Friday, December 9, 2011

Universals vs. Genera vs. Generals in Realism vs. Nominalism

I am continuing the discussion of realism and nominalism.  Here I make the distinction between genera (genus-species relationships), generals (e.g., Lockean abstractive processes), and universals.  I also explain the importance difference between a realist and nominalist on several key points of metaphysics and phenomenology.

Leon and I were discussing the basics of Duns Scotus, etc., and my own appropriations of the terms that adds some Aquinas.  Below is an edited portion of our discussion.

The terms are “real distinction,” “formal distinction,” and a third term I do not know off the top of my head.  We often just say a “mere distinction" or "merely formal."

Universals and genera are of the second kind.  I forgot what the perfections are, which are special cases of this ontological problem.  Note that genera and generals are distinct, whereas the latter refer to principles arrived at via an inductive process, e.g., Locke’s “abstraction” or logical induction.  The former refer to genus-species relationships, e.g., categorical logic.

Correct, generality (generals) is neither a real nor formal distinction.  Hence, a nominalist often thinks that generality is all we have—not universality or genera of being.

As for essence vs. thing, you are right.  If we are Hobbes-style nominalists and think that all things are (corpuscular) particularities, then we deny essences.  Recall that essences are a kind of universal, while quality is another kind of universal.  Also, "essentiality" is not the category of uniquity (uniqueness); the latter is "quiddity."  As I wrote in my post, nominalism gives up on substantial or essential identity—identity is a best a function of something.  It does not necessarily give up on absolute particularity.

As for whiteness, it is a universal and a general, but not in the same way for both.  Insomuch as whiteness has reality, it is a universal.  Insomuch as we experience or know whiteness, it is a general; we infer from experience that this encounter is of the category of whiteness.  Now, if we are not realists about universals, then we know merely the generality “whiteness.”  The problem here is that we no longer experience the real thing, but merely a generated appearance.  If one is a nominalist, one does not think that there is anything “under” this generation.

In my Peirce-Deweyan position, for instance, we add to the idea that whiteness is a universal and general.  We talk about the generation of the phenomenal quality, so we are talking about a generated quality.  However, since we think that generation is a real process, then the generated quality maintains a real, non-arbitrary relation to the thing experienced.  (Note that the “thing experienced” is not an entity or object, but I’m keeping it simple for now.)  A nominalist, on the other hand, can merely say that the phenomenal quality was generated, full stop.


  1. p.s. a real universal would be a constraint on generals. E.g., "red" refers not to an arbitrary particular given a name, but something that exists in reality independent of any naming, reference, etc.

  2. I would to add a follow-up lest I be named a correlationist. The very first counter I expect to what I've said is that I'm regressing into a "philosophy of access." False. There is a difference between restricting what can be said to *how* we can know it in advance, and insisting that explanation is in principle possible although not requiring it in advance. I affirm the latter but not the former. Else, what good is the principle of sufficient reason other than to reject ex nihilo causation, etc.?

    A nominalist that refuses to accept anything other than particulars and generals cannot say anything concerning the passing present, or perhaps present as past, as Leon notes. Such a thinker would have a hard time discussing the future except through induction, but then must find a way to circumvent the inductive fallacy of Hume--that we can never presume that the future will be like the past. Much of this can still be circumvented, I believe, by pragmatic or abductive procedures, but the result will be very limited. I cannot in abstract say how limited; we would have to discuss the work of particular thinkers.

    In sum, I am not arguing for correlationism qua frisking interlocutors for how they "access" what they claim in advance.

  3. Addendum to above.

    It would be far too easy to counter what I say by claiming that I am a correlationist who is is trying to shut people down by asking "how can you know?" in the usual post-Kantian way. Rather, I only insist *that* one could know in principle, but I do not insist that one demonstrate how.

    That's the difference between a Kantian transcendental deduction and an abduction; the former is a special kind of deduction, while the latter is an induction followed by a speculative deduction per contemporary formal logic. In sum transcendental arguments are not necessarily the same as abductive arguments.

    One might say, loosely, that we induce from sensation to perception, and then speculatively deduce from perception to interact with the world to investigate that perception.

  4. counter to someone playing devil's advocate.

    I never said that the speculative deduction was in any way limited by human being. I did imply that something must have made us think about the topic, which is trivially true. Rather, I insist that someone offer their criteria for making the abductive-deduction. That way we can evaluate between speculative views.