The Demarcation Problem: Falsificationism

I have books on a wide variety of topics from philosophy to the social sciences to technology for sale on Amazon, Apple Books, and Google Play Books! All of my ebooks are currently discounted to $6. Just search either for Andrew Chapman or for The Autodidact’s Toolkit.

In this lecture, I show why falsifiability is NOT the demarcation criterion. I do this by discussing seven topics:

1. The Demarcation Problem
2. The Relationship between Demarcation, Methodology, and Progress
3. The Failure of Verificationism
4. Falsificationism
5. Falsifiability as the Demarcation Criterion
6. The Central Problem with Falsificationism
7. The Implications of the Quine—Duhem Thesis

[NS-woze]

I study at Oxford and your lecture is at this level. Awesome

[colins]

This was an absolutely excellent lecture and presentation! Outstanding. I love your presentation style. So clear. You're brilliant at explaining things. This has really helped me to get my head around this topic and to understand the critical problem with Popper's falsificationism - which before, I had a uncomfortable feeling about but couldn't quite put my finger on. It makes a lot of sense after watching this. Now I feel so much more confident in writing an essay which is due in a couple of days! Thank you so so so much!

[MRLJ]

Your videos are fantastic. I'm studying artificial intelligence and got interested in the philosophy of science and epistemology after reading David Deutsch's books. All your videos nicely clarify and separate out all of these positions.

Thanks

[rogerlindsay]

Is there something I'm missing here? In the example purporting to show that falsification is impossible, focusing on the issue of whether variation in mass affects the rate of acceleration under gravity, the lecturer tells us that this is a good example because the hypothesis that mass affects rate of acceleration is known to be false. If falsification is impossible, how could this hypothesis, or any other, ever be known to be false?

[Vadjhars]

This channel needs more attention. Subbed.

[geoffreyjohnson]

How do we rescue falsifiability so that it is useful? Can we not define falsifiability as the ability to gather empirical evidence against a hypothesis, not that a hypothesis is proven false using empirical evidence? Using this more appropriate definition of falsifiability, is a hypothesis falsifiable with the ability to gather empirical evidence so long as we make a no unmeasured confounders assumption? Do my ideas fall under some other philosophy of science already in existence? Is my rescued idea of faslifiability given a new name? Some type of neo-falsifiability? To me this neo-falsifiability seems to be the demarcation principle we are after.

[chefstevekirsch]

Outstanding lecture. I just subscribed!

[JonSebastianF]

Whoah, this was SUCH a well-written lecture :D

[Rnankn]

Science is a method of systematic inquiry, usually the empirical method of using evidence to describe reality. This video conflates fields of technical and professional practice with method to advance a political agenda defending the hegemony of natural science and scientific materialism.

[grantdillon]

If we take the Hume-ian view about knowledge and belief, that we ought to proportion our confidence in keeping with the evidence, what does this objection to falsificationism then say about our ability to be confident in any scientific knowledge at all? What set of criteria, therefore, would be used to develop confidence in any knowledge at all?

I suppose it was my impression that the entire point of falsificationism is that, while the mere status of something being unfalsifiable doesn't mean it's untrue, most of the fabricated, deceptive, and outright BS claims will be structured to be unfalsifiable. Thus falsification, if nothing else, acts as an excellent razor to rule out BS. The question therefore becomes, if falsificationism is an unworkable epistemological tool, how might we be able to find confidence in our determination between truth and lies, between fact and fiction etc?

[muhammadwaqas]

So good explanation. Thank you so much sir...🍮

[hilinayinager]

Thanks, Andrew. You're awesome! And funny

[climatedeceptionnetwork]

2:30 - By now I'm thinking that we're condemned to pragmatism. I will follow further. This is very good.

[ArtisanTony]

Motive is more of a problem than anything. The motive of those wanting a demarcation criterion is suspicious to me. I have this funny thought that those who want to establish the criterion, want to be the creator and holder of it :) This comes when science is used to gain more than knowledge. If the motive is anything more than knowledge, you can bet the criterion itself will benefit some and hurt others. We are human so motive is present in any case. You had a motive in making this video. :)

[woutv.m.]

Great lecture. Helped me understand a thing or two and taking it's place two new things to question.
No hypothesis is ever falsifiable based on empirical evidence. But what about other types of evidence? Can logical evidence pull through for example?

[zerge]

Great lesson. But the question remaining is, what IS the demarcation criteria then?

[maxpercer]

hi andrew, do you have a video on a *solution* to the 'problem of induction'. its easy to pose the problem of induction, but i have yet to see a good solution. saying we use induction out of habit as Hume suggested -because the past has been stable or regular enough for the habit of induction to produce good results - is a descriptive solution. it's like saying a particular tribe engages in cannibalism. it isn't a justification for inductive inference, i.e. why are we justified to think we have reliable knowledge about the future based on the past.

[yosemitejam]

This sounds like verification can never really be addressed either (22:22)
If you can not isolate a hypothesis then verification can’t be addressed either. Seems like you’re just taking one side.

[لالهوةإلالهوتي]

Usually a hypothesis is put to the test under a certain set of conditions that are not exactly reproducible in every subsequent application. With this understanding I could accept a hypothesis (and start applying it elsewhere) or reject it but I wouldn't be making absolute truth claims in either case. So if I falsify it I am rejecting it not saying it is the truth. That's what falsification means to me

[PTJ_S]

Falsification is necessary but no sufficient.