the physical phenomenon under investigation; the law is formulated by the physicist. The law is expressed in words or symbols, or expressed iconically in diagrams and models; but the written law is an abstract, arbitrary representation of physical phenomena, and the iconicities of diagrams and models, some of which are said to represent invisible phenomena, are as much expressions of physicists' modes of thought and feeling, and their conformity to the conventions of physics, as they are expressions of physical reality. In Physics And The Mind, Roger Penrose writes:
Why is it that the physical world seems to obey mathematical laws in such an extremely precise way? Not only that, but the mathematics which seems to be in control of our physical world is exceptionally fruitful and powerful, simply as mathematics. 
The mathematician invented mathematics as the poet invented poetry, in response not only to features of the natural world but also to the workings of the mind. For centuries poets claimed that their responses, poems, were imitations of the natural world, but few poets made the mistake of projecting what they claimed to be imitations onto the world and then claiming that the world writes poetry.
Mathematics and physics, like poetry, derive from innate functions of mind that are expressed in abstract, symbolic, conventionalized codes. The natural world engages the imagination of the mathematician and the physicist as it does the imagination of the poet, but that world does not write poems or laws of mathematics and physics. 
When Penrose adds that one has the feeling that the mathematics needed to describe these things [natural phenomena] is out there, there is no doubting the sincerity of his feeling. Instead, one suggests that when Penrose finds exceptionally fruitful and powerful mathematics out there, it is because his mind put it there. Laws of mathematics are mathematicians' conceptual representations of physical phenomena; the laws are also mathematicians' representation of abstractions that they themselves create. Whatever truth there might be in the laws is conceptual truth, which is not the same as the physical realities the laws claim to represent.

If the laws of mathematics and science are seen in these terms, then the truth of mathematics and science is no more valid than the truth of poetry. Today, as in the past, scientists proceed by experimentation and investigation; poets, too, investigate experience and experiment with language, testing the value of every word in every poem they write. And scientists today as in the past also proceed like poets: by speculation, intuition, flashes of insight, dreams, and chance. The poet and the scientist have these means of discovery in common because they share the same functions of mind. And the mind, some scientists claim, functions like a computer.

Scientists are drawn to propose theories and models of brain and mind, and it seems inevitable that, in innocent hubris, scientists should choose one of their own inventions, the computer, as the current model. Earlier models were the clock, the railway network, and the fixed-line telephone exchange; a more recent model, discussed with no sense of its absurdity by Stephen Mithen in The Prehistory Of Mind is the Swiss Army Knife. Whatever single or composite model is eventually agreed, the evidence already available shows that the brain and its emergent system, the mind, are designed to process large volumes of different kinds of information, sometimes instantaneously, sometimes sequentially and sometimes retrospectively. The evidence also shows that, at different stages in life and in different states of mind at any of these stages, an individual will process the same kind of information in different ways; that is, brain and mind are designed to collate and interpret masses of information every day, but the methods of collating and interpreting can vary.

Writing poems is not part of that design, because a writing system is not an innate, natural faculty but an invention. For some poets the process of composition is one of inspiration; electrochemical activity in the brain is transduced as if spontaneously into language in the mind. But most poets can use the brain's language-processing systems only serially and slowly, and they feel frustration on those occasions when they form images, ideas, and words so much faster than they can write them down, that some words do not have time to enter the short-term or working memory but are forgotten between the writing of one word or phrase and another.

Writers might also feel frustrated when they find that their finished poems fail to express the multiplicity and intensity of sensations experienced during the process of writing. Most poets would agree that the finished poem is a compromise. Would a poetry-writing computer also agree? Such a computer, or any computer, would have a more efficient memory than the poet, whose memory is designed not only for accurate recall but also for misremembering, forgetting and confabulating when designated neural networks decay or evolve or combine. A computer cannot be programmed to work in these ways because no one knows enough about the functioning of mind and memory.

Poets know that it is possible to design a computer than can produce and interpret language in ways that are faster, more consistent, and extend across more semantic fields than they can. They know that a computer can be programmed to convert speech into writing or writing into speech, and to use its artificial language systems in order to translate one natural human language into another. Poets know that a computer could be programmed to operate with the kinds of randomness and unpredictability that are similar to those of the mind, and that it might be programmed to simulate fantasy and dreams, intuition and insight, and perhaps to write poems. But the poet cannot believe that a computer can create linguistic subtleties such as irony, satire, paradox and metaphor, or elliptical patterns of syntax and thought, or the kinds of creative ambiguity one finds in the dialogues and monologues of Shakespeare. And if a computer programmer has a concealed reason for compiling a particular programme, for example, to identify a colleague engaged in industrial or political espionage, would the computer be able to identify the programmer's ulterior motive? 
Poets note that the computer is a recent human invention, whereas the mind has been evolving for hundreds of thousands of years; they note that the mind is as old as consciousness and the brain is older by far than the mind, and they feel that it is impossible to devise a computer program that takes full account of these evolutionary time-spans.

The computer imitates the brain to some extent and the mind to a lesser extent; but the computer can never be a human brain or mind and thus can never replace humans, unless some day the brain and the mind are fully understood and the understanding fully encodable in the computer. Such total understanding, the poet believes, is impossible to achieve. Computer scientists will never devise a program that takes account of all the faculties and functions of mind, because we shall never fully understand the nature of mind. We shall never discover when and how and why consciousness and language evolved, because they emerged from non-consciousness; equally, we cannot know the relationship between consciousness and non-consciousness, or the extent of the mind's non-consciousness. (Whatever discoveries neuroscientists make in the future, we shall never fully understand the mind unless there are future discoveries, along with a more highly developed faculty of meta-consciousness, that is, our awareness of being aware.) But at present we know of no mental power that can take account of all faculties of mind, including itself, the accounting and observing power.

Poets do not wish - and do not believe that scientists wish - their minds and brains to be the same as a computers, except in the strictly limited sense in which people learning how to operate a new computer program will have to change some of their neuronal networks in order to meet the requirements of the new program; but every learning process requires us to change our neuronal networks in some way in order to accommodate the new information. The poet might feel that a great moral dilemma underlies the analogy of computer and mind: if there were an equivalence of human and machine, would the humans be treated as machines or the machines as humans? In fact, the acceptance by some scientists of the analogy of computer and mind is a measure of the dominance of science in our culture. 

One of the most effective philosophical rejections of the analogy comes in the book What Computers Still Can't Do: A Critique Of Artificial Reason, by Hubert Dreyfus:
To have a complete theory of what speakers are able to do, one must not only have grammatical and semantic rules but further rules which would enable a person or a machine to recognize the context in which the rules must be applied. Thus there must be rules for recognizing the situation, the intention of the speakers, and so forth. There are grammatical and semantic rules governing written standard English, and there are sets of rules, sometimes implicit rather than explicit, for recognizing the context and situation in which language is used; for example, courts of law, morning assemblies in schools, television chat shows, parliamentary debates. There can also be sub-sets of language rules; in one and the same newspaper there are different rules for the use of language in news reports, obituaries, book reviews and crossword puzzles. Few members of a language community give conscious thought to these rules or even know of their existence, but most members of a language community have some level of understanding of them.
Dreyfus states that a complete theory requires rules governing the intentions of speakers, but he knows that there can be no rules for such infinitely variable and unpredictable factors. He concludes:
But if the theory then requires further rules in order to explain how these rules are applied, as the pure intellectualist viewpoint would suggest, we are in an infinite regress. 
A person can break the normal rules of language and yet be understood by listeners or readers, who do not need a fixed set of rules in order to understand abnormal uses of language - indeed, no such rules exist. If computers are to do all that humans can do, then computers must be able to understand abnormal uses of language; but they can understand only what their designers have told them about.

No comments:

Post a Comment