Agent causation depends on the notion of action. An action is a
certain type of event. An event is (at least, it may be more) a change in a
state of affairs. Many (perhaps all) events have causal powers, and
event causation is a very common notion in natural science. Event
causation is understood in terms of antecedent conditions and covering
laws--a nomological-deductive model.
An action differs from an ordinary event in that an action
involves intentionality, and this brings in the agent. An agent performs
an action for the purpose of bringing about a certain state of affairs.
If this intention is rational and significantly free, then it is
personal. Agent causation is law-like in that is describable in general
terms ('x did A for the purpose of/to bring about S'), but (unless one
wants to adopt psychological or neurophysiological-reductive accounts of
intentionality) the operation of agent causation is not nomologically necessary.
Thus agent causation, depending on the concept of intentionality, is a
non-natural explanation. (Some argue that our concept of causation
derives ultimately from our conception of ourselves as causal agents, so
agent causation is a more basic concept than event causation.)
Agent causation may be a viable alternative explanation to a
naturalistic explanation for a certain phenomenon. Certainly in the
social sciences it holds a centrql place. And in those sciences in the
gray area between social and natural, such as archaeology, it is a
crucial concept. An archaeologist who uncovers an object with certain
interesting markings may consider a natural explanation (forces of
erosion + adequate time), but a non-natural explanation in terms of the
intentional production of an artifact, i.e. agent causation, maybe the best
explanation.
Clearly agent causation, since it doesn't depend upon antecedent
conditions and natural laws, is quite appropriate when answering the
question "What caused the Big Bang" and similar questions. But I wonder
if it might also play a role in explanations of phenomena other than at the
boundary conditions of the universe, such as the appearance of new phyla,
or the appearance of high-information-content genetic material.
Is this any different, Keith Miler asks, from divine providence?
Or, it might be asked, is this more than just a fancy name for a "God-of-the
gaps" approach? Maybe not. But since agent causation is a respectable
philosophical concept, and indeed a necessary concept if human freedom
means anything, it is not ad hoc. It would, however, cut against the
methodological naturalism of many natural scientists since it is by
definition non-natural.
Garry DeWeese
P.S. Sorry for the length of this post. Also sorry if this has been
beaten to death before; I've only been lurking on the asa list for some
six weeks.