I normally don't respond the to IDiocy of Uncommon Dissent, but John Lynch, may he rot in purgatory for a thousand years, has made me. As usual, I won't dignify it with links. If you are that interested you can find it.
There are two items: one is by DaveScot, who argues (!) that because Popper's falsification hypothesis means that until we find non-white swans, a hypothesis that swans are white stands, we should somehow assume that Intelligent Design stands as a scientific view. This is silly for a couple of reasons. One is that we have got counterinstances to the need for ID to explain complexity and functional novelty in biology. It's called "natural selection", and it is, as Haldane once famously noted, a complexity and improbability generator. So the universal claim that ID is required to explain these biological facts dissolves.
The other reason it is silly is because Popper's view of science is very far from being the accepted one in the philosophy of science. [Parenthetically, what is it with the ID types? Do they think that making philosophical arguments is a substitute for doing science? If so, they are a long way from anything scientists accept as correct about science. When sciences dissolve into philosophical debates, say, about methodology or the nature of theories, it is usually because there is no science to do in respect of that issue any more.]
Popper's view of science leaves out much, if not most, of the interesting parts of science. For one, the processes of scientific discovery. For another, the crucial role of classification. For sure, Popper presented this as a demarcation criterion for science rather than, initially, as a positive account of what science actually did, but the way he gets used, often by creationists and other antiscience types, is to argue that, hey, they haven't yet been falsified in some extenuated metaphysical sense, so they are still "doing" science. Research and data don't seem to come into it.
Popper's view of science has been supplanted by a number of later views, not least being the sociological accounts of Kuhn and Lakatos, which, being sociological, don't tell us what is science but only how it proceeds descriptively. Prescriptive views of science are much more nuanced than Popper these days, and they lack a simple slogan like the cry of "falsifiability!" They typically focus on the heuristics (rules of inference) and how they have developed overall and in particular disciplines. If you want to argue that ID is science, go read van Fraassen, or Hacking, or Giere, or Laudan and get back to me. And deal with issues of naturalism in science (which, unlike the way creationists frame it, means the naturalising of epistemology, in the philosophy of science, or treating knowledge acquisition itself as a natural process).
Then, for contrast, we get BarryA arguing that Popper is not pope. Wow, film at 11. It seems that he is vaguely aware that Popper is a very bad source of authority for what is science, with respect to ID. He appeals to inference to the best explanation (IBE), or abductive reasoning, which is effectively a Bayesian process of drawing the most likely inference based on the data and our knowledge. Problem with this, is if you do the sums properly, you get anything but ID as an IBE. Why? Same reason as above - natural selection trumps it. If you add into your priors that selective explanations reduce the unlikelihood of complex traits, IBE requires that we adopt it, rather than an unlikely, unknowable, uninvestigable designer who can do what seems to be physically impossible (i.e., "front load" biological systems with traits that will be useful in several billion years).
So, we are not that impressed by the ID appeal to Popper. Not that this will influence their thinking in any way...
- Log in to post comments
You forgot to mention that the ultra-naive form of falsification presented by Dave der Depp has very little to do with anything that Popper wrote.
That would be redundant. I mentioned it was at Uncommon Dissent.
Well Dave Scott clearly is not aware that there are non white swans. At least last time I was in Western Australia there were.
There is another side to this, and I think that it is more important: Not only is ID not necessary to explain complexity, etc.; ID is also not sufficient to explain - well, to explain anything.
For example, just to take one way which ID is not sufficient: Every example of design that we are aware of (dare I say, that we can even imagine?) is an application of laws of nature. To put it briefly, design is not sufficient without specifying how it might happen.
I think that this is better, because it is not subject to the misinterpretation of your remark as accepting that ID is falsifiable.
There doesn't seem to be any systematic attempts to make a "science of science", that would systematically observe and explore why and how different methods work and become accepted.
Meanwhile testability is a common recognized characteristic for theories among scientists, and falsifiability is a common model for how that works. AFAIK Kuhn isn't usually seen as describing science properly. (At least among physicists.) Someone who I have seen get accolade for describing science is Alan Sokal.
The string theory/landscape (or multiverse) debate is useful for such analysis. If string theory isn't testable, is string theory merely a mathematical tool, or is it physics proper? Since it in fact predicts things like the same black hole entropy as semiclassical models, it makes theoretical predictions but not yet uniquely distinguishing it from current accepted physics.
Some are dissatisfied with this, so naive falsifiability is clearly not enough for them. OTOH while Popper's version of test is eagerly searched for by those individuals, others are satisfied with less.
ID fits perfectly well into this test for theories - without a descriptive mechanism it isn't predictive (so not testable).
Btw, bayesian methods aren't unheard of in science. Cladists use them to suggest models, so does cosmologists. But then those "best of breed" (likelihood or parsimony) models are testable with more data. ID isn't...
As far as ID and a characteristic of scientific theories:
First of all, we don't need a precise boundary for describing scientific theories to see that some things fall outside of the domain. In the course of trying to think of something which is obviously not a scientific theory, something to serve as a prime example, without having to worry about concepts like "testability" or "falsifiability", ID happens to serve very well. Whatever a scientific theory is, ID isn't one. (So ID does serve a role, after all, the role of being obviously not a scientific theory.)
Secondly, moreover, ID is not a theory of any kind. There are other kinds of theories, other than theories in science. There are theories in history, in the arts, in the law, and so on. And ID is easily distinguished from what makes a theory in any discipline.
For example, a theory ought to have some kind of coherent structure in which the intended goals of the theory (for example, the things were are going to be explained) are somehow connected to the starting points of the theory (for example, the explanatory factors). The connection may be logical, mathematical, causal, historical, or whatever. ID does not have such a structure. There is no connection, and there is no interest in thinking about any connection, between "design" and "the structure of the vertebrate eye".
For example, a theory ought to distinguish somehow between those conceivable states of affairs which are less coherent with the theory and those which are more coherent. Here I mean "conceivable" in some wider sense than "consistent", so that "there is a largest prime number" is a conceivable "state of affairs" which is inconsistent with number theory. A would-be theory, such as ID, which treats all conceivable situations equally, fails this standard.
But I would generalize even further, to have a kind of "theory of theories" which says that, to count as a theory, there must be something that we can say about a would-be theory which makes it somehow-or-other like other theories. A unique thing, which is totally unlike everything else that we know of as a theory, is strongly suspect as a candidate for being a theory. And I would suggest that it is up to the would-be theoreticians of such a would-be theory that they tell us how the would-be theory counts as a theory. Even if it fails the two tests that I suggested above, a most generous observer can still ask: "Well, if it doesn't meet those standards, then by what standards is it to be called a theory?" Or: "If your would-be theory is a theory, is there anything in the world which is not a theory?"
(In effect, applying the two tests to the "theory of theories" by which the candidate asks for admission to the company of theories. And, I suppose we can do this recursively - perhaps saying that "something is a theory if there is a theory of theories which includes this as a theory".)
I may be mistaken, but as far as I can see, ID not only fails to meet my suggested (minimal) standards for a theory, but also there are no suggestions for any standards for theories which ID would pass.
There aren't any white swans, all swans that appear white are actually whick, which is a colour that will appear white until 2010, at which point it will appear black. I recall Popper arging that falsifiability was necessary for a hypothesis to be scientific, but I don't recall him ever arguing that falsifiability was a sufficient criterion.
It is funny that ID-ers would try to lean on Popper for support. I guess they take their philosophy as seriously as they take their science. As Torbjorn points out, ID is not falsifiable. It "explains" everything even more glaringly than Popper's usual punching bags.
What does ID prohibit? What risky predictions does it make?
Popper's views of science may need refinement, but I think they can safely be used to leave ID out.
OK already! Point taken.
I think I shall break a lance for Popper for whom I will always have a soft spot, reading Popper alongside Stephan Koerner on the philosophy of maths at the age of 25 is what first turned me on to the philosophy of science. Previously I had been a professional field archaeologist with a passion for the history of maths.
Not for the first time John you have been moderately rude about Karl Raimund in one of your posts, I sympathise strongly with your point of view and mostly agree with your criticism but I still think that you are unfair in your judgement of Popper or at least overly harsh. All of the more advanced and more modern philosophies of science that you quote Lakatos, Giere et al. all have a solid Popperian core as Thorben writes:
Before Popper this was not the case, as Popper published Logik der Forschung in the 1930s it was still the belief, and had been since the 17th century, that there were true laws of nature and that science was capable of discovering those laws and proving their truth. Popper changed all that and although the philosophy of science has moved on considerably since then I think there is a modern tendency to forget how much the discipline owes to Popper and to be unfairly dismissive of his achievements.
Sorry about the mini-rant but I felt a need to ventilate.
I'm not sure if this is addressing my comment on studying science by its own methods (a "science of science"), since the rest of that paragraph eludes my reading. But if it was I can note that I doubt the result would be a comprehensive or complete theory, since this doesn't happen elsewhere.
Thorben that should of course read Torbj�rn!! Schei� dsygraphia! With humble appologies to Herr Larsson.
I'll try that again! "Thorben" that should of course read Torbjoern!! Scheiss dsygraphia! With humble appologies to Herr Larsson OM.
Thony, you need to use the html 4 code: ö to get ö.
On your entirely justified rant: you are entirely justified. Popper is a lot more sophisticated than the populist notion of falsification (and even that went through a number of refinements. But, and this is why I rag on him in a public forum - he really does leave discovery, testability, and classification out of science. As a demarcation principle between science and nonscience I have considerable sympathy for falsifiability, but as a description of what it is science does it is entirely inadequate.
To: Torbj?Larsson, OM
I wasn't directly answering your comment on "science of science", but we seem to have similar ideas.
Where I differ is that I was thinking of something covering more than just science.
And that ID not only fails to be a science, but is not a theory of any kind.
It's OK, in my opinion, to have a theory which is not subject to experimental investigation ("falsifiability"). In history, we can't do experiments to check on a theory of World War I. But that doesn't mean "anything goes" - if the "theory" equally well accounts for the outbreak of war, or for Russia and Austria-Hungary becoming allies, or for an invasion by space aliens, then it's a failure as a theory.
So I was suggesting as one mark of a real theory that some things are more coherent with the theory than are others.
And, as a check on that, a "theory of theories" would have to say that some would-be theories are better fits as a theory than are others: Not everything that we can dream up is a theory.
And, finally - if there is anything at all that is not a theory, then ID is not a theory.
There was a serious attempt in the 1980s in Germany, instigated by Jurgen Mittelstrass, to create a Wissenschaft der Wissenschaft (a science of science) a combination of philosophy of science, sociology of science and history of science that was intended to be more than a sum of its parts. As far as I know it died a death largely because of inter-disciplinary rivalry. Each discipline being more interested in establishing the supremacy of its own research than on co-operating with the others.
TomS:
I agree that testing isn't describing science well (see Wilkins comment #15). Whether it can be used outside empirical sciences is a matter of taste perhaps.
I could possibly conceive of testable historical theories, for example correlating climate trends with crops and national economy. It would probably shore up the reliability of an area. But OTOH history are based on possibly forged and often subjective documents. We would not be able to specify the uncertainty meaningfully anyway.
Thony C.:
Interesting, thanks!
I always thought that Popper's demarcation criterion should be called "Popper's Razor" (on the lines of Occam's better known shaving implement) because it can so easily cut the bad from the good science. But it is not all of the philosophy of science.