Is complete knowledge an oxymoron?

[Jim__]

This is a follow-up column by WILLIAM EGGINTON on another column that was discussed in the R/T Forum: here. In this follow-up - a column mostly concerned with questions of the possibility of a deterministic universe - Eggington states that completed knowledge is an oxymoron:

Knowledge can never be complete. This is the case not merely because there will always be something more to know; rather, it is so because completed knowledge is oxymoronic, self-defeating. AI theorists have long dreamed of what Daniel Dennett once called heterophenomenology, the idea that, with an accurate-enough understanding of the human brain my description of another person’s experience could become indiscernible from that experience itself. My point it not merely that heterophenomenology is impossible from a technological perspective or undesirable from an ethical perspective; rather, it is impossible from a logical perspective, since the very phenomenon we are seeking to describe, in this case the conscious experience of another person, would cease to exist without the minimal opacity separating his or her consciousness from mine. Analogously, all knowledge requires this kind of minimal opacity, because knowing something involves, at a minimum, a synthesis of discrete perceptions across space or time.

The Argentine writer Jorge Luis Borges demonstrated this point with implacable rigor in a story about a man who loses the ability to forget, and with that also ceases to think, perceive, and eventually to live, because, as Borges points out, thinking necessarily involves abstraction, the forgetting of differences. Because of what we can thus call our constitutive ignorance, then, we are free — only and precisely because as beings who cannot possibly occupy all times and spatial perspectives without thereby ceasing to be what we are, we are constantly faced with choices. All these choices — to the extent that they are choices and not simply responses to stimuli or reactions to forces exerted on us — have at least some element that cannot be traced to a direct determination, but could only be blamed, for the sake of defending a deterministic thesis, on the ideal and completely fanciful determinism of “how we are” at the time of the decision to be made.

Far from a mere philosophical wish fulfillment or fuzzy, humanistic thinking, then, this kind of freedom is real, hard-nosed and practical. Indeed, courts of law and ethics panels may take specific determinations into account when casting judgment on responsibility, but most of us would agree that it would be absurd for them to waste time considering philosophical, scientific or religious theories of general determinism. The purpose of both my original piece and this response has been to show that, philosophically speaking as well, this real and practical freedom has nothing to fear from philosophical, scientific or religious pipedreams.

This last remark leads me to the one more issue that many readers brought up, and which I can only touch on now in passing: religion. In a recent blog post Jerry Coyne, a professor of evolutionary biology at the University of Chicago, labels me an “accommodationist” who tries to “denigrate science” and vindicate “other ways of knowing.” Professor Coyne goes on to contrast my (alleged) position to “the scientific ‘model of the world,’” which, he adds, has “been extraordinarily successful at solving problems, while other ‘models’ haven’t done squat.” Passing over the fact that, far from denigrating them, I am fervent and open admirer of the natural sciences (my first academic interests were physics and mathematics), I’m content to let Professor Coyne’s dismissal of every cultural, literary, philosophical or artistic achievement in history speak for itself.

[darkstar3]

1. Um, duh?

So we'll never explain every single mystery of the universe. BFD...next!

[damntexdem]

2. Wait! Do we have complete knowledge about not explaining very mystery of the universe?

Of course, then, those who DO have complete knowledge about every mystery in the universe (e.g., the fundie rightwing) tend to be those whose store of knowledge has been thoroughly demonstrated to be completely empty.

[Silent3]

3. "thinking necessarily involves abstraction, the forgetting of differences"

I don't buy that.

There's a big difference between provisionally setting aside and ignoring differences, and actually, completely and totally forgetting them.

I don't have to actually forget that nothing in the real, physical world is as precise and as simple as a mathematical triangle in order to use the Pythagorean theorem.

There may be other arguments against the possibility of complete knowledge, but "forgetting of differences" sure isn't one of them.

[Jim__]

4. His first argument is different from that.

His first argument is the example of heterophenomenology - the ability to describe another's experience in such a way that it becomes your experience. Doing it, changes the experience (the experience that you now have) from the other, to you.

On the second point, thinking requires abstraction, if we know everything there is to know about, say, each individual horse in the world, is there any benefit in describing common attributes that are largely similar? Since we already know all the detail, I'm not sure what we gain by summarizing certain attributes - never exactly the same - across the entire set.

[Silent3]

5. Even if you somehow knew every detail of every horse...

...in the world, if you wanted to start a business selling horse feed, and customizing each horse's food wasn't practicable or profitable, building a generalization would certainly help simplify your business, and you have to literally forget all the specifics you knew to accomplish that.

At any rate, that seems to be more an argument against the usefulness of having all knowledge, not the ability to obtain all knowledge. Lots of impractical things are nevertheless possible. Like arguing on the internet. :)

The stuff about heterophenomenology -- that seems to be much like recursion issue, like knowing something, knowing that you know it, knowing that you know you know it, etc., just modified with the quality of the experience of knowledge from different perspectives.

That sort of thing makes me react with a big "so what?", however. It doesn't seem to be much to hang hopes on for dressing up Divine Revelation or other aspects of religion as special forms of knowledge, if that's the point of the OP.

[Jim__]

7. "... you have to literally forget all the specifics you knew to accomplish that"

That is, by and large, the point of both Egginton's argument and Borges' story. Literally forgetting implies returning to a state of incomplete knowledge. And, no, it's not an argument against the utility of having all-knowledge, it goes directly to the point that the concept is self-contradictory. Clearly, if what you need to know to simplify your business is an abstraction of your current detailed knowledge, and your detailed knowledge blocks you from that abstraction, then your current state of detailed knowledge is necessarily incomplete. Your example actually reinforces the point.

If heterophenomenology is a form of recursion, then the point is made. We know that in sufficiently complex systems, recursion leads to paradox. Eggington's point about heterophenomenology is that it leads to paradox.

You are, of course, free to react to this as you will. I have not heard this argument, that complete knowledge is not possible, and I find it quite interesting. One of the things I've often wondered about is what are the limits of human knowledge. If Eggington is correct, then we can know that there is a limit to our knowledge. It actually answers a question I've long wondered about, and did not expect to be answered. The answer, of course, leads to more questions, but that's always the way.

It doesn't seem to be much to hang hopes on for dressing up Divine Revelation or other aspects of religion as special forms of knowledge, if that's the point of the OP.


???????? Care to elaborate?

[Silent3]

9. If I can temporarily set aside differences between...

...ideal mathematical triangles and real somewhat-triangular objects, and after I'm done doing that, I don't have to go through a intensive remedial education process so that I can re-learn about somewhat-triangular real-world objects, then I haven't forgotten anything.

If I knew every detail about every horse in the world before I decided to construct a generality about horses so I could design a general-purpose horse feed, and then when I'm finished I'm not stuck in a state where I think every horse in the world is identical until I'm taught everything all over again, then I haven't forgotten anything.

This particular argument about the supposed paradox of complete knowledge relies on falsely conflating the construction of processes with filtered access to knowledge with loss or absence of that knowledge. If such a process is merely a subset of a system with complete knowledge (or simply with greater knowledge than some internal processes are granted access) that is not forgetting, that is not loss or knowledge or absence of knowledge from the system.

In software engineering, there is a principle called data hiding. It keeps code easier to maintain, easier to understand, and less at risk of unintended side effects when modules, packages, objects, and methods expose no more data than other components need to effectively use these modules, packages, etc., and they take in no more data than they need to perform their functions. Software applications don't have to lose or "forget" anything at all simply because data hiding is applied extensively to the individual components out of which applications are constructed.

Care to elaborate?

Care to tell me if the OP is or is not meant to have any bearing of the subject of Religion/Theology, the forum within which it is posted? Are you really shocked or genuinely confused that someone tries to take a guess at some possible connection between the OP and the forum it has been posted in?

If the OP has no connection at all to such issues as I guessed at, why not just post the water is wet or that wombats are fur-bearing mammals?

[Jim__]

11. You used the phrase "literally forget."

I can only respond to what you say.

As to why this is in the R/T Forum, as I said in the OP, it's related to a previous thread in this forum.

[Silent3]

12. I used the phrase, but only sarcastically.

I don't buy that you'd have to literally forget, not at all.

[struggle4progress]

6. Coyne's post is remarkably ill-tempered. I'm not that excited by Egginton's arguments, because

I think many so-called "debates" wildly miss the mark: on the topic of "freewill," for example, I can't see how to make any sense at all of human activity and discussion, if one doesn't believe people have some choice; the question is then how much choice one actually has in given situations and how to make the decisions one can actually make; but, of course, it has been recognized for a century or more that human responses can be mechanical, and that mechanical responses can be quite informative in ways many of us may prefer not to consider in detail

Science is certainly concerned with the problem of accurate prediction, and I find it strange that many people seem unaware of the difficulties associated with numerical instability. In meteorology, for example, if current models are correct, there is simply no hope of predicting the weather accurately more than two or three weeks into the future -- not simply because we do not know the initial conditions well enough but because we could never know the initial conditions well enough: the tiniest changes in initial conditions grow exponentially in time, so that very very tiny differences eventually produce enormously different predictions. One is not really free to disregard such facts, if one considers oneself scientific: it does not mean meteorology is futile or worthless; it does not mean no one should study meteorology or that no one should try to construct better models that provide more illumination; it simply suggests that here, as in many other endeavors, there may be insurmountable limits to what we can actually accomplish; even if we make ever-better models forever and ever

One ought to distinguish the philosophical stances appropriate when one attempts scientific research from the philosophical stances appropriate when one attempts to evaluate our current state of knowledge. If one wants to make very accurate predictions, a deterministic hypothesis is quite useful; if one realizes that many models exhibit instability so that insurmountable uncertainty about initial conditions limits the testing of the model, then one should also recognize in many cases one cannot put a deterministic hypothesis to a strict test and hence cannot take the view that it is established

But the real theological issues lie elsewhere, I think. The relation between the scientist and the object of study is a subject-object relation, I-It. Many relations between people are also of the I-It variety, but they are not particularly satisfying (as human relations go); one gets something quite different from subject-subject I-Thou relations. It is one thing, and all too easy, to interact with another person as a mere object; it is an entirely different thing to really interact another person with regard for the person's subjectivity

[Jim__]

8. I agree Coyne's post is ill-tempered, but the opening remarks are also amazingly sloppy.

Coyne claims: William Egginton ... was exercised by recent research showing that when monkeys make a “decision,” their neurons register it before they’re conscious of it. Of course, nothing in the described experiment measured when the monkeys were conscious of their decision.

While the question of free will versus determinism may ultimately be unresolvable, it's still important. Not only is the question of free will important for religious considerations, it also has tremendous political implications. Many people seem to be convinced that the universe is deterministic. Yet, I've never met anyone who thinks thay are not free to raise their arm when they want. There is a lot of confusion around the issue, and discussion/debate can always help shed light.

If we accept determinism, it seems that there can't really be a subject/subject relation. If the universe is entirely deterministic, then, "conscious" beings do have their own subjective perceptions, but we cannot relate to anyone else as a subject.

[trotsky]

10. If so, then the traditional monotheistic god is impossible.

Well, OK, so that's not any kind of earth-shattering news. But it is awfully funny to take a theist's weak attempt at a critique of atheism and turn it right around against theism.