What this reveals is that ALL our causal inferences could be coincidences. That means causality, as we perceive it, could simply be nothing more than a coincidence. We can't know for sure. — TheMadFool

So you are saying that every time I reach for a cup and actually get hold of a cup is just a coincidence?

So you are saying that every time I reach for a cup and actually get hold of a cup is just a coincidence? — Rich

That's a strawman but yes, that's essentially it.

It's not a strawman. It's right to the point. Your philosophical approach just ignores your everyday experience in favor of ....

favor of .... — Rich

Truth or the lack of it.

So you consider ignoring your everyday experience as an approach to discovering the "truth"?

If one wishes to understand the nature of nature, one needs to observe patterns in nature. One such pattern is that when I reach for a cup to grasp it, I indeed great it. It is called developing body/muscle memory. With this one begins to understand how habits are formed by body intelligence.

So you consider ignoring your everyday experience as an approach to discovering the "truth"? — Rich

I'm simply putting everything in question here and I've justified my stance.

Everyday experience could be wrong. I know that sounds crazy, as your example demonstrates, but follow your reason and it leads you to the question I asked...it IS possible that everything is just a coincidence.

If the coin is loaded — TheMadFool

I've never heard of a loaded coin. Loaded dice, sure, the Romans used them. But loaded coin? I've heard of 2-headed or 2-tailed coins.

Michael Ossipoff

I've never heard of a loaded coin. — Michael Ossipoff

If the weight of the coin were to be unevenly distributed that would be a loaded coin.

If the weight of the coin were to be unevenly distributed that would be a loaded coin. — TheMadFool

Sure, but has it ever been (intentionally) done? Would it be able to significantly affect the probabilities of heads and tails?

Make the coin of aluminum, but have a bit of osmium concealed in it, at one edge, on one side.

Without examining the problem, my first guess would be that that would, in some way, very slightly affect the relative probabilities of heads and tails. But enough to help you win enough money to make it worth the trouble?

Michael Ossipoff

Casino dice are required by law to be as balanced and perfectly cubic as possible.

The pips, in order to have the same density as the rest of the die, are filled with the same material as the rest of the die, but just dyed a different color.

Ordinary store-bought dice are testably unbalanced and biased...as one would expect.

I once did an experiment in which I threw an ordinary store-bought die, about 3000 or 5000 times.

What I expected was that 6 would come up significantly more often than 1.

And indeed it did.

But, surprisingly to me, that wasn't its biggest bias:

The 6 and the 1 came up more often than the 5 and the 2, which came up more often than the 4 and the 3.

All of these biases happened to a degree that would be unlikely by chance. I don't remember what the significance levels were, but some of them were as good as having less than 1% probability by chance, and others were as good as having less than 5% probability by chance.

The big differences in the occurrences of those pairs of numbers can be explained by the die's dynamic unbalance. It was wobbling like a dynamically-unbalanced tire, making some rolling-orientations more stable than others.

Michael Ossipoff

What this reveals is that ALL our causal inferences could be coincidences. That means causality, as we perceive it, could simply be nothing more than a coincidence. We can't know for sure. — TheMadFool

No, it does not mean that.

Here's why -
1. The conclusion "We can't know for sure." was arrived at using a mathematical method of investigation which is different from a physical method of investigation and equivocating the different results between the two (as well as the combination of the two). Any good "causal inference" investigation has both the mathematical as well as the physical tools to generate a possible cause.

2. The statement here "there is no necessity that the coin will NOT show heads throughout a series of a million flips. " should read "there is no mathematical necessity". There could in fact be a physical necessity.

3. The statement "The only thing is such events are highly improbable." confuses again what the "events" in question are. Are the "events" mathematical or physical in nature?

And then, just logically, I think it refutes itself:
4. The statement is a premise as "That means all our inferences of causality (not a coincidence) are actually cases that are highly improbable." can be rewritten to state: "All (causal inferences) are (cases that are highly improbable.)" or "All p are q", where "p" is a (causal inference) and "q" is a (highly improbable case).
Let's use the Square of Opposition:

If the original premise "All (causal inferences) are (cases that are highly improbable.)" is held to be True, then these would follow:

- It's Contradiction must be False, so that "Some p are not q" or "Some (causal inferences) are not (cases that are highly improbable.)". To shorten - "are not (cases that are highly improbable.)" we could also state that "are not probable or of equal chances". In other words, it must be False that "some inferences are probable or of equal chances.". If there is ever an "inference" which is "probable" or "equal chances" for the outcome, then it must be false since only "improbable" outcomes can qualify for the "inference" to be true. Therefore it must hold that "all causal inferences" are "all improbable" since it is false that some are "probable". We find the hidden equivocation in the premise that states "improbable outcomes" are mathematically and physically equivalent to "probable outcomes".

- It's Contrary must also be False, so that "No p are q" or "No (inference of causality) are (cases that are highly improbable)". So by asserting "all causals are improbabilities" the simultaneous assertion is made that "No causals are probable". You can never have a probable outcome.

- The fact that "All (causal inferences) are (cases that are highly improbable.)" being held as True will then have the Implication that "Some (causal inferences) are (cases that are highly improbable.)". That means as the original presmies' Subcontraries must also both be true, so that "Some p are not q" is true, or "Some (causal inferences) are not (cases that are highly improbable.)" - if we rewrite that it states "Some (causal inferences) are not highly improbable."...meaning some causal inferences are probable or of equal chances. But wait...we already determined that all causal inferences are improbable with the premise itself?!?

Leading us full circle - the premise itself is a contradiction since all causal inferences MUST be improbable by their nature, yet by stating they are improbable means that there can be no causes that are probable or have equal chances for occuring.

However, there is no necessity that the coin will NOT show heads throughout a series of a million flips. The only thing is such events are highly improbable.

But improbable is NOT impossible.

That means all our inferences of causality (not a coincidence) are actually cases that are highly improbable. We can't say for sure that a causal event is NOT a coincidence. A series of a million heads in a row is improbable, yes, but could be a coincidence. — TheMadFool

You move here from a million heads being 'highly improbable', to 'not impossible', to 'all' inferences being 'highly improbable'. That's a magic trick.

Did you ever see or read 'Rosencrantz and Guildenstern are dead'? Stoppard's play, revived at the moment in London in the UK, riffs for some time on an amazing sequence of calls of heads being right. They debate what you're debating as a consequence, besides doubting their own identities (since part of the point is that in 'Hamlet' the two characters Rosencrantz and Guildenstern are interchangeable)

You move here from a million heads being 'highly improbable', to 'not impossible', to 'all' inferences being 'highly improbable'. That's a magic trick. — mcdoodle

I guess there's a philosophy behind it. I think it's called pragmatism - in essence we better take what we get. I know there's a shift from ''highly improbable'' to ''not impossible''. It does seem a little crazy to make that move as it implies zero total knowledge for mankind.

However, there is no flaw in my argument. Perhaps rationality taken to an extreme becomes insanity. I think we see this all the time. Spock sometimes seems crazy.

No, it does not mean that. — Uneducated Pleb

1) All causal inferences are coincidences
2) Contradictory: It is false that some causal inferences are not coincidences
3) Subaltern: Some causal inferences are coincidences
4) Contrary: It is false that no causal inferences are coincidences

Where's the contradiction?

The contradiction of 1 is

5) some causal inferences are not coincidences.

5 can't be proven. My argument demonstrates that.

The distinction between physical and mathematical necessity is not helpful. Mathematics, probability specifically, is fundamental to all causal arguments. I'm only familiar with Mill's method of demonstrating causality and ruling out coincidence (mathematically describable) is a central part of Mill's methods.

The contradiction of 1 is

5) some causal inferences are not coincidences.

5 can't be proven. My argument demonstrates that. — TheMadFool

The laconic version is better, thank you. I presented a poor argument for a contradiction using the square.

If we hold this true:
1) All causal inferences are coincidences. T

Implication - Some causal inferences are coincidences. T
Contradictory - Some causal inferences are not coincidences. F
Contrary - No causal inferences are coincidences. F
Sub - Some causal inferences are not coincidences. U (but assumed False)

Either it is true that "some inferences are coincidences" and that "some inferences are not coincidences" is true (contradiction), or it is true that "some inferences are coincidences" and false that "some inferences are not coincidences".

Here is my reasoning for the contradiction with #5, where you show it as unproven:
I do think it helpful to make the distinction between the mathematical and physical and how they interact. I based my objection on the grounds of the assertion in the coin flip example that there is no way to know, purely from the results of the flips, what caused a mathematically improbable result. Mathematically speaking, that is true. No physical knowledge is gained or denied, but physical knowledge is assumed a priori in order to even apply the equation. Concerned with only the flip results, maths will make no distinction between Odin and a loaded coin. Could be both, could be neither. We do know from the math-only approach that it is highly unlikely that a fair coin flipped 1 million times comes up with 1 million heads, but not impossible. *

(*I liken this to the argument that 1.99999999999999999999... is not 2.0 as it is mathematically possible to have 1.999999999999... but in physical reality it is 2.0.)

But, within the decision to use probabilities for the coin flip - we must make some physical assumptions. Our first assumption is that we are indeed using a coin and not a die, the coin has 2 different sides, and the recorder of the flip results is accurate, etc. just to name some of the first physical constraints that come to mind. If we use a fair coin, we know it should be roughly a .5 chance for each of two possible results. If we know we are not using a fair coin, then we can apply different equations where the results of those equations map more closely to the physical results (say 75% heads and 25% tails for the loaded coin). So there are physical reasons and mechanisms which change the mathematics that are applied and which then define subsequent results as being probable or improbable. If I use the equation parameters of the probability of a coin with a 1 face and a 2 face - but then start rolling a die with 1,2,3,4,5, and 6 - how improbable are the results of the equation modeling a coin? The question is - can you know all of the physical constraints? Probably not in most cases, but in some it may be possible (like a simple fair coin toss).
So "causation" can't be identified with mathematics alone. I think we can agree there, where we still disagree is to whether "causation" can be identified physically (or in combination).

The next objection has to do with the definition of a "causal inference" as an "improbable result". I am assuming we agree that there are critical differences between "causal inferences" and "causal mechanisms". A "causal inference" is basically an educated, but constrained, guess. The "causal mechanism" is the actual case, but they are two different classes of things. One can see and record the physical results of a "causal mechanism" and not know what the mechanism is and then assign an incorrect "causal inference" to those results. That would be the definition, I would think, that counts as a "coincidence" since there can be many more causal inferences than causal mechanisms. The inferences can be modelled mathematically to cohere with the physical. The results can show what is probable and what is improbable based on the the merits of the coherence.

So then we have to ask - can there be an inference which does actually describe a state or mechanism? The odds of an inference actually describing a mechanism, mathematically speaking, is not 0 (even in random chance). With that said, the vast majority of "causal inferences" could be incorrect and therefore counted as coincidence, but what mathematical or physical rule states that all inferences must be coincidence?

If that were true, then even religious/spiritual "causal inferences" would always be as much of a coincidence as any other posited "causal inferences", since there is 0% chance of any correct inference.

But, mathematically speaking, there is a >0% chance that there is at least one possible or existing "causal inference" which can or does accurately describe a "causal mechanism". In other words, one inference must be correct, but every inference is not necessarily correct.

That is the contradiction that I see - stating that all inference is coincidence I could then ask in what way is it meaningful to even say anything is "coincidence"? If there is at least one correct inference, then it is meaningful and possible to say that any particular inference is coincidence (as a result of mathematical and physical coherence), while another is not coincidence. It can also point to a currently unknown inference if all other inferences fail the criteria of the investigation.

So, if I may, let's state the following (if you agree based off of the reasoning so far): If we take the above as true, that means it is true that some "causal mechanisms" can be described by some "causal inferences".
Therefore:
1) All "causal inferences" are "causal mechanisms". F
2) No "causal inferences" are "causal mechanisms". F
3) Some "causal inferences" are "causal mechanisms". T
4) Some "causal inferences" are not "causal mechanisms". T

What this reveals is that ALL our causal inferences could be coincidences. That means causality, as we perceive it, could simply be nothing more than a coincidence. We can't know for sure. — TheMadFool

I can agree with some inference as coincidence and some as mechanism. A one-off result is different than repeated results. Even if we state we still don't "know" after thousands of trials, that is not the same level of ignorance inherent in the first trial. (I refer back to the 1.9999... vs. 2.0 distinction)

This matters to me because it puts in doubt the current paradigm of scientific knowledge. — TheMadFool

Does it? Or does it actually give the tools to doubt any paradigm that supplies "causal inferences" with 100% accuracy?

Science has been used to undermine religion by always insisting on naturalistic explanations of events and succeeding in this endeavor. — TheMadFool

How is that success measured for science?
How is it measured for religion?
As an aside, if religion posits a deity as 100% true, then how can there NOT be causal inferences that describe causal mechanisms?

However, if my argument is sound science, its entire content, could simply be a coincidence - highly improbable BUT not impossible. — TheMadFool

An argument is not "sound science", it is a only an argument - a hypothesis, correct? The results of "sound science" performed with the hypothesis would include both physical and mathematical investigation backed with repeated trials and have results that cohere with each other.