79. Science, Pseudoscience, & the Demarcation Problem | THUNK 

I am concerned about the problem of underdetermination. I have a series on it, but simply, it involves Kuhn's ideas to the next step, there is no way for us to tell of a particular errant result where the problem lies, be it in our experiment or one of the many beliefs that make up our background theory. We can never test a particular theory in isolation, merely in comparison to our background beliefs. This seems to make no theory falsifiable. Such a conclusion is in line with Kuhn's theories, but seems directly opposed to Popper's. If nothing else this is a clear point where Kuhn and Popper are miles apart if not a significant problem for falsification style solutions to the demarcation problem. Thanks for the video!

Sudo-Science made me laugh. Well done.

Kuhn is sometimes taught as a stepping stone to throwing out science all together or embracing new age or religious views. The irony is that science does evolve (making it more open minded) while these other systems seem remarkably static. Recent technology and AI are making teleological arguments come back in vogue. ( I like your episode on that subject) which is yet more irony since they are fruits of science.

I've been watching your videos since you started and I love them. Just wanted to let you know that. keep up the puns!

Quality content as always. Just want to let you know I get excited now when I see you've posted a new video. Strange that as I get older it's the people that can express and discuss great concepts and ideas coherently and concisely that brighten up my world, that have the capability to genuinely put me in a better mood and get the gears turning again when everything else seems to fail.I know you've addressed that you'd rather have quality over quantity in regards to your viewers but I sincerely hope you get both in the coming years. You deserve it Thunk!

I've always liked Kuhn's theory, since IMO seems to interpret the history of science very well. Science tends to believe it has everything sorted out and just kind of hand-waves away any anomalies. That is until there are too many anomalies and then a new scientific revolution is born. Yesterday's pseudoscience becomes today's science.

there is probably more than one difference between pseudoscience and science. Various traps to fall into, such as not being falsifiable; ignoring evidence; constructing experiments where the experimenter already has some bias towards a particular outcome. etc

I always love that the term "pseudoscience" appears to be used as often (and in the same contexts) as the word "heresy" has been by the Catholic Church. I am particularly entertained when some well-meaning "scientistic" person assures me that something has been "disproven by science" when there has never been a shred of research actually performed on the subject. I'm still looking for anybody who has done any peer-reviewed research on astrology and how it might function (and good astrology is no more, or less, predictive over the same timescale than meteorology.).

When I read scientific papers even if they make positive claims, I always read them as possibilities that are always possible to be disproven. I was always taught that science is falsifiable, so I see that it should always be seen that way. No model is 100% correct, we can always build on our current ones to the point of pushing all previous claims out of the water, as this could happen at anytime I don't see why anyone should take a paper as 100% empirical fact. So I side with Kuhn.

loved it

I find it entertaining that you cite the Mythbusters as I have watched many of their episodes because my children have enjoyed watching them blow things up. I have found it very instructional that whenever they have set out to "bust" a myth, they tend to play very fast and loose with environmental variables which provide for good (i.e. entertaining) television, but (IMO) poor science and when they want a myth validated, the variables are precisely accounted for.

I like Popperian logic and the analytic school, but lets say, if hypnotism is a somewhat empirical allusion to the notion of a subconscious, given it's success in the clinic and school yard magic tricks, surely then psychoanalysis is not totally unfalsifiable. I could concede that the psychology of object relations and developmental psychology could be seen as speculative areas for research and not scientific in method, but the method of psychoanalysis that posits and explores the unconscious makes use of the scientific method of phenomenology to secure depth. Otherwise one would be throwing the baby out with the bath water. Right about now I would reference "Against Method", by Paul Feyerabend, which has remarkably similar assumptions to the Lacanian School of thought in psychoanalytic theory. Slavoj Žižek is a concrete example of this, a philosopher moving beyond the landfill of historical textbooks, to go to the point where he may have conviction in the matter. He has oddly specific ticks and gesticulations of sort that allude to the fact that he is pent up with a historical memory of references to modern culture, movies and films, of the past and present, fact and fiction, and yet he avoids having an up tight personality in relation to them in the production of discourse with others using them as references and not for the purposes of other cultural studies, supposing that some subject of psychoanalysis is to be an emergent relation between people using arbitrary literature. THIS could otherwise be barred as an object of cognition, like jokes in a court room, or, as it is in reality, more or less excommunicated like Spinoza from the Dutch religious assembly. In short, this is because of the interpretation of psychosexual developmental theory, ruining the context of meta-psychological papers, for instance, the idea that knowledge construction could be an anarchic process. It would be hard to give a finite objective description of the scientific method without coming to terms with and solving difficulties in the theory of mind writ large, and problems like induction and hypothesis generation in the scientific method.

Are there any theories about the demarcation between science and math? Because according to both Popper and Kuhn, math is not science (which I agree with as a mathematician), but the line gets fuzzy in some circumstances. For example, my university claims fluid dynamics is math, a physicist friend of mine insists that string theory is math as well and though I myself think statistics /isn't/ math, other people think it is.

I hate to nitpick, but there is more than one demarcation problem. The demarcation problem between science and pseudoscience is one of the most well known, but it is not the only one. In philosophy, whenever there is a problem that concerns drawing a boundary on a spectrum between two categories, that's a demarcation problem. Another good example is the demarcation problem between observable and unobservable objects.

I think Popper called this "naive falsification"

"there's no shortage of disciplines that people claim are as good as or better than mainstream science like physics or medicine at prediction and accessing truth." Being employed in the field of archaeology (ok pun intended), I erm sort of make the claim this discipline is better than science at accessing truth (about prehistory).

Hey man, great video as usual! Wall of text incoming, hopefully it doesn't get too convoluted! Personally, I think the mistake here is a sort of semantic idealism. We have gotten so used to science being considered the "right way to find knowledge" that we've trapped ourselves into trying to determine a structural definition that encompasses a bunch of different things that don't actually belong together structurally. An analogy to make this a bit clearer: Imagine we had a word for all foods that are sweet. Let's call them "Dulcens". In this case, Dulcens are to Science as sweetness is to the effective acquisition of knowledge. Now, let's say that, one day, someone decides that they want to find the structural similarity between all Dulcens, so that they could easily determine which things were sweet and which were not without actually tasting them. So they bring in many fruits, some vegetables and an assortment of candies. Try as they might, they just can't find a structural similarity between all these things. There are some people arguing that only candies are true Dulcens, because they are all sweet, and that all other things are fake or psuedo-Dulcens. But others point out that this is a flawed statement, because there are a number of non-candies which are very sweet as well. From our perspective, the answer is pretty obvious: if we want to find a structural similarity between these things, it won't be based solely on the fact that they are sweet. That sweetness is merely a trait, and a functional categorization would be more appropriate based on the appropriate structures; define based on structure, not end result. But the people who use the world "Dulcens" are so used to using it as this end-all-be-all to sweetness that they want desperately to find a way to bring all the things which are sweet under one structural roof while excluding all things that are not. But this is an idealistic and unrealistic goal, because there are fruits which are not sweet and vegetables which are, which means that including a structure appropriate to fruits or vegetables will always make it over-inclusive. So, bringing the analogy back home, we are so used to calling all the highly effective methods of acquiring knowledge "science" that we've fallen into this impractical idealism which states that we can somehow put all these effective methods under a single name which also happens to represent a structural similarity. Popper's idea of science can be seen as the "candy" in our analogy, in that pretty much everything within it is a method that is highly effective at finding knowledge. But there are plenty of other methods which do not meet popper's criteria that also happen to be great at finding knowledge. These would be our fruits and vegetables, and we see them all the time in philosophy, sociology and economics. Even the "harder sciences" have some of these. Anyways, this leads us to the question, if we can't unite all these things structurally under one roof, how do we differentiate and define them? Well, there are a few options. The first would be to keep science as we know it in it's most basic form (stuff that works), but stop trying to define it structurally, and then have new names with which to structurally define all the different groups of methods. This allows us to keep the common usage of the term, but leaves us without it having a methodological criteria, meaning that it basically just becomes a category that we throw things into once we've confirmed they work. Another option would be to limit the use of the word "science" only to Popper's ideas, or perhaps to the scientific method, since these are the methods which have consistently been shown to work. By doing this, we have made structural criteria for what is science, while maintaining the idea that everything that is science works. However, we will also have to accept that there are things outside of science which work, and no longer be able to say things like "Once it works it becomes science". Whichever one we choose, we will not be able to have this idealized concept of science we often hold.

I think he also claimed historical materialism was a pseudoscience

The list you made on pseudo-science was great, until you mention the last item "a whole slew of weird stuff" that's good for your health. A lot of supplements have had (hopefully impartial) trials done on their affects, and simply haven't been FDA approved because the FDA has too much on their plate (pun intended). I agree a lot of stuff is marketed as "science" when it's not, but a lot of non-conventional medicines/supplements work, and have the scientific trials to back it up. You probably meant things like the metal wristbands that are supposed to cure arthritis, but I had to mention this since there are a lot of people who reject non-conventional medicines outright because they assume it's pseudo-science. Overall great video, it's a wonderful introduction into the Demarcation Problem.

Popper....should've been a balloon modeller.

Or we could take the Lawrence Krauss approach of "THE ONLY PEOPLE WHO READ PHILOSOPHY OF SCIENCE ARE OTHER PHILOSOPHERS OF SCIENCE!" *farts and inhales deeply*

Add dates.

Each demarcation seems to apply to a science that is somehow differentiated from the other. The former seems to be applicable to observations that are less complicated and require less inference; for example, there is very little “puzzling” over anything involved in Newtonian physics. With more complex phenomena that may require their own genre of science, puzzling is the better demarcation because their theoretical bases are more complex, and their hypothesis are constantly being tested beyond the scope of what one would do with the former group of sciences. For example, puzzling is useful for quantum mechanics, cognitive processes, social psychology, etc. In essence, the latter group of sciences, requiring puzzling as the demarcation, are still somewhat nascent and perhaps even inchoate.

God says, sudo fiat lux

There's a big problem though with established scientists calling everything that contradicts their theories a pseudoscience.

Hilarious puns!

Pseudo science everywhere.