Realism, truth and all that

A comment to my last post raised some questions about my views on realism and on truth. This post will be an attempt to respond to those questions.

Background

First, some background. I’m a mathematician and I have spent time doing computer science. I have no doubt that my experience in mathematics and computer science have greatly influenced my thinking on many topics.

Around 1990, or perhaps a year earlier, I asked myself the question of “how does learning work?” This arose, in part, out of pedagogical concerns. It seemed to me that the dominant ways of teaching mathematics did not work well with how we actually learn.

That started me in a project of attempting to understand human cognition. In some ways, this has been very successful. I believe I do have a reasonably good, if imperfect, understanding. But, in other ways, it has been a failure. Much of what I now understand about cognition goes against the conventional wisdom. And that makes it hard to explain to others.

I bring up this background, because I might occasionally mention what was my view before 1990. And by that, I mean what was my view before I started that investigation into cognition.

Realism

The comment to which I am responding begins with:

I’m curious to know what you mean when you say that you are a realist.

For many scientists, “realist” and “pragmatist” are almost interchangeable. I guess that’s a good starting point for my own views on realism.

People seem to use the word “real” as if it has a meaning that comes directly from God. But it can only mean what we take it to mean. I like Wittgenstein’s “meaning is use”. I don’t see how “real” can mean anything beyond what the conventional wisdom takes to be real. There isn’t any magic that gets us to some pure meaning that is independent from how people use the word.

I contrast realism with Berkeley’s idealism. Perhaps I should say that for me, realism is a denial of Berkeley’s realism and a denial of Kant’s transcendental idealism.

When Samuel Johnson was confronted with the question of Berkeley’s idealism, he supposedly kicked a stone and said “I refute it thus”. It is generally agreed that idealism cannot be refuted in that way. As best I can tell, there isn’t any way to refute idealism. Whether to go with some version of idealism or some version of realism seems to be a personal choice. For me, realism is the pragmatic choice. For Berkeley, given his religious views, perhaps his idealism was a pragmatic choice for him.

Yes, metaphysical realism (from philosophy) is tied to questions of truth, and often on the question of whether the laws of physics are true. I see that as a mistake.

According to the conventional wisdom, the laws of physics are descriptions of reality. But that’s surely mistaken. A baseball player or a tennis player or a basketball player can do very well at controlling the motion of the ball used in his game. I do not know of any evidence that a tennis player’s skill was improved by virtue of his learning Newton’s laws or learning relativity. Those laws, quite simply, are not actual descriptions of reality. Rather, the laws of physics are technical descriptions of scientific procedures that scientists have found to be useful.

When scientists change their procedures — say from Aristotle’s methods to Newton’s methods or to relativistic methods or to quantum mechanics, they change their technical descriptions. Whether or not those technical descriptions are said to be true would seem to have very little relevance to the question of realism.

Truth

Continuing my quote from that comment:

Often when people make that claim, it is an expression of where they stand on truth.
The correspondence crowd has the strongest claim I think, in that, for them there is a way that the world is, as you say, and we somehow receive more or less accurate reports of it.

So where do I stand on truth? If I had been asked before 1990, I would have gone with the correspondence theory. However, once I started investigating cognition, I had to attempt to understand how people discern truth. And it was not clear that they were following any correspondence theory. To go with correspondence as the basis for truth, I would have to identify what correspondence was being used and where that correspondence comes from. As best I can tell, there is no such correspondence. When I look at it closely, the correspondence theory seems to be circular. To a first approximation, it says that true sentences are true and false sentences are false. And if that’s all it says, then it doesn’t say anything at all

There seems to be broad agreement that natural languages are conventional. That is, there is a system of social conventions and social norms that set the basis for language. Those conventions establish how we should express descriptions of the world. But those conventions also set how we should evaluate whether a description can be said to be true.

I guess you could say that view of truth is that it is truth by convention. But I am not talking here about explicit truth conventions. Rather, the conventions of a natural language already set how we should understand truth.

Continuing with that comment:

I think some in the sciences fall into that category, but I think most are actually coherentists. Your previous example of the Newtonian equation F = MA taken as definitional of the factors fits the bill. Frankly, I can see the point for those who take that position.

I am not a fan of coherent theories of truth.

What can we say about mass otherwise that is not just a statement about how we categorize experience as well?

Mass is an abstraction. Whether or not mass is real does not seem to actually be important. What matters, is how we ascribe mass (or assign mass). The laws of physics tell us how to ascribe mass and some other physical quantities. By doing that they are, in effect, defining a correspondence between reality and sentences that use those physical quantities. As I see it, we cannot use the correspondence theory to decide the truth of those laws of physics, because it is those laws that actually setup a correspondence. Thus the laws of physics are prior to questions of correspondence truth relating to those physical quantities.

My preferred view is that the laws of physics are neither true nor false. We adopt those laws on pragmatic grounds, not on the basis of their truth. In effect, they become truth conventions for the truth of measurements made under those laws.

I don’t see many people in the deflationary camp, but I think there are some.

If the deflationary theory just says “we don’t need a theory of truth” then I don’t have any problems with that. If it says more than that, then I have failed to understand what it says.

Omphalism and truth

My views on truth are influenced by omphalism.

Of course, omphalism is absurd. However, it is an interesting attempt to be consistent with the evidence. On many questions of truth about the past, the omphalist will disagree with the traditional understanding. And those truth disagreements cannot be settled with evidence.

It seems that the choice between omphalism and the traditional view of the past can only be made on the basis of pragmatism. And this illustrates why science must be seen as a pragmatic endeavor, rather than a search for truth.

Much the same can be said about idealism vs. realism. The choice between the two must be made on pragmatic grounds. The disagreements cannot be settled by evidence, because idealism explains away the evidence.

For me, then, I see a world where major principles are decided on a pragmatic basis. Once widely adopted, these principles become, in effect, social conventions. Once we have those major principles, we use those to construct rules. And then we use truth to assess the adherence to rules. So we have a two-level system, with pragmatic social conventions for the major issues and rule following for the smaller issues that we get by subdividing the question.

As an example, I see heliocentrism vs. geocentrism as a pragmatic issue. One of the proposed tests of heliocentrism was the parallax of observing stars. Unfortunately, the parallax was too small for the measuring methods available at that time. But even if parallax were observed, that would only be evidence for a pragmatic judgment rather than a truth judgement.