I’d like to run an idea past all of you and get your
reactions. My apologies for the length
of this – I don’t know how to say this more briefly.
I’m going to use the word “random” here primarily in sense
of being a property of a sequence of events that makes the n+1st event
unpredictable from the first n events.
This probably conflates the words “chance” and “random” as they are used
in the SEP article, but I need to get a bit more comfortable with that distinction
before I start using it.
A distinction is often made between epistemic and
ontologically random processes. The
former means the events it generates are unpredictable but with enough
additional knowledge, it could be made predictable. The latter typically means that no amount of
additional knowledge could render the events predictable. Some statisticians and scientists have
regarded this distinction as of no practical importance since none of their practices
with actual data would be any different for an epistemically random or an ontologically
random process. However, philosophers
and some philosophically oriented scientists still regard the distinction as
important on ontological grounds. That
is, if all randomness is epistemic, then the world is deterministic. If ontological randomness exists, it is
not. So, they argue, much of how we see
the fundamental nature of reality hinges on this question.
I’d like to argue that the distinction is not
philosophically meaningful either.
Consider the example of gender selection for a child. (The argument applies to any genetic factor
in any organism that reproduces sexually, but, for clarity’s sake, let’s keep
it simple.) It’s certainly random. Is it epistemic or ontological? Well, the process is incredibly complex. Enormous numbers of sperm are swimming around
prior to ejaculation and the timing of which one is best positioned to win the
race is critical. And who can predict
which one is the best swimmer? Also,
there are roughly 100 million conceptions each year and (with the exception of multiple
births) these are independent of one another.
However, one could argue that with enough detailed knowledge, the uncertainty
as to which sperm wins could be eliminated; furthermore, complete knowledge of
timing could render the independence irrelevant. So a case could be made for this being
epistemically random. However, that
level of detailed knowledge is, for many reasons, inaccessible to human
beings. But perhaps it is accessible to
God, so from the divine perspective, gender selection is epistemic. But let’s suppose for the moment that God
does not normally direct the gender selection process; in Thomistic terms, he
uses it as a secondary agent, in this case, a non-deterministic one. So this means that even though the detailed
knowledge that would enable prediction might be available to God, He does not
use it. Let’s go even further and
suppose God created it to be a non-deterministic process. Then it follows that non-determinacy is a fundamental
property of that process. If that’s the
case, it would seem inappropriate to deny its ontological randomness. That is, if one defines ontological
randomness as “not epistemic,” gender selection is not ontologically
random. But if one defines it in terms
of the properties of the process God gave it, it is ontologically random. Thus, in this sense, it is both epistemically
and ontologically random. So the distinction
between them breaks down.
Of course, I still haven’t addressed the question of why I
think God created the process to operate non-deterministically. I
appeal to two principles: (1) inference to the best explanation, and (2) divine
goodness. The process of gender
selection along with diffusion, quantum collapse, mutation, the operation of
the immune system, and many others appear random. There is no observable evidence of a hidden
agent directing them but a straightforward, transparent explanation can be
given in terms of non-determinacy. So it’s the better explanation. Also
as Einstein put it, “The Creator is subtle but not malicious.” So I would call all of these processes
ontologically random without denying the possibility that knowledge may be
obtainable that reduces their uncertainty.