Are Theories Falsifiable? (Holistic Underdetermination)

[@louisng114]

It is like playing a sudoku puzzle with incomplete information.

[@CarneadesOfCyrene]

louisng114 Exactly.

[@Ansatz66]

This seems like a small problem being blown out of proportion. This isn't theories being shown to be unfalsifiable; it's just an overly high standard of falsifiability. Naturally when we say a theory makes a prediction we actually mean that a whole collection of theories makes the prediction, and if the prediction is falsified we have no way of knowing which theories are wrong, but that is not a philosophically serious problem. The important thing is that we have falsified the theories and we come up with a new collection of theories that matches our new evidence. Using a dark matter Vulcan to explain Mercury's orbit sounds bad, but if it actually allowed us to correctly predict the orbits of planets then there's no real problem. I'm no astronomer, but suppose that dark matter Vulcan and Newton really would give us a perfect description of all the movement of the planets for all time down to the finest measurements. In that case, I think it's fair to say that dark matter Vulcan is real from a certain point of view. There will always be multiple ways to explain and accurately predict the universe. If dark matter Vulcan is one of those then choosing to use it in your calculations is just a preference. The fact that there are multiple ways to fix a falsification does not make theories unfalsifiable; it's just the sort of choice that scientists deal with every day. This is what Occam's razor was designed to deal with. Keeping things simple until forced to make them complicated isn't a law of nature; it's just a rule of preference.

[@CarneadesOfCyrene]

Ansatz66 I think that the worry that I have is that too often science is used to claim that some specific theory has been shown to be true or shown to be false. If you are saying simply that scientific theories aren't there to be true or false just to be useful, then I have no qualms with you. You will run into problems with the demarkation problem from my understanding. How do you separate science from pseudo science without falsifiability? Why can't we posit supernatural entities to solve our contradictions?

[@Ansatz66]

Carneades.org "If you are saying simply that scientific theories aren't there to be true or false just to be useful, then I have no qualms with you." Science isn't about being true, false, or useful. No amount of testing will ever establish something as true or false. Instead of looking for falsity, science looks to find repeatable experiments that demonstrate a failed prediction of the theory. That's not the same as being false since the experiment could always be a hallucination or conspiracy. Instead of looking for truth, science just repeatedly tests theories in every imaginable way. If some credulous people start believing a theory just because it has survived years of testing, then that is their own business and has nothing to do with science. As for being useful, science can sometimes be aimed at practical goals, but it is often aimed at testing theories which will probably never have any practical value. Science can be its own reward, especially in fields like astronomy. "How do you separate science from pseudo science without falsifiability?" The kind of falsifiability described in the video is too strict if all you are trying to do is separate science from pseudo science. Pseudo science never makes predictions that you can test, so the issue of determining which specific theory caused a prediction to fail never arises. For the sake of demarcation, you can just define falsifiability like this: a theory is falsifiable if it makes a prediction that can be tested. Forget demonstrating the theory to be false; that's both impossible and more than we need.

[@CarneadesOfCyrene]

Ansatz66 At the end of the day, i think I really agree with you on this. I think science is best framed as a way to poke and prod at ideas and preconceptions. Really my objections go towards the scientific realists that are saying that science is stating something true about the world. I will have to think more on the demarcation problem, though that sounds like a sufficient answer for the moment.

[@canismajoris6733]

No, because you would need to prove that it actually is made of dark matter and exists. Otherwise you are just messing around with a mathematical model

[@ahmedbellankas2549]

1- If theory t and assumptions a1 and a2 then d; 2- it's not the case that d; 3- therefore, either not t or not a1 or not a2. 4- we know that a1 is true and a2 is true; C- therefore, t is false. - it seems that if we verfied a1 and a2 before verfying t and their truth value is keept constant at the time at which we test t and we know all that then we can test t in isolation. - so, it seems that if we take time into account and the fact that we verify our theories and others assumptions at different times then p3 is false.

[@gdn5001]

My only issue with the videos in the series is that while the problems of induction and underdetermination are a big problem for science, there also big problems for everyday common “knowledge”, history, and quite a few other fields as well. I feel like presenting them as solely an issue for science may be problematic as this will lead some to use the arguments exclusively against science without realizing that they’re also attacking other facets of what they hold to be “knowledge.”

[@uptide1214]

I'm not sure it's as big of a problem in how other fields are perceived at least around the people I know. They will be dogmatic about physics and in the same paragraph say that we have incomplete information about, say, history, and how it's simply our best guess. For the people around me, they will question tradition and religion before questioning science.

[@gdn5001]

@@uptide1214 Traditions and religion are nonsense.

[@frankconley7630]

Dude that was lovely. Thank you so much.

[@theultimatereductionist7592]

Please replace the word "theory" in every instance of this video with the word "statement" or "assertion". There exists nothing about "theories" that make them any less true or falsifiable than statements or assertions.

[@duder6387]

Yes that’s the problem. We can doubt even the foundations of logic before we doubt a theory.

[@juantubec]

Wait, so why replace it if it's not any less falsifiable? What does the difference in word choice make here?

[@MicahIsser]

Would a system of logic based on probability functions (like a Bayesian statistics) or a fuzzy logic have an easier time accounting for this problem of under-determination? Even if one piece of evidence can't falsify a theory, we still might want to quantify our process of reasoning. Can the process of 'losing confidence' in the correctness of a scientific theory (such as in Kuhn's sense of an old paradigm) be formalized through a mathematical notation?

[@CarneadesOfCyrene]

I have an entire series on Bayesian epistemology and how we can formalize belief change. While it may solve some problems, it has a few problems of its own: kzbin.info/www/bejne/j4PdaZebnq9obac

[@ObeySilence]

Modus tollens is A is B, C ist not B, therefore C is not A. How do you apply that to your example at 2:45? I´am not so versed in logic, but can you tell me how you make the variables match the formula?

[@CarneadesOfCyrene]

To be clear, what you state is not Modus Tollens. What you seem to be referencing is from categorical logic: All A are B, No C are B, therefore No C are A. This is a valid form of categorical logic, but not what we are referencing here. In this we are talking about propositional logic. Categorical logic deals with the parts of a sentence, while propositional logic deals with the sentences themselves. Modus Tollens says: P Implies Q, Not Q, Therefore not P. If you want a refresher on logic, check out my series on the 100 days of logic: kzbin.info/aero/PLz0n_SjOttTcjHsuebLrl0fjab5fdToui

[@davidlilley4637]

This video and the one you made on the cosmological principle are mischievous slights on Popper's epistemology. Popper's position is that knowledge grows by trial and error. There is so much rubbish spoken of epistemology. You can find many sites on KZbin that explain epistemology that come across as lecture notes passed down from generation to generation unchanged in centuries. "There is empiricism and rationalism, take your pick." Not so. Knowledge doesn't begin with an observation or an idea. It simply begins with a problem and follows the course P1, TS, EE, P2 repeat (where P1 is the initial problem, TS is the first tentative solution, EE is testing or error elimination and P2 is where you are after testing, when you have a better understanding of the problem and what doesn't work and what your next tentative solution, TS2, might be). Our knowledge of the world we live in and of ourselves grows via P1, TS, EE, P2 repeat and in getting from P1 to Pn we use thinking, observation and testing. Imagine how many labs, TSs and EEs have been involved in eliminating Ebola from West Africa. Solving the Ebola problem. Philosophy does BIG and its biggest problem has always been the epistemological problem "1. how does knowledge grow? and 2. how do we distinguish between truth and falsity?" The answer to the first question is via the scientific method P1, TS, EE, P2. The answer to the second question is via the "EPISTOMOLOGICAL ARSENAL" which includes logic, rational argument, nominalism (never ask "what is questions") and the falsification criteria to exclude pseudo-science. The falsification criteria is a quick knockout for all the quack pseudo-sciences like astrology and psychotherapy but we could go from the special theory to the general theory by replacing the word falsifiable with the word testable. There are three kinds of knowledge; belief, opinion and objective knowledge. No. There is only objective knowledge. Knowledge that has pasted the truth test, "correspondence with the facts", until it is superseded by a better explanation with greater truth-likeness, universality and explanatory value. Comments welcome.

[@CarneadesOfCyrene]

+David Lilley I have a number of concerns here. Let's take a look. First, it seems to me that there is an important distinction to be made between actions and beliefs. Between things that are done and things that are known. Usefulness and truth. Simply because a particular theory is useful to solve a problem, it does not mean that it is true (kzbin.info/www/bejne/o6DIkniDd52Gb7M), or that it is known. Second, you seem to completely ignore the failings of the scientific method. The problem of induction is a great place to start (kzbin.info/www/bejne/qZWblKuOm7V9i5Y) but the new riddle of induction (kzbin.info/www/bejne/oXioY3ydZ8mpl6c) and this series are equally problematic objections. You claim that we must use the scientific method to arrive at knowledge. But how can we justify the claim that the scientific method can lead us to knowledge? Using the scientific method itself is flatly fallacious (kzbin.info/www/bejne/f3uVlYqofpKDiqM) but this is the only tool you have offered for us to grow our knowledge. Third, logic itself is not above scrutiny (kzbin.info/www/bejne/jp-Tf6aVbamLd8k and kzbin.info/www/bejne/g6Svgn59msSUn9k) it seems to me the case for the rationality of science is pretty weak (kzbin.info/www/bejne/bHLZinaKh91qaLM), nominalism is an equally debated philosophical position, and as stated in this video, no theory is really falsifiable. My biggest concern here is that you fail to address the very arguments against your positions offered in this video. How do you think that anything can be falsified? how can we tell if it is our TS or our background beliefs that are at fault if the EE does not back up our theory? Finally, we generally consider knowledge more than just true belief (kzbin.info/www/bejne/mam6qXqPi7t5gJI), it has to be justified or warranted in some way. And even so, there are a number of problems for correspondence theories of truth, check them out here (kzbin.info/www/bejne/q5-TfJV9m9iXi68)

[@davidlilley4637]

+Carneades.org This is fantastic. Such a detailed reply. You give me a lot of homework to do before I can give a full reply. But my first pass has to be that all that you have stated above is incorrect. I hope that you like arguing.

[@davidlilley4637]

+Carneades.org I have now watched a number of your videos and not been impressed by any of them. You just appear to be making a pseudo-intellectual attack on science and its big brother epistemology whist giving the impression of omniscience with your bountiful use of big words and concepts.When you state that knowledge is justified, true belief (Aristotle not Plato) you are jumping back in time 2,400 years. We stand on the shoulders of giants and see further. Today, post Popper, we don't talk about belief or subjective knowledge in epistemology. We only talk about objective knowledge which is mostly scientific knowledge that has survived critical testing, corresponds with the facts and deserves a place on the shelf marked truth. But it is never absolute truth but just the best description we have at the moment. It may be replaced tomorrow by a better explanation having greater truth-likeness, universality and explanatory value. Any guess, theory, hypothesis, explanation or whatever you wish to call it is false or falsified if it doesn't correspond with the facts. We don't have background beliefs because we don't do belief. All the background explanations and data that we rely on is also objective knowledge. Not assumptions but truths. If, as you state, "logic is not above scrutiny", with what will we scrutinise it with. Logic and maths are one and the same thing. All maths is logical. But we can "prove" and have absolute certainly in maths (2+2=4) whilst we can only have answers having a high degree of truth-likeness in science. Nominalism is just accepting that words are labels that we create to reduce a description to a single word. We have to be nominalist to avoid the pitfalls of essentialism where you end up talking about words rather than the world and treat knowledge as encyclopaedic and not in progress. Philosophy does BIG. Why would they debate labels? Why do you accuse science of using inductive reasoning when Hume junked induction in 1739? And why would anyone introduce a new riddle of induction when induction is already history.When you state that the scientific method is flatly fallacious please remember that it is nothing more than trial and error and we only learn by trial and error.Thank you for your time. Comments welcome.

[@davidlilley4637]

Russell/Whitehead's "Principia Mathematica" and Gödel's incompleteness proof are far removed from my stating that all maths is logical. "Gödel's incompleteness theorems are two theorems of mathematical logic that demonstrate the inherent limitations of every formal axiomatic system containing basic arithmetic." I only discussed the scientific method, the truth-likeness, explanatory value and universality of science (problem solving in general) and not the usefulness of science. Apparently only 200 people have read Principia Mathematica and probably most did that some 100 years ago. Its only readable by logicians and the same with the Wikipedia reference you make (loaded with the word proof). For us mortals, we use maths everyday to "prove" this or that. Maths "proves" whilst trying to understand the real world in science or everyday problem solving can only ever achieve "best truth-likeness" for the time being. Lulie Tanant has the best KZbin on epistemology.

[@werrkowalski2985]

@@davidlilley4637 You have committed almost every single fallacy in the book.

[@drinkmug]

Interesting video Carneades.

[@CarneadesOfCyrene]

twightlight3 Thanks!

[@theultimatereductionist7592]

All you asserting is that "all approximations are false" in formal logic, because they are approximations. Please watch c0nc0rdance's video on "All Models Are Wrong". I hate that title. But what he says in that video is 100% correct. A map IS NOT the physical thing it describes. In order to have the map EXACTLY describe reality, the map would HAVE TO BE reality.

[@duder6387]

I know this is a little late but I have some responses. The issue is that we cannot even know if maps or models are approximations of reality. In order to know that we’re approximating reality we would have to know what reality is, which it seems we don’t. For example, how do we know that the map of the town in the video is an approximation of the town? Perhaps our senses deceived us and our map is completely wrong, perhaps the town changed shape while we were making the map and now the layout has changed completely, perhaps an evil deceiver made us think there was a town where there was none, and so on and so forth. The point is that we can never know if our models correspond to or approximate reality due to an infinite number of skeptical scenarios that prevent us from accessing reality itself.

[@AbCDef-zs6uj]

Is it just me, or does this seem something like a consequence of the rule-following paradox from Kripke/Wittgenstein?

[@yunoewig3095]

Regarding P2): What if we consider a certain list of scientific theories T1, T2, …, Tn, which is presumed to be an exhaustive list of all possibilities regarding the state of the world. A particular theory T can only be tested in relation to one of these other theories, and each of them will provide a different interpretation for the result of an experiment concerning theory T. If we could, however, establish that, under any one of those interpretations, theory T must be false in case the experiment in question produces a certain outcome, then T will have been shown to be falsifiable, and its tested falsehood does not depend on the truth of other theories, even though it cannot be tested in isolation. In fact, we had to consider theories T1, T2, …, Tn in order to be able to interpret the result of an experiment; however, the resulting falsehood of theory T depends in no way in the truth of those theories, for it has turned out to be independent of the particular interpretation that is given to it (although we had to consider those interpretations in order to arrive at this conclusion in the first place).

[@CarneadesOfCyrene]

Daniel Sampaio Interesting argument. I doubt that any such exhaustive list could ever be produced. Or at the very least such a list would be infinite. We are the bottom layer of a set of 5 virtual realities. We are the bottom layer of a set of 10 virtual realities. We are the bottom layer of a set of 100 virtual realities... All the explanations would be different, but they easily constitute an infinite list. We could do the same with ED scenarios and so on. I do not know how it would be possible to compare a theory to an infinite list of possibilities.

[@yunoewig3095]

Carneades.org «I doubt that any such exhaustive list could ever be produced.» I also doubt it, and even if it could be produced, I do not see how one could prove it to be so. I just offered this argument because I came up with it and thought it might be interesting. I am glad you agree with me.

[@CarneadesOfCyrene]

Daniel Sampaio It is definitely an interesting argument. And if you could ever show that such a list could be produced or in some why conclude that there were a finite number of theories, I think it could hold water. But until you can put that premise in place, it's just an intriguing exercise. :)

[@keystothebox]

I disagree with premise 2 in the first example requiring other theories. Theories can purely stem from epistemology. Even in the case with theory dependency (later ex p1), they can and are tested as a lower independent unit which gives traceable root cause to what Theory is inaccurate. In any case suspension of belief until there is reasonable evidence where truth is important is the only rational course. Additional premises such as adding Vulcan without evidence of its existence is irrational. I love your mathematical grasp of logic, and it looks like you are pretty solid there, though that is not the stumbling block I see you running into.

[@thisismyname9569]

Yes, theories are falsifiable. The auxiliary assumptions are part of the theory, though some may be implicit and not explicitly known by the tester . If experiment or experience disagrees with the theory, the theory is falsified. The question is then which PART(S) of the theory are wrong. If I drop a ball and it doesn't fall to the center of the earth, that doesn't falsify gravity, because of auxiliary hypotheses involving floors which are part of the theory. Any theory not including floors AS PART OF THE THEORY would be obviously false. That there is no Vulcan is part of the theory. If it is not found, that theory is falsified. That it doesn't exist as dark matter was part of that theory, which the astronomers of the time probably weren't aware of at first, but after some thought, might consider changing the theory to allow it, and then test the new theory. Hey, you asked.

[@touyubeusr]

Self-contradictory theories can be saved by thinking classical logic (which is just one more theory) isn't true.

[@CarneadesOfCyrene]

touyubeusr Exactly! Underdetermination is a real problem for any theory.

[@myothersoul1953]

No it can't because if logic doesn't work, then reasoning fails and nothing or everything can or can not be concluded from any anything else. That is nothing is reasonable.

[@georgesimpson1406]

We can save them all! "We are in a simulation with malevolent aliens causing illusive observation and thought!"

[@afacere736]

This is a massive straw-man presentation of Popper - he in no way said that falsifiability distinguished scientific theories from non-scientific theories. Popper acknowledged, that there are many sciences that were less capable of being falsifiable - only these were weaker sciences because of that. Likewise, Popper was not saying "a theory is scientific _if,_ and *only* _if,_ it is falsifiable" - that, again, is a massive misrepresentation of what Popper was saying. Popper's falsification must be seen in the context of what he was trying to achieve, which was a science that does not rely on induction (he thought Hume broke induction). Popper went on to devastate what he called historicism (in "The Poverty of Historicism" and "The Open Society and its Enemies"). He tried to paint a science that was purely teleological and therefore focuses, not on the historical conditions that give rise to the theories but rather, on discovering the unintended consequences of what theories are trying to do: make predictions. Falsification was never a solution, it was a perpetual method to search for any unforeseen consequences of scientific theories, an unending quest. This was the foundation of his position of critical rationalism. Popper was never talking about theories as being shown to be true, he sae scientific theories entirely as "the best we've got so far." Falsification was never about proving anything, it was always about becoming less and less wrong. Popper saw science as a torchlight in a dark forest, not beads pulled out of an urn. I'd suggest reading up on him a bit more, your skepticism might get on well with his style. So that "there's no way for us to tell which of those assumptions we should thow out" is made immediately obsolete with the next step of falsificaiton. And so on, and so forth. And on marches science.

[@y0n1n1x]

i understand your concerts on how popper actually deviates from what carneades called popper's definition of science, but if I just choose to ignore popper and simply work on the definition, I am still philosophising, and understanding the argument on it's own, and not popper.

[@y0n1n1x]

quote by popper: "In so far as a scientific statement speaks about reality, it must be falsifiable; and in so far as it is not falsifiable, it does not speak about reality."

[@ziggydeath9397]

Premise 3 is false; it is not the case that a theory cannot be tested in isolation. (Physics is not the only science.) Also, it is a false claim that *"G-d exists" is not falsifiable* and therefore unscientific. It depends upon the definition of G-d. If G-d is a physical being that exists at a certain longitude/latitude on earth and does not move from that position, that is a falsifiable statement. The definitions of G-d are manifold, some are falsifiable and some are not.

[@CarneadesOfCyrene]

What scientific theory can be tested in isolation of the background theories and data? You get a contradiction by having two pieces of information with conflict. You can always preserve one piece of information by discarding the other. Which you discard is simply a matter of personal bias. Let's demonstrate this with what you claim is a falsifiable statement: that "G-d is a physical being that exists at a certain longitude/latitude on earth and does not move from that position". This is certainly not a falsifiable statement since even if we arrive at a contradiction, we merely need to doubt some of our background beliefs in order to dispel the contradiction and leave the theory intact. So say you go to test this theory. You go to that longitude and latitude and you see nothing. Here are simply a few of the claims that you could doubt: That your devices can correctly calculate longitude and latitude. That all physical beings must be visible. That God is a human sized object (could be so large that while this is the center of God, the entire universe is in a gap between two of his smallest particles, or so small that no device we possess can detect him). That God is not visually identical to something else at that location, say a rock. That you understand the actual meaning of all of the words in that statement. That you are not dreaming That you are not in the matrix That you are not systematically and pervasively deceived. Just to name a few. The point is that there is no way to claim that it was the theory that we in fact falsified and not one of our background beliefs.

[@ziggydeath9397]

It is falsifiable regardless of what you do after it is shown to be false. If you revise your theory - as often happens, like Newton's first law, which was revised. "The law of inertia states that an object at rest stays at rest and an object in motion stays in motion with the same speed and in the same direction unless acted upon by an unbalanced force." It was changed to include "unless acted upon by an unbalanced force." It is not a personal bias. You may revise your theory like this one was, but every time that happens it becomes less scientifically strong. Besides, there are other realms of science - for instance what I practice: document analysis - where theories are not based upon other theories. They do exist in isolation and are falsifiable. The data set is limited and closed, unlike a universe where you may find out something about the extra small and the extra distant that you do not know now. In my line of work, you start out with all the information you need, and you develop theories about the closed data set you are working with. Those theories can be tested in isolation, and they are all falsifiable.

[@7DYNAMIN]

Good vid senpai!

[@COEXISTential]

I'm not a philosopher, much less a logic-chopper, so bear with me: If a given theory relies on the falsifiability of other theories to be falsifiable, and you can't know which one is incorrect, wouldn't that be true of all such theories, and as such the issue is not that it's not falsifiable, just that you don't know how to falsify it? But you allowed for beliefs to underpin a theory, as well as theories. These would be regress-stoppers (axiomatic/a priori), wouldn't they? It seems unlikely that any given theory would rely on two theories (or a theory and a belief) about the same constituents, for example, E=mc^2 is unlikely to rely on two theories/beliefs about 'c', and thus you can test the relevance of 'c' to a falsification of E=mc^2 (and likewise for 'E' or 'm' in isolation). What other theories underpin the belief that the speed of light in a vacuum is the fastest possible speed? And if we break those down (assuming they exist), do we get to a bedrock regress-stopper? Bearing in mind that we are where we are, scientifically, by holding our assumptions about theories constant and building new theories on top of them, and then tearing them down and starting again when all consequent theories have failed - indeed isn't it the case that when all logical interpolations from a cluster of theories have been exhausted that those theories are then up for being refined?

[@CarneadesOfCyrene]

COEXISTential This sounds like a response that Thomas Kuhn (Thomas Kuhn vs Presuppositionalism) would have been proud of. The question would be, how can we know if we have exhausted all logical consequences of a set of theories? When does the paradigm become so weak that it must shift? This problem will be stated more explicitly in the next video, but that's the basic problem.

[@COEXISTential]

Carneades.org I look forward to the follow-up, then. Thanks.

[@GEdwardsPhilosophy]

If they're not truth bearers in the first place, then no they're not. Although I don't agree with Popper's demarcation either.

[@MaximilienDanton]

In what way do you disagree with Popper's demarcation Gary?

[@GEdwardsPhilosophy]

Tjaart Blignaut I don't think falsifiability is sufficient for the science / pseudoscience distinction.

[@MaximilienDanton]

***** if you have written or made a video about it I'd be interested to hear/read your thoughts. If not, consider it a request for a future video if you're interested. :)

[@CarneadesOfCyrene]

***** Interesting. Do you have another way to distinguish way to solve the demarcation problem?

[@GEdwardsPhilosophy]

Carneades.org I outline my approach to demarcation about three minuets into this: Spirit Science? Shit Science!

[@TheRealisticNihilist]

I would take the digestion that newtons mechanics are still correct but they don't predict this phenomenon to be an ad hoc hypothesis.

[@CarneadesOfCyrene]

***** Why? Why should you deny them and not the claim that there is no such planet as Vulcan?

[@TheRealisticNihilist]

Carneades.org because it's posited ad hoc. I'm not saying that the hypothesis would be false, merely that it's not interesting unless it makes novel predictions which lens it to falsification as well.

[@CarneadesOfCyrene]

***** But perhaps the prediction that Newton's laws make is simply that there is another planet there. It was this very problem that led to the discovery of Neptune (the lack of a planet beyond Uranus could not explain what Newton's laws predicted, therefore the claim that there was not planet beyond Uranus was false). It does not seem ad hoc, if the actual prediction of the theory is that there is other matter there that affects the orbit.

[@TheRealisticNihilist]

Carneades.org Thera a confusion somewhere, I'm not sure where, though. Given a set of observations, we ask if the predictions made by a preceding hypothesis are compatible with this set of observations. If they are not compatible, we say the theory is falsified. Any positing of an unobserved phenomenon is ad hoc until the predictions made by the hypothesis are corroborated through observation. Positing a planet there would be as ad hoc as the theory of relativity or whatever, but we decide which is accurate based on the predictions of the new hypothesis. So, we say if relativity bends space, we should see gravitational lensing, but if that is unobserved yet posited to explain a particular phenomenon, then it's ad hoc as well. When the new hypothesis is introduced, it will falsify the old only if a set of observations predicted by the new are incompatible with a set of observations in the old.

[@CarneadesOfCyrene]

***** "Given a set of observations, we ask if the predictions made by a preceding hypothesis are compatible with this set of observations. If they are not compatible, we say the theory is falsified. " What shows that the theory is incompatible with the observations? The only thing that can do this is our other theories. To get deeper into the problem, what shows that our observations represent the way that the world is, only our other theories. We could just as easily throw our either claim and remain consistent. "Any positing of an unobserved phenomenon is ad hoc until the predictions made by the hypothesis are corroborated through observation." Your claim that we can't throw out certain theories in an ad hoc way, is just another theory in our web of beliefs. It's just as subject to doubt as any of our other theories. Why not throw it out instead? Also, note that we are not adding new theories, we are taking theories away. We are not positing that there is another planet, we are getting rid of the assumption that there are only 5 (or however many we assume there are at the time). Only with the assumption that there are only so many planets will the denial of the conclusion follow. The claim that if relativity bends space we should see graviational lensing, is not just based on the theory of relativity. And our assessment of whether or not we do see gravitational lensing is based on many more theories to boot. I'm going to deal with this problem more in the next video, but suffice to say that ampliative principles like not allowing ad hoc theories are insufficient to give us reason to throw out one theory over another, as there is no reason that we should not instead throw out the ampliative principles. If that does not make sense hopefully the next video will help out.

[@pawekopytek7596]

Do you also need to test the hypothesis that science is unverifiable in isolation from other hypotheses? If you really wanted to preserve the conclusion that science is verifiable, you could just throw out one of the premises and say that it must be false then.

[@ernestamoore4385]

Why is it that philosophy and logic spend great amounts of words to transmit a simple idea? The point of your whole video with fancy notation is that "no theory is falsifiable because theories always depend on the falsifiability of other theories..." just common sense. However, you conclude that "science doesn't work because theories can't be falsified". That is a wrong premise to start with. Science is more than falsifiable theories. How did we get to achieve discoveries like evolution, gravity, etc. if "science doesn't work"?

[@CarneadesOfCyrene]

The concern is that usefulness does not imply truth. Just because using a theory lets you do something, it does not mean that that theory is true.

[@ernestamoore4385]

No, the theories I refer to are beyond useful, they have been supported by data to be the best approximation to the truth. Science is about approximating truth because in science nothing can be proven 100% to be true. If you take off from this premise, then science does work. If you take off from the premise that science is about finding the truth, they science definitely doesn't do that (=science does not work). So it depends how you define the goals of science.

[@myworldmusic3426]

I totally agree. The logical representation makes it totally boring and confusing for those that are not familiar with first order logic. You have sacrified supposed rappresentational rigour for a clear intuitive explanation of the problem. The video was good until you start representing it in FOL.

[@quleughy]

Melantis Obtrusis He has already said that science is useful. He never said it wasn't. But how do you go from that to the idea that science approximates truth? This is the problem he posed to you. It appears to be the case that we can conduct a number of experiments upon which finding consistently verifiable results will allow us the useful ability to make better judgments based on the understanding we gained from them, but we can never truly know if that which is made useful to our understanding is the same as some objective truth to be learned out there. Consider the problem of induction: no matter how many observations one makes, at the end of the day they don't guarantee a correct assessment of what's going on; in fact since our understanding within science is completely dependent on the framework of the current paradigm within science, what we may be so sure is an approximation of truth for the science of our time may be completely laughable to future scientists working under different but more useful assumptions.

[@joaop.1238]

The point is that science does not "approximate truth". Science is just a set of beliefs that "works" to do some useful stuff. Anybody claiming more than that is being overly optimistic. The point of using logic was just to show that the only point that can be questioned in his argument is the premises themselves, which he defended. Or you could question logic itself but we know that that won't lead to anything productive.

[@gdn5001]

The only way I can see to solve problems of scientific justification are ampliative characteristics such as predictive power, fit with prior assumptions, and simplicity. The issue is that those ampliative characteristics also require justification. Or I could just make a Moorean argument for them and call it a day. Or I could toughen up and just bite the skeptical bullet. It seems at some level you either have to be a moorean or a skeptic. Ugghhhhh

[@gdn5001]

Like I feel like I can solve the problems with science by appealing to common sense at some level and working my way up, but that’s beside the point in conversation with a skeptic isn’t it. 😕

[@theultimatereductionist7592]

Mathematical modeling, not mere words & philosophy, is the only way to accurate describe physical reality.

[@unknownknownsphilosophy7888]

Tell that to Kripkenstein. How do you know your doing addition and quaddition when you do math? Look into Kripke's rule following argument and see if this comment you made makes any sense at all after you've understood it.

[@Google_Censored_Commenter]

P3 at 7:40 is false. It is very easy to test theories in isolation, IF you are a realist. Since realism is not a theory, is a necessary axiom. There is no alternative to it. And all the axiom states is that there must exist a world outside our minds, independent of us. So, if I observe a rock, I can make a theory that this rock exists in the world defined by our axiom. I might not know any other properties of this rock, but I can say it exists, because the axiom demands it. Next, I can make the observation that rocks interact. Somehow. Quantum Mechanics, gravity, friction, you name it. Now these are all scientific theories, and they might all be wrong you say. But all of that doesn't matter. It is undeniable that they interact, even if we cannot explain the why. And so there we go, we already got 2 theories that have been tested in isolation. There are probably more. And with those we can test every other theory no problem, since they're necessarily true. We don't have to doubt anything.

[@duder6387]

I know this is a year old but I have some responses. The axiom “there must exist a world outside our minds, independent of us” is a background belief and therefore it is impossible to test any theory in isolation with this axiom. If you use any axioms you are including background beliefs. Here’s the Stanford Encyclopedia of Philosophy: “ The idea here is that a single scientific hypothesis does not by itself carry any implications about what we should expect to observe in nature; rather, we can derive empirical consequences from an hypothesis only when it is conjoined with many other beliefs and hypotheses, including background assumptions about the world, beliefs about how measuring instruments operate, further hypotheses about the interactions between objects in the original hypothesis’ field of study and the surrounding environment, etc. For this reason, Duhem argues, when an empirical prediction turns out to be falsified, we do not know whether the fault lies with the hypothesis we originally sought to test or with one of the many other beliefs and hypotheses that were also needed and used to generate the failed prediction” Also it’s very easy to doubt the axiom that there’s an external world. Some examples of philosophers doubting the external word include Rene Descartes and George Berkeley. Also it is very much deniable that objects interact with one another. See David Hume. He found that it is impossible to determine that one thing causes another, we can only observe correlations between the two. For example, if I press my hand against a box and it moves I cannot conclude that my hand caused the box to move I can only conclude that there is a correlation between my hand pressing against the box and the box moving. Saying that my hand caused the box to move is unjustifiable.

[@Google_Censored_Commenter]

@@duder6387 You seem to be confused. Axioms are not supposed to be tested. They are necessary assumptions we make in order to draw other conclusions. Using background knowledge isn't a problem, actually, for it is necessary if you want to make any kind of claim of the world. "I think therefore I am" can be discarded as relying on background knowledge, namely, that assumption that the "I" is constant. That the thinker is the same I as the one that "is". As for causation being ultimately unjustifiable? I disagree. Deductively, sure. But deduction has never been useful in generating much knowledge anyway. The way we actually learn about causality is by looking at multiple correlations, not just one. So, if I push the box with my hand, and it moves, it might be that I moved it, or it might be that something else did. The wind. The rotation of the earth, a ghost who wants to mess with me. What we do is employ the falsification principle - we attempt to falsify the claim that my hand caused the box to move, and if we can't, we accept it. This acceptance is again not some "ultimate necessary truth" that stands forever and until the end of time. Merely until it is falsified. This means that we can never attain the absolute truth of the world, for it may always be falsified in the future. And some truths may even be unfalsifiable, but wholly unreasonable for us to believe, so we're forced not to commit to them. For instance, even if we did accept my hand made the box move that one moment in time 10 minutes ago, maybe reality is such that it is a unique one-time event. My hand can never, ever, move the box again. So despite it being a true causation, it is unfalsifiable, we can never know its truth, and as such are forced to not accept it. Too bad. But luckily reality is not so cruel, at least as far as we can tell. It's rather persistent from our observations. It passes virtually all of our intersubjective tests, suggesting we do live in a shared reality. Not just at one point in time, but all the time. (until now) Say hundreds of us flawed humans with our subjective quirks and biases and interpretations walk into a room together. We see the box in the room. We all confirm we see the box. We can all touch the box, hear sounds reflecting off the box. Measure the box's weight and so on. All of our minds receive the same data. Then we go to sleep, lose consciousness, wake up. All still in the same room. Same box. Same experiences. Despite our flawed subjective experiences, nontheless we are forced to come to the conclusion that the box must exist in an objective reality independent of us, but which we share space in. I do not see what falsifiable proposition you can make, that would falsify the shared reality hypothesis, that hasn't already been tested and ruled out for the time being. Ultimately, like I said before, background knowldge isn't problematic to invoke, as long as we recognize it is tentative in nature.

[@duder6387]

“ You seem to be confused. Axioms are not supposed to be tested. They are necessary assumptions we make in order to draw other conclusions. Using background knowledge isn't a problem, actually, for it is necessary if you want to make any kind of claim of the world.” That’s the problem. I know that axioms are assumed to be true, however that makes them unjustified. Carnedaes.org points this out in a lot of his videos. If we are allowed to assume certain propositions or definitions to be true then there is no limit to what we can assume to be true. I could create an entire system with the axiom “I am omniscient.” This is somewhat addressed in this video from 10:36 to 12:04. Where he tests the theory “God does not exist” against the axiom “The Bible is correct” and thereby falsifies the theory that God does not exist. “ As for causation being ultimately unjustifiable? I disagree. Deductively, sure. But deduction has never been useful in generating much knowledge anyway.” How then do you generate knowledge? Through induction? If so then David Hume comes through again with the problem of induction. Just because something has occurred multiple times in the past doesn’t mean it will occur again in the future. Also falsifiability was designed to be a deductive process so as to avoid the problem of induction. “What we do is employ the falsification principle - we attempt to falsify the claim that my hand caused the box to move, and if we can't, we accept it. This acceptance is again not some "ultimate necessary truth" that stands forever and until the end of time. Merely until it is falsified.“ The issue is that there is no way to tell whether or not we should falsify a claim. In order to perform science we must also assume that our equipment is working and that our senses are working and an infinite number of other claims before we can even get started. You can watch Carnaedes.org video “why science rests on faith”. This makes any claim unfalsifiable since we can always doubt the theory’s background beliefs. That is to say, if I try to falsify the hypothesis “My hand caused this box to move” and find that I cannot falsify it I could always muddy the waters by casting doubt onto my senses or my equipment or positing an infinite number of skeptical scenarios. This brings the hypothesis to the same level as the hypothesis “Ghosts are real.” There is no way to no if I have or have not falsified a claim making it unfalsifiable. This is also addressed in the video from 13:50 to 14:50. “ This means that we can never attain the absolute truth of the world, for it may always be falsified in the future. And some truths may even be unfalsifiable, but wholly unreasonable for us to believe, so we're forced not to commit to them.” I agree with you on this. I find science to be very useful and I think it is should be seen as the line of best fit. However I still doubt that theories are falsifiable. “ Say hundreds of us flawed humans with our subjective quirks and biases and interpretations walk into a room together. We see the box in the room. We all confirm we see the box. We can all touch the box, hear sounds reflecting off the box. Measure the box's weight and so on. All of our minds receive the same data. Then we go to sleep, lose consciousness, wake up. All still in the same room. Same box. Same experiences. Despite our flawed subjective experiences, nontheless we are forced to come to the conclusion that the box must exist in an objective reality independent of us, but which we share space in. I do not see what falsifiable proposition you can make, that would falsify the shared reality hypothesis, that hasn't already been tested and ruled out for the time being.” All the claims you listed here are unjustified. There is no way to tell if all the people are receiving the same data or that there are any people to begin with or that there’s even a box. There are an infinite number of skeptical scenarios that prevent this. Just because you assume the existence of the external world doesn’t mean that you can know what it’s like. There may be an external world but that doesn’t mean that the box is a necessary part of that world. If you say that knowledge comes from sense data then you’d be more like an empiricist. Also the hypothesis “the box must exist in an objective reality independent of us, but which we share space in” is unfalsifiable. You seem to assume that sense data is a way of getting information from the external world, you also seem to trust that your senses aren’t deceiving you, you also seem to trust that the other people aren’t lying or being deceived. These are all background beliefs that can be doubted making the hypothesis unfalsifiable. “ Ultimately, like I said before, background knowldge isn't problematic to invoke, as long as we recognize it is tentative in nature.” I agree that it’s tentative in nature. I don’t agree that it isn’t problematic. Anyways thanks for the response! Also the best defenses of science that I have seen come from the KZbinr AntiCitzenX I have some issues with his ideas but if you’re interested in this kind of stuff I recommend him.

[@Google_Censored_Commenter]

@@duder6387 You're literally dismissing everything I have to say as "unjustified" instead of responding to it. Let me make this very clear: I do not care about your standard of justification. So it is not a valid response. Got it? Cool. Now try again. You don't get to simply dismiss observations or even necessary axioms because they don't fit your criteria. We could play this dance forever, I could argue logic isn't justified, because it too rests on axioms. We get it. Munchhausen trilemma, nothing can be justified. Now move the fuck on. This is like arguing with christians who say "but do you know that with 100% certainty? Oh you don't? Then I don't have to believe it!" It's a stupid ass standard, and why I mentioned falsification as the true standard we should use. And yet even that, you dismiss as a "deductive process". This could not be more wrong. And it is not an inductive process either, this isn't a binary. Read some more Popper before you respond. Really, if I can tell you haven't I won't respond in kind.

[@MichaelRicksAherne]

I'm not seeing the point. Celestial mechanics failed to predict the perehelion, therefore it's false. General relativity - which does predict the perehelion accurately -- is more correct. Now, for the sake of practicality, we still use Newton's celestial mechanics nearly everywhere, because the situations in which it is false are rare. And also because the general relativity equations simplify to Newtonian ones in all other cases. But the point stands - Newtonian Mechanics as a *universally*-applicable theory is false. That's why falsifiability is important. By contrast, what does the "God theory" predict about the motion of planets? Or the motion of fluid? Or the motion of anything? Actually, are there any predictions the "God theory" makes?

[@duder6387]

I know this is a year late, but I was hoping to respond to your questions. Newtonian mechanics failed to predict things in the past. For example Uranus’ orbits were different from how Newton’s theories predicted them, this led to the belief that there was a planet messing with its orbit. They were correct this planet was Neptune. When they saw that Mercury’s orbit was off it made sense to believe that there was another planet messing with its orbit. Just because something fails to predict a phenomenon doesn’t make it wrong. The issue Carnaedes points out is that it is impossible to tell whether or not we should throw out Newtonian mechanics or if there really is some mass interfering with Mercury’s orbit that we haven’t observed yet such as dark matter which has mass but is invisible. Maybe Newtonian mechanics is predicting a new phenomenon that we just don’t know about yet. There’s no way to tell. Also the God theory as much more predictive power. The God theory can make predictions about anything and still make sense. Let’s assume that I am God. I make the prediction that the sun won’t rise tomorrow. People observe the sun rising and I say, “You are simply being deceived the sun actually hasn’t risen it just appears that way.” In this way I can make a prediction about anything while still being logically consistent. If you think this theory sounds stupid or counterintuitive think about how stupid and counterintuitive it was when people claimed the earth was round. Everyday people observed it to be flat and the scientists essentially said, “You are simply being deceived the earth is actually round it just appears flat.”

[@MichaelRicksAherne]

@@duder6387 "Also the God theory [h]as much more predictive power." Could you give an example of one of those predictions?

[@JTKroll12]

sounds like a problem happened

[@7DYNAMIN]

Holy fucking shit what a video