Truth, knowledge, skepticism, and stuff…

A standard philosophical definition of knowledge goes as follows:

(I know A) <=> (I have a justified belief that A) and (A is true)

As I write this essay, I’m in the middle of reading Robert Nozick’s discussion of knowledge and skepticism in Philosophical Explanations.  Nozick uses a very specific concept of “truth tracking” which leads to some very interesting results, but is essentially (it seems to me) just a particular approach to defining justification, one that leads to a coherent (I think), but sometimes quite peculiar conception of knowledge.  I find myself working around a rather different rejection of philosophical skepticism.  What follow are musings, not intended to represent a complete expression of a developed idea.

If some particular skeptical scenario (SK: SK => not-A) were true, then A would be false, and I would not know A.  However, A, so not-SK, and I do know A.  The skeptic’s objection fails, because SK is not true.

Truth is outside my direct experience.  It is not something I have direct access to.  To use Dewey’s language from Experience and Nature, it is not something I “have directly”.  All that I have directly is justification (evidence, inference rules, trusted sources, etc.)  There is no way to operationalize truth; I can only operationalize justification.  One reason that theories of knowledge remain so problematic is that knowledge, vs. justification and belief, has exactly zero influence on our behavior at time t = now().  We try to justify our beliefs, and to act on justified beliefs.  Later, we may come to believe that our prior beliefs were not (properly) justified, i.e., were false.  But at any given moment (outside of certain contexts of philosophical inquiry) the question, “Are my beliefs true, or merely justified?” is never a useful one.  The question, “Is this belief justified?”, however, is (always?) an important one.  Operationally, the question of “truth” always devolves to:  from perspective B (perhaps a later one), do the beliefs held to be justified from perspective A still appear justified?

There is no way to operationalize (general, philosophical) skepticism at all.  Skepticism places all logical possibilities on an equal footing.  All available evidence is discounted, and there is no way to distinguish one possibility from another, to prefer one over the other.  Skepticism cannot give any positive plan of action.

Philosophical skepticism of the sort I am referring to (what Hume referred to as Pyrrhonism) must not be confused with a critical evaluation of our methods of justification.  Critical evaluation of our methods of justification is completely operational and is critical to ensuring that they track truth as closely as possible.

P.S.  Nozick, I think, expresses an insight similar to mine above about truth and experience when he says (p. 232 in my paperback copy) “We have said that knowledge is a real connection of belief to the world, [which we call] tracking, and… this view is external to the viewpoint of the knower, as compared to traditional treatments, [though] it does treat the method he uses [to track truth] from the inside, to the extent it is guided by internal cues and appearances.”

P.P.S.  Anticipating certain gleeful but misguided reactions to my rejection of “skepticism”, I want to point out a deviation between a common vernacular use of the word “skeptical” and philosophical skepticism, to wit, the phrase: “Skeptical about God.”  Disbelief in God (or even just doubt) is usually based on a belief that you can trust the evidence to lead you to true conclusions about the world.  Doubt about God stems from observing the lack of positive evidence, and disbelief from observing that the available evidence is incompatible with the God hypothesis.  These methods of justifying belief are exactly antithetical to philosophical skepticism, which holds that no amount of evidence is ever sufficient to justify a belief.  Theists, in fact, often (mis-)apply a skeptical argument when they claim that “You can’t prove a negative result,” not realizing that, if true, this argument merely puts their God on exactly the same footing as being a brain in a vat manipulated by alien scientists in the Alpha Centauri system.

P.P.P.S.  (9/7/2013)  I need to correct my statement that Nozick's concept of "tracking" is just a special case of justification.  (He is at pains to distinguish them on p. 267, by which time I was in a position to realize what he meant.)  Tracking, like truth, is external:  our methods really do track truth.  Justification is internal: we believe our methods track truth, and are therefore reliable or justified.

Starting from Zero

I’ve never read Hegel, but I recently reread Russell’s chapter on Hegel in A History of Western Philosophy as background to some early Marx we were reading for the Democratic Socialists of America Boston local’s reading group. Per Lord R, Hegel starts from the idea that “the Absolute” is “pure being”, deduces that this means it has no qualities, and finds this self-contradictory because to have no qualities is to be nothing. This is the antithesis – “The Absolute is nothing” – and leads to the first synthesis: “The Absolute is Becoming.”

What struck me is that Hegel thinks he is starting from zero, using only one premise – that reality cannot be self-contradictory – but of course he really is applying many premises: that there is something that may reasonably be thought of as “absolute reality”, that “pure being” is a meaningful concept, which entails having no attributes, that to have no attributes (or qualities) is to be nothing, that the union of being and being nothing is becoming.

It seems there is a basic flaw in a philosophy which tries to start from zero, or from some very small set of premises, building up from these by synthesis; to wit, that we cannot do it. We always, in fact, have to start with a lot. Descartes’ radical doubt leads, perhaps, to “cogito ergo sum” (more precisely, perhaps, “...something is”, but to go further, he needs to introduce more premises, “undoubting” certain things. He selects premises that purport to justify the undoubting – clear and distinct ideas, and the goodness of God – but this is basically B.S., not justified by his method, as stated.

Really, what we do, always, is to start with everything we know, or think we know, and start juggling it about. We apply reasoning (rules of inference) that feel right to us and (perhaps) we try to codify these into a system of logic. We apply various tests in doing this. What tests depends on the person and what he or she brings to the process. Examples are empirical tests (how well do these premises and this logic predict future experience) and faith-based tests (do these premises and this logic lead to conclusions that conflict with “revealed” scripture?) We can apply tests, at any time, to some of our a priori beliefs about facts, and or to our rules of inference, but each test involves accepting other facts and rules, at least temporarily and contingently. We can never test the whole shebang.

This process is irreducibly ad hoc and messy, which is no doubt what has led many philosophers to reject it, and to try and find a purer and simpler alternative. But they were deluding themselves.

The actual process “works”, more or less, because our sensual and reasoning faculties have evolved by natural selection. If they failed badly at the task of extracting meaning from the universe, we would have become extinct, or at least these faculties would have. But this is no guarantor of “ultimate knowledge”, whatever that is. Aspects of our reasoning system may be selectively neutral, such that there is no effective, general difference in survival rates from reasoning one way or the other. These aspects, then, would not be fine tuned by natural selection. Like blue eyes and brown eyes, both may survive indefinitely in the species. I suspect this may be particularly true of the ways we construct abstract structures of meaning to explain our experience to ourselves.

Different ways of knowing can also serve the same person better or worse under different conditions. Something like what I called “faith based” learning is certainly the most satisfactory way of learning for a child, who will learn much too slowly if she must test every new factoid empirically, rather than trusting the wisdom imparted by her elders. In fact, realistically most of what we learn throughout all our lives, we learn from others whom we trust, either because of their social position as teachers, authors, or what-have-you, or simply because we believe they have no reason to lie. In complex situations, we have to apply other tests – such as when we find that the “experts” don’t agree with each other. And empirical experience is the ultimate arbiter, to which faith, for any reasonable person, must bow. Galileo’s telescope trumps the Pope.