Of what, exactly, is something said to be independent, when it is said to be independent of «ourselves»? This doesn't have to be construed as a question about the nature of selves or persons , or indeed, about the nature of any kind of entity (though it might be), for it will be readily admitted that when the realist claims that something exists/obtains independently of ourselves, he/she (most often) means to be saying not that it exists/obtains independently of the fact that we or «our minds» exist (which would mean that it exists/obtains independently of any mental property that we might have), but more modestly, that it exists/obtains independently of the fact that we have some specific mental feature or power M. As far as I can see (and historically, at least), there are three main candidates for playing the part of the designated mental feature M: (i) our capacity to know, (ii) our capacity to think or conceive and (iii) our capacity to perceive or experience.
Thus, in most cases, the realist who claims that something exists/obtains independently of ourselves is really claiming that it exists/obtains independently either (i) of our having the capacity to know it (i.e., of its being knowable by us), or (ii) of our having the capacity to conceive it (i.e., of its being conceivable by us), or (iii) of our having the capacity to experience it (i.e., of its being experiencible by us). This suggests that the selected mental feature or power M will more often than not be a relational one; that is, one which we are supposed to have with respect to a certain thing, in the sense that, e.g., while the state-of-affairs that P may be said to obtain independently of our possessing M with respect to P, some other state-of-affairs Q will be said to obtain independently of our possessing M with respect to Q. This will later prove to be of some relevance.
But for now, I want to consider how the notion of independence itself is to be construed. It will be convenient, for this purpose, to take dependence as the primitive notion, and to understand independence as the denial of dependence (and hence as a non-symmetric relation). Thus, to say that something exists/obtains independently of our possessing mental feature M is just to say that (it exists/obtains and) its existence/obtaining doesn't depend on our possessing M. And to say that the existence/obtaining of something depends on our possessing M is to say that a certain state-of-affairs depends on another (or in the semantic mode, that the truth of some judgement depends on that of another). Thus, the notion of dependence must here be construed as a relation between states-of-affairs or judgements.
Two natural readings of the claim that the state-of-affairs (judgement) that P depends on the state-of-affairs (judgement) that Q are (i) the counterfactual one, according to which P depends on QFoot note 1 iff (a) it would not be the case that P if it were not the case that Q and (b) if it were the case that Q it would be the case that PFoot note 2, and (ii) the modal one, according to which P depends on Q iff necessarily, if P then Q. In both cases, we are dealing with entailment relations: to say that P counterfactually depends on Q is to say that not-Q counterfactually entails not-P and Q counterfactually entails P, and to say that P modally depends on Q is to say that P necessitates Q (i.e., that not-Q necessitates not-P). As Lewis (1973: 167) remarks, when it is assumed that both P and Q are actually the case, P will counterfactually depend on Q just in case not-Q counterfactually entails not-P, since the other condition will trivially hold. On this assumption, modal dependence turns out to be strictly stronger than counterfactual dependenceFoot note 3.
At first sight, there would seem to be nothing to prevent either of these two construals of the dependence relation to be involved in some particular realist/irrealist dispute. But it has recently been argued (by Jenkins (2005)) that none of them is strong enough to capture what is «really» at stake in such disputes. According to Jenkins, Fact-realism about some subject matter must not be construed as the claim that the facts pertaining to this subject matter are modally or counterfactually independent from the mind, but as the claim that they are essentially independent from the mind, where essential dependence is taken to be stronger than either modal or counterfactual dependenceFoot note 4.
I'm now going to discuss this proposal in some detail. But it is worth mentioning at the outset that even if we were to grant Jenkins point that any «genuine» realist thesis must involve the claim that something is «essentially» independent of our possessing some mental feature, it would not follow that all forms of irrealism must similarly involve the claim that something is «essentially» dependent on our possessing some mental feature; for as was pointed out above, there are forms of irrealism which oppose the realist's «actuality» claim without opposing his/her «independence» claim. Hence, it is to be understood that only what I call forms of «Independence-irrealism» are at issue in the following. That being so, it seems appropriate to assume that we'll be dealing only with states-of-affairs which do obtain (or are taken to obtain by both the realist and the irrealist), and thus with cases where modal dependence is strictly stronger than counterfactual dependence. The latter can accordingly be put aside in what follows, since it could not be «strong enough» unless modal dependence also is «strong enough»Foot note 5.
Let me now make a simple observation about the dialectic of the debates opposing the realist and the irrealist. It would seem that anyone engaging in such a debate (or indeed, in any kind of theoretical debate) should strive to defend the strongest position one thinks is rationally defensible: the stronger the better. This means that the (Independence-)irrealist should aim at establishing the strongest form of mind-dependence he/she can (the stronger the form of mind-dependence he/she is able to establish, the better he/she should feel), while the realist should aim at denying the weakest form of mind-dependence he/she can (the weaker the form of mind-dependence he/she is able to deny, the better he/she should feel). If this is right then Jenkins'main contention, namely, that realism must be construed as involving not a denial of modal mind-dependence, but a denial of some stronger form of mind-dependence which she calls essential mind-dependence, can provide little more than a fall back position for any would-be realist. Accordingly, and correspondingly, it does challenge the (Independence-)irrealist to establish a stronger form of mind-dependence than any he/she may have wanted (or thought possible) to maintain.
I doubt that there is any clear and definite answer to the question which form of mind-dependence best captures the «true spirit» of all realist/irrealist theses. As I see it, various forms of mind-dependence are worth investigating, in various domains, and with respect to various mental features. It is quite obvious that one may well hold both that something modally (essentially) depends on our possessing some mental feature M, and that it doesn't modally (essentially) depend on some other mental feature M*. It should be no less obvious that just as there is nothing to prevent one from being both a counterfactual mind-dependence irrealist and a modal mind-dependence realist in a certain domain, there is nothing to prevent one from being both a modal mind-dependence irrealist and an essential mind-dependence realist in some other domain. The question whether someone in this position should be described as a «realist» or an «irrealist» tout court is largely a matter of taste and convenience.
That being said, it certainly makes sense to ask what is the strongest form of mind-dependence which could plausibly be maintained, relative to any given choice of (i) a kind of judgements or facts K and (ii) a specific mental feature M. And there is indeed some reason to think that the claim that something's being the case (the obtaining of some state-of-affairs) modally depends on our possessing mental feature M will not always succeed either in capturing the irrealist intuitions about the given subject-matter, or in defeating the corresponding realist intuitions. On the other hand, some forms of dependence might be so strong that virtually nothing could have this relation to anything other than itself (or to any mental feature thought to be relevant to the realism issue), thus threatening to trivialize any realist claim to the effect that something or other doesn't bear this relation of dependence to this or that mental feature. Clearly, to insist that the realist position be defined by its denial of any such strong dependence relation is to be unacceptably biased against irrealism. What I am going to argue is that, if the notion of essential dependence is going to be understood along the lines suggested by Jenkins, then appealing to it in this way runs a serious risk of expressing just such a bias.
Jenkins'characterization of the notion of essential dependence takes its inspiration from some work of Kit Fine (1994, 1995)Foot note 6. It is, however, a matter of some importance that this work doesn't at all trade in the realism/irrealism business, but is concerned with finding an intuitively satisfying way of understanding a general notion of ontological dependence, conceived as a relation that some entity bears to some other entity, when the existence of the first can be said to depend on the existence of the second, or on the second's being a certain way.
From this perspective, it is easy to show that it would be inadequate to understand the claim that some entity A ontologically depends on some other entity B as meaning simply that necessarily, A exists only if B exists, or the claim that A ontologically depends on B's having a certain feature as meaning simply that necessarily, A exists only if B has this feature. For this would imply (i) that everything ontologically depends on every necessary existent, (ii) that any two things which necessarily co-exist ontologically depend on each other, and (iii) that nothing can ontologically depend on some other thing's being F without also depending on this other thing's having any feature necessitated by its being F (or for that matter, by its bare existence). But it is intuitively clear that one may want to deny that Socrates ontologically depends on the number two, even though the latter exists in all worlds where Socrates exists. It is just as clear that one may want to deny that Socrates ontologically depends on its singleton, despite the fact that the latter exists in all worlds where Socrates exists. Likewise, one may want to hold that X ontologically depends on Y's being F (and hence, on Y's existence) without being committed to the claim that X ontologically depends on Y's being F-or-G, despite the fact that Y is F-or-G in all worlds where it is F, or to the claim that X ontologically depends on Y's being G-or-not-G, despite the fact that Y is G-or-not-G in all worlds where it exists.
What this shows, is that the notion of modal dependence (as currently understood in the possible worlds framework) is too weak to account for some of our intuitions concerning ontological dependencies. As has been observed by several authorsFoot note 7, what must be captured, is the intuition that when some entity A ontologically depends on the fact that some entity B has a certain feature, it should be the case that it is part of the nature or essence of A that necessarily A exists only if B has this feature. This is the idea that ontological dependence should be understood in terms of some notion of essential dependence (i.e., of an entailment relation somehow grounded in essence).
This obviously has some bearing on issues concerning Thing-realism/irrealism, since the Thing-realist claims that the existence of entities of a certain kind depends on our possessing some mental feature M, and this can be read as an ontological dependence claim; but it doesn't do anything to show that it is always inappropriate to frame Thing-realism in terms of modal dependence. However, Jenkins'view is that all realist/irrealist claims can and should be understood as involving essential dependence, which requires that the notion of essence be extended in such a way that it makes sense to talk of the «essence» of something's being the case (i.e., of a state-of-affairs'obtaining) and to say that something's being the case essentially depends on something else's being the case. The worry is not that there is no way in which this could be done, but that Jenkins'way of doing it results in a notion of essential dependence which is so strong as to make the corresponding realist claims almost trivially true.
It is certainly unclear what we should take to be the essence of entities of any kind. But it is even less clear (pace Jenkins 2005: 200) what the essence of something's being the case should be taken to be. For this is not to be viewed as the essence of any kind of entity: even if states-of-affairs are treated as entities, the question doesn't concern the essence of states-of-affairs, but the essence of the obtaining of a state-of-affairsFoot note 8. Jenkins (2005: 200) introduces the generalized notion of essential dependence by sayingFoot note 9 that its being the case that P essentially depends on its being the case that Q (e.g., on our possessing mental feature M) when it is part of what it is for P to be the case that Q be the case (e.g., that we possess M). From this we learn that the essence of its being the case that P is what it is for P to be the case, or what its being the case that P consists in, and that it may have «parts». Given that she later adds (2005: 206-207) that what it is for P to be the case should be identified neither with what makes it the case that P, nor with what it takes for P to be the case, nor with what it is in virtue of which it is the case that P, this doesn't leave much to rely on in trying to understand what she means to be saying here.
Even so, it could at least be noticed that this way of generalizing the relation of essential dependence between entities to turn it into a relation between something's being the case and something else's being the case, departs somewhat from the original idea in the following respect. To say that it is part of the essence of some entity A, that necessarily A exists only if B has a certain feature, is not to say or imply that B is a part of A or that B's having a certain feature is part of what it is for A to exist (or to be what it is). It is just a way of saying that the essence of A is such that it requires that necessarily A exists only if B has a certain feature. Accordingly, one would have expected the claim that its being the case that P essentially depends on its being the case that Q to have been understood as meaning that it is part of what it is for P to be the case that necessarily it is the case that P only if it is the case that Q (i.e., as meaning that the «essence» of its being the case that P requires that necessarily it is the case that P only if it is the case that Q). But on Jenkins' explanation, for its being the case that P to depend essentially on its being the case that Q, its being the case that Q must be part of what it is for P to be the case, and hence, of its «essence».
Hence, we are confronted with two different ways of understanding the notion of essential dependence. On the one suggested by Fine's work, its being the case that P essentially depends on its being the case that Q iff the «essence» of its being the case that P requires that its being the case that P necessitates its being the case that Q. On the one proposed by Jenkins, its being the case that P essentially depends on its being the case that Q iff what it is to be the case that Q is part of what it is to be the case that PFoot note 10.
On either interpretation, the realist who claims that its being the case that P doesn't essentially depend on our possessing mental feature M has the obvious advantage that he/she may eventually grant that its being the case that P may yet modally depend on our possessing M. But how reasonable is it to require that the notion of essential dependence involved in such a claim be construed in the way suggested by Jenkins?
Suppose the irrealist wants to advance the claim that it is «part» of what it is for P to be the case that we possess mental feature M. This is tantamount to the claim that its being the case that P is at least in part a mental fact (considering our possessing M as a mental fact), or that it reduces, in part, to a mental fact. Transposed in the semantic mode, this suggests that the judgement that P somehow makes reference to us and to feature M, i.e., that it is, in part, about our possessing M. But there seems to be no compelling reason to think that the irrealist, as such, must always be committed to such a strong claim (i.e., to think that it should be impossible to be irrealist about something completely non-mental in nature).
Here it may be useful to call attention to the fact that in most cases, the mental feature relevant to the realism issue will be a capacity to have some intentional attitude towards something's being the case. For example, it might be claimed that its being the case that P depends on our possessing the capacity to know that P, or on our possessing the capacity to judge (think, conceive) that P (where «our possessing this capacity» is short for «someone's possessing this capacity»). Reading this as involving Jenkins' notion of essential dependence yields the claim that its being the case that P consists, in part, in our possessing the capacity to know or judge that P. But it is hard to see how this could ever be the case. How could its being the case that P consists, even in part, in its being the case that we have the capacity to know or judge that it is the case that P? That would seem to require that the thought that P is in part a thought about our capacity to know/judge that PFoot note 11. But it is hard to see how anything can be such that being known or thought about, or being such that we can know or think about it, is literally part of what it is.
Suppose that part of what it is for it to be the case that P is that it be the case that we have the capacity to know/judge that P. It is at least arguable that this capacity involves a capacity to have a thought about its being the case that P. This is not (yet), however, to say that its being the case that P must in turn be part of what it is to have the capacity to know/judge that P (if only because one can have this capacity without ever exercising it), which suggests that our having the capacity to know/judge that P could be part of what it is for it to be the case that P, without being identical with it. Suppose that they are not identical. What, it can know be asked, is the other part of its being the case that P? And why should this other part be «mental» in nature? I don't see how such questions could be answered. Now suppose they are identical. Then, its being the case that P is our having the capacity to have a thought about it. But that doesn't specify any state-of-affairs (is there any way to specify what «it» refers to here?).
But whether they are identical or not, any thought about its being the case that P will at least in part be a thought about our capacity to know/judge that P, i.e., a thought about our capacity to have a thought about whatever this very thought is about. Perhaps there are such thoughts, and perhaps we can even have some of them. But it does seem unfair to saddle the irrealist with the view that whenever we think about a mind-dependent state of affairs we are in part thinking about our capacity to think about whatever we are thinking about, even if some irrealists may indeed have embraced some such views.
Moreover, it is not even clear that such «properties» as that of being thought about are «real» properties, or if they are, that they are «mental» properties. Does Vienna count has having a mental property, just in virtue of the fact that I am thinking about it? It would seem that the one who does the thinking is the one who has the mental property.
In any case, it seems doubtful that one who holds, e.g., that its being the case that 2 + 2 = 4 depends on our having the capacity to prove it, or that its being the case that John believes that P depends on our having the capacity to judge that John believes that P, has ever meant to be claiming that part of what it is for 2 + 2 to equal 4 is for us to be able to prove that 2 + 2 = 4, or that part of what it is for John to believe that P is for us to be able to think that John believes that P (at least in any sense implying that the thoughts that 2 + 2 = 4 or that John believes that P are thoughts about our capacity to know or judge them). It has been objectedFoot note 12 that on some views about mathematics, for it to be the case that 2 + 2 = 4 is for us to be able to prove it. This surely is a common way of expressing certain constructivist conceptions of mathematics. But as far as I can see, the views which tend to be expressed in this way are not really of the kind I am here discussing. The claim that its being the case that 2 + 2 = 4 is our being able to prove it is not to be understood literally, as saying that its being the case that 2+ 2 = 4 consists in our having the capacity to prove that 2 + 2 = 4, but as a lousy way of saying that it consists in our having the capacity to prove the formula «2 + 2 = 4» (or something like that).
At one point, Jenkins (2005: 204) mentions Berkeley's view that «truths about the physical world consist in truths about ideas in the mind of God» in support for the claim that irrealists generally consider themselves as being committed to essential dependence. I agree that the view just alluded to, according to which physical facts actually are identical to (some) mental facts, does involve essential dependenceFoot note 13. But on the face of it, it is just not the kind of position which would currently most often be associated with irrealism. For one thing, the irrealist as such is not necessarily committed to any form of reductionism (there must be room for non-reductive forms of irrealism); for another, the relevant mental facts will nowadays (most often) be taken to be facts about us. But more importantly, most standard formulations of realism/irrealism involve dependence with respect to intentional attitudes, and it strikes me as just implausible to suggest that part of what it is for something to be the case (for a judgement to be true, or a state of affairs to obtain) is for it to be (or possibly be) the content of any such attitude. Hence, it could not be objected that I am saddling Jenkins with an implausibly strong construal of the notion of essential dependence. The trouble doesn't have anything to do with the strongness of the relation in itself. It is reasonably clear, at an intuitive level, what is meant by saying that part of what it is for something to be the case is for something else to be the case, and I don't want to suggest that such claims are so strong as to never be true. The trouble comes when it is suggested that this kind of relation might hold between something and (a capacity for) an intentional attitude that one may have about or towards it.
None of this, however, is to deny that in many contexts, the irrealist may and will have in mind a relation stronger than modal dependence. One candidate is the notion of essential dependence, as understood in the way suggested by Fine, another is what might be called a relation of «constitutive» dependence. To take an analogy, there seems to be a sense in which, e.g., its being the case that Sherlock Holmes smokes the pipe strongly depends on someone's having said so. But no one would want to suggest, I take it, that (even part of) what it is for it to be the case that S.H. smokes the pipe (for S.H. to smoke the pipe) is for it to be the case that someone said so. In such a case, it is tempting to say that its being the case that S.H. smokes the pipe is somehow constituted by the fact that someone said that he smoked the pipe (but note that saying that S.H. smokes the pipe doesn't involve saying that one is saying so). The trouble with such cases, of course, is that they deal with the realm of fiction, where statements are often claimed not to be «strictly» true, while we want to consider cases where it could «strictly» be the case that P, even though its being the case that P «constitutively» depends on our capacity to know/judge that PFoot note 14. They may nonetheless give some idea of what is the intuition behind the notion of constitutive dependence.
A question might be raised, as to whether constitutive dependence really must be understood as being strictly stronger than modal dependence. The same question is raised by Jenkins (2005: 203), with respect to essential dependenceFoot note 15. According to her, there is no question that essential dependence entails modal dependenceFoot note 16, when modal claims are understood in the usual sense. But she contends that they are sometimes understood in such a way that modal independence fails to entail essential independence. She illustrates this by means of the following example. An irrealist about mathematics might be happy to admit that «2 + 2 = 4» is true even in worlds where nobody exists, while maintaining both (what Jenkins (2005: 203) takes as an essential dependence claim, namely) that «what it is for 2 + 2 = 4 to be the case is for us to construct mathematical structures of a certain kind» and that it is necessary that 2 + 2 = 4. The funny thing about this example, is that it doesn't involve anything like a non-standard or unusual understanding of modality: both the statement that necessarily, 2 + 2 = 4, and the statement that necessarily, 2 + 2 = 4 only if we have constructed a certain structure could be understood in the usual way, as meaning, respectively, that 2 + 2 = 4 in all possible worlds (including those where wo don't exist), and that we have constructed this structure in all worlds where 2 + 2 = 4. It is just that since the first statement is admitted as true, the second has to be rejected as false. But if there has been no shift in the notion of modal truth, then there must have been a shift in the notion of essential dependence itself.
What the example actually relies on, then, must be a new reading of the notion of essential dependence itself, one which departs from Jenkins'own «official» interpretation and conflates it with a form of what I call «constitutive dependence». On the face of it, it will easily be granted that one who claims that what it is for 2 + 2 = 4 to be the case is for us to construct mathematical structures of a certain kind is not likely to be implying that the fact that we have constructed such and such structures is part of the fact that 2 + 2 = 4 (or to put it otherwise, it is unlikely that a constructivist about mathematics will express his/her view by using these words, except perhaps as a matter of stylistic convenience). For it is clear (since Jenkins'essential dependence does entail standard modal dependence) that if he/she were, then he/she would be committed to the corresponding modal dependence claim.
This is not to deny that there may be some non-standard reading of the claim that necessarily, 2 + 2 = 4 only if we have constructed a certain structure (i.e., a non-standard interpretation of modal dependence), on which it will follow from (or at least be compatible with) the claim that its being the case that 2 + 2 = 4 «constitutively» depends on our having constructed the relevant structure, and/or fail to follow from the corresponding essential dependence claim. Thus, I'm willing to acknowledge that there may be some construal of the modalities on which the resulting (non-standard) modal dependence claims will not entail Jenkins'essential dependence claims. But this is irrelevant to the point at issue, for it does nothing to show that the irrealist who understands modal claims in such a way is nonetheless concerned with Jenkins's essential dependence (quite the contrary).
If this is right, then what Jenkins'example shows is that constitutive dependence can be construed in such a way that it doesn't entail modal dependence. I see no reason, however, to conclude that it must be so construed, or more importantly, to conclude that no specific constitutive dependence claim can entail the corresponding modal dependence claim. Either way, I think it is now clearer why to insist that the realism/irrealism debates must be put in terms of Jenkins' notion of essential dependence threatens to trivialize any realist position. Admittedly, insisting that they be put in terms of Fine's notion of essential dependence (or for that matter, in terms of constitutive dependence) would not have this consequence, but I fail to see any point in doing so.Foot note 17
Département de philosophie
Université de Montréal
C.P. 6128, succ. Centre-Ville
Montréal, Qc H3C 3J7
daniel [dot] laurier [at] umontreal [dot] ca
[Foot Note 1]
Saying that «P depends on Q» should be seen as a convenient way of saying either that the obtaining of the state-of-affairs that P depends on the obtaining of the state-of-affairs that Q, or that the truth of the judgement that P depends on the truth of the judgement that Q, as the case may be.
[Foot Note 2]
This is a currently standard definition of counterfactual dependence, which derives from Lewis (1973: 167) and is to be found in several places such as Sanford (1989: 182) and Collins (2004: 109).
[Foot Note 3]
The claim that P necessitates Q is equivalent to the claim that in all possible worlds where Q is not the case, it is also not the case that P, while the claim that P counterfactually depends on Q, on the assumption that P and Q are both actually the case, comes down to the claim that in all the closest worlds where it is not the case that Q, it is also not the case that P.
[Foot Note 4]
Presumably, Jenkins would be prepared to extend this to Thing-realism; but since she doesn't discuss any form of Thing-realism, I'm going to ignore this in what follows. Though I think that the same points basically apply in this case as well.
[Foot Note 5]
Where this notion could be useful is in the context of a defence of some realist thesis. For it would obviously be sufficient, in order to establish that some state-of-affairs doesn't depend (in some suitably strong sense) on our possessing M, to show it doesn't counterfactually depend on our possessing M.
[Foot Note 6]
Related work by Lowe (1994) and Correia (ms) is also relevant here, though Jenkins doesn't mention them. See also Mulligan, Simons and Smith (1984).
[Foot Note 7]
See e.g., Fine (1994, 1995), Lowe (1994) and Correia (ms).
[Foot Note 8]
This suggests that, if the discussion was to be put in the semantic mode, the question would have to be seen as a question concerning the essence of truth.
[Foot Note 9]
Jenkins'own gloss is in terms of «independence», but nothing hangs on this.
[Foot Note 10]
It has been suggested (by Mark McCullagh) that these two «readings» may not be substantially different. If this is the case, then the objections I am going to raise against framing the realist/irrealist debates in terms of Jenkins'essential dependence would carry over to framing them in terms of Fine's essential dependence.
[Foot Note 11]
As we have all learned at mother's knee, from the fact that its being the case that P consists (in part) in its being the case that Q, it doesn't follow that one cannot judge that P without (in part) judging that Q. But it does follow that any judgement that P is a judgement (in part) about Q. To take one objector's example, if Lois Lane thinks that Superman can fly, and its being the case that Superman can fly consists in its being the case that Clark Kent can fly, then her thought that Superman can fly is a thought about Clark Kent's ability to fly (without being a thought that Clark Kent can fly).
[Foot Note 12]
By Gregory Lavers.
[Foot Note 13]
Jenkins (2005: 204) also mentions the kantian view that «phenomenal reality is (in part) a product of our sensibility» in support for his contention. But here I shall deny that the view in question is best understood in terms of essential dependence; there is no clear sense in which «our sensibility» (or for that matter, «our mind») has to be seen as a part of its products. In my opinion, this kind of view is best construed as involving some sort of «constitutive» dependence (see below).
[Foot Note 14]
Unfortunately, I have no space here to discuss the question whether those who think that there is some metaphysically neutral notion of «minimal truth» would or should count true fictional statements as true simpliciter, in this minimal sense. The obvious analogy, here, is with some versions of constructivism in mathemathics, where the question may arise, as to whether true arithmetical statements should be seen as on a par with true fictional statements.
[Foot Note 15]
In this and the next paragraph, I'm indebted to Mark McCullagh.
[Foot Note 16]
This holds on either of the two ways of construing this notion that I have mentioned.
[Foot Note 17]
This paper has been read at the annual meeting of the Canadian Philosophical Association (York, may 2006). Special thanks to Mark McCullagh for his thoughtful comments.