Thursday, March 5, 2009

Haecceities and Haecceitism

Haecceities are primitive identities in a world. Haecceitism has to do with> primitive trans-world identities (allowing for de re differences between worlds without qualitative differences). My question regards the connection between the two. Prima facie, possession of haecceity implies haecceitistic differences between worlds. However, it has been pointed out that identity might be primitive with respect to a set of conditions but not others (Legenhausen (1989)), and showed that the two things can be kept distinct (Lewis (1986)), and in fact haecceitism can be true even if there are no haecceities. Adams (1979), one of the main recent proponents of primitive thisness, thinks haecceitism is also true, but feels compelled to provide an argument for it, additional to the existence of haecceities. Is counterpart theory the only way to believe in haecceities but not in haecceitism? Is it relevant whether haecceities are considered to be genuine properties (Duns Scotus), or just ‘aspects’ of things only separable via conceptual distinction (Ockham and other Scholastics, Adams himself)?
Generalising, it seems four possibilities are allowed; and if one introduces the distinction between moderate and extreme forms of haecceitism and/or anti-haecceitism (that is, as I understand it, primitive identity with or without essentialist constraints on the one hand, and non-primitive identity without or with the Identity of the Indiscernibles on the other) probably even more (8? I am not sure about extreme anti-haecceitism with primitive identities, maybe it only requires the Identity of the Indiscernibles to be a contingent truth). But which combinations are really possible/plausible? What conditions do they require exactly? What are people's intuitions/preferences?


  1. "Is counterpart theory the only way to believe in haecceities but not in haecceitism?"

    A trivial way to have haecceities without haecceitism is to assume a Leibnizian framework, where individuals are world-bound and de re statements are interpreted via identity (rather than a counterpart relation). Since every individual has all its properties essentially, in particular it is not the case that

    (H) Something could have different properties than those it actually has, whereas things remain qualitatively the same.

    If we assume that entities have haecceities (e.g. by being members of their singletons), there we have a structure with haecceities but without haecceitism.

    As a side note, Skow (2007) argued that Lewis' possible-world formulation of haecceitism

    (LH) There are qualitatively indistinguishable worlds that differ by the way they represent something de re

    is not equivalent to (H). So, if (H) is the proper modal formulation of haecceitism, Lewis' counterpart theory does satisfy haecceitism, even if it does not satisfy (LH).

  2. Thank you Alessandro (we have met in Krakow last year haven't we?).
    I am aware of Skow's work, but I think the point makes sense anyway. One just needs to formulate it in terms of identities of specific entities vs. possibilities for systems of entities. In particular, I have in mind a difference related to something I am currently working on: the ontological status of a quantum particle vs. quantum systems considered from the point of view of statistics.
    Certainly, in this case as with ordinary objects in general, the 'Leibnizian way' might be looked at with suspicion.

  3. Haecceities are not the only way for there to be primitive identity. The identity relation and the non-identity relation may be primitive. If this is so then a not= b is a primitive relation that is not grounded either in qualitative features of a and b, or in intrinsic non-qualitative features of a and b (haecceities). This kind of primitive identity is definitely compatible with the denial of haecceitism.

  4. James,

    From the first sentence of Matteo's post, it looks like Matteo is assuming that one cannot be committed to haecceities without being committed to primitive identities and vice versa.

    The argument for this, I guess, would be that a cannot bear a primitive identity relation to itself without also having the non-qualitative property of being identical with a, a property which would seem to be an intrinsic property of a.

    I'd be curious to hear which of the premises of this argument you would be inclined to reject in order to deny its conclusion. Would you say that a does not have the property of being identical with a despite bearing the relation being identical with to a? Or would you say a has that property but deny that it is both a non-qualitative and an intrinsic property? In which case, would you deny is intrinsic or would you claim it is qualitative?

  5. In answer to Gabriel:

    I agree that if a be identical to itself but the point is that the relations of identity and diversity among a collection of objects may be primitive and not grounded in intrinsic haecceities. So I am saying that the facts about identity and diversity need not supervene on the intrinsic facts including the non-qualitative ones. (For such an account of primitive identity in the context of mathematics see the paper by Hannes Leitgeb and I in the latest Philosophica Mathematica. For more on the same stuff see my paper in the Proceedings of the Aristotelian Society 2007.)

  6. Sorry for the typos in that first sentence. It should say 'I agree that a is identical to itself'. Good job too that I agree with that. I have in mind in my comment the difference between what John Stachel calls 'contextual individuality' and 'intrinsic individuality'. He argues, and I agree, that the individuality of spacetime points and quantum particles is contextual.

  7. I just wanted to offer some things that came to mind while I was reading.(I'm not going to attempt to endorse any form of haecceitism here, though I do find weak haecceitism coupled with strong essentialism an attractive position.)

    It is probably important to note that Adams supports primitive thisness, not haecceities. His reason is two-fold: haecceities (supposedly) exist without their objects and are independent and distinct from every qualitative property that the object that they belong to possesses.

    Primitive Thisnesses are, according to Adams, supervienent on the qualitative properties of an object; so they do not fall prey to the two previous objections.

    He claims that they are non-qualitative, but it seems to me that they are more or less condensed qualitative statements; Russelian definite descriptions come to mind. To my mind then, Adams has primitive identity rest upon qualitative properties. I wonder then if identity is as primitive as he thinks.

    Alessandro mentioned earlier that a Leibnizian framework makes all properties of an object de re necessary and hence, gets haecceitism without haecceities. Perhaps. And perhaps this is precisely Adams' way of seeing the matter. But I don't think Leibniz thought so.

    Following Cover & Hawthorne, I think that it is more plausible that a Leibnizian Complete Concept is better understood as 'Law of the Series' (rather than a 'definite description') that is more or less a function whose outputs are all the qualitative properties that the object has; timelessly, of course. But then Leibniz has non-qualitative haecceities that ground primitive identity and so doesn't have haecceitism without haecceities.

  8. Alessandro TorzaMarch 5, 2009 at 9:03 PM


    I am only familiar with your 2007 paper. If I remember correctly, in order to show that primitive identity does not necessarily involve haecceities, you mention the case of a graph G with two nodes and no edges. You argue that, when the graph theorist defines the structure G, the nodes are meant to be two and, therefore, distinct. Since there are no edges, the nodes' individuality is primitive. But no haecceities are involved in the definition of G, hence primitive individuality does not entail the existence of haecceities.

    I wonder whether your graph example trivializes the structuralist thesis that facts about individuals are reducible to more fundamental facts about structure. Sure, if a structure S is defined as a set D of individuals *plus* a (possibly empty) set of relations, it is trivially a structural fact that D is a collection of individuals, so we do not need to postulate haecceities. This seems to be the line of reasoning underlying your graph example. But if this is the notion of structure at stake, structuralism will be true by definition.

    On the other hand, if "structure" means only the set of relations defined on D, then the individuality of the members of D may not be reducible to structural facts, for example if S allows nontrivial automorphisms. In this case we do need to introduce haecceities, for example by expanding the structure with a singleton for each object.

  9. Alessandro TorzaMarch 5, 2009 at 9:13 PM


    my point about Leibniz meant to show that we can have haecceities w/o haecceitism, not the other way around as you say.

    Regardless, I agree that the orthodox Leibnizian would refuse haecceities, but that does not prevent us from modifying the orthodox view. The crucial point was just that we can have world-bound individuals (a' la Lewis) while de re statements are interpreted via indentity (a' la Kripke).

  10. James,

    The point I was trying to make is that the argument above seems to commit you to the existence of intrinsic non-qualitative properties if you are committed to primitive identities. Once you concede that, however, the two views seem to be just two different ways to describe the same state of affairs (a's being identical with a) into components. If this is the case, I can hardly think of any good reason to argue that either is more fundamental ("grounds") the other. (Incidentally, since the supervenience in question seems to be completely symmetric, I don't see what evidential weight your supervenience claim carries) Moreover, once you concede there can be intrinsic non-qualitative properties, on what grounds would you would ban properties such as being a? And, if you accept such properties, what would be your reasons for thinking they are less fundamental than identity relations?

  11. Reading this thread (probably too quickly) reveals that I don't quite understand how folks use the terms "haecceity," "thisness," and "primitive identity." Perhaps someone can point me in the direction of something like a definition of these terms, even if vague and imprecise?

  12. Gabriele

    Basically I think intrinsic primitive identity suggests that states that differ only by the permutation of the objects in them should be counted as distinct (violating diffeomorphism invariance in GR and permutation invariance in QM) and so should be rejected in favour of contextual primitive identity.


    Hannes and I argue that the relations in question should be taken to include the relations of identity and diversity. So we don't have a problem with structures with non-trivial automorphisms because the fact that say i not= -i is a primitive fact about the structure as a whole.

  13. James,

    I am familiar with your work, and I follow what you say here. However, I must admit that I don't see what it might mean - at least for a material object - to have ungrounded but extrinsic identity (that is, identity which is not derivative on qualities/relations but doesn't also 'belong to the object'). The only answer I can think of, supported by what you say here and by your structuralist approach in general, is that an object's identity is given by the place it occupies in the structure, so that object permutations don't entail exchanged identities. But this commits you to locating the primitiveness of your identity/diversity relation in primitive intrinsic features (non-qualitative, I would add) of places, doesn't it? If so, then given that the only motivation for this is that you want to avoid haecceitism, then maybe simply resisting the claim that intrinsic primitive thisnesses imply haecceitism (or, maybe, a sort of haecceitism we don't like) would be a better option. This is in effect the point of my post.


    I am not sure I agree with your reading of Adams. He expressly says that he endorses primitive thisnesses not corresponding to any qualitative difference (although he accepts a form of moderate haecceitism allowing for *some* essentialist constraints on trans-world identities). Hence, I don't see how he is committed to claim that thisnesses 'rest on qualitative properties', and the grounds for your Russellian interpretation in general.

    Terminology (so also answering Jonathan):
    the primitive thisness of x means that the fact that x is itself cannot be analysed. Haecceities are USUALLY intended as full-blown, purely metaphysical properties determining this sort of facts. But primitive thisnesses need not be grounded on such things, whence the sort of 'haecceities are not the only way' remarks such as James' above. However, while that haecceities are properties was Scotus' view, already many medieval philosophers didn't agree (e.g., Ockham), and claimed that every complete particular 'just has' its unique identity, and the latter is not 'caused' by anything. I definitely assume this 'Ockhamist view', but should have been more careful in my post. (Maybe one could allow for intra-world/trans-world combinations additional to those I mentioned in that post, based on what form - if any - of primitive identity one assumes?).
    As for primitive identity vs. primitive thisness, I don't think there's any the difference, at least once one bears in mind that - strictly speaking - identity is a logico/linguistic relation while thisness, the property of 'being this object' is clearly metaphysical (so that, for instance, it could be true that 'a=b', meaning that two names refer to the same object, but certainly not that two objects have the same primitive thisness). Perhaps my difficulty in translating the primitiveness of James' contextual primitive identity into a clear claim about material things stems from the fact that he speaks about an identity relation holding in a formal system such as, for instance, those of graph theory?

    In conclusion, so far we have James rejecting all forms of haecceitism and Adams supporting some mild form of haecceitism. But both James and Adams seem to allow for primitive thisness of some sort (although not as a full-blown property). All other options are still to be explored, but it is already interesting to notice that primitive identity is 'striking back', as it were; and to ask, more specifically, whether a moderate form of haecceitism should be looked at with suspicion by the scientifically-minded. For example, is it bound to get the quantum statistics wrong?