Believe me I believe you that you are professional philosopher, I just needed these lectures for my exam from Philosophy of Science. And, I may say I`m thrilled, you are good explainer. Keep on ...
Of course Newton can be wrong, but the point of this series isn't to 'get to The Truth of the philosophy of science'. The point is to explore the major ideas in the philosophy of science, true or false. At this stage in the series I'm presenting the ideas that were dominant in the first half of the 20th century. Later ideas will contradict these, that's how dialectic works. But you can't just jump to the current thinking if you want to understand the history of ideas.
I did not mean that I knew with "absolute certainty." I simply meant that I have sufficient justification to qualify my belief as a knowledge claim. Nobody truly "knows" anything about the external world, nor can they ever. And that's the problem with these discussions about induction - people want perfect knowledge about the external world when such a thing is philosophically unobtainable.
4:20 The difference is that it is necessary for a first and second children to exist for a third one. This is not the case with copper conducting electricity.
Another point worth considering is the nature of the inductive reasoning itself. The reason I "know" that the sun will rise tomorrow is based on much more than the mere fact that the sun merely "rose yesterday." I have an entire theoretical foundation for how the Earth travels around the sun and spins on its axis, all of which has likewise been cross-confirmed and tested accordingly. We don't just take empirical samples in a vacuum and leave it at that. We build theories.
"There is no "law like" hypothesis..." So you don't think 'all copper conducts electricity' is law-like? Or is it you don't think that's a hypothesis? "One hypothesis is just as valid as any other hypothesis.- until it is tested... A single test is not powerful in supporting a hypothesis - but is a step..." Please stick around for the lecture on Karl Popper. I'm curious what you'll think about his ideas.
More specifically, the chicken should have realized, as a corollary, is that the food being delivered MAY have required the farmer (or something) but that the coming of the food required something to bring it. The Hume thing's flaw is that one presupposes that there is a past and a future and that they have a meaningful difference, when it's necessary to cognitively isolate cause - effect from event1:(possible mechanisms):related-event2. Did that make sense?
I still fail to understand why this is a "problem." No one ever said that induction had to be "perfect." Only that, if the inference is valid, then you can make good decisions accordingly. And if the inference is not valid, then eventually you'll make a decision that turns out bad. When that happens, you then simply modify your theory accordingly as best as you can. Induction only really "works" as long as you also embrace fallibilism and falsifiability.
The chicken is still fed every day. When the farmer shows up and the chicken's head gets cut off, the chicken gets fed...to people. Horrible puns aside, I like to think about this stuff a lot and it's good to see new videos from you.
Yes there is. If you posit an analytical statement about a regularity, then prior to it being tested sufficiently, it is a hypothesis in a sense. In this case, after testing it graduates to law status instead of theory status.
Newtonian gravity may not qualify as a theory of gravity, but as an implicit theory about the ability of mathematics to describe the forces of nature. And even if seen as a theory of gravity, you can apply David Deutsch's concept of "Invariable theories."
I think that would have been news to Newton. The distinction you're employing isn't one he would have recognized. He explicitly argued that it is not the job of science to explain anything, but merely to describe nature. But he surely would not have said there is no such thing as a scientific theory. I'm going to do a whole lecture later on the topic of scientific explanation. I'll get into this in more detail there.
How would you analyze the following hypothesis for being law like: All X does Y... all copper conducts electricity... all plums drop from trees... all lottery tickets pay money...
My response to Hume would be that while we don't know that our theories about the world are correct, it's still smart to assume that they are. We don't know that jumping off a cliff will lead to death, but our experience of the world, applied in the best model of reasoning we have, would suggest that we should expect this, and that jumping off a cliff would be a bad idea.
Confirmation is the term used by the philosophers of science that he's talking about. In fact, I just read the chapter on confirmation in Hempel's Philosophy of Natural Science.
This comes down to how one defines "describe" and "explain".Let's say I have a Rube Goldberg-style mouse trap.I "describe" how the system works so as to drop a net on a mouse.That is also "explaining" how the trap would catch a mouse.It is a semantics thing.A law though is like simply stating "these parts of the trap, when released, always accelerate downward at a rate determined by their color".It is embedded in the larger system's "description" or "explanation" if you'd like.
"what should be considered a current law changes with the state of knowledge." So we should never study anything that isn't currently believed? How are we to learn from our mistakes if we don't understand the history of ideas? Newton's comment about 'standing on the shoulders of giants' seems relevant here: science isn't just a current set of beliefs, it's a process. And that process is nonlinear. (Kuhn argues it's actually circular. More on that later.) Surely it behoves us to study history.
It explains the magnitude of the force of attraction between masses. It states that every point mass in the universe attracts every other point mass with a force that is directly proportional to the product of their masses and inversely proportional to the square of the distance between them. It just doesn't explain what you want it to explain. It was never intended to explain the cause of gravity, just the magnitude of attraction between objects which we call gravity.
You're right. Bouncing off multiple mentions of gravity below, the interested reader would do well to search for something like gravity is a law not a theory
"I was talking about the real world and not a self-contained logical system." The whole idea behind the HD method is that we can use logic as a foundation for our theory of evidence. The problem is that logic seems to allow things to count as evidence that we don't intuitively think should count as evidence.
Part 4: A "confirmed" data point then supports the hypothesis (or theory). Both of your examples are equally valid... In Unobtania women never talked about their pregnancies - no one knew... It was a molecular biologist that, in looking at the genome came up with a hypothesis that every person was a third child... He did exhaustive interviews and uncovered confirmation after confirmation (notice "confirmations" are individual data points) which provided unceasing support for his hypothesis
My argument is that we don't know how the world really works, so we should go with the way it seems to work, because that's the only model we have. We don't know how things really are, so we should act in the way which will produce the most desirable results assuming our interpretation is correct. So I guess I would agree with your view that our view of the world is defined according to utility, if that's what you're saying. Whether or not this justifies religion is debatable.
Part 3: Forget "confirm" - it will get you into trouble 1. confirm - To make valid or binding 2. confirm - To support or establish the certainty or validity of; verify 3. confirm - To make firmer; strengthen: You should be talking about 3 but you mix up 1 and 2... An individual data point can be confirmed - but that is only "confirmation" of that data point - not the hypothesis... A "confirmed" data point SUPPORTS the hypothesis... If confirmed a data point IS VALID - (that one point)
so youre saying you see more problems with 1+1 = 2 (standard natural numbers, with addition and comparison) than with seeing sun rise every day and concluding that sun will rise every day in the future as well?
Which I guess would imply that this type of metaphysical stuff should be defined based on what actions are appropriate given the knowledge we have. So I guess instead of concerning ourselves with the way the world is, we should focus instead on what we should assume to be true in order to produce desirable actions. This would also seem to be the strongest argument in favor of religion.
Why does copper conduct electricity? One could say that the conduction of electricity occurs when a mobile ion is present in a compound. That would, in an oversimplified way, give us a mechanism. Then you can go beyond copper and talk about everything that could conduct electricity.
...cont It is a law in that it is still an analytical description of a regularity sans mechanism, but it lacks the force of being the limits to our understanding in an area. It might still be called a law, but I do not personally think it should be considered as such.
Newtonian gravitational theory describes the force of interaction between two bodies - it didn't intend to explain the cause of the observed force. Zarkoff45 is right in this case...
I stated that it mean the following: 1. Obviously, if significant exceptions are found, then the law does not hold. 2. If we find the mechanism that explain the regularity described by a law, then it is only a law in a historic sense. If it is currently explained, then it lacks the force it once had. For instance, let's say it was once a law that "red things have a temperature determined by the saturation of red".Once we understand the mechanism behind this, then ...cont
I tend to agree with A C Grayling's definition of rationality as a proportioning of belief to evidence, which would make the chicken rational until the data changes: watch?v=mbi44ObWuvk (the relevant part is 40:35-42ish) Or, alternatively: "It's not possibility that matters, it's probability. So until you give me a good reason to think that my belief is not just possibly false, but probably false, I'm not changing anything about what I believe or what I think I know." -John W. Loftus
Hume's point was that we don't actually know if any of this stuff is really valid. What we're assuming by using statistics is that the world will always work the way it works in the past. There is no way to justify this without using a circular argument.
"You create out of whole cloth a type of theory you call "law like" - no such thing." No, he didn't, smart guy. This is a term Nelson Goodman used in 'Fact, Fiction and Forecast'. SisyphusRedeemed is a professor teaching the works of the analytic philosophers of the 20th century. That doesn't mean he agrees with everything each of them said. In fact, if you've been paying attention *at all*, you'd see that he has refuted most of the philosophies he's already taught (logical positivism, anyone?)
Laws do not explain and often represent a limit of our understanding. A law is a regularity that is observed. It is kind of a complicated and difficult to observe fact. Theories use laws within their explanations often. You will notice that once the mechanism of gravity was somewhat explained by relativity, that it was called the theory of general relativity.
"At this stage in the series I'm presenting the ideas that were dominant in the first half of the 20th century." Wait, there really was no distinction between law and theory in the first half of the 20th century? So, Einstein and Dirac didn't know there was a distinction between a law and a theory? That might explain how the Copenhagen interpretation came about. One would have to be pretty clueless to propose a mechanism like consciousness.
And when I say "observed", I mean what I say - in the lab and field. Direct observations (in controlled settings) are pretty much the most "factual" piece of information there can be.
"Scientific theories involve mechanisms that tell us how x does y." That's not true of all scientific theories. Newtonian gravity had no mechanism, and it was criticized harshly by the Cartesians for that reason. Surely Newton's theory of gravity was scientific, even though it lacked a mechanism.
I didn't say natural selection was a fact. I said it was a fact that it had part in creating the diversity of life. But my main point was: you're arguing about terminology, not meaning, and your particular use of "fact" and "theory" are not very clear. I'd stick with what most people use, with the generally accepted meanings. There is nothing wrong with saying "Darwin's theory" to mean "the theory of evolution by natural selection", it's not linguistically problematic in the least.
Newton did not propose a theory of gravity he proposed an universal law of gravity. In the language of science, the word "law" describes an analytic statement. It gives us a formula that tells us what things will do. Meanwhile, the word "theory" is used to describe an explanation of why and how things happen.
Einstein's geometric gravity is a theory because it imagines a mechanism by which gravity operates, a curvature of space-time, but Newton had no mechanism, just a mathematical description of what was happening.
that was the point. your issues are only relevant to IR, because DR does not even to attempt to address possibly invalid axioms. DR is only for (assumed) true premises or axioms. its like complaining that you cannot use scissors for writing. the best you can do w/o premises are absolutes (A=A etc.). on the other hand, i havent really seen IR applied w/o DR. induction gives you premises, then you deduce. so using IR adds uncertainty to DR and is therefore objectively more questionable.
I can't speak to Einstein and Dirac, but I can speak for the positivists and their contemporaries. And they did not use this distinction in the way you are using it.
Induction is implying the homogeneity of nature, which in turn eludes to invariants of nature. So the chicken problem boils IMHO down to the question of whether a hypothesis applies to a structural invariant of nature or just a fleeting "configuration" of nature (farmers are not invariants of nature).
But that's just my point: Newtonain gravity is scientific, but does not explain, does not provide a mechanism. Hence, not all scientific theories are explanatory.
"Newtonian gravity had no mechanism" To be fair, gravity is both a law and a theory, depending on which is being spoken of. The Law of Gravity (universal gravitation) is what Newton devised. A theory of gravity is what Einstein brought to the table. However, this could be wrong and needs revision, as all theories are always subject to... ;)
"we should focus instead on what we should assume to be true in order to produce desirable actions." No. I'd just stop at out of sample prediction accuracy as a measure of utility. That cuts out religion and is in line with how scientists actually think.
when it comes to baldwin: the statement implies existence of a first and second child, leading to a contradiction with the statement. it can be disproved deductively through contradiction. its like saying that every ordering of a set labels every element as third. i think a better example would be: people commenting on youtube videos are stupid, giving a specific youtube video with an overwhelmingly stupid comments section. its not even that difficult to find :).
"it is a suggestion that is supported by the evidence" Actually no, it's an observed fact, like "objects with mass attract each other". Natural selection and other evolutionary processes have been *observed* to create diversity of life, that's not just a suggestion, an educated guess. I wasn't saying "all diversity of life has been explained by evolutionary theory". Words DO matter.
It is the theory of evolution by natural selection. That things change over time, which is what is meant by evolution, is observable and so can be called a fact. Natural selection must be included to call it a theory because it provides the mechanism and as such can be tested.
I do not believe that Newton used the distinction of law vs. theory when first describing his universal law of gravitation. As of current views on these things, there is a distinction between a law and a theory in science, and Newton's LAW of universal gravitation is the former. Both laws and theories, however, can be tested via the scientific method. And what should be considered a current law changes with the state of knowledge.
"Facts - are brute truths discovered through observation" Absolutely true (if absolute truth exists)... Newton was thought to have discovered Absolute Truth in his "laws."... It is only in the twentieth century where we discovered that even the most rock solid Law is only provisional... Physics was thought to be dead because everything had been discovered... Now we know that a theory is never "confirmed" - just supported... Of course, that depends on how you define "confirmed"
You don't know enough to be having this conversation. When you analyze data, then you are extracting information about the past. This is how stats are used to test hypotheses, etc. You can test your models, etc based upon predicting cases that were not included in your calibration data set. This s important, and if you don't know why then read more. If you then make the assumption that the future will behave similar to the past, and use states to make probability statements about ...cont
I'm not sure I see the peoblem with your baldwin analogy. If you have a theory that all people are third children, and the only person you have met is a third hild, then that is evidence that all people are third children.
....cont probability statements about the future and objects outside of your data sample in order to guide your actions, then you are beginning to get into what Hume was speaking of. It's really a dumb thing to be concerned with.It is simply bemoaning the human condition in this universe.It forms an interesting loop to that affirms itself.How do we know that assuming uniformity is necessary and useful?In the past when we did this,it was found to work well.All that evidence does count.
Basically, it's not "irrational" for me to believe that the sun will rise tomorrow, and it's also not irrational for the chicken to believe that it will be fed tomorrow. What would be irrational is to believe something which is NOT most probable, as opposed to what IS most probable. Did the chicken just watch the farmer butcher another chicken? What's rational to believe has just changed, because the data has changed. (Rationality = probability, not 100% certainty.)
@@humeanrgmnt7367 How do you use deduction to figure out whether I'd be rational to start stockpiling coal and wood for a future with no more sunrises?
"we pretty much know it is true" You're missing the point. Imagine we didn't know this to be true. Imagine we've discovered our first piece of copper and we speculate that it might conduct electricity. We test it, then a few more, and sure enough, all of them conduct electricity. We then propose the law-like hypothesis 'all copper conducts electricity.' It's a model for scientific hypothesis generation.
"all copper conducts electricity' is law-like"... Only because it is trivial - we pretty much know it is true (for the copper forms we have tested)... How about: "all forms of hydrogen are non-metalic" - is that "law like?" And when you discover metallic hydrogen - was the original hypothesis balderdash? What about: Every particle of matter in the universe is attracted to every other particle of matter in the universe by a FORCE proportional to (etc..). Is that balderdash - it's wrong...
It is not an observed fact - that is why there are alternative explanations - like sexual selection, genetic drift, etc. Objects attract each other is an observed fact - WHY they do so (the mechanism) is a suggestion and not central to the facts (and Newton got it wrong, even though objects still do attract)).... I am not saying natural selection is not working or that it is not the prime mechanism - just that is is not an observable FACT (as causing diversity of life)...
I say that Hume was doing great until 11:45 "the future is under no obligation...". That's flawed. And that very flaw is what the chicken, and many avenues, fall prey to (no pun intended). That the chicken would have experience with it's own actions, like "food comes" and then "I eat the food" doesn't imply that food coming causes eating. But that the eating of food had a logical requirement (presence of food) is as far as the chicken should have gone.
Again, it's the philosophy of science. You have a problem with the so-called great philosophers, but you are pointing the blame on one of the messengers. The term "confirmation" is a major point of contention in analytic philosophy; the fact that you think you have it all solved just makes you look arrogant silly.
Actually chicken didn't think like we do. She was more likely over thinking. Doing the exact thing that religious people tend to do. Aka, jump into conclusion on false premise. Evidence around chicken only pointed out that farmer feeds them every time it comes. It never even implied that this is how world works. That was the chicken's own false conclusion. But chicken was right on the part that farmer would get fed. Not to mention that with more information we can get more accurate predictions.
All x do y isn't really a scientific theory. Scientific theories involve mechanisms that tell us how x does y. Why and how does copper conduct electricity? Answering that question would give you a theory. If all copper conducts electricity, that's a fact, not a theory, isn't it? The theory of evolution doesn't simply state that living things change over time, it gives us the mechanism of natural selection. That mechanism is tested with evolutionary algorithms. watch?v=VuP3NhDu6Hk
Only if Newton's theory is of gravity. It's real power comes from another source than as an explanation - it is a mathematical description, not an explanation, of an observed fact and the description is, in David Deutsch's term, Invariable - you can't change the equations and get an accurate description. This is why mechanistic theories supported by mathematical predictions are powerful, invariable theories. So, I still have to reject your definition of theory.
"But this is exactly the problem that Hume was bringing up. The concept of "predictive ability" is technically meaningless." This is simply false. There is a huge body of applied math and statistics that addresses this. If you are simply stating that we do not know beforehand whether our model will predict an unknown/future occurrence then that is just the obvious and does not actually change how people decide on models based upon things called cross validation, out of sample error rate, etc.
Part 1: You seem confused. There was nothing wrong with your second example (third child) and there was nothing wrong with the evidence - it supported the assertion (that every person is a third child). In fact in the country of Unobtania it is absolutely true. The people who live there have a genetic expression where the first two offspring die at birth, the third survives and thereafter the woman becomes sterile. This situation can be verified by conducting research. See Part 2
"I think that would have been news to Newton. The distinction you're employing isn't one he would have recognized." So, does all your philosophical speculation on the truth in science depend on having a Newtonian view of science? Newton got a lot of things wrong, he was an alchemist who interpreted the Bible. His description of what science is is also wrong. Being the big genius of his time doesn't give him papal infallibility.
Part 5: In your examples BOTH results are Good Scientific "Confirmation"... Copper has equal support (one) with third child (one)... There is no bad science here... Keep testing fresh examples of copper and third child and see if the hypotheses hold up...
If you're are going to try to use logic and philosophy to explain scientific theories, then you need to be more precise about what a scientific theory is than that. According to Peter Ellerton Newton didn't have a theory of gravity, that didn't come until Einstein: watch?v=ENZRZ9lAYAo
"it is a fact that natural selection and other evolutionary processes create diversity of life." It is not a FACT that natural selection, etc. create diversity of life - it is a suggestion that is supported by the evidence... If we found s better explanation for speciation we would discard natural selection - you can't do that with facts... Words DO matter.... Creationists like to say that "Evolution is not a fact - it's just a theory..." (I know they mean theory as in a guess...)
But this is exactly the problem that Hume was bringing up. The concept of "predictive ability" is technically meaningless. Although I do somewhat dislike the idea of inventing another world in order to alter our model of the world as it appears to achieve desired ends. That does seem a little dishonest.
"Facts - are brute truths discovered through observation. Common descent of all life is one such fact,.." Common decent is NOT one such fact... Change in biological expression and diversity over time is a fact - just observe the fossil record... Common decent is the conclusion of one particular theory - Decent With Modification... "it is a fact that natural selection and other evolutionary processes create diversity of life" Natural selection is not a fact - it is a proposed mechanism...
And you only have knowledge about what has happened. I do not believe anyone can claim knowledge about the future. We can be able to make bets, so to speak, though on outcomes based upon probabilities and act accordingly, however.
You are talking about something unrelated, it seems. What would be related is "what should we believe?" and I would go with beliefs with the greatest predictive ability. You must first have a model of the world before you think how to act. You are skipping to the end.
I disagree with the chicken analogy. I don't think it's really being a very good empiricist by simply observing the small world of her pen. Rather, a good empiricist should be constantly pushing the boundaries of what it observes as we've done with space probes and microscopes and global communication. For instance, the chicken should fly the coop and observe other farmers' interactions with chickens. Maybe she's been studying chicken anatomy when she sees a farmer munching on a piece of meat that she can now identify as a chicken leg. If it only employed these methods, I think the chicken could do quite well for itself.
I think the idea is that the "chicken" represents us and the" coop" represents the limits we have to observing our universe. Think to what Hume said about the laws of nature, we will never be able to actually observe them, only see the effect they have. We literally cant fly out of our coop, and if we cant, what are the implications on what we can know.
Come OFF it.. you're trying to correct the usage of "fact" and "theory", while just making it twice as confusing. Facts - are brute truths discovered through observation. Common descent of all life is one such fact, and it is a fact that natural selection and other evolutionary processes create diversity of life. But evolutionary theory is, unlike facts, is *an explanation* of how those processes work, it makes predictions, allows for calculation, measurement, comparison.
Part 2: You create out of whole cloth a type of theory you call "law like" - no such thing... I believe you are misled because you talk about confirmation..... You misuse confirmation - as if it is kind of like "prove" or something... If you go back and replace every place where you use "confirm" (etc.) with "supports" (etc) you will find that both theories are the same... Testing one piece of copper gives as much confidence in the conductivity of all copper as you have for third children
That is fine - as long a he stays in philosophy and doesn't move into science... What he is saying is silly... There is no "law like" hypothesis... One hypothesis is just as valid as any other hypothesis.- until it is tested... A single test is not powerful in supporting a hypothesis - but is a step... The hypothesis of the third child is supported by his single data point...
Believe me I believe you that you are professional philosopher, I just needed these lectures for my exam from Philosophy of Science. And, I may say I`m thrilled, you are good explainer. Keep on ...
Of course Newton can be wrong, but the point of this series isn't to 'get to The Truth of the philosophy of science'. The point is to explore the major ideas in the philosophy of science, true or false. At this stage in the series I'm presenting the ideas that were dominant in the first half of the 20th century. Later ideas will contradict these, that's how dialectic works. But you can't just jump to the current thinking if you want to understand the history of ideas.
I did not mean that I knew with "absolute certainty." I simply meant that I have sufficient justification to qualify my belief as a knowledge claim. Nobody truly "knows" anything about the external world, nor can they ever. And that's the problem with these discussions about induction - people want perfect knowledge about the external world when such a thing is philosophically unobtainable.
4:20 The difference is that it is necessary for a first and second children to exist for a third one.
This is not the case with copper conducting electricity.
Kind of beside the point, but yes, the third child hypothesis could be falsified by pure logic, so it’s a bad example.
Good to see you back! The philosophy of science is one of my favorite sub-disciplines in philosophy.
Glad to see you're back Dr. Merriam!
Another point worth considering is the nature of the inductive reasoning itself. The reason I "know" that the sun will rise tomorrow is based on much more than the mere fact that the sun merely "rose yesterday." I have an entire theoretical foundation for how the Earth travels around the sun and spins on its axis, all of which has likewise been cross-confirmed and tested accordingly. We don't just take empirical samples in a vacuum and leave it at that. We build theories.
6:55 isn't this Hume's problem of causality/causation?
"There is no "law like" hypothesis..."
So you don't think 'all copper conducts electricity' is law-like? Or is it you don't think that's a hypothesis?
"One hypothesis is just as valid as any other hypothesis.- until it is tested...
A single test is not powerful in supporting a hypothesis - but is a step..."
Please stick around for the lecture on Karl Popper. I'm curious what you'll think about his ideas.
More specifically, the chicken should have realized, as a corollary, is that the food being delivered MAY have required the farmer (or something) but that the coming of the food required something to bring it.
The Hume thing's flaw is that one presupposes that there is a past and a future and that they have a meaningful difference, when it's necessary to cognitively isolate cause - effect from event1:(possible mechanisms):related-event2.
Did that make sense?
I still fail to understand why this is a "problem." No one ever said that induction had to be "perfect." Only that, if the inference is valid, then you can make good decisions accordingly. And if the inference is not valid, then eventually you'll make a decision that turns out bad. When that happens, you then simply modify your theory accordingly as best as you can. Induction only really "works" as long as you also embrace fallibilism and falsifiability.
The chicken is still fed every day. When the farmer shows up and the chicken's head gets cut off, the chicken gets fed...to people.
Horrible puns aside, I like to think about this stuff a lot and it's good to see new videos from you.
Yes there is. If you posit an analytical statement about a regularity, then prior to it being tested sufficiently, it is a hypothesis in a sense. In this case, after testing it graduates to law status instead of theory status.
Newtonian gravity may not qualify as a theory of gravity, but as an implicit theory about the ability of mathematics to describe the forces of nature. And even if seen as a theory of gravity, you can apply David Deutsch's concept of "Invariable theories."
I think that would have been news to Newton. The distinction you're employing isn't one he would have recognized. He explicitly argued that it is not the job of science to explain anything, but merely to describe nature. But he surely would not have said there is no such thing as a scientific theory. I'm going to do a whole lecture later on the topic of scientific explanation. I'll get into this in more detail there.
How would you analyze the following hypothesis for being law like:
All X does Y...
all copper conducts electricity...
all plums drop from trees...
all lottery tickets pay money...
Very nicely done!
My response to Hume would be that while we don't know that our theories about the world are correct, it's still smart to assume that they are. We don't know that jumping off a cliff will lead to death, but our experience of the world, applied in the best model of reasoning we have, would suggest that we should expect this, and that jumping off a cliff would be a bad idea.
Confirmation is the term used by the philosophers of science that he's talking about. In fact, I just read the chapter on confirmation in Hempel's Philosophy of Natural Science.
This comes down to how one defines "describe" and "explain".Let's say I have a Rube Goldberg-style mouse trap.I "describe" how the system works so as to drop a net on a mouse.That is also "explaining" how the trap would catch a mouse.It is a semantics thing.A law though is like simply stating "these parts of the trap, when released, always accelerate downward at a rate determined by their color".It is embedded in the larger system's "description" or "explanation" if you'd like.
I'm so glad the experts on this agree with me and not SisyphusRedeemed, it gives me a sense of triumph.
"what should be considered a current law changes with the state of knowledge."
So we should never study anything that isn't currently believed? How are we to learn from our mistakes if we don't understand the history of ideas? Newton's comment about 'standing on the shoulders of giants' seems relevant here: science isn't just a current set of beliefs, it's a process. And that process is nonlinear. (Kuhn argues it's actually circular. More on that later.) Surely it behoves us to study history.
It explains the magnitude of the force of attraction between masses. It states that every point mass in the universe attracts every other point mass with a force that is directly proportional to the product of their masses and inversely proportional to the square of the distance between them.
It just doesn't explain what you want it to explain. It was never intended to explain the cause of gravity, just the magnitude of attraction between objects which we call gravity.
You're right. Bouncing off multiple mentions of gravity below, the interested reader would do well to search for something like
gravity is a law not a theory
"I was talking about the real world and not a self-contained logical system."
The whole idea behind the HD method is that we can use logic as a foundation for our theory of evidence. The problem is that logic seems to allow things to count as evidence that we don't intuitively think should count as evidence.
Part 4:
A "confirmed" data point then supports the hypothesis (or theory).
Both of your examples are equally valid...
In Unobtania women never talked about their pregnancies - no one knew...
It was a molecular biologist that, in looking at the genome came up with a hypothesis that every person was a third child...
He did exhaustive interviews and uncovered confirmation after confirmation (notice "confirmations" are individual data points) which provided unceasing support for his hypothesis
My argument is that we don't know how the world really works, so we should go with the way it seems to work, because that's the only model we have. We don't know how things really are, so we should act in the way which will produce the most desirable results assuming our interpretation is correct. So I guess I would agree with your view that our view of the world is defined according to utility, if that's what you're saying.
Whether or not this justifies religion is debatable.
Part 3:
Forget "confirm" - it will get you into trouble
1. confirm - To make valid or binding
2. confirm - To support or establish the certainty or validity of; verify
3. confirm - To make firmer; strengthen:
You should be talking about 3 but you mix up 1 and 2...
An individual data point can be confirmed - but that is only "confirmation" of that data point - not the hypothesis...
A "confirmed" data point SUPPORTS the hypothesis...
If confirmed a data point IS VALID - (that one point)
so youre saying you see more problems with 1+1 = 2 (standard natural numbers, with addition and comparison) than with seeing sun rise every day and concluding that sun will rise every day in the future as well?
1+1=2 is completely different knowledge than 'will the sun rise tomorrow.' apples and oranges. Analytic vs. synthetic.
Which I guess would imply that this type of metaphysical stuff should be defined based on what actions are appropriate given the knowledge we have. So I guess instead of concerning ourselves with the way the world is, we should focus instead on what we should assume to be true in order to produce desirable actions.
This would also seem to be the strongest argument in favor of religion.
Why does copper conduct electricity? One could say that the conduction of electricity occurs when a mobile ion is present in a compound. That would, in an oversimplified way, give us a mechanism. Then you can go beyond copper and talk about everything that could conduct electricity.
...cont
It is a law in that it is still an analytical description of a regularity sans mechanism, but it lacks the force of being the limits to our understanding in an area. It might still be called a law, but I do not personally think it should be considered as such.
Newtonian gravitational theory describes the force of interaction between two bodies - it didn't intend to explain the cause of the observed force.
Zarkoff45 is right in this case...
I stated that it mean the following:
1. Obviously, if significant exceptions are found, then the law does not hold.
2. If we find the mechanism that explain the regularity described by a law, then it is only a law in a historic sense. If it is currently explained, then it lacks the force it once had. For instance, let's say it was once a law that "red things have a temperature determined by the saturation of red".Once we understand the mechanism behind this, then ...cont
I tend to agree with A C Grayling's definition of rationality as a proportioning of belief to evidence, which would make the chicken rational until the data changes: watch?v=mbi44ObWuvk
(the relevant part is 40:35-42ish)
Or, alternatively:
"It's not possibility that matters, it's probability. So until you give me a good reason to think that my belief is not just possibly false, but probably false, I'm not changing anything about what I believe or what I think I know." -John W. Loftus
Hume's point was that we don't actually know if any of this stuff is really valid. What we're assuming by using statistics is that the world will always work the way it works in the past. There is no way to justify this without using a circular argument.
well said.
"You create out of whole cloth a type of theory you call "law like" - no such thing."
No, he didn't, smart guy. This is a term Nelson Goodman used in 'Fact, Fiction and Forecast'. SisyphusRedeemed is a professor teaching the works of the analytic philosophers of the 20th century. That doesn't mean he agrees with everything each of them said. In fact, if you've been paying attention *at all*, you'd see that he has refuted most of the philosophies he's already taught (logical positivism, anyone?)
Laws do not explain and often represent a limit of our understanding. A law is a regularity that is observed. It is kind of a complicated and difficult to observe fact. Theories use laws within their explanations often.
You will notice that once the mechanism of gravity was somewhat explained by relativity, that it was called the theory of general relativity.
"At this stage in the series I'm presenting the ideas that were dominant in the first half of the 20th century."
Wait, there really was no distinction between law and theory in the first half of the 20th century? So, Einstein and Dirac didn't know there was a distinction between a law and a theory? That might explain how the Copenhagen interpretation came about. One would have to be pretty clueless to propose a mechanism like consciousness.
And when I say "observed", I mean what I say - in the lab and field. Direct observations (in controlled settings) are pretty much the most "factual" piece of information there can be.
"Scientific theories involve mechanisms that tell us how x does y."
That's not true of all scientific theories. Newtonian gravity had no mechanism, and it was criticized harshly by the Cartesians for that reason. Surely Newton's theory of gravity was scientific, even though it lacked a mechanism.
I didn't say natural selection was a fact. I said it was a fact that it had part in creating the diversity of life.
But my main point was: you're arguing about terminology, not meaning, and your particular use of "fact" and "theory" are not very clear. I'd stick with what most people use, with the generally accepted meanings. There is nothing wrong with saying "Darwin's theory" to mean "the theory of evolution by natural selection", it's not linguistically problematic in the least.
Newton did not propose a theory of gravity he proposed an universal law of gravity.
In the language of science, the word "law" describes an analytic statement. It gives us a formula that tells us what things will do.
Meanwhile, the word "theory" is used to describe an explanation of why and how things happen.
Newton never explains what gravity is...only what it does.
Einstein's geometric gravity is a theory because it imagines a mechanism by which gravity operates, a curvature of space-time, but Newton had no mechanism, just a mathematical description of what was happening.
Don't assume laws of nature based upon people's or agent's actions?
that was the point. your issues are only relevant to IR, because DR does not even to attempt to address possibly invalid axioms. DR is only for (assumed) true premises or axioms. its like complaining that you cannot use scissors for writing. the best you can do w/o premises are absolutes (A=A etc.).
on the other hand, i havent really seen IR applied w/o DR. induction gives you premises, then you deduce. so using IR adds uncertainty to DR and is therefore objectively more questionable.
I can't speak to Einstein and Dirac, but I can speak for the positivists and their contemporaries. And they did not use this distinction in the way you are using it.
Induction is implying the homogeneity of nature, which in turn eludes to invariants of nature. So the chicken problem boils IMHO down to the question of whether a hypothesis applies to a structural invariant of nature or just a fleeting "configuration" of nature (farmers are not invariants of nature).
But that's just my point: Newtonain gravity is scientific, but does not explain, does not provide a mechanism. Hence, not all scientific theories are explanatory.
You know he's a professor in the philosophy of science, right?
"Newtonian gravity had no mechanism"
To be fair, gravity is both a law and a theory, depending on which is being spoken of. The Law of Gravity (universal gravitation) is what Newton devised. A theory of gravity is what Einstein brought to the table. However, this could be wrong and needs revision, as all theories are always subject to... ;)
If the future must mimic the past then why did the chicken die?
"we should focus instead on what we should assume to be true in order to produce desirable actions."
No. I'd just stop at out of sample prediction accuracy as a measure of utility. That cuts out religion and is in line with how scientists actually think.
and of course, what Sisyphus said also applies.
when it comes to baldwin: the statement implies existence of a first and second child, leading to a contradiction with the statement. it can be disproved deductively through contradiction. its like saying that every ordering of a set labels every element as third. i think a better example would be: people commenting on youtube videos are stupid, giving a specific youtube video with an overwhelmingly stupid comments section. its not even that difficult to find :).
"it is a suggestion that is supported by the evidence"
Actually no, it's an observed fact, like "objects with mass attract each other". Natural selection and other evolutionary processes have been *observed* to create diversity of life, that's not just a suggestion, an educated guess. I wasn't saying "all diversity of life has been explained by evolutionary theory".
Words DO matter.
It is the theory of evolution by natural selection.
That things change over time, which is what is meant by evolution, is observable and so can be called a fact. Natural selection must be included to call it a theory because it provides the mechanism and as such can be tested.
Doesn't mean the observable facts will hold into the future; there is no way you can know that.
I do not believe that Newton used the distinction of law vs. theory when first describing his universal law of gravitation. As of current views on these things, there is a distinction between a law and a theory in science, and Newton's LAW of universal gravitation is the former. Both laws and theories, however, can be tested via the scientific method. And what should be considered a current law changes with the state of knowledge.
"Facts - are brute truths discovered through observation"
Absolutely true (if absolute truth exists)...
Newton was thought to have discovered Absolute Truth in his "laws."...
It is only in the twentieth century where we discovered that even the most rock solid Law is only provisional...
Physics was thought to be dead because everything had been discovered...
Now we know that a theory is never "confirmed" - just supported...
Of course, that depends on how you define "confirmed"
It is a law. It does not explain. It describes.
Muito bom adorei
science is a model of reality that just happened to have the same results so far.
You don't know enough to be having this conversation.
When you analyze data, then you are extracting information about the past. This is how stats are used to test hypotheses, etc. You can test your models, etc based upon predicting cases that were not included in your calibration data set. This s important, and if you don't know why then read more. If you then make the assumption that the future will behave similar to the past, and use states to make probability statements about
...cont
we need a Jason Lisle Chicken too.
It was a law. A law and a theory in science are two different things.
I'm not sure I see the peoblem with your baldwin analogy. If you have a theory that all people are third children, and the only person you have met is a third hild, then that is evidence that all people are third children.
....cont
probability statements about the future and objects outside of your data sample in order to guide your actions, then you are beginning to get into what Hume was speaking of.
It's really a dumb thing to be concerned with.It is simply bemoaning the human condition in this universe.It forms an interesting loop to that affirms itself.How do we know that assuming uniformity is necessary and useful?In the past when we did this,it was found to work well.All that evidence does count.
Basically, it's not "irrational" for me to believe that the sun will rise tomorrow, and it's also not irrational for the chicken to believe that it will be fed tomorrow. What would be irrational is to believe something which is NOT most probable, as opposed to what IS most probable. Did the chicken just watch the farmer butcher another chicken? What's rational to believe has just changed, because the data has changed. (Rationality = probability, not 100% certainty.)
It's certainly not rational to believe the sun will rise tomorrow. It's based on inductive reasoning.
@@humeanrgmnt7367 If induction is not part of the foundation of what constitutes "rationality", then I don't know what is.
@@Flyborg then I'll tell u- deduction.
@@humeanrgmnt7367 How do you use deduction to figure out whether I'd be rational to start stockpiling coal and wood for a future with no more sunrises?
@@Flyborg you can't. that's why inductive reasoning isn't rational. deduction only works with analytic knowledge- algebra and arithmetic.
"when you discover metallic hydrogen - was the original hypothesis balderdash?"
No, it's just wrong. That happens in science, from time to time.
"we pretty much know it is true"
You're missing the point. Imagine we didn't know this to be true. Imagine we've discovered our first piece of copper and we speculate that it might conduct electricity. We test it, then a few more, and sure enough, all of them conduct electricity. We then propose the law-like hypothesis 'all copper conducts electricity.' It's a model for scientific hypothesis generation.
"…makes the other Baldwindash," surely? :D
"all copper conducts electricity' is law-like"...
Only because it is trivial - we pretty much know it is true (for the copper forms we have tested)...
How about: "all forms of hydrogen are non-metalic" - is that "law like?"
And when you discover metallic hydrogen - was the original hypothesis balderdash?
What about:
Every particle of matter in the universe is attracted to every other particle of matter in the universe by a FORCE proportional to (etc..).
Is that balderdash - it's wrong...
It is not an observed fact - that is why there are alternative explanations - like sexual selection, genetic drift, etc.
Objects attract each other is an observed fact - WHY they do so (the mechanism) is a suggestion and not central to the facts (and Newton got it wrong, even though objects still do attract))....
I am not saying natural selection is not working or that it is not the prime mechanism - just that is is not an observable FACT (as causing diversity of life)...
Google
universal law of gravitation
and
difference between a theory and a law
I agree...
I say that Hume was doing great until 11:45 "the future is under no obligation...". That's flawed. And that very flaw is what the chicken, and many avenues, fall prey to (no pun intended).
That the chicken would have experience with it's own actions, like "food comes" and then "I eat the food" doesn't imply that food coming causes eating. But that the eating of food had a logical requirement (presence of food) is as far as the chicken should have gone.
It's 'flawed' because you're a scientist and just do not want to accept it.
The second hypothesis is actually baldwindash.
(sorry)
General advice for the chicken: Bayes all the way, baby!
Again, it's the philosophy of science. You have a problem with the so-called great philosophers, but you are pointing the blame on one of the messengers. The term "confirmation" is a major point of contention in analytic philosophy; the fact that you think you have it all solved just makes you look arrogant silly.
Actually chicken didn't think like we do. She was more likely over thinking. Doing the exact thing that religious people tend to do. Aka, jump into conclusion on false premise. Evidence around chicken only pointed out that farmer feeds them every time it comes. It never even implied that this is how world works. That was the chicken's own false conclusion. But chicken was right on the part that farmer would get fed.
Not to mention that with more information we can get more accurate predictions.
All x do y isn't really a scientific theory. Scientific theories involve mechanisms that tell us how x does y. Why and how does copper conduct electricity? Answering that question would give you a theory. If all copper conducts electricity, that's a fact, not a theory, isn't it?
The theory of evolution doesn't simply state that living things change over time, it gives us the mechanism of natural selection. That mechanism is tested with evolutionary algorithms.
watch?v=VuP3NhDu6Hk
Only if Newton's theory is of gravity. It's real power comes from another source than as an explanation - it is a mathematical description, not an explanation, of an observed fact and the description is, in David Deutsch's term, Invariable - you can't change the equations and get an accurate description. This is why mechanistic theories supported by mathematical predictions are powerful, invariable theories.
So, I still have to reject your definition of theory.
"But this is exactly the problem that Hume was bringing up. The concept of "predictive ability" is technically meaningless."
This is simply false. There is a huge body of applied math and statistics that addresses this. If you are simply stating that we do not know beforehand whether our model will predict an unknown/future occurrence then that is just the obvious and does not actually change how people decide on models based upon things called cross validation, out of sample error rate, etc.
4:53 More like Baldwin-dash :3
Part 1:
You seem confused. There was nothing wrong with your second example (third child) and there was nothing wrong with the evidence - it supported the assertion (that every person is a third child). In fact in the country of Unobtania it is absolutely true. The people who live there have a genetic expression where the first two offspring die at birth, the third survives and thereafter the woman becomes sterile. This situation can be verified by conducting research.
See Part 2
"I think that would have been news to Newton. The distinction you're employing isn't one he would have recognized."
So, does all your philosophical speculation on the truth in science depend on having a Newtonian view of science? Newton got a lot of things wrong, he was an alchemist who interpreted the Bible. His description of what science is is also wrong. Being the big genius of his time doesn't give him papal infallibility.
Part 5:
In your examples BOTH results are Good Scientific "Confirmation"...
Copper has equal support (one) with third child (one)...
There is no bad science here...
Keep testing fresh examples of copper and third child and see if the hypotheses hold up...
If you're are going to try to use logic and philosophy to explain scientific theories, then you need to be more precise about what a scientific theory is than that. According to Peter Ellerton Newton didn't have a theory of gravity, that didn't come until Einstein:
watch?v=ENZRZ9lAYAo
"it is a fact that natural selection and other evolutionary processes create diversity of life."
It is not a FACT that natural selection, etc. create diversity of life - it is a suggestion that is supported by the evidence...
If we found s better explanation for speciation we would discard natural selection - you can't do that with facts...
Words DO matter....
Creationists like to say that "Evolution is not a fact - it's just a theory..." (I know they mean theory as in a guess...)
But this is exactly the problem that Hume was bringing up. The concept of "predictive ability" is technically meaningless.
Although I do somewhat dislike the idea of inventing another world in order to alter our model of the world as it appears to achieve desired ends. That does seem a little dishonest.
..or just pick the oldest of the baldwin brothers. that should work as well :PPP.
"Facts - are brute truths discovered through observation. Common descent of all life is one such fact,.."
Common decent is NOT one such fact...
Change in biological expression and diversity over time is a fact - just observe the fossil record...
Common decent is the conclusion of one particular theory - Decent With Modification...
"it is a fact that natural selection and other evolutionary processes create diversity of life"
Natural selection is not a fact - it is a proposed mechanism...
You don't actually "know" the Sun will rise tomorrow. You don't know anything for certain about the future.
And you only have knowledge about what has happened. I do not believe anyone can claim knowledge about the future. We can be able to make bets, so to speak, though on outcomes based upon probabilities and act accordingly, however.
You are talking about something unrelated, it seems. What would be related is "what should we believe?" and I would go with beliefs with the greatest predictive ability. You must first have a model of the world before you think how to act. You are skipping to the end.
I disagree with the chicken analogy. I don't think it's really being a very good empiricist by simply observing the small world of her pen. Rather, a good empiricist should be constantly pushing the boundaries of what it observes as we've done with space probes and microscopes and global communication. For instance, the chicken should fly the coop and observe other farmers' interactions with chickens. Maybe she's been studying chicken anatomy when she sees a farmer munching on a piece of meat that she can now identify as a chicken leg.
If it only employed these methods, I think the chicken could do quite well for itself.
I think the idea is that the "chicken" represents us and the" coop" represents the limits we have to observing our universe. Think to what Hume said about the laws of nature, we will never be able to actually observe them, only see the effect they have. We literally cant fly out of our coop, and if we cant, what are the implications on what we can know.
Irving Ceron Funny enough, it seems the LHC is proving Hume was right.
+Xander Patten You fail to understand the analogy. The coop is our capabilities and all of your examples are within the coop.
SVU Man You fail to understand my critique of the analogy.
No I don't. All of those examples are still within the realm of the coop. None of those transcend the coop.
Come OFF it.. you're trying to correct the usage of "fact" and "theory", while just making it twice as confusing.
Facts - are brute truths discovered through observation. Common descent of all life is one such fact, and it is a fact that natural selection and other evolutionary processes create diversity of life. But evolutionary theory is, unlike facts, is *an explanation* of how those processes work, it makes predictions, allows for calculation, measurement, comparison.
Part 2:
You create out of whole cloth a type of theory you call "law like" - no such thing...
I believe you are misled because you talk about confirmation.....
You misuse confirmation - as if it is kind of like "prove" or something...
If you go back and replace every place where you use "confirm" (etc.) with "supports" (etc) you will find that both theories are the same...
Testing one piece of copper gives as much confidence in the conductivity of all copper as you have for third children
Bingo.
That is fine - as long a he stays in philosophy and doesn't move into science...
What he is saying is silly...
There is no "law like" hypothesis...
One hypothesis is just as valid as any other hypothesis.- until it is tested...
A single test is not powerful in supporting a hypothesis - but is a step...
The hypothesis of the third child is supported by his single data point...