Tuesday, June 06, 2023

Evening Note for Tuesday, June 6

 Thought for the Evening: Kinds of Account of Causation

Causation is a kind of dependency, so a cause must be prior to its effect, it must have some relevance to the effect, and that relevance has to have some kind of strong modality associated with it. These are the key elements that any account of causation requires, and which are generally recognized as essential to causation: priority, relevance, and modality. ('Modality' is sometimes called 'regularity'.)

There are two kinds of priority that come up in accounts of causation: logical and temporal. Logical priority (L) has, in the West, at least, been historically dominant; Plato, Aristotle, Lady Mary Shepherd, William Whewell are all examples of L-theorists. Temporal priority comes into its own with Hume. Hume faced the problem that his basic empirical theses made an L-account impossible, but he still needed to have some way to distinguish cause and effect. This requires some kind of asymmetry between the two. The only relation in Hume's empiricism that had an asymmetry that would work was temporal priority (T). (It's an interesting question whether there is a non-logical, non-temporal relation of asymmetry that a very different philosophical view might allow, but for our purposes it's moot, because all of the major accounts seem to be either L-accounts or T-accounts.) T-accounts slowly spread through the nineteenth century, and then quickly spread through the twentieth century, to such an extent that an entire generation of philosophers seem to have just assumed that all accounts of causation have to be T-accounts; but over the past fifty years there has been a slow and growing push-back against this dominance of the field by temporal priority.

A common (but not inevitable) issue distinguishing L-accounts and T-accounts is simultaneous causation; L-theorists often argue against T-theorists by arguing that at least some causes are simultaneous with their effects. Strictly speaking, L-theorists, just as L-theorists, are not committed to any position about simultaneous causation; since the distinctive feature of L-theory is holding that the priority of cause is logical, the position doesn't require any particular view about how causes and effects are related in time -- indeed, you can have an L-theory that takes there to be causes and effects that have no temporal relation to each other at all, you can have one in which there are no causes simultaneous with their effects, you can have one in which all causes are simultaneous with their effects, etc. However, if there are any causes simultaneous with their effects, this causes a serious problem for T-theories, which use non-simultaneity to distinguish causes from effects. (Historically, this issue has been greatly complicated by the fact that even very intelligent people keep confusing simultaneity -- which is a relation of overlap -- with instantaneity -- which is not a relation at all and thus not any kind of overlap -- and thus assuming, falsely, that simultaneity of cause and effect requires that the effect instantaneously result from the cause.) However, it may be possible for a T-theorist to allow simultaneity if they can somehow get the temporal asymmetry in a more indirect fashion; simultaneity of causes and effects gives one a reason to prefer L-theories to T-theories, but might not strictly refute T-theories.

There seem to be two kinds of relevance of cause and effect. One kind takes the relevance to be intrinsic to the cause and/or effect; an example of a common kind of proposed intrinsic relevance is the idea that causation consists in an action of the cause in the effect or on which the effect depends for being an effect. Again, historically most accounts of causation have been intrinsic relevance (I) accounts. The opposite view is that the relevance of cause and effect to each other is due to something other than the cause and effect themselves. This is extrinsic relevance (E). Again, the best known version of an E-theory is that of Hume, who takes the relevance of cause and effect to be a matter of custom or habit of mind.

In general, L-theorists have tended to be I-theorists and E-theorists have tended to be T-theorists.  (The reverse is not, I think, strictly true; that is, it's at least unclear that I-theory pushes one toward L-theory or that T-theory pushes one toward E-theory. There are a lot of TI-theorists. There are relatively few clear examples of LE-theorists, although it isn't a logically impossible position; if you held that 'cause' and 'effect' were simple roles in a model, that would be an E-theory, and while most such accounts have assumed that causes are temporally prior to effects, nothing seems to prevent holding for logical priority instead.) 

Modality is in many ways the most unruly element. There are many kinds of modality: alethic (necessary), epistemic (known), doxastic (believed), deontic (obligatory), temporal (always), locative (everywhere), etc., and you could in principle have an account of causation that thought about the modality of causation in any of these ways. Possibly one way to make this jungle more orderly is to divide modalities into mind-independent modalities (N) and mind-dependent modalities (D); that's imperfect and sits awkwardly across the categories (e.g., you can ave mind-independent and mind-dependent accounts of most modalities); also, it's possible that you could have a view in which some cases are N and some cases are D. But it's at least something that one can have some handle on. Historically, most accounts of causation are N-theories; Hume, of course, gives a D-theory.

Assuming all this, the full possible set of options for giving an account of causation is:

LIN -- LID -- LEN -- LED -- TIN -- TID -- TEN -- TED

Aristotle is LIN; Hume is TED. Everybody is on a spectrum more or less somewhere between them.

Various Links of Interest

* Thomas Johansen, From Craft to Nature: The Emergence of Natural Teleology (PDF)

* Alex Byrne, The origin of "gender identity"(PDF)

* Nathan Rothschild, On Why Thumos Will Rule by Force (PDF)

* Elizabeth S. Radcliffe, Hume on the Psychology of Public Persuasion (PDF)

* Samuel McIlhagga, Britain is Dead

* David Polansky, A Brief History of the Bourgeoisie, or We Are All Bourgeois Now

* Thomas H. Lee, The Judicial Power -- Admiralty Clause (PDF)

* Tobias Hansson Wahlberg, The Truth about Social Entities (PDF) -- I am wholly unconvinced by this account, in which there are in a sense no social entities, but it's a nice way of trying to make it plausible.

* Sharri Irvin, The art of rules, on conceptual art, at "Aeon.co".

* Brendan Hodge, For Joan of Arc, history 'bent' toward justice, at "The Pillar"

* William E. Carroll, The Condemnations of Paris and the Christian origins of modern science, at "Catholic World Report"

* Steve Schale, Anatomy of a Murder: How the Democratic Party Crashed in Florida -- I suspect the problem discussed in this article is one that will be with us a long time, and will keep tripping up both parties, with minor variations.

* G. M. Trujillo, Jr., Aristotle on Friendship, at "1000-Word Philosophy".

Currently Reading

Charles Dickens, Bleak House
Pope Leo I, Sermons