The Demarcation Problem and Truth

Josh, I'm not following you. You said:

Yes, scientific theories can be false or true, and nonscientific claims can be true or false.

I agree with that.

Then you said:

The difference between the two is that science offers an empirical way to tell which is which, while non-science doesn't.

This doesn't seem compatible with what you said above, though. If scientific theories can be true or false, then how can science give us an empirical way to tell which is which?

I'm not just talking about the difference between good science and bad science, BTW, I'm talking about current theories that are not true but are considered good science. For example, right now, because they aren't compatible with each other, we know that either relativity theory or quantum mechanics (or both) are wrong. Yet these are both considered good science (and rightly so).

If you want to know what I think on this subject, I'll be very clear right here. I think the natural sciences are, in general, pretty useful for giving us models that make testable predictions. Whether these models are true or not is a metaphysical question - one that science can't answer. Good science doesn't necessarily result in truth (as I argued above with the quantum mechanics and relativity example). Bad science (or non-science) doesn't necessarily result in falsehood. Very little of this has anything to do with my original post on demarcation.

If this means I'm lumped in with the postmodernists, then so be it. You are free to label me in anyway you want. The question, however, is not what I am to be called, the question is whether or not what I'm saying is true or not.

The difference between the two is that science offers an empirical way to tell which is which, while non-science doesn't.

I would contest that from the other side Macht did. Post hoc ergo propter hoc is decidedly not a scientific way to tell truth from falsehood, but it is empirical nonetheless. (And popular in some quarters.)

My point here is that "science" isn't just a word, it represents something actual, and that actual thing has a method.

The devil is in the details. There are ways to test theories against the evidence that are scientific, and there are ways to test theories against the evidence that aren't. Part of the key, I think, is that the methods of science have been themselves developed through the same sort of empirical-based process as any other scientific theory--or so I strongly suspect. Which is a bit circular, but eh, so what?

Well, part of the so what is the problem of justification. Because there is a certain symmetry of science vis-à-vis other "ways of knowing". Science is well justified on it's terms, but various religions are justified on their terms, and wooery tends to be well justified on it's terms (such as they are). Where I disagree with many so-called postmodernists is the conclusion that science is equally valid as various religions, woo, etc. That conclusion simply doesn't follow from the observation on symmetry I made above, and I reject it. I do not know if the people arguing woo as being equally valid as science are misrepresenting postmodernism or not; but don't lump me in with them. And I do not think I am alone in my philosophical position, whatever it may be called.

The first thing to understand is that I am not treating science as an assemblage of knowledge, but as a process by which inquiry proceeds.

That means that a scientific theory is one that is amenable to scientific testing. It can be testable and false or testable and true.

Newtonian physics is false in that sense, but indisputably scientific. The proposition that invisible, intangible pink fairies explain quantum gravity is not testable, not scientific. It could be true (I suppose), but it isn't science.

QM and relativity are both wrong in at least some sense. String theory may be able to unify both, or something else will. The process isn't over, so suggesting that "good science doesn't necessarily result in truth" is incomplete at best. We haven't yet isolated a unified theory of physics, but we have shown that some things cannot be that theory. That's what falsifiability (and the demarcation problem) buy us.

The fallacy of postmodernism (also implicit in your treatment of demarcation) is treating lack of certainty about the truth as absence of certainty about untruth.

Science doesn't give us absolute certainty about all truth claims. It does allow us to place limits on what could be true. All claims don't deserve equal treatment. The same can't be said of political ideologies. That's why it's so wrong to claim that the demarcation problem is about politics. It's about what processes allow us to separate truth and non-truth.

Josh,

My treatment of demarcation had nothing to do with certainty or lack thereof (implicitly or explicitly). That is simple false. And I don't claim that all claims deserve equal treatment. Nothing in my post implies that, either.

I have no problem with you disagreeing with my argument that the demarcation problem is primarily about politics - if you want to talk about that, we could look at the actual motivations of people who have written on the topic. But the conclusions you are drawing from my post about what I believe are just plain wrong.

And by "simple," I mean "simply."

One way of dealing with the demarcation problem, perhaps, is to assert that concepts that are illuminated by post-modernist critique, e.g. concepts regarding aesthetics or political norms, are not scientific concepts, while concepts that degrade in meaning when subjected to the same critical methods, e.g. concepts of natural selection, plate tectonics, or quantum mechanics, are scientific.

My original point was that your treatment of science is difficult to distinguish from the postmodern (and relativist) treatment of science. My additional point, one I expanded on here, was that demarking science from non-science is important because science is capable of distinguishing truth from nontruth. That problem is not political.

To say that demarcation is a political problem is either to deny that science distinguishes truth from nontruth, to assert that there are other empirical processes for distinguishing truth and nontruth, or that distinguishing truth from nontruth is itself purely a political problem (or that the distinction isn't meaningful pure relativism).

I suspect that you don't think any of those things, so I don't see why you would claim that demarcation is a political problem.

The fallacy of postmodernism (also implicit in your treatment of demarcation) is treating lack of certainty about the truth as absence of certainty about untruth.

Ah. So if I understand correctly, your basic contention is that scientific falsification implies/means a theory is an untruth. (Well that, and we shouldn't teach non-science as science.) I would agree. Certainly, past a certain point the epicycles required to save a theory from falsification become ridiculous, and a theory becomes dead, Jim. The difference between that and absolute certainty is slight. I do see problems in justifying this position philosophically, but that doesn't mean I don't hold it.

All claims don't deserve equal treatment. The same can't be said of political ideologies.

Oh, I don't treat fascism with the same respect I treat libertarianism. But yes, political ideologies do not have the same nature as the truth claims science tests.

That's why it's so wrong to claim that the demarcation problem is about politics. It's about what processes allow us to separate truth and non-truth.

... in the empirical domain. Unless you're a physicalist, in which case that qualification is superfluous. But of course the empirical domain is rather, um, important regardless. And I do see it as a rather bad sign that the demarcation problem is so political outside of science. Science should be those methods that work in practice, corroborating theories that predict well, not those methods we would like to work, corroborating theories we would like to be true. When science becomes politicized, empiricism does seem to fall by the way-side.
However, I do think that some politics within science is useful, helping guard against premature consensus.

Josh, I'm not sure I can accept how you've approached the question. I read you as saying:

1) Demarcation as a political act is wrong.
2) Postmodern descriptions of science are bad
3) Postmodernists talk about politics alot
4) If you believe the demarcation problem is political, then you share a postmodernist view and
5) Since po-mo is no good, you're demarcation claim is no good.

I see a number of problems with this, the greatest of which is that it's not clear to what you are referring when you say "postmodern." Can you clarify that?

Beyond that, I think there are also some problems in your position that derive from what seems to be an overly defensive posture. (Plus, you'll see from the chat Janet and I had about Popper that I don't see the demarcation problem being as simple as you suggest.)

But, would you say that if someone is postmodernist, then you disagree with them? (And again, and more, what do you mean by postmodernist? That Derrida made a quip about Einstein that Sokal couldn't understand?) Or would you say that, if someone says what counts as science is not defined through technical criteria alone then you disagree with them?

I don't mean this in an antagonistic way, Josh, and I very much enjoy the discussion above, and conversations about the demarcation issue in general. I'm just not sure you're reasoning above is sound. If we're all after the same thing, and I think we are -- how to encourage good science, and how to keep bad reasoning out of public funding, and the schools, let's say -- then I just don't think your argument will ultimately help. By reference to empirical data, the argument that demarcation isn't political simply doesn't stand up to the historical evidence.

(I wonder too, now that I wrote this, if the problem also lies with the fuzzy sense of what we mean by "political," but I leave that for the next commenter.)

Ben

Ben, you don't seem like a stupid person, but your 'reading' of Josh seems to me an obviously absurd distortion. Josh doesn't declare Macht's view as guilty by association to theories he does not like. Rather he is claiming that the thesis that demarcation is primarily understood as political is difficult to reconclile with a view that science has a way of approaching truth. The denial of the latter seems to be associated with deeply mistaken views, sure. But Josh's point really is not about pomo anything as such - it's just a word, perhaps better avoided altogether.

I must confess to being a little irritated to see Sokel criticized for 'not understanding' Derrida. When D (or anyone else) spouts off in his way about science topics not understood, he himself likely doesn't understand what he is saying. You see, there is nothing to understand, no 'there' there; it's what Frankfurt calls 'bullshit'. This sort of thing is simply an intellectually unhelpful, childish rhetorical technique whereby one appropriates the 'authority' of scientific theory to one's own view.

There is to be sure no way really to know what goes on in anyone's head, but when someone tries to make some kind of point using science they clearly do not understand, I see little reason to do anything but laugh. I wasn't born yesterday, you know. Look, any scholar can say anything no matter how absurd, then turn arround and claim to be misunderstood. I don't buy it.

"I wonder too, now that I wrote this, if the problem also lies with the fuzzy sense of what we mean by "political," ..."

I was kind of loose with this term in my post. But I think it was pretty clear that I was arguing 2 things. First, if something is called "science" it tends to have a special status in our society. Second, the demarcation problem almost always comes up in the context of politics. This is true if we look at the media. And it is true for many philosophers who have written on the subject. I sometimes made the distinction between political, social, and rhetorical motives in that post, but I was kind of sloppy, using the term "political" for all three at times. What I didn't do was talk about truth or certainty - at all.

"Beyond that, I think there are also some problems in your position that derive from what seems to be an overly defensive posture."

This seems to be the main issue. The fact that I posted this on a pro-ID blog means, for many people, that I must have some ulterior motive. Never mind the fact that I don't think ID is good science or even science at all at this point (although I think it could be in the future). Never mind the fact that the whole tone of my post was that it is a bad thing that people are more concerned about the politics and rhetoric of demarcation than the questions "is it rational?" or "is it well-supported?" (Also, never mind the fact that postmodernists would tend to think ID is dead wrong in trying to legitimize its beliefs through science.) But the whole point of Josh's post was to link the crazy postmodernists to the crazy IDists, so what I actually wrote seems to be irrelevant.

(And this isn't meant to be antagonistic, either. A lot of anti-ID people want to make it look as if every ID sympathizer is as looney as DaveScot. And, quite frankly, I get sick of that. That's every bit a rhetorical move as the demarcation problem often is and I'd like to see the discussion rise above that.)

I find myself generally more friendly to postmodernism than others, and I suspect that the "microfascism" article I quoted in the previous post on this topic gave a rather poor account of pomo, but that is the sense of the term I'm working from for the moment.

I also confess that the account of demarcation I've given here is drastically oversimplified (though I'm not convinced that it's as complicated as other people sometimes make it). Pete Seeger says that any time you get a group of people together, that's politics and in that sense, you are right that science is political.

What I object to about saying that demarcation is a political problem is that we aren't using the term in Pete Seeger's sense. I see it used here as a synonym for "partisan."

Something that's purely political is not built on empiricism, it's built on power and promoting self-interest. My reading of Macht's original claim that "the motivation for asking whether something is science or pseudoscience is primarily social/political in nature" is that people are trying to draw arbitrary lines to keep the rabble out, that this thing we call "science" exists only in furtherance of some social/political agenda.

It may be I'm misreading macht, but the sense I get from a lot of creationists is that they see demarcation exactly that way, as arbitrary rules imposed by mean guys in lab coats or tweed jackets. Which was the same message sent by the "postmodern" critique of evidence-based medicine.

The demarcation problem is definitely more complicated than what Popper envisioned, and it may well be that the practice of science, like scientific knowledge itself, is only an approximation of some ideal form.

"It may be I'm misreading macht, but the sense I get from a lot of creationists is that they see demarcation exactly that way, as arbitrary rules imposed by mean guys in lab coats or tweed jackets."

I've written on the demarcation problem quite a bit at Telic Thoughts and I've said repeatedly that this is not how I view it. I've made it very clear that I think scientists rely on wisdom and judgment and that these things come with experience in the lab, performing tests, interpreting results, etc. These things can't be described by strict rules or methods. This doesn't mean that scientists haven't developed methods and procedures that work very well, though. It just means that for any rule or method that a philosopher of science can dream up, there may be times that a scientist might decide that applying that method or rule in a given situation is inappropriate. The problem is that wisdom and judgment are hard to pin down precisely and this is why the demarcation problem hasn't been solved (and won't be solved, IMO). But there seems to be the need by many people to say that science can be distinguished from non-science by some simple rules (like falsification or methodological naturalism or whatever). And (I argue) this need most often shows up in the context of politics. That is what I was claiming in that post. I didn't say or imply that things like falsification or MN are arbitrary. On the contrary, I think that there are probably a dozen or so things (like falsifiability, the ability to make accurate predictions, consistency with other scientific theories, simplicity, etc.) that we look for in good scientific theories. But if any particular theory is missing one or two of these things, that doesn't automatically make it non-science.

Macht:

First, if something is called "science" it tends to have a special status in our society. Second, the demarcation problem almost always comes up in the context of politics.

Let me say first that I disagree with the second point. I suspect you just haven't been reading the right parts of the scientific literature, issues of demarcation were big in arguments over ecological theory, cladistics, string theory, and elsewhere. These are cases where the philosophy of science and details of demarcation have fed back into the practice of science.

To the extent that demarcation is raised as an issue in politics, I think your error is in failing to see the linkage between your first and second points. It's true "What [you] didn't do was talk about truth or certainty - at all." The reason for your first point, the reason that science has a special status, is that science has a methodology that separates truth from untruth. That special status makes science and the knowledge gained from the scientific process very useful in political and social discourse. To say that demarcation is "often" a rhetorical move is simply bogus.

My point was not to link ID to pomo and thereby reduce it to a previously solved problem, but to show how that bogus way of thinking about science pervades both creationist and post-modernist thinking.

I think that there are probably a dozen or so things (like falsifiability, the ability to make accurate predictions, consistency with other scientific theories, simplicity, etc.) that we look for in good scientific theories. But if any particular theory is missing one or two of these things, that doesn't automatically make it non-science.

I would first note that the ability to make predictions relies on a phenomenon being bound by natural law, so naturalism is a consequence of predictability, and the ability to make predictions is a necessary and sufficient condition for falsifiability. So we've just reduced three of your four examples to one. And I don't know that anyone would claim that something is scientific only because it is simple. The role of simplicity in the social process by which scientific ideas are accepted and disseminated through the scientific community, but I don't know that anyone would treat it as a demarcation criterion.

If you can show that there is nothing which is sufficient and necessary for something to be science, then demarcation would be sufficiently arbitrary to be raw politics. If there is something that is necessary and sufficient, then it isn't political.

I'm not convinced that falsifiability is quite right, but I think it's basically right. Janet wrote, in the Page 3.14 debate referenced above, that "By the scientists' lights, he really captures something important about the *spirit* of the scientific endeavor." That's right. The details need to be fine tuned, but he's on the right track. It isn't either/or. If something lacks falsifiability (or some more sophisticated derivative thereof) it isn't science. It may well be that science-as-practiced only asymptotically approaches that ideal, but scientists recognize that ideal in Popper's description. I feel comfortable saying that something lacking those traits wouldn't be science because it wouldn't be able to do what science does.

Science, unlike the biological world, is teleological, it exists with a goal of separating truth and nontruth. Without falsifiability (or something very like it) it cannot do that. That's why these things matter for a lot more than politics.

Josh, this is really getting frustrating because it seems like you are trying to divert the conversation away from the issue at hand. I will gladly discuss with you whether the demarcation problem is mostly political - it's an interesting topic, I think, and worth talking about - but I will not do so until you at least recognize that you were (not, may have been) wrong in using me as an example of the type of thinking that you were trying to link to postmodernist thinking. I've repeatedly denied it and I've repeatedly explained that the post in question doesn't imply that thinking at all. The most I've gotten out of you is "It may be I'm misreading macht." You are misreading me - you are taking things that I've never written and things I explicitly deny and charging me with believing them. At the very least, I hope anybody following this exchange realizes I don't believe the things that Josh charges me with believing.

I guess what I'm saying is that I don't see how demarcation can be political, not based on strict rules, and still non-arbitrary. I feel that the act of arguing that demarcation is political puts it within a sphere of arbitrariness.

Therefore, I don't see how we can separate the issues that you are trying to distinguish. My intent is not to misrepresent you, I see an inconsistency in your argument and am trying to tease it out. I will state for the record that my comments about creationists in general do not apply to you in particular, and that I don't mean to suggest that you are a postmodernist. That hadn't been my point before and it isn't now.

But Josh, I think the crux of this back-and-forth with Macht is that you are taking for granted that which you want to prove. I think your argument is built inside itself. You say that "the ability to make predictions relies on a phenomenon being bound by natural law," but we don't know what natural laws are except through the very scientific process that you're saying is bounded by those laws. It's post facto -- you assume that our scientific process is right (and better) because it finds these laws, although we don't know it's right until we find the laws. Science proves that science is right?

(Also, thanks RJ for not thinking me stupid. My wife and I argue about this alot, and I'll let her know I've gained an ally!)

Thanks, Josh. I think I understand better what you were saying. I've tried to explain here why I don't think I'm being inconsistent in my views.

Science proves that science is right?

Yes. Those methods that work are kept.

This is the nature of methods and models. They seem circular if one forget the distinction between nature and model and that observations lift up data to the model. The model or method may relying on circular definitions, but the situation isn't.

One simple example is the following description of Hooke's law: Linear-elastic systems (materials) are described by Hooke's law F = - k*x. Those systems (materials) that are described by Hooke's law are called linear-elastic. So lin. el. -> law, law -> lin. el.

What are lacking in this description is the arrow from observations of material behaviour that describes the force and the elongation: (F,x) -> lin. el. & law.

This is the same problem when one claim it is circularity in method evaluation. Good scientific process -> laws, laws -> good sci. pr. But really observations -> good sci.pr. & laws. Which BTW is why we can go on and improve both methods and theory, while a real circular situation would take us nowhere, no pun intended.

Josh,

Any thoughts on my post?