Customer Reviews

Minds, Brains and Science (1984 Reith Lectures)

5 star:		(4)
4 star:		(3)
3 star:		(0)
2 star:		(1)
1 star:		(0)

 

 

Searle is an interesting philosopher for me to read, because I was trained in neurobiology, and Searle is a philosopher who thinks like a neurobiologist. On the other hand, I am a neurobiologist who thinks like a philosopher.

Although the book discusses several classical problems such as the problem of freedom and free will, the mind-body problem, right and wrong, etc., for me the two most interesting chapters were the one on the mind-body problem, and the one on cognitive psychology.

Here Searle proposes a thorough-going biological and physical explanation that, as a neurobiolgist, I've always liked myself.

You really need to read these two chapters to understand all the details, of course, but I'll briefly summarize his idea, and you can decide if it makes sense to you.

Basically, Searle says there really is no mind-body problem. This dichotomy occured because philosophy completely misunderstood the entire issue. There is no mind-body problem, because the mind depends on the brain, and on the neural workings of the brain, and there is no reason even to say that consciousness itself is separate from the brain itself.

Searle points out that we explain the properties of normal matter, such as a steel ball, which has mass, weight, is impenetrable, is magnetic, and so on, by reference to its atomic and molecular properties. There is no reason to posit any intevening layer of "rules" or theory.

It's the same with the mind-body problem. Mind depends on neurons. All our behavior depends on neurons. There is no reason to posit this intermediate entity of consciousness or of mind which is separate from the underlying biology. There is no doubt that consciousness exists, but there's nothing special about it, and although Searle doesn't claim it can be reduced to neural functions yet, he leaves no doubt that classical views about the mind and consciousness are fundamentally flawed.

Anyway, I can certainly sympathize with this point of view, and would like to make a point of my own. I've studied the brain, and when you see people with tiny, focal, strokes in the language area of the brain who have no detectable impairment except they can no longer use articles or conjunctions in their speech, or people with temporal lobe damage who can easily name an object when you show it to them, but who can't tell you its function, and vice versa, where there are people who have temporal lobe damage in an adjacent area with exactly the reverse syndrome--they can tell you what its for but can't name the object--in other words, the naming function and the definition function seem to be separate in the temporal lobe, and the two areas must communicate in order to be able to do both, or at least the information is stored separately and you need access to both--you very quickly get the idea that if it's not in the brain, it's not anywhere. There are legions of other neurological cases where people have lost very specific or general functions depending on the source and extent of the damage to the brain.

Furthermore, it's becoming clearer as a result of research that there is no single part of the brain that gives rise to consciousness. Consciousness relates to different functions located in different parts of the brain being integrated in time through a finely controlled and switched system of neural communications pathways. Thus, consciousness is not a unitary entity at all, although it might seem so to our own introspective minds. More accurately, it is a unified process that occurs through the integration of diverse brain areas and brain functions.

Anyway, Searle's biological reductionism and determism isn't very different from how neuroscientists think, and I give him credit for being willing to discuss the subject in those terms and propose such a radical solution (from the standpoint of most philosophers) to the mind-body problem.

Dealing with some of the problems in philosophy that persist, even in our "post-modern" times, this book by John Searle of the U.C. at Berkely provides a quick, easily read survey of some of the issues about minds, bodies and artificial intelligence that are of special relevance today. Searle is especially keen to restore a commonsense view of things and so his philosophy seems particularly down-to-earth with regard to some of the knottier problems.

His notion that consciousness (the stuff of minds) is to brains as digestion is to the stomach (a function of it) and that there are various orders of explanation that can be invoked for the same phenomenon go a long way toward enabling those who are stuck in the mind-body conundrum to get beyond it. In some ways he offers an updating of Wittgenstein who, similarly, offered a way of getting beyond such "problems" though Wittgenstein reduced it all to a matter of how we talk while Searle wants to say that this only answers the question in part. Unlike Wittgenstein, who dismissed the idea of theoretical explanations superceding ordinary language, Searle wants to reaffirm the importance of such explanations, and to offer a way to develop them. In many ways his proposals make quite a bit of sense.

However, I remain struck by his argument against the possibility of what he terms the claims of "strong artificial intelligence" proponents. He describes this view (page 28) as "saying that the mind is to the brain as the program is to the computer hardware" and elaborates by noting that "on this view, any physical system . . . that had the right program with the right inputs and outputs would have a mind in exactly the same sense that you and I have minds." Thus, "strong AI," as he repeatedly terms it, is the view that minds are in no way unique to creatures like us (with organic brains like ours) but are merely the function of the right sort of programs on the right kind of hardware operating in the right sort of way. That is, inorganic machines, like digital computers, can be made to have minds like ours (with the same kinds of features ours have).

Searle is very much opposed to this view and in this book brings to bear his most famous argument against it: the Chinese Room thought experiment. In a nutshell it holds that a computer, insofar as it is no different from someone inside a sealed room following purely formal rules in responding to written questions submitted in Chinese (a language he doesn't know) who appears to be responding AS THOUGH HE KNEW CHINESE, so too, can that computer appear to have intelligence, to understand its inputs, without really doing so.

His core argument against "strong AI" is that this is ALL a computer can do, but that simulation in this way is not what we mean by intelligence at all. In fact, he rightly notes, when we think of intelligence, we think of understanding, what he variously calls intentionality and/or meaning (semantic content). But the computer modelled on his Chinese Room evidences no understanding but only a rote process of mechanically producing certain squiggles, according to some pre-established rules that look like they reflect an understanding.

The crux of this is a syllogistic argument he also reproduces in this book:

1) Brains cause minds.
2) Syntax is not sufficient for semantics.
3) Computer programs are entirely defined by their formal, or syntactical, structure.
4) Minds have mental contents; specifically they have semantic contents.

On the basis of the above he offers a number of conclusions but his first is telling: "No computer program by itself is sufficient to give a computer a mind. Programs in short are not minds, and they are not sufficient by themselves for having minds."

While thinking highly of Searle's work, I believe his Chinese Room argument is deeply flawed. I can't offer a full account here for reasons of space, but the point is that, while he has correctly noted that the kind of intelligence we have (and that we might be interested in recreating on a computer) is a conscious intelligence, his argument that computers can never offer a medium for recreating this hinges on a mistake. He says in point #2 that "Syntax is not sufficient for semantics." He amplifies this by noting that this is "a conceptual truth," i.e., it is true by definition. "It just articulates our distinction between what is purely formal and what has content," he adds. But the fact that the two notions are conceptually different does not entail a claim that one type of thing cannot give rise to another.

Searle wants to say that computers can never provide an adequate platform for the phenomena of consciousness on this basis but his argument doesn't demonstrate this. In fact, despite all his efforts with the Chinese Room thought experiment, one can still envision a complex of computer-driven programs that replicate all the various functionalities we find in our own minds which, if combined in the layered and integrated way they are in humans, can presumably yield the kind of consciousness we have. Of course, this is not to say this CAN be done technically, only that Searle's argument that it can't be founders on the meaning of the term "sufficient" that he uses in the second step of his argument. He wants us to accept that though step #2 arises from a purely conceptual claim, it can be construed to have an empirical implication. In this I think he errs.

All of this said, however, I want to add that this is a very fine book. It is concise, clear, and profound in its thinking. Searle is certainly right in his recognition about what "intelligence" is, as far as he takes it, which may be about as far as a philosopher can. The rest is up to the scientific workers in the field, those who are experimenting with and designing new ways to operationalize minds in machines.

SWM

How can we reconcile our perspective as "conscious, free, mindful, rational" creatures in an impersonal universe composed entirely of "mindless, meaningless" physical particles? How, indeed, can we suppose that these ineffable mental states which compose consiousness can affect material things and yet elude physical description? It's easy enough to imagine a meaningless universe of particle soup. But adding consiousness to the mix often seems like the dreaded "nomological dangler" - what Ockham's Razor attempts to sever from a complicated theory.

John Searle attempts to solve these questions and more in this intruiging book for the non-technician such as myself. I found his ideas concerning the consciousness problem intriguing, lucid and very well formulated. These constituted the first lecture and helped him take on computer-affectionate behaviorists and cognitive scientists in the second and third.

In the fourth and fifth lectures, Searle reflects on the nature of action and the difficulties inherent in the social sciences. These build up to his last lecture where he confronts the freedom of will problem. Unfortunately, he makes this problem devastatingly clear without presenting a solution to my satisfaction. The nature of action forces one to believe that we are not walking somnabulists or marionettes. But the physical sciences haunts us yet with the prospect of determinism (or indeterminism, which is equally devastating for free will). The author ends the lecture with a ponderous question mark. At least he's honest!

I recently attended a lecture by a university professor on 'Philosophy of the Mind'. Thinking back to these BBC Reith Lectures which I heard in 1984 (on radio, not TV, by the way) I asked him about models for the brain other than the computer, such as water mills. He seemed blissfully ignorant that there had been any. Reading this book shortly afterwards, I was reminded that not only mills but also hydraulic engines, switchboards, and telegraph systems have, at various times, been used as models for the brain. This should give caution to anyone who thinks it's obvious that the brain is a computer.

The text thankfully retains the combination of conversational style and intellectual depth of the original lectures. It's illuminated by examples such as the artificial intelligence researcher (John McCarthy) who believes a machines as simple as thermostats can have beliefs ("it's too hot in here"), which is both unforgettably absurd and seductively radical.

An obvious problem in this field is that people are disposed to declaim at each other rather than listening. Searle and Dennett are still going hammer and tongs over the former's 'Chinese Room'. I can't help thinking that if the parties really wanted to settle these disputes they'd try to agree on definitions for their terms. Admittedly this may not always be easy - - for example 'understanding' is a strange word if you think about its etymology - - but would surely be more constructive than abusive cross-talk.

The book is impressively wide-ranging in its 100 pages and includes, for example, the clearest exposition of the arguments for and against free will which I've ever read.

There's some occasional slackness, eg what exactly are the 'powers' by which we are supposed to judge any machine against brains, before ascribing mental states to the former? But, line for line, it's by far the most valuable book I've read in the field.

"Minds, Brains and Science" collects John Searle's 1984 Reith lectures. These lectures focus on some of the essential issues in the philosophy of mind and related areas, such as artificial intelligence.

Searle plunges directly into the classical mind-body `problem'. Searle suggests four features of mental phenomena that have traditionally made them difficult to integrate with our physical conception of the universe. These four features consist of: consciousness, intentionality (intentionality in the technical, philosophical sense whereby mental states are said to be representational, to be 'about' things), subjectivity and mental causation. Searle believes that the supposed mind-body problem disappears once we realize that all mental phenomena are caused by neurobiological processes. A special, biological sort of causation is put forth in which mental phenomena are both caused by brain processes at the neuronal level and simultaneously realized in that same system. This is similar to the way in which the liquidity of water (a surface or macro-level feature) is both caused by and realized in the system that is made of hydrogen and oxygen micro-elements. Thus, "there is a cause and effect relationship, but at the same time the surface features are just higher level features of the very system whose behavior at the micro-level causes those features." This is different from the straightforward notion of cause and effect (e.g., billiard balls clacking into one another). It is the adoption of the `billiard-ball' model of causation that leads to an apparent dualism between mental and physical phenomena.

Searle follows up this discussion with an attack on the strong A.I. position. According to the strong A.I. view, the brain is not LIKE a computer, the brain IS a kind of computer, in the sense that it takes in a given set of inputs, performs rule-governed operations on the input and produces meaningful output. Brains and minds stand in the same relation to each other, as computer hardware does to its programs. For Searle, this computational theory of mind is deeply flawed. He uses his famous thought experiment (the Chinese room) to prove his point. A thought experiment (or gedanken) is a way of imagining a hypothetical scenario in order to investigate the nature of things. Thought experiments have a history not only in philosophy but also in the physical sciences (e.g., Einstein's street-car approaching the speed of light, the `elevator thought experiment' showing the ways in which gravity and acceleration are isomorphic, the leaning tower of Pisa and Galileo's law of falling, etc.).

What Searle deduces from his Chinese room thought experiment is that brains are made of the right kind of `stuff' to cause intentional, semantic states whereas computer programs are solely formal or syntactical in structure. Thus, for Searle, it would be a gross error to say that when a computer runs a language comprehension program, it understands in any real sense of the term (simulation does not equal duplication). The computer is a machine that manipulates formal symbols; it does not have semantic states. For Searle, the computer is at best a metaphor for the way in which the brain works, a metaphor that is no more `valid' than previous ones (telephone switchboard, steam engine, etc.). The brain is first and foremost a biological organ.

The Chinese room argument has been much maligned in some circles (Dennett coined the term `intuition pump' in referring to it), but it is such a well-known thought experiment that everyone interested in philosophy of mind and cognitive science should have some level of familiarity with it. Searle uses it as a foundation to mount a more large-scale critique of the cognitivist/computational position.

Searle's other lectures discuss the reasons why the psychological/behavioral and social sciences have not enjoyed the same astounding successes as the strictly physical sciences such as physics and chemistry. He concludes with an insightful lecture on the problem of free will. Our scientific conception of the physical world is seemingly incompatible with our commonsense notion that we are conscious, freely acting agents. The obsolescence of the Newtonian clockwork universe does not alter this deep incompatibility says Searle. Quantum mechanics and particle indeterminacy are not enough to safeguard free will, because (i) indeterminacy at the particle level does not translate into indeterminacy at the macro-level and (ii) "it doesn't follow from the fact that particles are only statistically determined that the human mind can force the statistically-determined particles to swerve from their paths." Searle's conclusion is that our notion that we have free will is a kind of illusion, built into our brains by evolution. In principle, an `ideal observer' (of the sort postulated by Laplace) could predict all of our actions, but for all practical purposes we behave as if we had free will. Perhaps, as Steven Pinker once stated, "free will is the idealization of human beings that makes the ethics game playable."

Overall, "Minds, Brains and Science" is a rather enjoyable read, despite some of the weaknesses in Searle's argument. It is short and accessible and worth reading for its many ideas.

This is the kind of book that you can finish in a day and makes you think about things like whether computers will think or not. I personally found it interesting and a good introduction to things like cognitive science.

john searle is a great addition to the studies of philosophy because he adds a modern perspective which is easily identifiable and agreeable. Particularly his room experiment and discusions on consciousness which compliments nicely with other philosohphy topics. The author puts a nice modern spin on many concepts and does so in an entertaining way.

This book is written in what I call "No. 2" language - abstruse, dense, and unclear. I've been assigned much better textbooks in my college philosophy courses, for instance "Questions that Matter: An Invitation to Philosophy" by Ed L. Miller.