David Hume on Causation & The Problem of Induction


You put the kettle on the stove, turn on the gas,
and in a little while, the water in the kettle boils. So, can we say that turning on
the gas caused the water to boil? Or is it just that all we can say is that there
were two events: gas goes on, water boils, but that there is no real
connection between the two? It seems like a simple question but it goes
deep into our understanding of the world. It was one of the many deep questions asked
by David Hume, the great Scottish philosopher whose 300th birthday we’re celebrating with a
month of special editions of the Philosopher’s Zone. Hi, I’m Alan Saunders, and to tackle this
aspect of Hume’s thought, we’re joined today by a distinguished Hume scholar:
Helen Beebee, professor of philosophy at the University of Birmingham in the UK. Hume’s central thought that really all we
get to find out about the world is just regularity, it’s just one thing following another, and then
next time, one thing following another again, and somehow out of that we conjure up beliefs
about what causes what—that’s been absolutely central to the debate about causation
as it’s gone on for the last three centuries. So he set the agenda? Yes, he did. Yes. Hume wrote two books of what we would
now call epistemology and metaphysics: “A Treatise of Human Nature” and “An
Inquiry Concerning Human Understanding”. So, let’s look at what he
was actually trying to do. He wanted to provide a science of man–I suppose
these days we should say a science of humanity– didn’t he? Yes, he advertises himself right at the beginning
of the “Treatise of Human Nature” as providing a science of man. And I guess the thought
there is he was very impressed with Newton, and very impressed with experimental philosophy
more generally, what we would now think of as science, but in those days experimental philosophy
and philosophy were kind of all the same thing. So the idea that if you want to find out
about the nature of reality, the way to do it is to go out and have a look, see how things work,
do some experiments. People had started doing that when it comes to the world around us, but
Hume’s thought was let’s apply that to our own case, to moral subjects
as he called human beings. So what he wants to do is to figure out
how it is that the mind works using a similar kind of experimental method. Rather
than sitting in the armchair and trying to figure stuff out a priori, we actually, not
exactly run some experiments, but we look at how it is that the mind actually works
going on, our observations of what sorts of things we believe and how
we come to believe them. Was he just trying to be the Newton of the
social sciences, or are there actually targets against which he is writing? Oh he very definitely has targets. I guess his main target is what
he calls “speculative metaphysics”. So he says, at the conclusion of all his kind of
researches into how the mind works, “Works of speculative metaphysics should be committed
to the flames, because they contain nothing but sophistry and illusion”. And what he has in mind there is anyone who
claims to be able to figure out the ultimate nature of reality, there are things that
just lie beyond our cognitive reach. He certainly has religion in his sights. So
theologians and philosophers who claim to know things about the nature & existence
of God, that’s all speculative metaphysics, that’s all got to go. And when it comes to causation, the view
that he mainly has in his sights is the thought that we can now a priori, just by reflecting
on concepts, just by reflecting on the nature of ideas, we can come to know things a priori
about the causal structure of the world. Well, let’s look at what he has to say
about causation, about cause and effect. Now he thinks that we can’t come to conclusions
about this “a priori”, which basically means before the facts, without
the facts, doesn’t it? Right, yes. For example, there’s a well-known principle
“ex nihilo nihil fit”, which basically translates as something like:
nothing can come from nothing. Now that’s a principle that prior philosophers
had wheeled out, typically in arguments for the existence of God. The thought is that it’s an a priori principle
that every event has a cause, because otherwise that would be something coming from nothing.
And now, well, if every event has a cause, then that couldn’t go on forever, backwards in
time, so there must somehow be a first cause that was somehow a cause of itself, or a
necessary existent so that it didn’t need a cause, what’s that gonna be?
That’s gonna be God. Looking closer at causation, let’s take
the time-honored case of the billiard ball. The idea is that we see the white ball
hit the red ball and the red ball moves. But we don’t actually see the white
ball “cause” the red ball to move. All we see are two separate events and we
can’t actually see the connection, if there is a connection between the two. So the world is, as it were, just one damn
thing after another. We can’t see or prove any connection between events.
Is this Hume’s view? Well, it’s certainly Hume’s view that we
can’t prove any connection between events. And it’s also true on his view that we can’t
see any connection in the sense of there’s no sensory impression as Hume would put it,
of causation. There’s nothing, as you say, when you look at the two billiard balls you
just see the one hitting the other and then the other one moving, you don’t see any connection.
So there’s no sensory impression of causation. And Hume thinks that creates a problem because
he thinks that all of our ideas, which we could think of as concepts,
come from impressions. So if we’re going to have the idea of causation
at all, which clearly we do, somehow we have to have an impression of causation
or necessary connection as he calls it. But that impression can’t come from experience.
You get the impression of red just by looking at a red thing. You can’t get the impression
of necessary connection just by looking at things happening in the world because you
just won’t find any such impression there. Now, what the consequences of that are on
Hume’s view is a wildly controversial issue. So some philosophers think that the consequences
that Hume thinks that causation just is a matter of constant conjunction, that’s all
there is out there in the world, and when I say one thing caused another, all I mean
is things of the first kind are constantly conjoined with things of the second kind. So every time a white ball hits a red ball
in just those sorts of circumstances, the red ball will move off in the same way. That’s all there is to saying
that one thing caused another. On another view, and I guess in some
moods this is my view of what Hume’s doing, Hume thinks that causation is something
that we kind of “project” onto the world. So, out there in the world—or at least as
far as we can know—all there is is just one thing happening and then another. But then we have this impression of necessary
connection which is just the creation of the human mind which we somehow impose on reality.
So, we kind of invent the causal structure of the world. On another view, Hume actually does think
that there is such a thing as causation out there in the world, he really does think
that there’s a necessary connection between causes and effects. It’s just that he thinks that we can’t detect it.
But he thinks that when we talk about causation that really is the thing we’re talking
about, some kind of real connection out there in the world
between causes and effects. He’s just skeptical about our ability
to detect or know about that relation. So there’s a huge range of views to choose
from about what it is that Hume actually thinks. And which side do you come down on? I kind of come down in the middle,
on the, sort of, projection side I guess. Hume thinks that the causal structure of the
world is something that we impose upon it. But I suppose I’m a bit inclined to think
that actually the metaphysics of causation, the nature of causal relations or what it
is that the word ’cause’ refers to, is not that much of a concern for Hume anyway really.
What he’s really interested in is how these beliefs come about, what their
epistemic status is, and so on. And we can agree on all of that stuff and
still leave it open what he actually thinks the word ’cause’ refers to. So there’s a sense in which we don’t really
need to resolve that dispute about what Hume thinks about the metaphysics of causation
in order to agree about what he thinks about the things that he really cares about, which is
the kind of the psychology and the epistemology. I’m interested in Hume’s view of
what’s called the association of ideas. Now, if you’re an empiricist, if you think that
most of our knowledge of the world comes through our sensory impressions as you’ve
said, you need to find some way of explaining how these impressions build-up into a view of
the world. And that’s where association comes in. But for Hume, what does it amount to? So for Hume, I guess association amounts
to the thought that anything that’s going on in your mind, unconsciously or instinctively,
is going to be a matter of the association of ideas. So ideas are faint copies of impressions. You
see red and thereby have an impression of red and now your idea of red, the kind of
the thing that you used to think with when you’re thinking about red,
is derived from that impression. But the most important part of association for
Hume is the relation between causes and effects. So the question—or at least I think
the question for Hume—is: How is it that we start out with all of these impressions and
the corresponding ideas, and we get from that to a situation where we actually come to have
beliefs or expectations about what’s going to happen? So for Hume, a belief is a, kind of, lively idea.
And the question is, okay, now we’ve got the idea, where does the liveliness come from? How is it that we absorb all of this stuff,
and out pop these actual expectations that are not just responses to what’s in front of
us but actually views about what’s going on outside of our experience; my expectation
that the sun’s still going to be shining when I go outside, that my car’s going to stop
when I put my foot on the brake and so on? How is it that we get to have
those things in our mind? Hume’s answer is: reasoning from causes
to effects, which he thinks of as just an associative relation. Every time you sit in your car and you put
your foot on the brake, the car stops, and once that regularity has been established
in your experience, the next time you put your foot on the brake, you just come
to expect that the car’s going to stop. You’re not doing any fancy piece of reasoning
there, there’s just a habit or a transition in the mind that takes you from the regularity
in your past experience to an expectation about what’s going to happen. So that’s a kind of really brute, crude psychological
mechanism, it’s one that just happens instinctively. Well yeah, a brute psychological mechanism—does
there have to be any logical connection between the ideas that we associate?
Or are we all like Pavlov’s dogs? We come to associate a ringing sound with
food, even though there’s no logical connection between bells and dinner. Yeah so, I think Hume thinks we are
a little bit like Pavlov’s dogs, or at least in our unsophisticated moments we are. One of the things that Hume is at pains to
prove, and that goes back to his target being people who think that you can have a priori
access to the causal structure of reality, one of his central claims is, look, there is no
logical connection between causes and effects. For anything you can figure out a priori,
anything could cause anything. So you think when you put your foot on the
pedal the car is going to stop. But, a priori, anything could happen if you
put your foot on the pedal. The car could speed up, it could turn into
a pumpkin, you could rise up in the air and start spinning around. There’s no “logical” reason why putting your
foot on the pedal is going to make the car stop. So, if you think about Pavlov’s dogs again, I mean
of course, Pavlov’s dogs are in an experimental setup that someone has rigged up
so that the bell-ringing isn’t really a cause of the dog getting the food,
but maybe the dog thinks it is. So, we like to think that putting your foot
on the pedal really is a cause of the car stopping, but as I say, that’s not because we’ve
intuited or experienced any causal connection between them, and it’s certainly not because
there’s some kind of logical connection between them. The first time you get in a car, if you’ve
never seen a car before, you’re going to have no idea what’s going to happen
if you stick your foot on that pedal. So is this an important
discovery on his part? I mean, does it make us think differently? Um, yes. Well I don’t know if it makes us think differently
now, but certainly at the time. I mean he got a lot of his material here from
Malebranche, it must be said. This is the French philosopher
of the 17th century. Right, yes. So Malebranche shared the view that we couldn’t
perceive any causal relations with Hume except that Malebranche went off in a very
different religious direction, thought that God was actually the cause of everything. Seeing as we can’t see a causal relation between
the billiard ball hitting the other billiard ball and the second billiard ball moving,
well there can’t be a causal relation there, it must be caused directly by God. All of
these things are being caused directly by God. Of course, Hume didn’t take that line. So, he got the idea that we can’t
perceive causation from Malebranche. Nonetheless, the thought that there really
isn’t anything there that we could possibly detect, I think is quite a sort of striking
and revealing philosophical result, and it does urge caution on us when we think that
we know things about the causal structure of the world, to be kind of reminded if
Hume’s right about this and I think he is, that really what’s in front of us, what we see
is one thing and then another thing happening. On ABC Radio National, you’re with the Philosopher’s
Zone, and I’m talking to Helen Beebee from the University of Birmingham about David
Hume’s views on causation and induction. Helen, we’ve dealt with causation, let’s now
turn to the related question of induction. First of all, what is induction? Well, induction means different
things to different people. In the context of contemporary philosophy,
inductive inference is normally seen as going from regularities that you’ve experienced
in the past to some universal generalization. So, As have always been followed by Bs in the past,
therefore all As are Bs or have been followed by Bs. For Hume, inductive inference–he’s less concerned
I think about the, kind of, very general case, he’s more concerned about: all right, all As
have been followed by Bs in my past experience, now I’m confronted with an A, what’s the mechanism
by which I come to expect that a B is going to happen? So he’s more interested in the single case
I think, rather than the kind of general laws. So in contemporary philosophy the problem
of induction is thought of as a kind of problem in philosophy of science: How is it that
we can know what the laws of nature are? Whereas I think Hume was more worried about:
How can you be sure your car’s going to slow down? So, the problem of induction is now the
problem of justifying that inference. How do we go from “All As have been
followed by Bs in the past” to “All As are Bs” or “the next A is going to be a B”? What licenses that inference? And that’s the problem that Hume’s commonly
thought to have both raised and also shown to be completely insolvable. And it doesn’t have to be time-related, it can
be a generalization about the present as well. All the crows that I know are black, so from that
I draw a universal law that all crows are black. Yes. So, Hume’s concerned because he thought
of inductive inferences as basically inferences from causes to effects. He was thinking of inductive inferences being As
followed by Bs, because that’s the cause-effect relation. Whereas since Hume, people have generalized
out and thought of it not necessarily as being a matter of all As are followed Bs, but as
you say, all As are Bs, or all crows are black, all swans are white is the
famous example, and so on. So, from the point of view of the problem
of induction it doesn’t matter which way you do that, but Hume is only worried about As
being followed by Bs, because he’s worried about inferences from causes to effects. And ‘all swans are white’ is a notoriously good
case of the inadequacy of induction because you go to Western Australia and there
you’ve got black swans. Absolutely. We might try, rather shonkily I think, to
justify induction by saying that it’s worked in the past, so there’s good reason for
thinking that it will work in the future. What’s wrong with that? Yes, that is one way you might try and do
it, and it’s quite similar to a way that Hume considers and dispenses with. And this way
is going to fail for just the same kind of reason, which is that, of course, the very claim
that it’s worked in the past so it’s going to work in the future is itself
a form of inductive inference. So if you want to try and show why inductive
inference is all right, you’d better not appeal to something which is itself an
inductive inference in order to do that. That’s just not gonna satisfy someone
who’s already suspicious about whether inductive inference is okay. Karl Popper, writing in 1934—this is the
great 20th century philosopher of science— referred to the problem of induction
as “Hume’s problem”. For him, it was essentially insoluble. He thought that it is simply incorrect to
assume that just because the sun has risen every day so far, it will rise tomorrow. From that emerges his philosophy, which is
basically that we should seek to disprove rather than to prove our theories. But was Popper’s problem
really Hume’s problem? Is the problem that he sees before him
the problem that Hume saw before him? I think Popper sets off a genuine disagreement
between him and Hume, which is to do with whether we actually use inductive inference. So, Popper not only thought that we should
follow the method of conjecture and refutation –you throw up a hypothesis and then you test it.
But he actually claimed that that’s how we really do go about our ordinary lives,
that is how we reason. And that’s how dogs reason as well. Now that stands in complete opposition to
Hume who thought that inductive inference, inference from causes to effects, was the thing
that is the inference that we rely on all the time. So Popper thinks that’s just false. It’s not
that we expect the car to stop on the basis of the fact that it stops when we put our
foot on the pedal every time in the past, it’s more that we have a hypothesis that in
general the car stops when you put your foot on the brakes, and now we test that hypothesis
in a rather high-risk fashion by going around and driving down the motorway and hoping
that this time our hypothesis is not refuted. In general, Popper thinks, in science, it’s
good when a hypothesis gets refuted because that shows you that it must be false. In the case of driving your car, you might
worry about whether it’s such a good idea for your hypothesis to be refuted or if
that’s something you should be hoping for. So he genuinely disagrees with Hume over whether
or not inductive inference is the form of inference that we use to go about our daily lives
or indeed in scientific investigation. Whether the problem of induction as characterized
as a problem about justification is Hume’s problem, is something that I’m not so sure about.
In fact, I think it’s not Hume’s problem at all. I think Hume’s problem is to do with the
psychology of our beliefs or expectations. How is it, as a matter of fact, that we get
from observing regularities in the past to coming to form this expectation
about what’s going to happen? What’s the mechanism that does that? So what you seem to be saying is that, for
Hume, the problem of induction is actually a psychological issue?
Yeah, that’s what I think. At least when he’s discussing what’s now
called the problem of induction, I don’t think he’s discussing an issue about justification;
he’s discussing an issue about the psychological mechanism where you go from a situation
where you have no beliefs or no expectations to a situation where you do have expectations. He’s aware that there’s a major skeptical problem
here, but I don’t think that makes him a skeptic. I mean, his response to the skeptic
such as it is, is basically to say, well, if we all followed your advice and suspended
belief, rather than having these expectations in trusting them, we just kind of said,
“well I have no idea what’s gonna happen” we’d all die, very quickly. I mean, that’s his response. You know, you
try and live like that, you won’t even last a day. So basically we have an input-output
model of the mind. Sense impressions are the input, belief
is the output, and Hume is interested in how we get from the one to the other. Yeah, that’s the idea, yes. So in the case of our billiard balls–our As
and Bs again–once I’ve acquired sufficient experience of As being followed by Bs, that’s
the thing that generates the mechanism. So now next time I see an A, what happens
is that I’ll end up with an expectation that a B will happen. And what is that mechanism? Well obviously that would be a question
for a neuroscientist, Hume doesn’t know how to answer it, he just says it’s habit or custom. We just get used to these things going together
–this is associationism again—these things have gone together in our past experience;
we come across a new A, we just by a sort of brute psychological mechanism, which we share
with dogs, come to expect a B on the basis of that. I mean, again, that’s not going to
satisfy the inductive skeptic, right? Because the skeptic is gonna say, “Well, we
have no more reason to think that the dinner’s going to nourish us than it is
that it’s going to poison us”. I mean, all the evidence we’ve got to go on
is that it’s nourished us in the past, and that doesn’t justify any claim about the future.
And Hume’s gonna say, “Okay that’s a fair point, but now I’m going to eat my dinner,
and I’m going to be the one that isn’t going to die of starvation”. Hume gave up philosophy more or less
and became a historian. The way you presented this aspect of his philosophy,
it sounds as though it’s not actually leading him anywhere except back into the world,
back into carrying on with your life. I mean, do you think that’s why he gave it up? I wouldn’t want to speculate
about why he gave it up. But I guess–I mean if you think of Hume as
really thinking that what he’s doing is giving a science of man, then insofar as he thought
that he’d got that pretty much right, I guess, and if people were not responding to him
kind of in his own terms, if you see what I mean, if they were kind of insisting on
pushing him on kind of abstruse metaphysical speculation which is what he thought we shouldn’t
be doing, then I can kind of imagine that he might have thought, “Well okay, I’m going
to wait until someone actually comes along with some hard experimental data that
shows me that my psychological views are all false, and then maybe I’ll get interested
again” and I don’t know, maybe nobody did. Well these are tangled issues but Helen Beebee
thank you for untangling Hume’s view of them for us. Thank you very much. Helen Beebee is professor of philosophy
at the University of Birmingham in the UK. More about Hume on our website: The Philosopher’s Zone is produced by Kyla
Slaven, the sound engineer is Charlie McKune, and the music today was from the Scottish
composer James Oswald, who died in 1769.

26 thoughts on “David Hume on Causation & The Problem of Induction

  1. Does anyone distinguish between a hypothesis and a strategy? Are strategies scientific if they are constructed to be implemented (change conditions) not to test conditions for accuracy of predictions?

  2. Anyways. thanks a lot for this interview. Helen Beebee sound awesome. Have to look her up.

  3. Hume claims that we cannot observe cause and effect. Not only is that claim demonstrably false, it goes against Newton's law of cause and effect in his Principia, published when Hume was just 16 years old.

  4. The 'brute psychological mechanism' is a form of normalcy bias and on some level it is the essence of logic in itself… no?

  5. The woman in this video should have used my favorite Hume quote from An enquirer concerning Human Understanding. “We have said, that all arguments concerning existence are founded on the relation of cause and effect; that our knowledge of that relation is derived entirely from experience; and that all our experimental conclusions proceed upon the supposition, that the future will be conformable to the past.” Furthermore, “To endeavor, therefore, the proof of this last supposition by probable arguments, or arguments regarding existence, must be evidently going in a circle, and taking that for granted, which is the very point in question.”

  6. Great discussion here about Hume and causation. Helen Beebee’s explanations were great. I would be very interested to know what Hume would have thought of artificial intelligence. That is a kind of brute force type of association. I wonder if Hume influenced some of the early AI scientists.

  7. haha love it…Man corrects himself – humanity….Women corrects him correctly the term is man when we are talking about humanity – it's not sexist it's the terminology….what a sexist pig I am haha

  8. Was literally boiling water as this started and had a moment of wtf.. LMAO

  9. Helen Beebee, a rare example of a philosopher who uses simple language to express large ideas, (a refreshing change from what is so often the opposite form of communication).

  10. "go out there and experiment"
    Drop an apple, or, kick a dog, see what happens!

  11. the problem of Induction demonstrates the inherent falllability of scientific theory due to limitation of data gathering and hence theory development

  12. for normal people, all these induction,theory,causation etc etc is too far-fetched…they just live their normal days on the dogmas that they embrace from their teachers,leaders,religious authorities etc etc

  13. What caused causation? It's the same concept behind the question, "If God created the Universe, then who or what created God?" Eventually, the idea of something being self-existent will emerge in the chain of logic and reason, and that's because of the apparent fact that causality cannot have an external cause. A similar question would be, "What created creation?"

    So many things appear to be circular, or self-contained. We make logical arguments to support the use of logic, and we use words to define "word" (or use language to describe language). Existence simply exists; it could not have been created by something else that existed. That is why "God" is called, "I am that I am." That is the description of Absolute Truth. There can be no Truth outside of all that is True; there can be nothing outside of Everything. That is what Logic demands. If that isn't True, then Logic is without foundational merit. In John 1:1, it reads, "In the beginning was the Logos (translated as "the Word"), and the Logos was with God and the Logos was God…" That is where the word "Logic" came from. In order words, Logic is both with God and is God, according to the book of John.

    I'm especially gifted at frustrating atheists.

    One can either slide into the portal of infinite regress, or, they can stop the chain at the point of self-existence. Whether one thinks that the Universe is self-existent or that God is self-existent makes little to no difference to me. As far as I'm concerned, the material Universe is the material Body of God the Father that is Mind. I see Truth as the "only begotten Son" because the only thing that exists is Truth. Any opposition to Truth is simply the creation of more Truth.

Leave a Reply

Your email address will not be published. Required fields are marked *