What's the difference between causation and correlation?

@Ultracrepidarian,

Ultracrepidarian;69569 wrote:

opium induces sleep because it is made of molecules that induce sleep?
Fascinating, I would like to know more. What kind of molecules? What are they called? What do they look like? What is the difference between these molecules and the molecules of water? between these opium molecules and milk, alcohol? What kind of sleep does opium induce? Is it deep sleep? Restful sleep?

Sure, you can try to answer these questions without telling me anything. but more than likely I'll learn something about the nature of opium and why it induces sleep. It won't be shuffling ignorance.

If you want to know about the molecules of opium, and why they induce sleep, I am sure that a chemist can explain it to you, or, probably, it is to be found of the internet. Why not look it up if you are interested? I don't know the details, and, anyway, why would you want a chemistry lesson on this forum?

@Aedes,

Aedes;69520 wrote:

But "in the real world", we have functional truths and functional causation. As Satan said in his reply to me, "negligible" means that something can be ignored, but that does not make it disproved. That is correct. But if all our decisions and assertions were beset by negligible (but not disproved) doubts, we couldn't function. In other words, there needs to be a functional, "useful" concept of causation whether or not it measures up in the most abstract, exacting terms.

The usefulness of the concept of causation suggests to me that there is something 'real' about it. Otherwise, we could get by with the concept of correlation alone. Admittedly, causation cannot be conclusively proved, but it is a non sequitur to claim that it therefore does not exist. In fact, there is overwhelming statistical evidence that it does. We constantly find that, out of a range of theoretical outcomes of an event, the statistics are strongly skewed towards one particular outcome: heated metal almost always expands, water usually flows downhill, and so on. These 'strong' correlations are, as pointed out earlier, different in kind from 'weak' correlations such as one clock striking just after another.

Then there is the question of prediction. We all act (even after philosophical reflection) as if we believe that strong correlations will continue to apply in the future, i.e. that there is causation. Sometimes this involves taking inconvenient or arduous precautions against some expected danger. If it is said that we could not function without assuming causation, the obvious question is: why not? The functionality of causation has to be taken seriously. We should think not only about absolute proof but also about evidence, which is just as valid a philosophical concept.

Satan

@ACB,

ACB;69579 wrote:

The usefulness of the concept of causation suggests to me that there is something 'real' about it. Otherwise, we could get by with the concept of correlation alone. Admittedly, causation cannot be conclusively proved, but it is a non sequitur to claim that it therefore does not exist. In fact, there is overwhelming statistical evidence that it does. We constantly find that, out of a range of theoretical outcomes of an event, the statistics are strongly skewed towards one particular outcome: heated metal almost always expands, water usually flows downhill, and so on. These 'strong' correlations are, as pointed out earlier, different in kind from 'weak' correlations such as one clock striking just after another.

Then there is the question of prediction. We all act (even after philosophical reflection) as if we believe that strong correlations will continue to apply in the future, i.e. that there is causation. Sometimes this involves taking inconvenient or arduous precautions against some expected danger. If it is said that we could not function without assuming causation, the obvious question is: why not? The functionality of causation has to be taken seriously. We should think not only about absolute proof but also about evidence, which is just as valid a philosophical concept.

Satan

It isn't a tautology if, as is true, the freezing point of water is 0 centigrade.

I wonder what you mean by, "a total explanation of everything"? A total explanation of every single thing in the universe, or, a total explanation of the whole universe considered collectively? Never mind what "total explanation" would mean, and never mind if we got that straightened out, whether there could possibly be such a thing.

But, as I have pointed out, we can know what the explanation of why water freezes is, without knowing what the explanation is of the explanation (or even assuming that there is one). Much less knowing what the total explanation of everything is, whatever that may mean, or, much less, knowing whether there is such a thing. As Joseph Butler wrote, "Everything is what it is, and not another thing".

@Satan phil,

It's not a tautology to say "the freezing point" either, because it's NOT always 0 centigrade. It depends on the solutes and on the atmospheric pressure. "Freezing point" is a shorthand for the temperature at which a chemical phenomenon termed "freezing" occurs, but one could express the same thing using words other than freezing -- it would just be a lot more cumbersome.

@Aedes,

Aedes;69582 wrote:

It's not a tautology to say "the freezing point" either, because it's NOT always 0 centigrade. It depends on the solutes and on the atmospheric pressure. "Freezing point" is a shorthand for the temperature at which a chemical phenomenon termed "freezing" occurs, but one could express the same thing using words other than freezing -- it would just be a lot more cumbersome.

In standard conditions, the freezing point of water is 0 C. That is certainly not a tautology.

@Satan phil,

Outside of a laboratory the freezing point is very close to 0, but not exactly 0 unless you've got a thermally closed system and distilled water at 760 torr atmospheric pressure. So unless you want to get into answering "what is 0" based on the number of significant figures, you CANNOT say that the freezing point is truly 0 outside of a lab.

@Aedes,

Aedes;69585 wrote:

Outside of a laboratory the freezing point is very close to 0, but not exactly 0 unless you've got a thermally closed system and distilled water at 760 torr atmospheric pressure. So unless you want to get into answering "what is 0" based on the number of significant figures, you CANNOT say that the freezing point is truly 0 outside of a lab.

Does this kind of thing really have anything to do with the issue? After all, it's only an example. I suppose it is easier to talk about than the philosophical issue, though.

@kennethamy,

Suppose you had a dream, and in your dream you picked up a rock, observed it, and concluded it was hard because of its molecular makeup. You would indeed be wrong, and if you had the same dream a million times, you would still be wrong. How do you know you are not dreaming now? Maybe reality is just a more elaborate dream created by an infinitely more powerful mind, and the only true causation has to do with the fact that we perceive anything at all.

@Neil D,

Neil;69599 wrote:

Suppose you had a dream, and in your dream you picked up a rock, observed it, and concluded it was hard because of its molecular makeup. You would indeed be wrong, and if you had the same dream a million times, you would still be wrong. How do you know you are not dreaming now? Maybe reality is just a more elaborate dream created by an infinitely more powerful mind, and the only true causation has to do with the fact that we perceive anything at all.

Eh, why would I be wrong if I dreamed that the rock was hard because of its molecular make-up? Suppose I dreamed that my house was on fire, and when I woke up, guess what? my house was on fire, and it was on fire when I dreamed it was on fire.

@kennethamy,

kennethamy;69589 wrote:

Does this kind of thing really have anything to do with the issue?

It does, because I think it's perhaps MORE accurate to speak of a "freezing point" than it is to speak of "0 degrees C", which is a concept that requires elaboration unto itself. And incidentally, the centigrade scale was created in order that 0 be the freezing point of water and 100 be the melting point. At least in the phrase "freezing point", we can rephrase that in terms of a physical phenomenon.

@Ultracrepidarian,

Ultracrepidarian;69569 wrote:

opium induces sleep because it is made of molecules that induce sleep?
Fascinating, I would like to know more. What kind of molecules? What are they called? What do they look like? What is the difference between these molecules and the molecules of water? between these opium molecules and milk, alcohol? What kind of sleep does opium induce? Is it deep sleep? Restful sleep?

Sure, you can try to answer these questions without telling me anything. but more than likely I'll learn something about the nature of opium and why it induces sleep. It won't be shuffling ignorance.

Those will be descriptions, not explanations.

Aedes;69572 wrote:

Our use of the word "cause" in anything but abstract philosophical banter IS a practicality. If someone goes to the mechanic and ask why your "Check Engine" light turned on, they are inquiring about cause. If the police investigate a murder, they are trying to discover the human whose volitional agency caused a death. Our actions and are conversations are riddled with the concept of cause. So cause, therefore, has a functional meaning that does not require that it be absolute.

That's practical life. What does that have to do with the truth? Do you subscribe to the belief that useful equals true? Otherwise, why do you keep appealing to practice when I'm talking about theory?

@Satan phil,

Satan;68889 wrote:

What exactly does it mean for A to "explain why" B happens? What is the difference between a description and an explanation? What's the difference between saying "metal expands after you heat it" rather than saying "metal expands because you heat it"? What's missing in the former that's present in the latter?

be-"cause" is the different from "after", as it ascribes causality ?

@memester,

memester;69688 wrote:

be-"cause" is the different from "after", as it ascribes causality ?

All I ever see are "afters" I never see "causes". After I flip the light switch, the light comes on (usually). I don't see any cause. What's the difference between B happening after A and A causing B? How can we test the difference?

@Satan phil,

Satan;69686 wrote:

Do you subscribe to the belief that useful equals true?

Sort of. The way I'd phrase it is that the only true that humans have access to is that which seems sufficiently true. We have no access to absolute truth -- talking about absolute truth belongs in church.

@Aedes,

Aedes;69705 wrote:

Sort of. The way I'd phrase it is that the only true that humans have access to is that which seems sufficiently true. We have no access to absolute truth -- talking about absolute truth belongs in church.

I subscribe to "we know what we make". Humans make logic and humans define truth so we can know the truth about logical systems absolutely. Humans didn't make the universe so we can't know anything about the universe absolutely. The only caveat is that since we make logic up, it's rather arbitrary. That's where the appeal to usefulness comes in. It may be arbitrary but it's useful. Under my view, there are still things that are absolutely true and absolutely false, tautologies and contradictions. I know it is absolutely true that it will either rain or not rain tomorrow. I know it is absolutely false that it will both rain and not rain tomorrow. Of course, that's knowledge about logical possibilities, not the world.

@kennethamy,

kennethamy;69313 wrote:

You don't mean that there is nothing that causes that rock to be hard, do you? You would be mistaken if you thought so. The rock is constituted of molecules that act in a certain way so that the rock is hard. A sponge's molecular structure is much different, which is why the sponge is soft.

We know that those attributes are present together, claiming that molecular structure is what causes hardness is not a necessary claim. It suffices to say that molecular density ect. is always present with hardness so that if one is present, it is in general safe to assume the other.

I can't claim that nothing causes the rock to be hard nor can I claim that I know precisely what does at a totally fundamental level, I can only observe the relational patterns that occur and draw practical conclusions/predictions.

If this is causation:"that X is always present with Y and nothing else is necessarily present(and X is more fundamental than Y) means that X causes Y", then I agree that it is fine to say that the molecular structure of the rock causes it to be hard.

However, this does presuppose that there is no presently undetected Z(or ignores the possibility until the Z is detected and explored). That is fine, as it is more practical and sensible to only take into account what is known, however; it is important to keep the possibility of Z in the back of our minds, recognizing that causality is not an absolute known.

@Satan phil,

Satan;69701 wrote:

All I ever see are "afters" I never see "causes". After I flip the light switch, the light comes on (usually). I don't see any cause. What's the difference between B happening after A and A causing B? How can we test the difference?

Is it fair to say that the light switch SWITCHES the light between on and off? You SEE the light switch SWITCHING. Agree or disagree?

On a side note, you test the difference by controlling the variables.

@Zetetic11235,

Zetetic11235;69723 wrote:

We know that those attributes are present together, claiming that molecular structure is what causes hardness is not a necessary claim. It suffices to say that molecular density ect. is always present with hardness so that if one is present, it is in general safe to assume the other.

I can't claim that nothing causes the rock to be hard nor can I claim that I know precisely what does at a totally fundamental level, I can only observe the relational patterns that occur and draw practical conclusions/predictions.

If this is causation:"that X is always present with Y and nothing else is necessarily present(and X is more fundamental than Y) means that X causes Y", then I agree that it is fine to say that the molecular structure of the rock causes it to be hard.

However, this does presuppose that there is no presently undetected Z(or ignores the possibility until the Z is detected and explored). That is fine, as it is more practical and sensible to only take into account what is known, however; it is important to keep the possibility of Z in the back of our minds, recognizing that causality is not an absolute known.

There are no "absolute knowns" whatever those are. But I suspect you mean "certainty" which is the impossibility of error. Since human beings are fallible, it is unlikely that it is impossible that they will err. In fact, there is the saying, "to err is human". But, of course, to know is not to know with certainty. So you need not worry about that.

@kennethamy,

[QUOTE=Satan;68880]Can anyone demonstrate that a cause is more than a mere correlation without engaging in this fallacy?[/QUOTE]

Post hoc, ergo propter hoc is a logical fallacy only insofar as causation has any meaning at all separate from correlation; it does not have any such meaning. Post hoc propter hoc is really the basis of our concept of causation, which is correlation of phenomena. When we can clearly define the interaction of those phenomena, we may find (create via the defining) new phenomena which intervene, so that what was once thought of as causation is then thought of as only correlation. The two only differ by degree and depend of perspective; what is first said to be causation becomes mere correlation when the definitions involved become more precise. E.g. John notices that whenever X happens, Y always follows. After years of study and consistent empiric observation, John concludes that X causes Y. However, a few decades later, John II discovers that in rare instances when C is nearby, X does not cause Y, and also that sometimes Y appears without X. New observations are made, new processes are discovered, and the objects involved are defined more specifically, and so the relationship between X and Y becomes a correlation. Because all phenomena are infinitely complex, we can with equal reason level the charge of post hoc propter hoc error against any causal relationship; i.e. any claim that X causes Y will eventually be invalidated by the discovery of new aspects to X or Y, which reveal that the relationship is not so simple, and that C or Q are in fact involved in a more complex causation: and so on ad infinitum. Every causation is in fact correlation, and most, if we survive long enough as a civilized species, will probably be revealed as such. Causality is a representative of the vain bias that's plagued science throughout history: i.e. that the present understanding is the correct understanding.

On a related note, there really is no causation anyway, there's only succession; i.e. one event follows another, but does not cause it, unless by cause we mean follow. If that's the connotation, then fine, but it seems to me that 'causation' is used in an erroneous fashion: i.e. implying some force separate from the events, objects, etc. involved. The same problem is evident in the way we think about force and matter. We had first to divide what is essentially a monistic event into two components: static matter (which exist only in the mind, as a memory, because reality is always changing) and then change itself, as an abstraction; this division is necessary in order to think, speak, write about the event. To define a monism is impossible; it has to be divided into parts in terms of which the whole is understandable. If you disagree, try to describe something without words, numbers, figures, or other standards, measurements, fixed definitions, etc. We then relate the two parts, matter and energy, in our explanation of some event, and impress upon the unschooled listener a false idea: i.e. that there is actually one thing 'causing' another thing, when in fact there's just a monism in flux. In order to avoid this error, I like to use the term succession, which allows for a division of the event into parts, but not into objects with causes and effects: just a series of events which succeed and precede one another. If we do away with the erroneous idea of causation, replace it with succession, and acknowledge that the division into parts is purely arbitrary and pragmatic (not reflective of a real state of affairs), then there is no need for a distinction between causation and correlation.

Any of these events, arbitrarily plucked from out of the monistic world, could rightly be called the cause of all those which follow, or a cause of them in any case. Whether something is one cause or the cause of some phenomenon depends purely on the standard being used, which is arbitrary. In other words, the idea of linear causation is untenable. Non-linear causation, while accurate (in that everything is a cause of everything that follows), is not interchangeable with the common 'causation' today; in that common sense of the word, we are attempting to produce something useful; it's a means of navigating the world, not a truth (like all of our knowledge). The idea that rain on a bicycle causes rust, though it fails to recognize the intervening processes which might more properly be called the cause of rust, will do just as well in keeping our bicycles working. We know exactly what we need to know; the standards are created by us; that rain on a bicycle causes rust is sufficiently true for our purposes, as a recognition of the other processes (oxidation e.g.) is not needed for our aims.

@BrightNoon,

The price of tea in China continues to cause my bike to rust.
Exposure to the elements continues to cause my bike to rust.

All the members in my household wake up around the same time.
By waking up, I cause everyone else in my household to wake up at around the same time.

Philosophy Forum

What's the difference between causation and correlation?

Quick Links

My Account

able2know