The Formal Distinction
- Posted by Andrew Bailey on Tuesday, August 07, 2007 at 9:05 AM
| 
0 Comments |
Nearly every medieval philosopher who treated universals employed (at one time or other) the so-called formal distinction. I shall here state this distinction as clearly as I can (following the early Scotus), offer examples of it, and then give one of Ockham’s arguments against the coherence of such a distinction.
It’s easiest to state the formal distinction by noting its relation to two other sorts of distinction: real and conceptual. In short, the real distinction is one found in objects; the conceptual in minds.
Conceptual: x and y are conceptually distinct iff it’s possible that x be signified (brought to mind) and y not (and the other way `round). (the variables x and y here range over particular objects considered under a particular mode of presentation)
Real: (For creatures x and y) x and y are really distinct iff God could create x without creating y. Theological concerns motivate the quantifier restriction in Scotus’ theory of formal distinction (so it applies only to created things). But we can do without this restriction by stating the distinction like this: x and y are really distinct iff it is possible that x exists while y doesn’t.
Examples: Everything is really identical to itself. The Morning Star is really identical to the Evening Star. There’s no world including Venus in which Venus doesn’t exist. But nonetheless, one can bring Venus (under one mode of presentation) to mind without bring Venus (under another mode of presentation) to mind. So the Morning and Evening stars are conceptually distinct.
The formal distinction is supposed to fall somewhere between the real and conceptual distinctions. The real distinction is based entirely in objects, so to speak. It has grounds in the way things are, not merely our thought of them. The conceptual distinction is based entirely in the mind. It has grounds only in our thoughts of things. The formal distinction, on the other hand, has grounds both in objects and in our thoughts of them.
Formal: x any y are formally distinct iff x any y are really identical (not really distinct) but x and y have non-overlapping formal definitions.
This analysis of formal distinction employs the notion of formal definition, so I’ll say something about what that is. Formal (Aristotelian) definitions are supposed to pick out the essence of a thing (these will include the thing’s genus and difference). But sometimes the very same thing can be picked out with more than one such formal definition. Put differently, the very same thing falls under two distinct natural kinds.
The key thing to note here is that formal definitions are more than just modes of presentation. They say something about what kinds of things objects really are.
Scotus’ pet example: the will and the intellect. The two are really identical (God couldn’t create a will without thereby creating an intellect), but can be picked out with non-overlapping formal definitions. distinctio formalis ex parte rei,
Ockham thinks the formal distinction is rubbish. So he makes fun of it. But he also gives arguments to justify his scorn. Here’s one of them:
1. If x and y are formally distinct, then they’re really identical (definition of formal distinction)
2. If x and y are really identical, then for all F, Fx iff Fy (indiscernibility of identicals)
3. x has the property of being formally distinct from y.
4. So y has the property of being formally distinct from y.
5. But this is absurd. So it couldn’t be that x and y are formally distinct.
Ockham’s argument is a good one, I think (as far as these things go), but it rests on an assumption that I don’t think Scotus should grant. It’s clear that the indiscernibility of identicals is true when it comes to numerical identity (that is, identity simplicitur, identity full stop, identity as such). But it’s not clear that the indiscernibility of identicals holds when the identity in question is anything less. In short, Scotus has good grounds to deny premise (2) while still holding onto the indiscernibility of identicals (with respect to numerical identity).
