Thursday, December 29, 2011

Determinism Doesn’t Imply Causal Necessity

A key part of the argument for causal dispositionalism in Mumford and Anjum’s Getting Causes FromPowers is the case against causal necessitarianism (chapter 3 in the book).   Causality is commonly thought to imply a necessary connection between cause and effect:  the authors say this is a mistake, and that the proper modality of causation is dispositional.   Causes dispose toward their effects - they don’t guarantee them.  The insight here is that other factors can prevent or interfere with the expected manifestation (and in everyday experience, they often do).   In fact, such prevention is always possible in causal situations, and if one moves to evade this fact by stipulating that prevention or interference is impossible, then the resulting necessity is not really coming from the causal process itself, but is being imposed in another way.

To see this, suppose I specify the causal factors involved in some manifested effect, and then someone points out another factor which could possibly interfere (despite my match being dry, a proper striking motion made and sufficient oxygen being present, a gust of wind might prevent the match from lighting).  Can’t I modify my scenario to specify that the threatening factor is absent (the wind is calm)?  Leaving aside the potential problem of listing an absence as a causal factor, the objector might present another possible interferer (a passing car might splash water on the match as it is being struck).  So, then I, in turn, specify that there is no nearby traffic, and so on.   In fact, no finite list of factors will ever suffice to rule out every interferer (however unlikely).  And by the time one is led to propose a “catch-all” condition, covering the whole state of the universe, we’re really not talking about a process of causal production anymore.

The authors note something interesting here.  They say that their argument against causal necessitarianism does not mean they are ruling out determinism.   This was a helpful observation for me because I have been guilty of confusion on this point.  One might think “determinism” means “causal determinism” which means “causal necessitarianism”.  However, determinism can be specified in other ways (including what might be the most common conception – see below).  Then causal dispositionalism could be compatible with determinism.  There is a causal process, and while it doesn’t necessitate effects, necessity is imposed in another way.

Note, that for the moment, we are leaving aside the idea of irreducibly probabilistic causation.  Such causation is likely a feature of our world (in fact I think the a posteriori case for it is nearly airtight), and therefore determinism is false.  But disentangling these ideas remains philosophically valuable.

As Mumford and Anjum say:  “The core idea in determinism is the fixity of the future by the past (p.75)” If one wanted to build a model of a deterministic world, causal necessitarianism is probably not the best tool, since the causal process doesn’t promise to cover all the possible loopholes - for instance if there are such things as uncaused events, then they would not be addressed.

It seems to me that the most common notion of determinism (probably inspired by classical mechanics) is this:  given a specification of all facts, and given comprehensive deterministic laws of nature, then the future is fixed.  There is no reason here to even mention causation – it adds nothing to the scenario.  One could be a Humean about causation and still endorse the deterministic picture.  And given the fact that in this physics-inspired vision the mathematical depiction of laws is symmetric with regard to time, it would be equally true to say that the past is fixed by the specification.  This is inconsistent with causation, which is not a symmetric process.  One might believe that the mathematically specified physical laws comprise a model of a causal world, but the laws themselves don’t constitute a theory of causation, and may very well be inconsistent with the idea of a causal process.

On this last point, I recalled a paper I had read a few years ago by Carl Hoefer: “Causality and Determinism: Tension, or OutrightConflict.”  In this paper, Hoefer defines a deterministic world specifically as one governed by deterministic micro-physical laws, and then goes on to argue that this definition is inconsistent with the presence of causation, using several philosophical theories from the literature as examples of how causation might be characterized.

Tuesday, December 06, 2011

A Note on Events and Causation

Presently I’m reading GettingCauses From Powers by Stephen Mumford and Rani Lill Anjum (and have finished six chapters out of ten).  I expect to blog more about this book, of which I think very highly.  I just wanted to very briefly comment on events, inspired by the treatment they are getting so far in the book.

Years ago, influenced by reading (later) Russell and Whitehead, I acquired the notion that (all else equal) there is an attraction to an ontology which gave a leading role to events rather than one primarily featuring substances (or objects) and their properties.  There seemed to be more potential for explaining the dynamic aspects of nature (including mind).

But while there has been an active modern debate on the nature of events, the most common depictions don’t seem to offer specific advantages to an event-focused ontology.  To greatly simplify, it seems philosophers would model events either as property exemplifications, in which case they are in danger of seeming much like static facts or states of affairs; or else events would be associated with spacetime locations, in which case they are little distinguished from objects, which are the quintessential occupiers of spacetime.  (The SEP article on events is here; an IEP article with additional focus on the theories of Kim, Davidson, and Lewis is here).  These sorts of models of events don’t seem to bring differentiated resources to metaphysical theorizing.

The goal of the Getting Causes From Powers book is to develop a theory of causation based on dispositional properties, or powers.  While powers play the leading role, their theory incorporates an intriguing view of events (at least causal events:  they don’t take a position on whether there are other sorts).  Specifically, causal events, which are manifestations of powers, are temporally extended processes.  The authors reject as misguided the typical “two-event” conception of causation, where cause is temporally prior to effect, in part because no one has a compelling account of how you get from one to the other.  Rather causes and effects are simultaneous – they are two aspects of a process which brings about a change.  Very Whiteheadian!