The Dispute
In an earlier entry that addressed Lukas Novak's argument against bare particulars I said the following:
The notion of a bare particular makes sense only in the context of a constituent ontology according to which ordinary particulars, 'thick particulars' in the jargon of Armstrong, have ontological constituents or metaphysical parts.
[. . .]
LN suggests that the intuitions behind the theory of bare particulars are rooted in Frege's mutually exclusive and jointly exhaustive distinction between concepts and objects. "Once this distinction has been made, it is very hard to see how there might be a genuine case of logical de re necessity." (115) The sentence quoted is true, but as I said above, the notion of a bare particular makes no sense except in the context of a constituent ontology. Frege's, however, is not a constituent ontology like Bergmann's but what Bergmann calls a function ontology. (See G. Bergmann, Realism, p. 7. Wolterstorff's constituent versus relation ontology distinction is already in Bergmann as the distinct between complex and function ontologies.) So I deny that part of the motivation for the positing of bare particulars is an antecedent acceptance of Frege's concept-object distinction. I agree that if one accepts that distinction, then logical or rather metaphysical de re necessity goes by the boards. But the Fregean distinction is not part of the motivation or argumentation for bare particulars.
My claim that bare particulars are at home only in constituent ontology raised the eyebrows of commenter John and of LN, who writes:
I cannot see why the notion of a bare particular should make sense only in a constituent ontology. A bare particular is a particular which has none of its non-trivial properties de re necessarily. This notion is quite intelligible, irrespectively of the way we go on to explain the relation of "having" between the particular and the property, whether we employ a constituent or functional or some other approach (of course, saying that it is intelligible is not saying that it is consistent!). If Bill agrees that once one makes the sharp Fregean distinction between concepts and objects then there is a strong motivation against conceding any de re necessity, then he should also agree that making this distinction provides a strong motivation for claiming the bareness of all particulars.
Resolving the Dispute
I believe that this is a merely a terminological dispute concerning the use of 'bare particular.' I am a terminological conservative who favors using words and phrases strictly and with close attention to their historical provenience. To enshrine this preference as a methodological principle:
MP: To avoid confusion and merely verbal disputes, never use a word or phrase that already has an established use in a new way! Coin a new word or phrase and explain how you will be using it.
Now, to the best of my knowledge, the phrase 'bare particular' enters philosophy first in the writings of Gustav Bergmann. So we must attend to his writings if we are concerned to use this phrase correctly. Now in the terminology of Wolterstorff, Bergmann is a constituent ontologist as opposed to a relational ontologist. In Bergmann's own terms, he is a "complex" as opposed to a "function" ontologist, Frege being the chief representative for him of the latter style of ontology.
"In complex ontologies, as I shall call them, some entities are constituents of others." (Realism, p. 7) "In function ontologies, as I shall call them, some entities are, as one says, 'coordinated' to some others, without any connotation whatsoever of the one being 'in' the other, being either a constituent or a part or a component of it." (Ibid.)
Bergmann, then, is a constituent or complex ontologist and his introduction of bare particulars (BPs) is within this context. BPs are introduced to solve "the problem of individuation." A better name for this problem is 'problem of differentiation.' After all, the problem is not to specify what it is that makes an individual an individual as oppose to a member of some other category; the problem is to specify what it is that makes two individuals (or two entities of any category) two and not one.
How does the problem of individuation/differentiation arise? Well, suppose you have already decided that "some entities are constituents of others." For example, you have already decided that ordinary particulars (OPs) have, in addition to their spatial parts, special ontological parts and that among these parts are the OP's properties. Properties for Bergmann are universals. Now suppose you have two qualitatively indiscernible round red spots. They are the same in respect of every universal 'in' them and yet they are two, not one. What is the ontological ground of the numerical difference?
On Bergmann's way of thinking, one needs an entity to do the job of individuation/differentiation. Enter bare particulars. And pay close attention to how Bergmann describes them:
A bare particular is a mere individuator. Structurally, that is its only job. It does nothing else. In this respect it is like Aristotle's matter, or, perhaps more closely, like Thomas' materia signata. Only, it is a thing. (Realism, p. 24, emphasis added)
Bare particulars, then, have but one explanatory job: to ground or account for numerical difference. They are the Bergmannian answer to the question about the principium individuationis. But please note that the positing of such individuators/differentiators would make no sense at all if one held to a style of ontology according to which round red spots just differ without any need for a ground of numerical difference. For a relational ontologist, OPs have no internal ontological structure: they are ontological simples , not ontological complexes. Here is Peter and here is Paul. They just differ. They don't differ on account of some internal differentiator. Peter and Paul have properties, but these are in no sense parts of them, but entities external to them to which they are related by an exemplification relation that spans the chasm separating the concrete from the abstract. And because OPs do not have properties as parts, there is no need to posit some additional ontological factor to account for numerical difference.
I think I have made it quite clear that if we use 'bare particular' strictly and in accordance with the phrases' provenience, then it simply makes no sense to speak of bare particulars outside the context of constituent ontology.
Unfortunately or perhaps fortunately, I am not the king of all philosophers and I lack both the authority and the brute power to enforce the above methodological imperative. So I can't force otber philosophers to use 'bare particular' correctly, or to put it less tendentiously: in accordance with Bergmann's usage. But I can issue the humble request that other philosophers not confuse the strict use of the phrase with their preferred usages, and that they tell us exactly how they are using the phrase.
Novak's usage is different than mine. He tell us that "A bare particular is a particular which has none of its non-trivial properties de re necessarily." On this usage my cat would count as a bare particular if one held the view that there are no non-trivial essential properties, that all non-trivial properties are accidental. But for Bergmann a cat is not a bare particular. It — or to be precise, a cat at a time — is a complex one of whose constituents is a bare particular. My cat Max is a Fregean object (Gegenstand) but surely no Fregean object is a Bergmannian bare particular. For objects and concepts do not form complexes in the way BPs and universals form complexes for Bergmann.
On a Fregean analysis, the propositional function denoted by '___ is a cat' has the value True for Max as argument. On a Bergmannian analysis, 'Max is a cat' picks out a fact or state of affairs. But there are no facts in Frege's ontology.
To conclude: if we use 'bare particular' strictly and in accordance with Bergmann's usage, one cannot speak of bare particulars except in constituent ontology.