Anscombe's "Causality and Determination"

The Learning Zone Reading Groups
1

Continuing the discussion from Davidson: Actions, Reasons, and Causes:

Anscombe’s “Causality and Determination” is central to my own thinking on causation. I’ve citing or linking to it repeatedly across many different threads and debates, concerning first causes, God arguments, determinism, materialism, and free will.

Anscombe is puzzling over what causation actually is, and concluding it is not material implication, modal necessity, nor statistical correlation. She uses the Wittgensteinian approach to rules in order to show that we can’t define causation precisely but rather “recognise it when we see it”. She distinguishes between causation and determination, and argues that determinism is an unnecessary or impossible goal for physics. This gives a way to resist the classical picture of the universe as a “clockwork mechanism,” opening room for free will without needing compatibilism.

The metaphysical payoff for me is to deflate cosmological/first-cause arguments; those arguments that link causation to the beginning of spacetime and invoke God, panpsychism, or Penrose largely fall to Anscombe’s account once determinism is questioned. Similarly, we can reject the maxim “everything in the universe has a cause,” a notion illegitimately imported from one domain into another.

My own extension of Anscombe’s position is to treat causation as neither purely “in the world” nor purely “in the word,” but as having a foot in both, being more a way of talking about the world, a grammar or logic that is presumed, with no further justification than its usefulness, but that dissipates when one attempts to examine it.

Do we need to say more? Please, read the article. It’s not difficult.

3 Likes
Reading Group: Davidson: Actions, Reasons, and Causes
2

That’s a significant contribution if her view is correct.

Did you mean to repeat “in the world” here? Correction - I read too quickly, world/word.

Isn’t it sometimes said of deflationary arguments that they tend to dissolve a problem linguistically without resolving the underlying issue?

How much should the average person concern themselves with causation? ( I read Hume and Russell on this some years ago and moved on)

1 Like
3

Is Anscombe a realist or a deflationist?

4

The average person shouldn’t be doing philosophy at all. It’s an unhealthy habit, leading to all sorts of maladaptive practices. Go weed the garden.

The word is not the world. No doubt the theologians will be along soon to point out that their book claims the word came first.

Is she “realist or a deflationist” in regard to what? What are you asking? (…and see here the sort of maladaptive practice mentioned earlier…)

5

In relation to causation.

I think this is probably correct. I’m going to go back to my advocacy papers for the Minister. :wink: Will drop by later to see if causation matters…

1 Like
6

If causation doesn’t matter, then there’s no reason to work on advocacy; unless you think that the minister might change their mind because of your work, you might as well go get a coffee.

The issue here is more about what sort of thing a cause is. Or better, if we might determine the minister’s actions.

1 Like
7

Anscombe’s points

  1. Causation isn’t exceptionless universalization. A can be the cause of B even if there are times when A occurs and B doesn’t: in the real world one has to consider prevention and interference.

  2. Determinism is when given some conditions only 1 outcome is possible. Causation is not determinism. A Geiger counter can be set up to cause an electric gate to open. Here we have indeterminism and causation. :thinking:

  3. Ethical freedom requires indeterminism. Compatibilism is incoherent or makes freedom meaningless.

She’s proposing a probabilistic/statistical model of causation. A causes B when A increases the likelihood of B. For example visiting a hospital makes it likelier that I catch a cold but it doesn’t mean I will catch a cold.

8

Not quite. She’s using these as a foil against those who see causation in strictly deterministic terms. The words “Coca-Cola” appearing in the shaken-box is her clincher against a probabilistic reduction: a genuine causal explanation for the pattern needn’t be derivable from the statistical laws governing the particle motion at all - precisely the opposite of saying causation is a statistical pattern. If causation just were statistical correlation, there’d be nothing for a further, non-statistical cause of the Coca-Cola pattern to explain.

Her account is much wider than that. Causes might be given in terms of laws, probabilities, regularities or none of these. A cause succeeds in virtue of the place it takes in the task at hand - and here I’m avoiding the use of “language game”.

Glad that you looked at the paper.

1 Like
9

I must’ve mistaken her critique of standard views of causation as the position she holds. Does she propose a theory of her own or is she critiquing existing views?

I’ve read a little on Hume’s theory of causation as psychological: One sees B follows A, a constant conjunction, and so when we see A it creates an expectation of B.

I thought some have abandoned the idea of causation, preferring a minimalist metaphysics.

10

More method than theory. She patiently shows how each of the proposed definitions of causation is inadequate, without offering an alternative theory. it’s not that there are no casual laws, and never probabilistic causes, but that whatever tight definition is offered, there are exceptions. That’s not to abandon causality, but to be honest about it.

11

Science seems to have settled on a probabilistic model of causation. Laplace’s \frac{x+1}{y+2} for inductive inferences recognizes its chanciness.

12

What suffers here?

  • Notions of science that rely on causal laws
  • Notions of determinism refuting free will, along with those forms of compatibilism that see physics as deterministic
  • Notions of our actions being determined by our physiology
  • Laplacean views of the absolute predictability of the universe, or even of discrete parts thereof
  • First causes and cosmological arguments
  • The causal closure of physics, considered more generally.

That’s a fair portion of those articles that are the regular fare of these here fora, and many that are presently in play.

1 Like
13

I looked in the corner to see if there is an elephant there, and there wasn’t. So for Laplace, the probability of my seeing an elephant next time I look is (0+1)/(1+2) = ⅓≃33%

Needs work. But it’s not a bad idea.

(On consideration, I haven’t looked to see if there is an elephant in the next room, so the probability of there being an elephant there is 50 - 50.)

1 Like
14

Causation is certainly mediated by our conceptual schemes, but that does not reduce it to a grammar. The very possibility of successful causal reasoning presupposes real intelligible relations in the world. Mathematical physics resides precisely in the application of mathematical reasoning to observable and predictable phenomena. That is more than a language game.

And perhaps the reason its explanatory force appears to dissipate when we ask why it works is that mathematics marks the terminus of explanation. To ask why the relationships described by physics are as they are is, at some point, to try to explain the very conditions that make explanation possible. We are no longer explaining the world, but asking for an explanation of explanation itself.

15

It has long been held that QM upends the classical notion of determinism. I wonder of QM and causality.

16

Classical physics is not determinists without the addition of certain assumptions. Consider:

  • Norton’s Dome.
  • Initial conditions needed by determinism are a mathematical idealization, not something ever actually given.
  • No causal influence from outside the modeled system.
  • No singular potentials: the force between two point masses goes as 1/r². As separation r→0, the potential energy →−∞ and the force →∞.

No need to go Quantum.

17

@Wayfarer
Mathematics is descriptive of phenomena if the phenomena in question is quantifiable.

@Banno QM is distinct from Norton’s dome. The latter exploits a gap in the calculus.

18

which is the province of mathematical physics isn’t it?

19

As opposed to what? Being determined by a soul or other immaterial mind substance?

20

Mathematical physics is probably an exciting domain. There’s something mathematical that can be taken up by physics e.g. Reimann geometry. I think I meant to say the laws are mathematical. Sorry.