Monday, May 31, 2010

What it is Like to be a Bat in Philosophy's Belfry

As you might suspect from my somewhat juvenile humor in titling this essay, I shall address and, to my satisfaction, rebut the core arguments of Thomas Nagel's essay, "What is it like to be a bat?" (available many places online should that link fail). My humor should in no way be construed as disrespect; Nagel's essay has now stood for decades as the shield in the van of those who resist the notion that the human mind is reducible to purely physical operations. It is a subtle and well crafted essay, and an absolute must for anyone seriously interested in philosophy or science of the mind.

I published the substance of this repudiation in a brief post on the-brights.net forum. Here I shall elaborate a bit more.

Nagel's paper has been endlessly discussed and naturally I am not familiar with all such discussion. It is quite possible that others have presented my core argument but that it has not caught my attention. Please direct me to any such should you know of them.

The subject question is succinctly stated in the fourth paragraph:


... Any reductionist program has to be based on an analysis of what is to be reduced. If the analysis leaves something out, the problem will be falsely posed. It is useless to base the defense of materialism on any analysis of mental phenomena that fails to deal explicitly with their subjective character. ...


In other words any physicalist explanation of human cognition must account for our phenomenology, how it is generated, and what it accomplishes in terms of the other activities of the mind reduced to brain functions. The question then is whether such is possible. Is there, or even can there be a strongly plausible preferably testable explanation for our common if unsharable experiences of what we see, hear and feel "inside our heads" are generated by the biology and chemistry which we do know, or can plausibly come to know, comprise the physical mechanism of our brains?

Nagel's answer is "no". To his credit he does not claim his title query and exposition as a positive proof. But his paper presents a potent argument that there must be an answer to it before physical reductionism can be taken seriously.

Despite positing that there may one day be a science which can answer questions of phenomenology, Nagel quite clearly believes that nothing resembling contemporary science can address his question. From the opening:


Without consciousness the mind-body problem would be much less interesting. With consciousness it seems hopeless. The most important and characteristic feature of conscious mental phenomena is very poorly understood. Most reductionist theories do not even try to explain it. And careful examination will show that no currently available concept of reduction is applicable to it. Perhaps a new theoretical form can be devised for the purpose, but such a solution, if it exists, lies in the distant intellectual future


And otherwise throughout the paper there are telltales that Nagel believes that physical explanations can never account for the phenomenal. In 1974 there was considerably more justification for such pessimism than can be rationalized today. We have not only models, a la Dennett, but cognitive psychology mapping interactions between the perceptual and phenomenal, and neurophysiologists tracking the generation of phenomenal experience in living brains with probes, high resolution fMRI and other techniques.

But it is Nagel's "what is it like" question and his powerful justification for it which raises his paper above a mere question begging blank assertion that the physical cannot account for the subjective. Even with the advances of science mentioned above, Nagel cannot be ignored. Even Dennett's attention to "What is it like ..." in "Consciousness Explained" seems to have misapprehended the motivation of Nagel's question.

Nagel opens the substance of his salvo by demonstrating that, meaningfully and objectively, "there is something that it is like to be a bat", just as every sentient person knows intuitively that there is something it is like to be him- or herself. Even accounting for the large differences in intelligence, any creature which exhibits behavior more sophisticated than autonomic reaction to stimulus has some "experience" ... it adapts to changes in its environment in complex ways.

Now his thesis that "what is it like" applies and must be answered for any sufficiently sophisticated animal and each human individually, but there is ample reason why Nagel chose a bat as his illustrative example. It is mammalian and sufficiently complex that we feel some sympathy - we are intrigued. But at the same time it is an animal profoundly unlike us, a nocturnal flier, but most significantly it possesses a sensory apparatus - echolocation - completely unlike our own senses.

Nagel has some fun demonstrating that no matter how sympathetic we are any attempt by us to imagine what its like to be a bat is doomed to failure. At best we can imagine what it is like for us pretending to try to be a bat. There is simply no way that we, equipped with human senses, can meaningfully correlate our own experiences with those of a bat. Of course the same is true for any creature with different senses or significantly different lifestyle and correspondingly different motivations. And while Nagel doesn't stress it, at bottom it also applies to our fellow humans. No matter how much we commiserate we can never truly have the what is it like experience of someone else; at best we can learn what it is like for us to experience very similar events and circumstances.

But further on comes the hammer blow. The "what is it like" question must be answerable for reductionism to have merit. Reductionism means equating the thing being reduced to an assemblage of simpler, better understood things and their interactions. The key is "equate". An equality. An "is". If there meaningfully and objectively is a "something it is like", then to be reducible there must be a what that something is like, preferably more immediately physical. But the question "what is it like" must be answerable.


As Nagel puts it:


But I believe it is precisely this apparent clarity of the word 'is' that is deceptive. Usually, when we are told that X is Y we know how it is supposed to be true, but that depends on a conceptual or theoretical background and is not conveyed by the 'is' alone. We know how both "X" and "Y " refer, and the kinds of things to which they refer, and we have a rough idea how the two referential paths might converge on a single thing, be it an object, a person, a process, an event or whatever. But when the two terms of the identification are very disparate it may not be so clear how it could be true. We may not have even a rough idea of how the two referential paths could converge, or what kind of things they might converge on, and a theoretical framework may have to be supplied to enable us to understand this. Without the framework, an air of mysticism surrounds the identification.


So the profound impact of Nagel's paper lies not in the impossibility of answering the question "What is it like to be a bat" in terms of human experience, but in the well justified demand that reductionist physicalism must be able to demonstrate that there is an answer, preferably one where the response is notably closer to the physical.

Is there an answer? Yes. And it is both trivial and informative.

While I have not yet here expounded on my decorations of Dennett's model, that answer is an elementary exercise in mechanics within it. But we needn't go that far to answer Nagel's quandary, and doing so will illuminate some aspects of how consciousness works for later discussion.

The single most interesting word in Nagel's title question is "like". All the other words are simple in context. Even "be" is elementary in context and we can ignore its ontological perplexities though peripherally pertinent. But what is "like"? What does it mean?

"Like" means to abstract the similar properties of two entities. It implies a comparison. "Something it is like" then asserts that for a being of minimal neurological sophistication, i. e. capable of having experiences, its contemporary instant experience has properties similar to something else. Nagel's gauntlet is the demand that we physicalists find out what that other something is.

So what can experiences be compared to? There is only one thing an experience can be compared to: other experiences. And since in the contemporary experience in question occurs privately in the head of a mute animal without means of exporting the properties of its experience, there is only one source of other experience with which to compare its contemporary experiences: its memory.

Here then is the answer to Nagel's title question: what is it like to be a bat? The answer is: it is like the memory of having been a bat.

The mysterians and willful obscurantists will undoubtedly regard that answer as inadequate. But it does fulfill the requirements of Nagel's paper. There is an answer. The X and Y of the reductionist formula have been met, there are two things different and distinct, immediacy and memory, intersecting in the "like" and resolving the "what" or "something" of the title and its source reasoning.

And it tells us something which is not intuitively obvious at first glance. It powerfully infers that memory is a requisite of whatever mentality we would classify as having experience.

To fortify against the first round of objections a few things should be noted. It implies and asserts that newborns will have little of consciousness until they accumulate experience. It predicts that there will be a positive correlation between the time or rate of accumulation of relevant experiences versus the sophistication of the adult in its niche.

Likewise it needs to be noted that memory is much closer to being understood at the purely physical level. Edelman won his Nobel for exactly that work. There are volumes of scientific detail and experimental validation for our understandings of how memory is laid down in living brains. And conceptually, the idea of memory is well understood; the computer you are reading this on contains billions of bytes of memory. We understand memory conceptually and are well on our way to understanding it as chemistry in cells and brains.

The only question then is how does that memory capture past experience and compare them to contemporary experience. But that of course relies upon how the subjective contemporaries are generated. We physicalists assert that it is simply chemistry in and among the cells of the brain, captured for memory by the processes Edelman and others have illuminated. The mysterians, of course will dispute that.

But the title challenge of Nagel's essay has been met. The catcalls of naysayers are simply the statement that the mechanics of phenomenology have not been explained - which is largely true in terms of scientific rigor, not true conceptually for Dennett's followers. But having supplied the missing solution to the philosopher's inquiry, such quibbles are themselves reduced to the question begging that Nagel so brilliantly avoided.

-- TWZ

Thursday, May 27, 2010

Some of My Reading

... from the top of my head, books I've read that directly concern the intent and topic of this blog.

Consciousness Explained --- Daniel C. Dennet
The Mind's I --- Daniel C. Dennett and Douglas Hofstadter.
Godel, Escher, Bach, the Eternal Golden Braid --- Douglas Hofstadter
The Emotion Machine --- Marvin Minsky
The Stuff of Thought, Language as a Window into Human Nature --- Steven Pinker
How the Mind Works --- Steven Pinker
The Ego Tunnel --- Thomas Metzinger
Wider Than the Sky --- Gerald Edelman
Second Nature, Brain Science and Human Knowledge -- Gerald Edelman

-- TWZ

Welcome to My Ruminations on Consciousness

Welcome.

I am creating this blog as a locus for posts and discussions (I hope) about consciousness and how the mind works.

I am nobody ... a college dropout who has been a computer programmer doing business applications for the past 25 years. But I have a deep and agile mind, the programmer's infamous "puzzle solver" mindset, and a long history of reading about science as well as "doing" technology ... and I don't think consciousness is really a problem.

In regards to consciousness, I am puzzled by why so many people still regard it as a puzzle.

While my reading of philosophy and questions of mind had before been shallow, a few years ago I read Daniel Dennett's "Consciousness Explained". It went click. It was lacking in much detail ... which I think can be filled in, but the overall picture seemed correct.

Since then I've read much more on the philosophy and science of mind, and with some elaboration on Dennett's model from years of argument, reading, and contemplating upon it, I think I can answer most of the outstanding questions that the philosophers prattle about regarding mind and consciousness and can rebut most criticisms of Dennett's model, or retort by my elaborations. To my satisfaction, anyway. I think all the mystery is gone. All that's left is work - a lot of it, which sadly I am in no position to undertake. But I can blog.

Which doesn't imply that I think we can build a human level conscious machine today. Maybe not tomorrow. But it won't be long. Of course my understanding is that a machine with consciousness is a trivial problem. Consciousness is not the problem - it is a very shallow phenomenon as it must be having arisen only lately in evolutionary time. The real problem of human level consciousness is the question of information synthesis, organization, representation, and retrieval interacting with our motivations as well as consciousness.

But solve the problem of representing information across multiple levels of scope, abstraction and multiple aspects and creating it from raw sensory information of the world, and then building human-like consciousness will be little problem. Of course consciousness does play a part in that, as I hope to elaborate in coming posts, but it is only a part of that problem. Understanding consciousness does not crack the AI problem; it is only part of it and a small part at that.

As said, I am nobody, and undoubtedly some of the results of my ruminations on consciousness will be wrong. But I hope that discussing my understanding may stimulate those who, now or in the future, may be in a position to do something about defining and hopefully settling the philosophical issues with scientific rigor.


----

The name of this blog, "Machine in the Shell", is of course a take off to the beautiful classic anime "Ghost in the Shell". That name derives from Gilbert Ryle's (Dennett's mentor) scornful description of dualism, "ghost in the machine", which Arthur Koestler appropriated for his book. GitS posits a subtle form of substance dualism. that individual consciousness or "soul" is an ultra-complex non-reproducible wave form. The name of this blog, by contraries, is the assertion that consciousness is simply a portmanteau for a body of matter properly organized to carry out various information processing functions in relation to its environment. The unremarkably mundane conception of a machine and its shell is intended as a stark statement of how shallow I think consciousness is.

----

ColonelZen is simply a handle for Terrence W. Zellers. Many years ago I was half watching a football game, half reading an article on the internet called something like "The Zen of Kernel Maintenance" and half doing web things (I'm good at math too!). For some reason yahoo would not let me create an account with my normal handle, tzellers, but there was an ad on the football game. For KFC. And ColonelZen was born. As partial justification, while I've never served in the military, I have indulged a bit of zen practice on occasion.


Hopefully, before long, I will start posting some meat on what I think is going on with consciousness, and perhaps others will join in and we can have fun arguing various ideas about phenomenalism, qualia, the Chinese room, mysterianism, zombies, property dualism and other arcana.

-- TWZ