Causation: Is it real?
Hume says that all there is to causation is constant conjunction. We can never experience the necessary connections we think we feel about causal power or force and so causation is a mental concept only.
Is he right?
Is he right?
Comments (54)
First, and I may have had this conversation with you, "necessary connection" is sometimes parsed as saying something about possible worlds. I don't agree that that's the right avenue to take here. If a billiard ball hits another, and the ball that was struck goes off with a particular velocity after being struck, all that matters is whether in the actual world, given things just as they turned out to be (so no counterfactuals need apply), there was a necessary connection between the first and second ball--that is, only in that actual world, with no counterfactual conditions, the second ball's velocity had to be as it is after being struck by the first ball, or in other words, the second ball's velocity wasn't random, acausal, sui generis.
Aside from that, we could be saying that sans omniscience, we don't actually know re the actual world, no counterfactuals, whether the second ball's velocity was random, acausal, etc. The problem with that is that it's a certainty concern, and I think that certainty concerns are misconceived, especially when it comes to empirical matters.
So yes, causality is real, we can know causality, but no, we don't have certainty for that, we can't know any causality for all possible worlds.
The first ball striking it.
How would you describe the distinction there, exactly?
And what caused the first ball to strike it?
I didn't stipulate that, so
It could have been hitting it with a cue, or another ball hitting or, or maybe the table was tilted, or whatever.
I'm not actually a determinist--I think that some phenomena could be random, but aside from that, if we're talking about a long causal chain, only the first step would be a cause because? I have no idea what the reasoning for that is supposed to be.
"A" versus "the" isn't at all a clear semantic distinction to me, aside from the context where we're talking about an event that may have had multiple causes rather than just one. It certainly doesn't suggest anything like "only the first event in a long chain" to me.
The second sentence doesn't really make sense to me, either. Why wouldn't the first ball be necessary?
How is it an explanation if you witnessed one ball hitting another. You'd only find an explanation useful if you didn't witness it.
Is your idea just an explanation for the existence of your post?
Every effect is also the cause of something else and causes are effects of other causes. It becomes incoherent to use terms like cause and effect. There is just causation, or maybe a better term is "relationships".
Exactly the point, multiple causes. Where is 'the cause'?
Hume says it is constant conjunction. The mental relationship that happens when you see something always happening in the same way.
That is called "recognition".
I love it when philosophers try to make up these complex-sounding words and phrases for something that we already have a simple term for what they are talking about.
And why do philosophers feel the need to reference long-dead philosophers as if these long-dead philosophers had access to something we don't. It's actually the other way around. We have access to modern scientific knowledge that they didn't.
You philosophers always ignore the hard questions that are asked (It's because philosophy never answers questions. It asks therm. Science answers them). There was more to my post than what you replied to.
I mean multiple simultaneous causes, not sequential.
In other words, two billiard balls can strike a third at the same time.
At any rate, I think it would be proper to first settle on an understanding of what causation is which adequately encompasses all of modern (and ancient) understandings of what it holds the potential to be. This prior to appraising whether or not it is real—and in which ways it might so be.
To give a maybe incomplete list, there’s Aristotle’s four types, there’s the logically conceivable retro-causation, and there are bottom-up and top-down forms of causation (neither of which occur in relation to duration: they each occur at the same instant of time addressed). So yes, Hume’s definition of causation is a bit outdated by modern standards, at least imo.
To my mind, though, all these conceivable forms of causation can be adequately defined via the notion of dependency; hence: When the presence of (set of givens) A (be they entities or processes) is existentially dependent on the presence of (set of givens) B (be these entities or processes), B is the cause to A as effect.
Yes, I’d very much like this curt definition of causation to be questioned for potential flaws; it would help me in better discerning where its deficiencies might be (if any; crossing my fingers here).
Still, tentatively granting this definition of causation, I then would be of the opinion that causation then necessarily exists. Deciphering the details of what causes what being a different matter altogether.
Sure, but that wasn't what I was referring to above re multiple (simultaneous) causes being the only situation where I can see an "a" versus "the" distinction amounting to anything.
If this thread is strictly about Hume’s notions of causation, I’ll likely abstain. No biggie.
The immediately temporarily antecedent action(s) or event(s) that produce a particular subsequent event.
I am working on Hume's two definitions of causation so I would prefer not to leave Hume out of it completely.
Why only the immediate ones?
Because that's what had to obtain for the effect in question.
Say that A causes B, which causes C. Well, if A caused B but B didn't subsequently produce C, then A is irrelevant to C, even though A causing B might be identical in both cases.
Got it. I don’t know the angle your approaching this topic from, but if this helps out:
There’s a weird paradox that can emerge from Humean causation when it is envisioned to be devoid of all instances of agency (here knowing that Hume himself did sponsor the necessity of agency … I don’t recall that he provided a positive account of how this all works, but I do recall that he concludes that both agency and determinism are equally necessary aspects of the world … it’s been a while though).
The paradox:
Given that each cause is itself the effect of a previous cause, a causal chain can be represented in the following manner:
… e/c – e/c – e/c – e/c … etc. This where “e” stands for “effect” and “c” stands for “cause”.
It doesn’t matter how complex the chain or web of necessarily conjoined instances of e/c becomes. In all instances, it produces a reality devoid of change—for there is no link which is not perfectly determinate and, thereby, immutable. This logical derivation of a perfectly static reality stands in rough parallel to Zeno’s paradoxes.
Discerning what given causes what effect here becomes fully arbitrary and fully contingent on the subjects that so discern, which a) are themselves fully enmeshed into this perfectly changeless reality and, paradoxically, b) cannot experientially be in the absence of change.
Ignoring the awareness of subjects that, here, arbitrarily discern links between causes and effects, what logically results is a changeless space wherein no cause or effect can be validly distinguished—wherein all that is becomes a changeless block with continuous presence devoid of valid instances of causation.
In other words, premising a world of efficient causation devoid of agency can, I think quite validly, result in an objective reality fully devoid of causation.
But this is contrary to our lived reality … everything from personal experience to our scientific enquiries.
If this makes sense and you’re so inclined, feel free to make use of it.
Why temporarily? Are you looking to create some sort of distinction between causal chains/events?
Lol, oh. I should have caught that. Still, “immediately temporal” implies the same thing.
Say that A causes B, which causes C. Then we can say that A causes C just as easily a A causes B.
Does A always cause B?
In a deterministic world, yes.
I have no ideas what you or Hume are talking about.
Time is change. Causation is change and thetefore the essence of time. Causation is also meaning as effects mean, or represent, or carry information about, their causes.
We witness the cause happening before the event - temporal priority, we witness a physical proximity between the two objects - spatial contiguity, and we witness these things happen the same way all of the time - constant conjunction.
Those are the three things we observe in what we conceive to be some necessary connection between events or some 'hidden power' in the objects that causes events to happen. We don't actually observe anything else so when we attribute cause we do so by means of inference alone. Hume believes that all forms of knowledge from induction is suspect.
That is what “temporal” means, the “immediate” implies a distinction between the one proceeding cause and all the relevent proceeding causes. That is a curious distinction to make, so I inquired.
I'm a nominalist. There's only one A and one B.
As Jamesk noted, I meant temporally, and I'm pretty sure I typed temporally, but I was posting from my kindle and it often "autocorrects" to something I don't actually want.
"Temporally" because the events occur "in time." If I don't specify "temporally antecedent" folks might have something they consider a non-temporal antecedent in mind instead, like the "If" clause of a conditional.
Am I just reading into your use of “immediately”?
Reading what into it? :confused:
That it was intended to create a distinction between the one proceeding cause and all the proceeding relevant causes.
Lol, I suppose the answer should be obvious enough to me by now, sorry for the confusion.
What about when a criminal confesses to a crime? The evidence is the effect and the criminal's actions is the cause. Is the criminal desribing an inference or an actual experience when he recounts the crime in detail which explains the evidence perfectly?
What about your own intent being the cause of changes external to you. In essence you are a power of cause and directly experience your will moving your hands to type a post. Or are you inferring that your will, or intent, is causing your hands to move?
Wouldn't Hume say that the mind is the cause of ideas? Can ideas exist without a mind? Think of a cause as the prerequisite conditions for some emergent property.
Another way of thinking about it is energy transfer/flow.
The criminal has provided the explanation of the event not the cause.
Quoting Harry Hindu
Agent causation means that I do x, the problem is that there will be as many if not more external, deterministic factors going on to make me do x. So when I raise my arm it is my mind telling me to do so, however my mind has been pre-influenced to already do it.
Quoting Harry Hindu
I don't think so. Hume says that our ideas come from impressions of the senses or from associations of ideas, so I need to have seen an apple to have an idea of one but once I have the idea I can play around with it in my imagination.
1. Temporal priority: Indeed the general consensus, to put it mildly, is to have the cause precede the effect. The accepted definition of causation is such but I hear quantum physics has issues with this. Which means the definition of causation is atemporal in some way. Can't wrap my head around this. How would an atemporal definition of causation look like?
One of the fallacies of causation we're supposed to guard against is, well, ''reverse'' causation which , simply put is to think A causes B but infact it's the other way round. The way we avoid this fallacy is to find out the temporal sequence of A and B. If A precedes B then, we're supposed to conclude, B cannot be the cause of A.
There's a thread on quantum weirdness and claims about effects preceding causes. What I'd like to know is how they avoid the fallacy of reverse causation. Looks like it's not a fallacy in the quantum world but what definition of causation is being used? I'd like to know that.
2. Spatial contiguity: Well, this is, literally, the space dimension of the whole matter. Cause and effect must be spatially limited. If I slap the wind and you, 1200 miles away, feel pain then you wouldn't hold me responsible for your suffering. Would you? This makes sense but unlike 1. Temporal priority, it isn't necessary as a definitional element. Causality could be independent of space. As a simple example take gravity. It's effects reach over billions of kilometers.
3. Constant conjunction: This is a necessary part of the definition of causality. If A and B aren't correlated then A and B could be just an acausal coincidence.
1. Yes indeed the Quantum theory upsets causation as it does everything else. It does look like the effect can come before the cause, but we can forgive Hume on this one.
2.There is no causation at a distance. I may not hold you responsible at 1200 miles distance because the explanation of the event is not the cause of the event.
3.Hume puts the major emphasis on constant conjunction, it is the regularity and uniformity of events that lead us to conceive of necessary connections and causal powers.
Our whole understanding of causality is based on this basic premise.
One could, however, look at it from another angle.
Imagine, if ''classical'' causality proves that A causes B based on the premise that cause precedes effect.
There could arise a situation, say in Quantum Mechanics, where B precedes A in time. This means two things:
1. We've committed the reverse causation fallacy
2. Effect actually precedes cause
Which of the two is the case?
If it's 1 then we're still in the domain of ''classical'' causality.
If it's 2 then we need to redefine causation. How does one remove time from a definition of causality?
Do we look at the mechanism of the event. We could know that cause A is endowed with features that explains its effect B. Then when B precedes A then it is truly a case of effect B preceding the cause A. Afterall we don't know the process that leads from B to A but we do know and understand how A causes B.
I don't know what Hume had in mind but I think he wants to say that causation is a mental construct rather than something real. A constant conjunction of events is very tempting to humans whose minds are pattern seeking.
There is no logical necessity in the pattern but there is one and we see it and think of it as causation.
Of course one could question Hume's view on this based on the fact that we can, after understanding a causal chain, fine-tune the results. Isn't that what science is all about.
The fact that we can do that seems to favor a view that causation is real and not just a mental construct. How else can we explain our ability to guide and modify causality?
Exactly what Hume means when he defines causation as 'an event followed by another , where the appearance of the former always conveys the thought of the latter. '
Quoting TheMadFool
Science would love to make an equation for causality if it could, it still believes that one day it will be able to. If we had such an equation we could then accurately predict events on the first experience of them, a priori. We also feel some sympathy with objects under causal influence. We build machines and form sentiments to them because they 'do' something. Big machines that do more are assigned gender, usually the female, but again machines are inert and any personality they have is projected on them from ourselves.
This is a ridiculous response. To say that our ideas come from something is to say that they are caused by that something. You are also saying that mind isn't necessary for the existence of ideas - that ideas can exist without a mind. Nonsense.
In other words, you and Hume cannot escape the notion of causation because it would be incoherent to do so.