Doubting Didymus

The proper unit of evaluation is not the theory, but the research programme: a family of similar theories sharing a 'hard core' of propositions surrounded by a 'protective belt' of auxilliary hypotheses.

To evaluate, one looks at 'problem shifts', or how a programme responds when confronted with contradictory evidence. A problem shift is 'progressive' if each proposed theory in succession prdicts everything its predeccessor did, as well as new novel facts in addition, which are eventually confirmed by evidence. This represents either growth of knowledge, or growth of predictive conjectures. If the shift does not predict new facts, but only "eliminates anomalies through verbal tricks" it is degenerative, and thus 'content reducing' i.e. no bloody good.

I'm fond of this metric myself. In my recent studies I've been using it to evaluate political science theories. I find it is a perfectly servicable way to evaluate theories whether falsifiyable or not, and I suggest that it is a way to sidestep the dilemma you have formulated as "We cannot exclude an unfalsifiable theory for failure to be "well-supported".

PoodleLovinPessimist

This seems like a reasonable enough approach to evaluate existing research programs and eliminate those that have gone awry or exhausted their subject matter. But how do you use these criteria to get a new research program off the ground? Consider relativity or plate tectonics: Without a body of researchers, there's no research program to evaluate.

This is precisely the complaint of ID: "Since everyone excludes our theory a priori, we cannot get a research program off the ground to evaluate. Since we're fighting political oppression, of course the few brave researchers have to spend their time fighting oppression to get enough of a research program off the ground for a fair evaluation."

Political "science" is perhaps valuable, in the sense that many nonscientific endeavors are valuable. I am not at all convinced, though that political "science" is actually science, precisely because its "theories" do not appear to be empirically falsifiable. I'm no big expert in PoliSci, though, so I could definitely be wrong. But I see no reason not to apply the demarcation criterion to political science, with the proviso that if it fails the criterion I do not then hold that it is without value, and it might well be the case that other methodologies borrowed from falsifiable science might prove valuable.

PoodleLovinPessimist

You (or Lakatos) have also embedded the falsification criterion into this definition: What does it mean to be confronted with "contradictory evidence"? What does it mean for a theory to "predict" something? Without the falsification criterion, what do "contradictory evidence" and "predictivity" mean?

Clutch

Well, those are good questions. They're just not the questions raised in this thread.

Suppose someone asked, "Can you fool some of the people all of the time?"

Some smartass might come along and say: Wait a minute. There's more than one thing that question might mean. It could mean, At every time, can you fool some people or other? But it might also mean, Is there some specific group of people who can always be fooled? Let's make sure we know which question we're trying to answer before we spin our wheels arguing over which answer is the correct one!

That's essentially what I'm doing here.

What you seem to be doing is analogous to replying: Oh, but who counts as people, and how can we decide what it is to be fooled?

Good questions, I guess. But not really to the point.

PoodleLovinPessimist

My point is that the questions you're raising in the thread are not very good. They're pretty good, and if we were dealing with a debate within the scientific community, they would be excellent questions.

And, I'm trying to show why I don't consider them good questions: Because they admit to a more-or-less reasonable interpretation which allows a theological and non-naturalistic "explanation" to be considered "science", or they could be considered as excluding a concept like ID on the basis of prejudice and bias.

Yes, there exists some interpretation of your questions under which ID fails. I myself would apply such an interpretation. But there's a difference between having a favorable interpretation, and being precise enough to render a differing interpretation to be entirely absurd.

I obviously don't think your OP meets that burden.

If you want to just declare my criticism invalid, I suppose you can do so--everyone has a right to his own opinion--and we have no other choice but to agree to disagree. <shrugs>