Causality & Laws of Nature in response to Wittgenstein & Hume
I just listened to an old podcast on The Partially Examined Life website about the Tractatus and Wittgenstein's view on science. Wittgenstein put forth a Humean view of causality in which A just happens to always be followed by B, even though both are contingent, such that our expectation that B will follow A in the future is merely one of past habit, which need not hold. C might follow A next time.
The example given is Hume's sunrise every day. The podcasters were gushing about Wittgenstein's view of necessity just being a series of contingent events where the sun always rises, even though it could not rise on any given day.
I find this view of causality to be extremely impoverished. Let's take a coin flip. We say it's 50/50 whether it will be heads or tails. Now if we came across a coin that had landed heads for hundreds of billions of days in a row (the sun rising), then we wouldn't think this was because of some incredibly low statistical event had occurred. We would think that the coin had been rigged to always land heads. As such, we would predict that the coin would continue to land heads because of the rigging. And this is how physicists treat the sun continuing to shine everyday.
They go on to discuss laws of nature as just being logical propositions related to empirical observations of particulars, and nothing further. So gravity is an equation derived from making a bunch of observations of falling objects.
Again, this is an extremely impoverished view of laws. General Relativity talks about gravity in terms of not just particulars, but spacetime itself being bent. And this applies generally across the entire universe, to the point of determining the eventual fate of the cosmos regarding further expansion or contraction.
If laws of nature are merely logical propositions regarding sets of particulars, then why would we expect them to have such far reaching consequences? When Newton arrived at his law of gravity, people were surprised and astonished that there would be a force that pulled the same on cannon balls, feathers, and heavenly bodies. That was not expected.
Furthermore, the relationships between different fundamental concepts in physics, such as acceleration, gravity, energy mass, space and time combines the cosmos into an extremely deep and astounding order that goes beyond noticing that you can apply an equation to some particulars.
The really big question for Humean causation is why would we expect the universe to be contingently ordered to such an astounding degree? Why would it stay ordered for billions of years when it can at any time be otherwise? The sun could blink out tomorrow, and gravity could become repulsive, and so on. Would you expect that kind of contingency to result in the universe we see around us?
The example given is Hume's sunrise every day. The podcasters were gushing about Wittgenstein's view of necessity just being a series of contingent events where the sun always rises, even though it could not rise on any given day.
I find this view of causality to be extremely impoverished. Let's take a coin flip. We say it's 50/50 whether it will be heads or tails. Now if we came across a coin that had landed heads for hundreds of billions of days in a row (the sun rising), then we wouldn't think this was because of some incredibly low statistical event had occurred. We would think that the coin had been rigged to always land heads. As such, we would predict that the coin would continue to land heads because of the rigging. And this is how physicists treat the sun continuing to shine everyday.
They go on to discuss laws of nature as just being logical propositions related to empirical observations of particulars, and nothing further. So gravity is an equation derived from making a bunch of observations of falling objects.
Again, this is an extremely impoverished view of laws. General Relativity talks about gravity in terms of not just particulars, but spacetime itself being bent. And this applies generally across the entire universe, to the point of determining the eventual fate of the cosmos regarding further expansion or contraction.
If laws of nature are merely logical propositions regarding sets of particulars, then why would we expect them to have such far reaching consequences? When Newton arrived at his law of gravity, people were surprised and astonished that there would be a force that pulled the same on cannon balls, feathers, and heavenly bodies. That was not expected.
Furthermore, the relationships between different fundamental concepts in physics, such as acceleration, gravity, energy mass, space and time combines the cosmos into an extremely deep and astounding order that goes beyond noticing that you can apply an equation to some particulars.
The really big question for Humean causation is why would we expect the universe to be contingently ordered to such an astounding degree? Why would it stay ordered for billions of years when it can at any time be otherwise? The sun could blink out tomorrow, and gravity could become repulsive, and so on. Would you expect that kind of contingency to result in the universe we see around us?
Comments (134)
But the claim was that the sun could cease to rise (shine) tomorrow. That it continues to rise is just a contingency that has always held to this point.
I don't see how that's different from the coin always landing heads. It could land tails, but it just doesn't. That sounds no different than probability, except we wouldn't know what sort of probability to assign to a star ceasing to shine, since we haven't observed that.
In the podcast, they were noting that Wittgenstein's analysis showed that scientific laws are mistaken when presented as reasons for why things always happen a certain way. The sun doesn't shine every day because of gravitational pressure resulting in fusion of atoms, that's just what a bunch of atoms in the sun happen to do for billions of years. But they might not tomorrow. However, in our universe, it just happens to be the case that they will (or so we think).
Thus Wittgenstein/Hume can preserve necessity (if B does end up always following A), while not introducing any mysterious causality. That sounds absurd.
A physical law is a mathematical relation that has been observed and confirmed, between some physical quantity values.
The fact that a physical law has always been known to obtain doesn't mean that it will continue to. For one thing, new physics could be discovered.
For example, Newtonian mechanics turned out to be incomplete, when relativity and quantum mechanics were found to give a better description of what happens physically.
There have been a number of instances in which known physical law seemed to be violated or events inconsistent with it. For example: The black-body-radiation's energy vs wavelength curve; the Michaelson-Morely experiment's result; the planet Mercury's seemingly anomalous rotation of apsides.
Those things were later consistently explained by new physics.
But of course there remain other unexplained events. No one has satisfactorily explained ball-lightning. The universe's expansion is evidently accelerating, contrary to expectation, and implying that there's physics that isn't known. Dark energy isn't explained.
Based on previous experience, it can reasonably be expected that our universe's physical laws are consistent, but not known, and that the events in the paragraph before this one might be explained by physics that isn't known yet.
Presumably it couldn't be proved that a physical world is inconsistent, because it could always be saids that new physics might consistently explain any observed inconsistency.
Why should a universe's physics be consistent? A self-contradictory universe would be impossible,
Quoting Marchesk
Perhaps you find it impoverished because you are used to thinking that causality is something that it is not?
Quoting Marchesk
See, I think that sounds perfectly sane. I think the reason you think it sounds absurd is because it goes against what you thought causality is (but is not.)
History has a way of constraining possibility. So it is true that the world seems to be fundamentally causal in this fashion. We can describe some general law that must be obeyed by every particular material event. Regularity gets locked in by a context.
However physics also now tells us that at the fundamental level - once history and context have been stripped away - then action seems to become a-causal or indeterministic.
Take the decay of an atom. It is a Poisson process. In any instant, the probability of it happening or not happening remains the same. There is no temporal sequence in which we can say that the pressure to decay built until it was an inevitability. The decay remains a spontaneous or "un-caused" event.
Of course, there are still the distal or contextual causes of the decay. Someone or some history would have had to prepare the atom that does the decay. But our conventional deterministic notion of causality breaks-down at the quantum limit. The determinism we assign to the world is revealed to be an emergent feature - a statistical property of large numbers.
So this does not support what you call a Humean view of causation - that it could all be just one mass of amazing coincidences. There is good reason to think in terms of general laws that cause particular events to happen. The weight of history, the weight of context, is a real thing. Completely predictable statistical regularity does develop on the large scale.
But it also has to change our understanding of what we could mean by law or causal determinism. We are only justified to talk about the general constraints on spontaneity. We have to accept that anything could happen next - at the primal or small scale level. Fluctuations rule.
Yet then, the expectation that "anything could happen" recedes to the degree that some regularising history has developed. Once we are dealing with large numbers, the probability that A is followed by B - that the sun will rise tomorrow - becomes "almost sure". We may as well call the probability to be 1, absolute certainty.
Of course, our belief that the sun will rise is bolstered by our having a rational explanation. We know that stars last billions of years, and earth rotates as it orbits the sun. We have a mechanical story of the causes involved.
But again, mechanics is just an account of a system in a state of "absolute constraint". All the degrees of freedom, all the spontaneity, have been adequately suppressed. We can throw away any doubt and model the system as a clockwork.
So mechanics is just a limit state description of fluctuations gone to equilibrium. It is not the way the world fundamentally is - at the small or primal scale. But it is certainly the way the world has pretty much become once it has cooled and expanded enough to be completely constrained by its own history.
Nah, I think it goes against any adequate explanation of necessary relations between A & B.
You are missing the point. First you are not comparing like for like. And this might lead to to your dissatisfaction. There are contingent reasons why the coin is 50/50, whilst the sun coming up is near certain. If a coin comes up heads a thousand times, there is still a 50/50 chance it will come up tails next. Not so with the sun. A lot would have to happen for the sun to NOT rise, for reasons we can offer in evidence. Induction tells us that the flipping of a coin is not like the dawning of the day
Wittgenstein incidentally reminds us the the sun never comes up and asks what would it look like if it was the earth moving and not the sun, as indeed it is. They might look the same, but through complex induction and observation we know better.
But the point about habit was Hume's. In observing billiard balls for the first time, we have no a priori reason to say what will happen when one ball strikes another. Will it bounce back, will the balls break; will the second ball move; or will the second ball change into a bunch of petunias??
Hume demands that we can only observe and record. The only way we can have knowledge about the universe is to see if our observations repeat, and by habitually we can come to conclusions a posteriori. That is perfectly satisfactory as in fact that is all we have ever done.
The laws we devise are consequent on this and not things that the universe is compelled to obey. It's just the way things are. Making physical laws is just a short hand to assist us to describe our understanding, and as such are contingent on the continued observations we make.
Kant doesn't think Hume can do this without causality being a structure of our cognitive capabilities. It's not that we observe B always following A and then come up with the concept of causality out of habit, it's that we're wired to filter the world that way. We expect causality to be a feature of the world like space & time, because that's how we experience the world.
Quoting charleton
The worry here expressed by both Plato and Kant is that skepticism is the result, not knowledge. Sensory impressions alone can't give us knowledge. There must be something that structures our experiences, whether it be Kantian categories of Plato's forms/remembrances.
That still doesn't answer the question as to why the sun would rise hundreds of billions of times in a row. The claims is that there is no reason for the sun to continue to shine, it just does. This is at odds with scientific explanation, which posits reasons why the sun shines, and thus it's perfectly valid for us to expect it to continue to do so. This isn't because of habit, it's because of gravity and nuclear physics.
That's a much better attempt than mere regularity. Regularity renders everything as brute. The sun could stop shining for no reason, but it just continues to shine for no reason. The fact that we can come up with good explanations for many necessary situations belies this account of causation.
It's only when we get down to the quantum level, or are dealing with entropy that the causal explanations turn into probabilistic explanations, and we arrive at brute posits. But there's no reason to do that for phenomena we can provide causal explanations for. As such, Humean causation is impoverished.
It's not even a question. There is no more reason 'why' the "sun rises" than why there is a universe in the first place!
However what the Enlightenment of philosophy and science has provided us with is the answer "HOW is it that the sun appears to rise each morning." For the answer to that is; it does not! The earth goes round!!! Please consult Isaac Newton for more information; all sought by induction non fingo hypothesis. If you don't like it there is nothing I can do for you. I'm interested in facts. What you want is a church I think. If you are more satisfied with the answe "god did it", you are welcome to it.
It is not at ANY 'odds" with scientific explanation. You seem to misunderstand what science does.
For example, suppose that while out bird-watching Bob suddenly declares
A. "All swans are white".
As Hume might agree, in spite of appearances A is NOT an empirical proposition about swans. And if we were to have a clear understanding of what Bob meant by this sentence we would either have to continue to observe Bob, or we would have to ask Bob for further information. He might for example reply "Having seen twenty swans, I have given up searching for a black swan and have decided to go home"
Likewise if group of physicists declares that "particle A always follows particle B", we can tell what they mean by watching their behaviour when they perform future experiments.
Like with any rule or principle of necessity, what we mean by causality cannot be verbally represented but only behaviourally demonstrated, similar to how a mathematician cannot linguistically represent what he means by "infinity", for it is a rule pertaining to the behaviour of the mathematician and it is not an object that the mathematician is pointing at.
The truth is that ONLY sensory impressions give us all the knowledge we will ever have. If that leads you to skepticism you'll just have to lump it.
The only exception to this is Kant's idea that we are structured to understand space and time, but nothing about that predicts what the universe is actually like - for all of it we have to build on what we can perceive.
You don't understand what the question "why the sun would rise hundreds of billions of times in a row?" means. That's the problem. When you ask a question such as "why X at point in time t?" you are asking "how can we calculate that the event X, and not some other event Y, will occur at point in time t based on events that occured before the event X?" That's all that is being asked by such a question. And such a question may or may not have an answer. This is because it presupposes that the event X can be predicted based on the events that preceded it. That's not always the case.
Quoting Marchesk
That's not what the claim is. The claim is that causality is a human invention. We connect the events. They are not connected themselves. We do this because we want to predict the future. But then, that does not mean we can connect them any way we want. Observations limit the manner in which we can make connections.
There may or may not be a reason for the sun to continue to shine. But you have to understand what the word "reason" means.
Quoting Marchesk
It isn't. You are merely confused. And the reason why it is "valid" for us to think that the past will repeat in the future is because we have evolved in relatively stable environments.
Quoting Marchesk
It is because of observations + habit. Our method of reasoning is a habit. This habit has evolved in relatively stable environments.
History does not constrain possibility other than in the epistemological sense i.e. our method of reasoning relies on history to tell us what is most likely to occur in the future.
Quoting apokrisis
Repetition isn't necessarily locked in, constrained or caused by context. Repetition, like any other kind of sequence of events, simply is. It is simply something that occurs.
The reason the universe appears to be fundamentally causal is because we live in an environment that is very stable (i.e. that does not change too fast.)
Quoting apokrisis
Lack of causal relations isn't always due to epistemology (i.e. lack of history and context.)
Quoting apokrisis
An event X is said to be a coincidence if there is no event Y that preceded it that can be used to predict it. With that definition in mind, events are not necessarily coincidental. Nonetheless, it is events that are fundamental and not laws that we create based on them.
Speaking of my own case, it isn't obvious to me that i call a banana a "banana" due to mental habit, but i might offer it as psychological explanation. Likewise it isn't immediately evident to me that my understanding of colliding snooker balls is based on repetitive familiarity, for me that would again be a psychological thesis.
Wittgenstein made this point in the blue book, that we often interpret our personal actions through the lens of rules or mechanisms as a post-hoc justification or explanation of what we did when defending our actions to others, even if we we did not consciously follow a rule or experience obeying a rule when performing our actions.
Also recall that the Later Wittgenstein even went further to say that one's conscious experience could not be generically described in terms of "following a rule" or "obeying a law", for rules can be variously interpreted and hence cannot be grounded in other rules, while words such as "following", "being guided by" and "obeying" imply no particular experience.
It is therefore only meaningful to say whether or not our experiences are obeying a particular rule if we can be shown in the individual case what constitutes obeying and disobeying the rule by something external to our immediate imagination and supposed "use" of the rule.
Hence putting aside all thesis about what causality is and referring only to lived experience, how do we even arrive at Hume's problem?
I don't normally behave in a matter that is epistemically coherent with the sun coming-up tomorrow because I have a reason to believe it, and neither am i aware of following a habit. I simply act in a certain fashion and then the sun comes up, without me having a reason for my behaviour. And the scientist after giving all of his verbal justifications acts similarly, without reasons.
Right, but the point was that Kant saw a big problem with Hume's view of causation, which was that it led to widespread skepticism, and made science impossible. So Kant's objective was to save science by reintroducing causality and other necessary categories as structures of human thought.
The consequence is the unknowable things in the themselves, but at least we're still able to do science confidently within our human filtered objective (or intersubjective) world of common experience.
Good point. There needs to be something else to show why A necessarily follows B, but D only follows C by accident. Or to show how correlation differs from causation.
It's true that all conceptions of causation have difficulties. I lean toward an underlying relationship between phenomena, because it all came from a common starting point in or prior to the Big Bang, whether that was the quantum vacuum or what not
It's not that there are brute particulars that happen to always behave a certain way, it's that all the particulars are related in a way that necessitates their common behavior. And that's why physics has been so successful in unifying phenomena, such as electricity and magnetism.
In short, there are fundamental underlying relationships to the cosmos that explain the observed regularities. That's the causation, however it works.
But cosmologists do ask and attempt to answer the question as to why the observable universe exists, and how it came to be the way it is. Saying there is no reason why is settling for skepticism before all possible science and metaphysics has been explored. How do we know there isn't a why?
Quoting sime
I don't agree with that notion of mathematics at all. And as for the white/black swans, all that example shows is that induction means we can be wrong about what we take to be universal. That doesn't mean there aren't universal laws.
Plato, Kant and plenty of others have disagreed with radical empiricism. Sensory impressions alone cannot give you any knowledge. You must be able to conceptualize your impressions.
That's not at all what I mean by asking the why question. I mean the causal reason for why B always follows A, not how to calculate a prediction that B follows A. That's the difference between interpretations of QM (for example), and the shut up and calculate folks.
And I do understand it just fine, thanks.
And why have we evolved in relatively stable environments? You realize that in order for biological evolution to happen, the physics have the universe has to be a certain way? It all goes back to the Big Bang. Pretty far reaching stuff. Offering evolution as an answer to why we presuppose causality is just begging the question of why evolution would exist at all in a merely contingent universe.
Quoting Magnus Anderson
No, it's not a habit. It's an evolved faculty for making sense of a causal world, just like eyes are an evolved organ for using light as a means to perceive objects.
The reason for your behavior is because you evolved in a causal environment, where it makes sense for you to understand the consequences for actions that can lead to death or reproductive success.
I've heard otherwise. That young children quickly develop an intuition for object permanence and casual expectation.
Obviously you have to conceptualise, but you can only conceptualise FROM sensory information which is the source of ALL knowledge - quite obviously.
Locke suggests we start as a Tabula Rasa, I do not exactly agree with that, yet without the sensations we have nothing to work on.
The knowledge we start with is very basic; such as where to find milk from a nipple, what is up and down, and hot and cold, maybe. But without experiencing those things even those primitive instincts fail us.
I have no idea what your objection or solution to this rather obvious reality is.
Yes and no. They do not ask why, they DO ask how.
If you want to know why ask a priest, as they have all the answers ready made.
I'm not going to get into a semantic argument over when to use why and when to use how. I take scientific explanations to be causal reasons for the regularities we observe. Humean causation undermines that, which was Kant's concern.
Not quite obviously, or Plato and Kant wouldn't have objected to that and come up with their own schemes for how knowledge is possible.
Quoting charleton
Yeah and people like Steven Pinker and Noam Chomsky have argued otherwise.
Quoting charleton
The objection goes back to Plato in which he argued that the flux of the world presented by our senses cannot be a source of knowledge in and of itself.
The solution in modern terms is that our brains have built-in structures or modules for learning how to apply concepts to sensations to form knowledge.
The question here is whether human beings could learn a concept like causality just from experience, or whether the brain is wired to develop along those lines in response to experiences. A similar argument has been made for language learning. You could apply it to math as well. Of course we learn basic math, but we also have the capacity for learning math somehow, which most animals don't (some birds and apes demonstrate siimple arithmetic abilities).
I'm not accepting Kant's version of causality. Rather, I'm critiquing Wittgenstein & Hume's.
Hume's view does not lead to skepticism and it does not make science impossible. Hume's view is merely a very accurate description of what was already there. In other words, he merely described how science works. You, and many other people, on the other hand, are mystics.
Quoting Marchesk
There is a difference between the two but it's not the kind of difference that you think it is. A simple way to put it is that every causation is correlation but not every correlation is causation.
Quoting Marchesk
It's precisely that there are only "brute particulars that happen to always behave a certain way". Laws are merely human inventions that are based on a selection of these brute particulars. Any other way of thinking is already a form of dogmatism and absolutism.
Quoting Marchesk
No. There are events that happen in a certain order. Each one of us is aware only of a selection of these events. Based on that selection of events, we invent laws. The purpose of laws is to allow us to go beyond what is known to us. This makes it possible for us to predict what's going to happen in the future and to take preventive measures.
Of course the past constrains the future in terms of what is possible. If you break your leg, you won't be running any races. History is the accumulation of a whole lot of events that limit the scope of the future in a definite way.
Quoting Magnus Anderson
OK, now for your counter-argument....
It does, because you have have no justification for coming up with predictive models. Nothing happens for any reason. Just because the sun's always shone doesn't mean we have any reason for coming up with a mechanism for it shining tomorrow. And as such, there's no reason to apply predictive models to the past before human experience. Maybe the universe was entirely different. Maybe the sun popped into existence along with human beings.
Quoting Magnus Anderson
Then scientists are dogmatists and absolutists, because they certainly go beyond brute particulars just happening to behave a certain way to overarching theories explaining how living things came to exist, or stars formed, or how stellar fusion results in heavier elements, which gravity acts upon to form rocky planets and so on.
The future is under no obligation to mimic the past. It can but it does not have to. The important thing is that the future is not compelled, forced, obliged, caused or otherwise constrained by what happened in the past to be a certain way. Rather, it is how our method of reasoning -- and reasoning is a process by which we guess the unknown -- works. It is based on the premise that the future will be maximally similar to the past. The reason our method of reasoning proves to be successful is because the environment we live in is sufficiently stable.
The two aren't mutually exclusive. The fact that the universe is a mass of particulars, and not a mechanism that generates these particulars, does not mean that there is no reason to create theories.
The problem you have is that you cannot accept that predictions and theories are fallible. You cannot live with this fact. You think that if something is fallible that it is necessarily useless. That's not true.
No, the problem is that I don't think you can get from brute particulars to any sort of theory, nor do I think brute particulars would behave in any sort of necessary relationship.
No, I don't think brute particulars are enough to provide the basis for any theory. That's the fundamental problem with empiricism.
Can you translate General Relativity into set theory?
Better yet, can you transform Evolutionary Biology into data sets? I'd love to see how natural selection falls out of that.
Yeah sure. Different argument.
Quoting Magnus Anderson
So now you show you don't get that to be constrained just means to be constrained, not to be determined?
Saying the past shapes the possibilities of the future is quite different from saying the past determines the future.
Quoting Magnus Anderson
Huh? Sure, reason is about inference to the best explanation. And we find that causal thinking works.
But I was then talking about our ontic commitments. The various ways we can view causality.
My point was that constraints-based causal thinking works better than talk about absolute laws or mechanical determinism. So my stress is on the evidence for a fundamental indeterminism in nature - the quantum facts. And then how that gets resolved by a constraints-based or contextual understanding of why the world seems classically determined on the whole. Classical regularity and predictability emerges due to large numbers and an emergent regularity that is probabilistic.
So are you doing anything other than not understanding that I am talking about a different explanation for the regularity of the Cosmos at the classical scale of observation?
Quoting Magnus Anderson
Well duh. Why is it sufficiently stable? Has change been regulated by a past that is an absolutely stable context?
(Or in fact not absolutely stable if we take into account the retrocausality that seems to apply at the quantum scale of individual events - https://en.wikipedia.org/wiki/Delayed_choice_quantum_eraser)
It's possible. However, it would require a lot of work. One of the reasons we create theories is in order to make knowledge independent from experience. Once you come up with a theory, anyone can use it. It does not matter whether or not they have the experience necessary to come up with. All they have to do is to follow the instructions.
Not at all. Explain how? Show what Kant says.
I understand that. I understand the difference between absolute and relative limits. The thing is that you do not understand my point. My point being that it is strictly speaking incorrect to say that the past constrains the future. It does not. That might be how we speak. But that's not correct.
Quoting apokrisis
I can agree with that.
Quoting apokrisis
But then you say something like this and I cannot help but think that you're doing something wrong. You need to understand what a "why" question means before you ask one.
Eh heh.
OK, I hear your assertion and await the supporting counter-argument. What could be more accurate than saying the past constrains the future?
It is inaccurate to say the past absolutely determines the future - that there is no actual quantum grain of free spontaneity.
And it would be even more inaccurate to say the past leaves the future completely undetermined, or radically free and spontaneous. On the whole - as you agree about stability - the future seems pretty classically predictable.
So why is my constraints-based view of causality incorrect when - strictly speaking - it covers both the classical determinism and the quantum indeterminism?
Quoting Magnus Anderson
I was hoping you might answer the question rather than question the question format.
But sure, if you think it needs rephrasing, will this pass your test? Will you answer now?
Quoting apokrisis
OR - you could answer my question "Tell me the non perceptual source of knowledge of which you speak!"
Cognition is the non-perceptual source of knowledge.
But Einstein notices a connection between acceleration and gravity, and posits the acceleration of objects through curved space as the gravitational force. So now you've moved from a shorthand for particulars to a very general principle. And not only are objects part of the principle, but light itself, which we have measured. Large distortions in space result in gravitational lensing. And furthermore, length and time get tied into this, along with frames of references.
That sort of scientific theory, like natural selection in biology, goes very much farther beyond noticing similar behavior among particulars over time. It explains why the particulars behave in a similar manner. That's fundamental to scientific theories.
Put a new born baby in a sensory deprivation chamber and see what you get.
You have not really used joined up thinking.
And nothing can be perceived without cognition. Remove a newborn baby's neocortex and see what knowledge they will learn.
Not really. All science is based on evidence. We are just getting better at it. Newton gives way to Einstein, who in turn may well be shown to be inadequate. Einstein's work is observable. If not then its not valid.
You are making my case for me, if you would but know it.
You can't ignore the role of theory in science. Positing concepts to explain phenomena is as central to science, as is putting those ideas to the test.
Not really, because you need the rational faculties to make sense of the empirical data. How we obtained or develop our ability to reason is a separate matter.
But the point is that there is no perception without also conception. Perceiving presumes "a world" for a start.
So the exact mechanism is very relevant to any theory of epistemology.
Psychological science makes many striking points of this kind. Any image stabilised on our retinas quickly becomes invisible. Our eyes have to dance with micro-saccades to keep the fixed and constant aspects of the world in sight.
So what is going on there? For some reason we are set up to tell what parts of the world are still by introducing motion into our view of the world. We in fact perceive the stability of the things "out there" in contrast to the instability that we can generate "in here, ourselves" as the necessary contrast.
There just is no "simple looking and seeing what is out there". It all starts with the organising presumption of our conceiving of "a world" in which "we perceive".
Even knowledge of the mechanism has to be perceived as evidence. By saying that there are OTHER sources of knowledge has to be false.
Quoting apokrisis
I never even implied that. I said what I said, that the source of all our knowledge is from what we perceive. There is simply no avoiding this.
But I even gave exceptions contra Locke, that we have limited instinctual "knowledge".
The problem is 'simple looking' is not a problem for me, as accepting the limits of perception is the only clear way to begin to over come them.
But there is something in the Kantian position that to get the ball of inference rolling, an abductive leap has to be made.
So the strenuous way you are arguing here certainly makes it seem your goal is to deny this. Evidence just accumulates and becomes our ideas. We are Lockean tabular rasa, Behaviourist association machines.
Quoting charleton
So you are saying perception does have a foundational element of conception to it. Isn't Marchesk saying the same thing? Isn't this a glass half full/half empty argument here?
Let's set aside the rationalism/empiricism debate, since this thread is about whether Hume/Witty's version of causality is adequate.
Resetting the issue: according to Hume, the only reason we think that B will continue to follow A is that it has so far in the past. Thus, our expectation that the sun will rise tomorrow is based on nothing more than it having risen before.
However, science says we have confidence the sun will rise tomorrow because it still has matter it can fuse due to it's relatively intense gravity. And furthermore, this will continue for a few more billion years until it can't fuse any more elements, and then it starts expanding and turns into a red giant. There is a reason the sun has been shining for billions of years.
As such, science isn't just cataloging Bs following As, it's looking to provide explanations for B following A. That's what Hume's account leaves out.
IOW, science operates under the assumption that there are causal explanations to be had. It's possible this doesn't always turn out to be the case, depending on how one interprets QM, but for a wide variety of phenomena, it has so far.
But Hume also says we have no logical reason to suppose the constant conjunction will continue. He presents a skeptical view of the future, and thus undermines prediction.
The past does not constrain the future. Rather, our method of reasoning is founded on the premise that the future will be maximally similar to the past.
The reason such a method of reasoning is useful is not because the past constrains the future. Rather, it is because the events, within the environment that we live, happen to mostly unfold in a manner that fits models created using such a method of reasoning.
Quoting apokrisis
It is inaccurate to say that the past determines the future in any sort of way, relative or absolute. Instead, what is accurate is to say that there is a relation between two points in time. Relation is nothing but a measure of similarity/difference. For example, a later point in time can be very similar to an earlier point in time. The degree of similarity can vary. It can be absolutely similar (i.e. no difference in all relevant aspects) or relatively similar (i.e. minor differences in some of the relevant aspects.) Similarly, a later point in time can be very different from an earlier point in time. It can be absolutely different (i.e. no similarity in all relevant aspects) and relatively different (i.e. minor similarities in some of the relevant aspects.)
Quoting apokrisis
It is not inaccurate to say that the past leaves the future completely undetermined. What is inaccurate is to say that the future breaks free from the past in any sort of way, relative or absolute.
The presence of relation between two events, i.e. a degree of similarity or difference between them, requires no underlying mechanism. Rather, it is intelligent organisms that require such a mechanism because they want to predict the future so that they can more effectively attain their goals (whatever these goals are.) Mind you, such mechanisms can only be useful in relatively stable environments. But that does not mean that stable environments are a product of some underlying mechanism. They aren't. It is an unnecessary assumption.
Quoting apokrisis
It makes things unnecessarily complicated.
Quoting apokrisis
It does not matter whether you ask "why" or "how". These two types of questions are very similar to each other. In fact, I'd say they are two sides of the same coin. I'll try to explain what's wrong with them in another post.
Why should Humean account (as you describe it) be confined just to Newton's theory of gravity? This contrast between Newton's and Einstein's theories that you keep invoking is quite puzzling. Both are empirical theories that are aimed at explaining certain categories of observations. Relativity has a wider scope of application and, where it can be applied to the same observations as Newtonian physics, it fits some of them better.
Nevertheless, you could paraphrase Einstein's theory in terms of what you call the Humean account just as readily as you did it with Newton's theory: Relativity is just a shorthand for objects [and all of the observables to which the theory applies] just happening to behave [as far as we know] as the theory says they should behave, as a matter of brute fact, without any underlying metaphysical powers of causation constraining or impelling them.
Now, I am curious: how would you distinguish such a "Humean" universe from one that is "enriched" with your favored metaphysics?
But while I puzzle, here's a longish talk about Wittgenstein and Turing that might amuse.
https://vimeo.com/241850881?ref=fb-share&1
I think science is implicitly realistic, even though people figure out ways to talk about in non-realist terms. The Newton example wasn't meant to say that Newton was Humean in his account. He was not. It was just an example of going from particulars to general law. The problem with Newton's realist account of causality is that he couldn't explain gravity as a force acting at a distance, but Einstein could.
The reason for thinking science implies or assumes realism is because unobservables and general laws are posited as part of the theories.
His argument is more subtle, suggesting that a priori reasoning does not predict the outcomes of causal interactions and that as humans we have since time immemorial simply had to OBSERVE and conclude from observations causality.
If you have something better, let me know.
And Kant's argument was that we couldn't have come up with causality by just past observation. It wouldn't be something that could occur to us as a concept.
If Hume had come up with a skeptical argument for space or time, the same Kantian critique would apply. Habit or custom cannot create a fundamental concept.
It's fight picking, without due consideration for considering what I have said.
1) the source of all knowledge is perception. This is irrefutable
2) If there is an exception it would have to be an abuse of language which considersinstinctive intuition as the same thing as knowledge; it is not.
3) If we were utter tabula rasa then we would be rendered incapable of understanding percpetion, and all sensory data would be white noise.
4) Kant's notions ofCategories of an inherent substrate understanding of time and space act to give us a ground of understanding upon which to build our interpretation of the world might be helpful in this, but his understanding of psychology is far too logical/ cerebral and anthropocentric in my view, and does not account for how simple animals function too.
I have more respect for Hume than Kant. Kant's view is tainted by his theism as this gives him licence to impart humans with whatever quality God wanted to bestow on them - however incredible. Humans are an example of a special creation, able to summon up objective knowledge without so much as an agreement. Hume was a skeptic and does not allow himself to enter into any such flights of fancy.
And even better yet, you read Kant. This is a discussion, not a book reading club.
Quoting charleton
...seems too strong.
I think this is because knowledge and perception are already such loaded words. Knowledge suggests direct realist truth. And perception is properly a conception-loaded process - so already crossing over into idealist territory - in psychological science.
We probably agree on the essential issue. The mind only exists as a modelling relation with the world. So all information that shapes states of conception are ultimately derived from that relating.
But note the stress on the information resulting from the relating, not actually from "the world". The source is not the world in any direct unmediated sense. It is the possibilities of the relation that are the source of any knowing going on.
There is no knowledge without perception, not even Kant disagrees with this. I'm not saying perception is knowledge. But you seem to what to avoid that what we perceive is the source and fudge it by saying it comes from a 'relation'. Even if that we not a fudge, it would still not avoid the irrefutability of the statement, as there is no relation without that which is perceived, and ipso fact the perception IS the relationship between the world and the knowledge.
:D
So where you are fudging things is talking about "the knowledge" and not "the knower". You concede my point about the relating, but seek to reify "the facts".
Sneaky.
He might have clarified Hume in this respect he did not add to him. Hume's work assumes that we understand causality. It is just Kant's hubris to claim that he had a special idea my making causality a basic category, which is, with space and time, not "knowledge as we accept it" but the ground of understanding upon which knowledge is built.
Perception is the relation, obviously.
Since you have mentioned the "knower", you would agree that something lies between the knower and the thing to be known about, and that relation is perception.
Well which do you want to say it is?
My position is that the relating creates the division into knower and known. So I don't in fact claim there to be a "knower", let alone "the knowledge". If we focus on the relating, we just see a process of experience becoming structured a certain way.
So talk about perception as a relation is talk of a relation that points in both directions. It points also to the kind of mind that would be necessary to see the world in a certain fashion. It points to a particular habit of interpretation.
Hume and Kant and many others could appreciate that there is something fundamentally knotty about epistemology. We are too wont to simplify that.
This is an abuse of language.
I get that you see the world through the eyes of a reductionist ontology. You don't abide with holism or systems thinking.
Your loss.
Quoting apokrisis
Wrong.
Are you on drugs?? Uh huh!!
Quoting charleton
Are you only a recent visitor to our planet?
And if I were a visitor, the god assumption would be a reasonable one. There are plenty of god botherers out there and several of them on this Forum.
As usual, you missed the fact you also had to mention the "you" that has "the experience".
Quoting charleton
Yep. The cognition does have to come from somewhere. The "you" has to be accounted for.
And clearly I account for it in evolutionary fashion. I don't invoke a soul.
I agree. Though I understand the practical and emotional reasons for seeking this mechanism. What is conceptual thinking is inherently 'mechanistic' ? It seems that we automatically model the world in the sense that we notice violations of our otherwise unconscious expectations. These expectations can become professional/scientific and hardened into 'laws.'
Quoting Magnus Anderson
In my view, they would be ontological reductionists only by denying the reality of what wasn't the mechanism itself. Or if they closed their minds to other structures in experience. I suspect that we are all reductionists whether we like or not, but I like philosophy that strives against our tendency to clamp down on a particular mechanism.
As usual, you miss the fact that subject (that which we say perceives) and object (that which we say is perceived) are both part of the experience. Subject-object relation is nothing but a relation between two objects of experience. You are confusing direct knowledge (what has been experienced) with indirect knowledge (what hasn't been experienced but can be assumed, expected, guessed, predicted, retrodicted, inferred, etc based on what has been experienced.) Your position is that all knowledge is indirect knowledge. That's not true.
In order to attain our goals, we must predict all of the relevant events. In order to be able to predict them, they must be predictable. That's why we love predictability. If there is a mechanism that generates every event that we observe then this means that, if we knew how this mechanism operates, we could predict these events with absolute precision i.e. without ever making a mistake.
The fact is that it is us who create such a mechanism. And we do so based on our direct knowledge i.e. based on what we have observed in the past. Such a mechanism can be formed to the extent that there is regularity, or more precisely homogeneity, within the observed events. If our data isn't regular or homogenous then no such mechanism can be formed. You can proceed to collect more data with the hope that there is a hidden order you are not aware of but this isn't a guarantee that further experience will make your data homogenous. Finally, no matter how homogenous your data is, further experience can always destroy its homogeneity.
The only way around this is dogmatism or absolutism: you just declare that the universe works according to some mechanism regardless of any evidence.
Quoting 0rff
Reductionism isn't a bad thing per se. Reduction is a very useful tool. It allows us to create models of reality which in turn allow us to make predictions. We cannot make predictions without reduction.
But that is what I am saying. That is what makes knowledge indirect.
We don't see the world directly - which would be a simple dyadic relation of self and world. We see it indirectly via a system of sign. It is a triadic relation. In-between me and the world are the states of sensation which represent to me the world as I could best understand it in terms of my goals and needs.
So the red I experience stands on the side of the experience. The world isn't actually coloured. It just scatters patterns of radiant energy. At the first step of the retina, this energy is transduced into a neural signal, some change in firing rates of a network of cells. The result is that I see the red of the post-box. It serves to sum up the world in a way that is ecologically useful.
Red and green are barely any different as actual quantities of energy. Red light is about 700 nanometres in frequency. Green is about 540nm. So even if the world reflected an even balance of both, an eye that just registered the energetic content of light would struggle to tell different "colours" apart.
But the visual pathways are set up to turn that slight contrast into a binary signal. Retinal cells are designed so that a "green" receptor is switched off in the presence of "red" light. They signal the absence of green in the presence of red, as well as signalling green in its presence. So a slight difference in frequency is turned into an absolute on-off mental response.
So yes, there is a world of energy that drives the sensory processing. There are real patterns out there. But what I am drawing attention to is the transduction step which creates - for us - an internal realm of signs. And that triadic relationship - where we use constructed experience as our map to navigate the world - is a Kantian epistemology.
It is the reason for saying we can't know the world directly. Experience is grounded already in a pragmatic structure. Right down at the first step - neural transduction - energy has been turned into signal. A particular logical operation has been applied. Wavelength has been turned into yes-no contrast. Everything real or physical about the world has been discarded as information to leave just the signalled output of a mechanical operation. A retinal ganglion cell summed its inputs and changed its firing rate in some way.
Quoting Magnus Anderson
Note that when I talk about reductionism in contrast to holism, it isn't about information reduction. I am of course a big believer in that. As I have just argued, that is what neurology is set up to do. The brain doesn't want to see "everything". It wants to reduce all the available energy of the world to a simple yes-no choice about what to do. And its starts doing that at the very first step of the neural processing hierarchy.
So the holism vs reductionism deal is about Aristotelean causality. Reductionism is taking the view that only bottom-up material and effective cause is "real". Holism is the larger view which accepts top-down formal and final cause as also "real". So it is about making models of reality which are just bottom-up, versus making models of reality which are about the triadic interaction between the bottom-up and the top-down (triadic, as the real is what emerges from these two complementary directions of cause).
Scientifically, both reductionist and holist models would be reductionist about the world. They are both models after all - maps of the territory. And the best maps are the simplest. They discard the most information.
Reductionist or bottom-up models are fine for most routine human purposes as we are only concerned with how to harness the material/effective causes of the world. We don't need to worry about the formal and final causes of things because that is going to be provided by us humans. We will supply any necessary design and reason when it comes to actually doing something with the modelled knowledge.
However when we come to describing nature itself, then we do have to include formal and final cause in the story we tell. We do have to speak to all four of Aristotle's causes.
And before we can make such a claim, i.e. that we don't see the world directly, we must have a clear understanding of what it means to see the world directly as opposed to indirectly.
For example, what does it mean to say that an observation that an apple on a table is red is not a direct knowledge of the world? Most people will agree with this statement, i.e. that an observation that an apple on a table is red is not a direct knowledge of the world, but how many will be able to explain what that statement means? Very few, right?
I make a distinction between facts and interpretations. Facts refer to what has been experienced in the past e.g. an observation that an apple on a table is red is a fact because it is something that has been experienced in the past. Interpretations, on the other hand, refer to what hasn't been experienced in the past but has been assumed in the past e.g. an assumption that there is a tree in a garden when noone is looking at it. Facts are direct knowledge of the world, interpretations are indirect knowledge of the world.
It is clear to me that you operate with a different understanding of what it means for knowledge to be direct as opposed to indirect. What I am asking you is to make this explicit.
If you couldn’t get that from my post then I’ll leave it there.
Experience on its own is neither subjective nor objective. Rather, it is objects of experience that can be either subjective or objective. To say that an object of experience is subjective or that it is objective is to say that it belongs to the category designated by the word "subject" or that it belongs to the category designated by the word "object". To determine whether any given object of experience belongs to the category designated by the word "subject" or to the category designated by the word "object" one must first understand the membership rules of these categories. The problem with saying that all experience is subjective (i.e. that every object of experience belongs to the category designated by the word "subject") or that all experience is objective (i.e. that every object of experience belongs to the category designated by the word "object") is that it does nothing but change the rules of these two categories so that one category becomes all-inclusive (which means there are no longer objects of experience that it excludes) and another all-exclusive (which means there are no longer objects of experience that it includes.) One must first understand the definition of these categories before one proceeds to determine which one of the two categories any given object of experience, or even every object of experience, belongs to. Given my understanding of the definition of these two categories, my position is that experience as a whole, i.e. the set of all objects of experience, does not belong to any of the two categories. Rather, some objects of experience belong to the category designated by the word "subject" and some objects belong to the category designated by the word "object".
There is no need for Charles Sanders Pierce kind of obscurity.
Everything thought, believed, uttered, and/or written is existentially contingent upon the attribution and/or recognition of causality.
Hume's 'problem' of induction is existentially contingent upon the idea of logical possibility. It is grounded upon logical possibility alone. If there is no empirical evidence to warrant belief that the sun will shine in the morrow, there is even less to believe that it will not. Logical possibility alone does not constitute sufficient reason to believe.
Objects are not subjects. Subjects are not objects. Sensibility requires drawing and maintaining a meaningful distinction between the two. Calling one - the other - renders that distinction absent, and renders the terminological use... incoherent and/or self contradictory. Aside from that...
I find that dichotomy unable to take proper account of that which is neither.
Objects and subjects.
Truth. Meaning. Thought. Belief.
All four consist in/of relationships whether they be recognized and/or attributed. Relationships are neither objects nor subjects. Some relationships are existentially contingent upon human awareness. Others aren't. None of the above are.
Actually your objects might be my subjects and vice versa.
If your objects are your subjects there's a problem. If my objects are my subjects there's a problem. If my objects are your subjects there's a difference in frameworks. So what?
This implies that you are applying a false objectivity by ignoring the ubiquity of varying frameworks.
Aren't you stating a fact that we see signs? How did you get at that fact if not through signs? How else can one get at facts?
You're lost.
And yes, I can mean to speak of the facts of the world. That is the realism in the indirectness.
If direct knowledge is everything I have experienced in the past, then what would indirect knowledge be?
What's the term "a" doing in the second sentence?
That which you haven't experienced but assumed in the past. For example, I haven't experienced that dinosaurs lived on Earth but I have assumed, i.e. imagined based on the available evidence, that they lived on Earth.
What's happening here is I am not expressing myself precisely and you are taking advantage of this to make conclusions that do not follow.
Perception is either direct, indirect, or both. That all depends upon one's criterion for what exactly counts as perception. Historically, it has included everything from simple thought and belief to complex conceptual notions.
Are you saying that you cannot differentiate between a memory (e.g. what you wrote to me just a few minutes ago) and an imagination (e.g. what you may write to me in the future)? Or are you asking by what mechanism do we know what is a memory and what is an imagination?
Quoting creativesoul
So you think there is no difference between seeing that it is raining at some point t in time (direct knowledge) and assuming that it rains at some point t in time (indirect knowledge)?
Perhaps the latter, although the term "mechanism" isn't part of my normal dialect.
I'm asking what the difference between an imagination of past experience and a memory thereof is. I suppose the question could be put differently:What is the difference between imagination and memory, and more importantly how can we distinguish between them?
Quoting Magnus Anderson
Sure. The former(seeing) is direct perception. The latter(inferring) is indirect(by historical accounts). Both are experience.
To say that there is "indirectness" is to imply that there could be "directness". What would it be like to have "direct" awareness or knowledge, as opposed to "indirect" awareness or knowledge?
Say some being has direct awareness of everything, and I only have indirect awareness of everything, (which, BTW, leads to an infinite regress, as how can I access anything, including the signs, if I'm always indirectly getting at it all? I must be able to get directly at something, which is the signs themselves). If we both end up stating the same facts, then what is the difference? If we end up getting at the SAME facts anyway, whether it's "indirect" or "direct", then what is the difference? And don't we have "direct" access to the "signs" themselves? The "signs" are just as real as everything else.