An essential property is a property that an object possesses in every possible world in which that object exists. An individual essence is a property (or set of properties) that an object possesses in every world in which that object exists, and that no other object possesses in any possible world. Call the claim that some artifacts possess an individual essence ‘artifactual essentialism’. I will argue that artifactual essentialism is true. In doing so, I will be responding to two recent arguments by Penelope Mackie against artifactual essentialism (Mackie (2006), esp. ch. 3.). In “Individual Essence Properties”, I will rehearse the qualifications that any property must meet if it is to constitute an individual essence, and in “Artifacts and the Recycling Problem” and “Artifacts and the Tolerance Problem”, I will rehearse Mackie’s arguments against artifactual essentialism. In “Artifacts and Weak Unshareability?” and “Artifacts and Strong Unshareability?”, I will show why both of these arguments fail. In “Mona Lisa’s Essence”, I will defend the interesting claim that some artifacts possess an individual essence. In the final section I will entertain some objections to my proposal.
This is a preview of subscription content, log in to check access.
Buy single article
Instant access to the full article PDF.
Price includes VAT for USA
Subscribe to journal
Immediate online access to all issues from 2019. Subscription will auto renew annually.
This is the net price. Taxes to be calculated in checkout.
I will be using the singular ‘property’ as a shorthand way of talking about property sets, noting that a property set may, of course, include more than one property.
Mackie (2006, pp. 19–22) also wants to exclude other sorts of IEP candidates from consideration: Plantingan world-indexed properties (e.g. Rex Stout possesses the property of ‘being the author of Some Buried Caesar in this (the actual) world’ in every world in which Rex Stout exists), and ontologically primitive haecceities. For the sake of simplicity, I will categorize these properties as “trivial” essential properties as well.
I understand artifacts to be objects whose properties have been intentionally modified by rational agents in order to possess: (a) some practical and/or aesthetic function(s) (i.e. a function of being used in a certain way or artistically appreciated in some way), and (b) a doxastic function (i.e. a function of causing rational agents to come to believe certain things about the object’s practical and/or aesthetic function(s)). I understand artworks to be a special subclass of artifacts proper. Cf. Dipert (1993, ch. 9).
Mackie does not explain how she understands the notion of a “plan.” Following Mele, I take a “plan” to be an “action-plan,” the representational content of an intention. To take a very simple case, the representational content of one’s intention to hit the Enter button on one’s computer is a prospective representation of one’s hitting the Enter button on one’s computer. For our purposes, we can think of the plans according to which artifacts are created as mental “blueprints” or “instruction manuals.” Because such mental “blueprints” are differentiated by their content, many different artifacts can be created according to the same plan. For more on action-plans see Mele (1992).
It warrants pointing out that for the Recycling Problem to get off the ground, Mackie is committed to the following two assumptions: first, that S1and S2 are, in fact, numerically distinct objects, and second, that the parcel of matter that composes S1 is, in fact, the same parcel of matter that composes S2. For the sake of argument I will be granting Mackie these two assumptions, but will note that there may be good reason to doubt them. For example, one might have the following worry about the second assumption, supposing we grant the first: even though S1 and S2 possess the same underlying matter, the parcel (or hunk) of matter that composes S2 might be distinct from the parcel of matter that composes S1 because one might think that for the same reason you cannot disassemble and reassemble an artifact, you cannot disassemble and reassemble a parcel or a hunk of matter. Mackie appears to need some kind of explanation for this apparent asymmetry. I am grateful to an anonymous referee for pointing this out to me.
One way to deny this conclusion is to deny that the accessibility relation between possible worlds is transitive. However, because I think that is too-high of a price to pay, my criticism of the Tolerance Problem will not rest on it, as we will see below. However, to pursue that line of thought, one may consult Salmon (2005).
It might also warrant pointing out that even if my objection to Mackie’s Recycling Problem fails, this does not mean that Mackie has shown that no artifacts can have weakly unshareable essential properties. For even if her argument goes through, this would only show that one specific property (MP) cannot be both essential and weakly unshareable. This would not entail that no other property is both essential and weakly unshareable.
There are other interesting ways in which the Actional and Material Senses of ‘compose’ differ. For example, “originally composed” in its Material Sense appears to be an “at-a-time,” or stative notion. On the other hand, “originally composed” in its Actional Sense appears to be an “over time,” or eventive notion. My thanks to an anonymous referee for pointing this out to me.
See, for example, Kim (1976).
Readers familiar with particular aspects of the philosophy of art literature may notice that my invocation of the act-type/act-token distinction as it relates to the causal history of artifacts recalls a disagreement found in the work of Greg Currie and David Davies about how to understand the ontological status of a piece of art. Whereas Currie has argued that artworks are action-types (1989), Davies has recently criticized this view and has argued that we have much better reason to understand artworks as action-tokens (2004, see esp. chs. 6 and 7). Although I have the same distinction in mind, I am making a different use of it. Whereas Currie and Davies are concerned about how we should understand the kind of thing a piece of art is, I am not concerned with the ontological category under which artworks fall. Rather, I am presently concerned with what kinds of historical properties artifacts possess. My claim is that artifacts possess the historical property of having been composed by a certain set of action-tokens, not only a set of action-types.
It should be noted that possessing the property picked out on the former reading entails possessing the property picked out on the latter reading, but not vice versa. Piers Rawling brought this to my attention.
In other words, many numerically distinct events at a world can fall under the same general causal law.
In the interests of full disclosure, however, I should confess that I find that the former reading of P to be a more intuitive way of picking out token events (i.e. by their unique causes, and not by the general natural laws under which they fall). Others may disagree. But even so, this ambiguity reveals an important problem for Mackie’s argument, for I have identified a property that is weakly unshareable, namely the possession of a certain token action history. However, one might object that such a property is trivial because the property is being identified by the very thing it leads to and is therefore non-qualitative. But this worry is assuaged if we keep in mind that when we stress the importance of the causal history of artifacts, we are not simply claiming that those historical properties individuate artifacts. Rather, (as David Davies points out and as we will see below) we are claiming that the causal histories of artifacts “significantly determine our sense of what the resulting work is and what its essential properties are” (2004, p. 118). Because the causal histories of artifacts are important in this respect, we would be remiss to classify their possession as “trivial.” I am grateful to an anonymous referee for raising this point.
I am very grateful to an anonymous referee for encouraging me to clarify my argument in this section and offering some very helpful advice in the process.
Of course, another way to rebut the Tolerance Problem is to reject the Tolerance Principle and endorse a view (call it ‘compositional mereological essentialism’) that if an object has even a slightly different compositional makeup at its origin in another possible world, it’s a different object. I do not endorse this view, but it is not as counter-intuitive as one might initially think. One could argue, for example, that there is no principled difference between claiming that 1% of an artifact’s actual, original matter could have been different but that, say, 50% of its actual, original matter could not have been different. In either case, the compositional mereological essentialist can claim, you do not have 100% of the actual, original matter, so why should it matter whether you exchange one percent of fifty? I am grateful to Craig Warmke for making this suggestion.
What I am calling the Compossibility Problem is a modified version of Forbes’s ‘Four Worlds Paradox’ 1985, ch. 7) which has been modified by Mackie (2006, ch. 3) to make this very point: that to deny the Restriction Principle results in the putative contradiction noted above. I am grateful to an anonymous referee for pressing this point, and from whom I have borrowed some language in formulating this problem.
This promissory note will be redeemed in the next section, fn. 25.
Indeed, some philosophers have argued that some artifacts possess essential properties. See, e.g. Davies (2004), Denkel (1995) and Dipert (1993). However, as we have seen, to claim that some artifacts possess an essential property is not to claim that some artifacts possess an individual essence, for a property may be essential to an artifact without it being both weakly and strongly unshareable.
For more on the physical structures of artifacts, see Denkel (1995)
I am aware that there are important questions having to do with how we should individuate beliefs, desires, and intentions. Such questions usually center on whether these kinds of mental attitudes are individuated by their ‘wide’ or ‘narrow’ content. For my purposes, here, I need not take a firm stand on this issue, but I should say that I do conceive of these attitudes, at least with respect to the creation of artifacts, as being general or non-specific. That is, when an artisan sits down to create a artifact, she does not (or, at least typically does not) intend to create some specific artifact—the content of her plan to create a chair, for instance, is not so fine-grained as to include in its representational content some specific chair. Rather, her intention is (at least typically) a more coarse-grained intention to create some non-specific chair. This is true even if her general, non-specific plan to create a chair contains further sub-plans for how to go about building a chair. I therefore understand types of beliefs, desires and intentions to be individuated by their coarse-grained content, which is all I think is necessary to make sense of why the deliberative histories of artifacts can be essential to their identity. I therefore set aside the contentious issue of how to individuate token intentional attitudes that possess the same type of representational content. For more on intentions and their content, see Mele (1992). For more on the deliberative histories of artifacts, and especially their intentional histories, see Dipert (1993, ch, 3), and Davies (2004, ch. 7). I am grateful to an anonymous referee for encouraging me to clarify this point.
For more on the deliberative histories of artifacts, see Dipert (1993)
Here, I make good on a promissory note from “Artifacts and Strong Unshareability?”. There, I claimed that the Compossibility Problem fails to produce a contradiction (thereby failing to evince the Restriction Principle) if the property in question is weakly unshareable. But property set (i–iii) is weakly unshareable. Therefore, the Compossibility Problem poses no threat to my rejection of the Restriction Principle in the present case. One may object that such a property is not essential to the Mona Lisa (I will try to deflect that criticism below), but it is clearly weakly unshareable, which is what is needed (along with my rejection of the Restriction Principle) to insulate my account from both the Recycling and Tolerance Problems.
My claim here rests on the assumption that the very events that occur in the actual world also occur in other possible worlds. Some may demur at this claim. Some may claim, for example, that in those other worlds, what we have is a numerically distinct event, very similar to the event in the actual world, and falling under the same natural laws. However, it would take a much longer treatment of the modality of events to address this point, space that I do not have available here. But because there seems to be nothing obviously wrong about such an assumption, I will help myself to it. I return to this point in fn. 36. I appreciate an anonymous referee for pointing out this assumption.
There are, of course, very serious and difficult issues having to do with artifactual identity through deterioration and/or restoration. However, these are difficult problems for any view of the metaphysics of physical objects that claims that artifacts can persist through change, and so those problems present no unique problem for my claim that some artifacts possess essentially, at some appropriate level of description, a physical structure.
Although I have co-opted Davies’ argument in defense of my claim that the Mona Lisa’s token action history is essential to it, it warrants pointing out that ultimately, Davies concludes that the artistic performance is the artwork, not just an essential property of it. Although I am sympathetic to this view, I remain agnostic about what sort of thing an artwork like the Mona Lisa is. Even so, I find his argument to the conclusion that an artwork’s action history is essential to it persuasive and so I happily use it here. For a fuller defense of the claim that an artwork’s token action history is essential to it, see his 2004, chs. 4–7.
For arguments against the claim that what is essential to an artwork is its type of action history, see Davies (2004, ch. 6).
Very briefly, here is one reason to suspect that the Mona Lisa’s deliberative history does not supervene on its token action history. If the Mona Lisa’s deliberative history supervened on its action history, da Vinci could not have, in another world, performed the same token mental and physical actions (e.g. decisions and bodily movements) but had beliefs and desires with different content at some point in the deliberative history. This seems implausible, at least on the face of things. If so, then an artifact’s deliberative history does not always supervene on its action history and would appear to be as independent as I have suggested. However, even if this is wrong, I should stress that I am comfortable with that—my goal is to gesture toward a plausible artifactual IEP. I am happy if some of my gesturing hits the mark.
I am grateful to an anonymous referee for this journal who raised this objection, and who attributed this sort of case to Gregory Currie.
One might wonder why in this case I have said that a different action history gives us a distinct painting but in the case where da Vinci does not care who is sitting for the portrait, I have implied that a different action history (i.e. his painting of Schmona instead of Mona) does not necessarily give us a distinct painting. The answer has to do with da Vinci’s artistic intentions. Not all of his actions in creating the Mona Lisa are essential to it. Suppose for example, da Vinci quietly tapped his big toe for 3 s one day while finished the nose. Certainly we do not want to say that this action is essential to the Mona Lisa. The reason is simply because (we can suppose) such an action was irrelevant to da Vinci’s artistic intentions. In the first case above, we have stipulated that da Vinci doesn’t care who is sitting for the portrait as long as whoever is sitting there is indistinguishable from Mona, so in a counterfactual situation in which he paints Schmona, we can treat this like the tapping of the foot. This explains the difference.
My thanks to Craig Warmke for helping me to formulate this reply.
To briefly return to a point I made above in fn. 26 about the trans-world identity of events: in this passage, Davies appears to think of the very same events that occur in the actual world as occurring in other possible worlds when he rigidly designates event ‘e’ and makes reference to “any possible world in which it [e] occurs.” This appears to imply that e occurs in this world as well as in other words, and that the event in question at another world is not just some numerically distinct event, very much like e, falling under the same natural laws.
One promising strategy for approaching these issues is found in Davies (2004). His strategy is to make a distinction between what he calls “doings” and “token-actions strictu sensu” (p. 171–173). To put things roughly, “doings” are action-tokens whose individuating conditions have been relaxed so as not to include the specific time at which the action occurs as essential to its identity. By understanding the action-history of the Mona Lisa as a “doing” we can retain the claim that the token-actions by which it was brought about are essential without having to be committed to the dubious claim that the exact time at which it was created is essential. Notice, however, that even if we construe an artifact’s action history as a “doing,” the property of possessing that doing would still be weakly unshareable, for every doing occurs at most once at a world.
I am grateful to Steve McFarlane, Michael McKenna, Piers Rawling, Michael Robinson, and Craig Warmke for their incisive comments on and helpful criticisms of previous drafts of this paper. I would also like to thank an anonymous referee for this journal, who has been extremely helpful in providing very insightful and gracious comments.
Currie, G. (1989). An ontology of art. New York: St. Martin’s.
Davies, D. (2004). Art as performance. Malden: Blackwell.
Denkel, A. (1995). Artifacts and constituents. Philosophy and Phenomenological Research, 55, 311–322.
Dickie, G. (1971). Aesthetics. Indianapolis: Bobbs-Merrill.
Dipert, R. (1993). Artifacts, artworks and agency. Philadelphia: Temple University Press.
Forbes, G. (1985). The metaphysics of modality. Oxford: Clarendon.
Kim, J. (1976). Events as property exemplifications. In M. Brand & D. Walton (Eds.), Action theory (pp. 159–177). Dordrecht: D. Reidel.
Mackie, P. (2006). How things might have been. Oxford: Oxford University Press.
Mele, A. R. (1992). Springs of action: understanding intentional behavior. Oxford: Oxford University Press.
Salmon, N. U. (2005). Reference and essence (2nd ed.). Amherst: Prometheus Books.
About this article
Cite this article
Warmke, B. Artifact and Essence. Philosophia 38, 595–614 (2010). https://doi.org/10.1007/s11406-009-9226-0