You Really Do Imagine It: Against Error Theories of Imagination
2014; Wiley; Volume: 50; Issue: 1 Linguagem: Inglês
10.1111/nous.12060
ISSN1468-0068
Autores Tópico(s)Epistemology, Ethics, and Metaphysics
ResumoBefore Kripke and Putnam, most philosophers accepted that a posteriori identities were contingent; it was, after all, very easy to imagine the falsity of any a posteriori identity. It seems fair to say that Kripke (1971, 1980) revolutionized how we understand the relationship between epistemology and modality. He argued persuasively for the necessity of identity and hence that even identities discovered a posteriori, like water = H2O, turn out to be necessary, not contingent. This is now the received wisdom. Let us assume that philosophers were convinced to accept a correct conclusion: there are a posteriori necessities of the kind that Kripke and Putnam cite. When we unpack Yablo's answer to this question, and examine how he thinks we ascertain that a situation is impossible, we will see that while Yablo concedes the answer to the first question may be "yes," he answers "no" to second. In this paper I argue that Kripke and Yablo are both mistaken about imagination. The answer to both questions is "yes:" it is easy to imagine impossible situations, including situations deemed impossible only a posteriori, even after we acknowledge that they are impossible. I first establish that Kripke in fact answers "no" to the first question (§1). I will then explain why I think that is a mistake, why it is plausible that we can imagine the impossible (§2). That discussion will also lay the groundwork for examining Yablo's elaborations of Kripke's views; I will argue that Yablo's view of imagination also doesn't stand up to scrutiny (§3).1 At this point I will have shown that Kripke and Yablo hold implausible views about imagination. One might think we nevertheless must accept this implausible view because it offers the best hope for retaining imagining as a guide to metaphysical possibility in the face of a posteriori necessities. If we routinely imagine the impossible, how can we continue to take imagining to provide evidence for metaphysical possibility? I will sketch a positive proposal that shows how (§4). Though we can imagine the impossible, my proposal supplies independent grounds (i.e., grounds other than the very modal conclusions at issue) for denying that imagined impossibilities are evidence for possibility. I close by showing how my analysis allows a tidy resolution to Kripke's influential discussion of the zombie argument for dualism at the end of Naming and Necessity (§5). The view I attribute to Kripke is an error theory about imagining. The picture I take Kripke to have in mind goes something like this. Perception tells us about actuality, and we perceive only what is true (understand 'perceive', 'see', 'hear', and so on as success verbs). Similarly imagination tells us about metaphysical possibility, and we imagine only what is possible.2 Think of imagination as a telescope through which we survey genuine possibilities—call this the telescopic view of imagining.3 On the telescopic view of course you cannot imagine the impossible. Kripke realizes that, given his conclusions about the necessity of some a posteriori truths, it's going to seem like we can imagine impossible situations. The telescope initially seems to reveal impossible situations. He accepts that he needs to explain away these apparent imagined impossibilities. Kripke's familiar strategy reconstrues these cases: we do not really imagine something impossible, rather we imagine a possible situation that is similar enough to the impossibility to be mistaken for it. This is an error theory about imagination; Kripke's version of it consists of at least three claims: Kripke's Error Theory explains why he thinks we needn't answer (Q1) affirmatively. We mistake what we see in the telescope: the situation we imagine is a genuine possibility, we are just a little confused about what situation it is. We make mistakes about what we imagine. In this section I want to establish that Kripke endorses the Error Theory. There are several reasons for doing so. First, Kripke's writings have been quite influential, so it will be instructive to see whether, and how clearly, he articulates the Error Theory. Second, I have found that philosophers divide on which question, (Q1) or (Q2), deserves priority. Some think a negative answer to (Q2) offers the plausible position: what we take to be impossible affects what we imagine, and (Q1) holds interest only insofar as it bears on (Q2). Others' intuitions pull in the reverse direction: if a situation is impossible, there is nothing to imagine, regardless of what we might confusedly think. Even if we mistakenly think a situation is possible, then we still cannot imagine it. Whichever way you lean, it is important to recognize that "no" answers to (Q1) and (Q2) represent independent views. Third, objections to the Error Theory outlined in section two serve as a framework for objecting to Yablo's way of refining Kripke's initial position. We will discuss Yablo's view in section three. But whatever we imagine counterfactually having happened to [the table] other than what actually did, one thing we cannot imagine happening to this thing is that it, given that it is composed of molecules, should still have existed and not have been composed of molecules. We can imagine having discovered that it wasn't composed of molecules. But once we know that this is a thing composed of molecules—that this is the very nature of the substance of which it is made—we can't then, at least if the way I see it is correct, imagine that this thing might have failed to have been composed of molecules. (Kripke 1980, pp. 126–27, emphasis mine)5 Here Kripke clearly endorses Unimagined. Eddie does not (because he cannot) imagine this very table composed of an ethereal entelechy because its "very nature" is to be composed of molecules. Perhaps before he realized the table was made of molecules he might have thought he could imagine it composed of an ethereal entelechy. But especially now that he realizes that the table is made of molecules, he realizes he wasn't imagining, because he can't imagine, that it is not. (The reader might wonder: does Kripke have to answer "no" to (Q1)? Perhaps in the past we could imagine the impossible, but only because we didn't realize it was impossible; that is why Kripke includes provisos like "given that it is composed of molecules." Once we realize a situation is impossible we can no longer imagine it; though the answer to (Q1) is "yes," the answer to (Q2) is "no." If that is the way you read Kripke, then you take Kripke to share Yablo's view, which we will come to in section three. In this section I focus on the "no" answer to (Q1).) [The materialist] has to hold that we are under some illusion in thinking that we can imagine that there could have been pains without brain states… He has to show that these things we think we can see to be possible are in fact not possible. He has to show that these things we can [seem to] imagine are not in fact things we can imagine. (1980, p. 163, emphasis mine). This is another clear endorsement of the Error Theory. Although Kripke endorses the Error Theory, in several passages imagination and error theory play a less prominent role.6 The reason is that Kripke works with two interwoven threads. Kripke's primary concern is to argue for the existence of a posteriori necessities. Should we say that it is possible for gold to have ten fewer protons, that possibly Hesperus ≠ Phosphorus, or that this table could have been made of ice? No, Kripke argues. The second thread is our focus: what we can and cannot imagine. Kripke shifts between these two threads, not overly concerned to distinguish them, because his primary concern is the first thread, with what is possible. When we attend to the second thread, however, it is plain that Kripke feels it's his burden to say something about our apparent ability imagine gold with ten fewer protons, Hesperus not identical to Phosphorus, and this table made of ice. He responds with the Error Theory. Confusion about Kripke's notion of epistemic possibility sometimes obscures the fact that he endorses the Error Theory. A proposition is epistemically possible if we are unable to rule it out relative to some epistemic standard: the proposition is true for all we are justified in believing, or for all we know, or for all we know for certain. An alternative interpretation claims that Kripke doesn't endorse the Error Theory: we aren't mistaken about what we imagine because we really do imagine epistemic possibilities.7 That is why Kripke concedes, "We can imagine having discovered that [this wooden table] wasn't composed of molecules," because we really imagine the epistemic possibility.8 No error theory applies. Our problem is we mistake genuine (metaphysical) possibility for mere epistemic possibility, but it is quite understandable if ordinary folk (not to mention pre-Kripkean philosophers) fail to appreciate that kind of technical philosophical distinction. I doubt whether this alternative interpretation is what Kripke intends, and, regardless, it fails to avoid the Error Theory. The point of epistemic possibility is to make Confusion (the second claim of the Error Theory) plausible, not to avoid the Error Theory. Kripke explains epistemic possibility as imagining "…appropriate qualitatively identical evidential situations [in which] an appropriate corresponding qualitative statement might have been [true]" (1980, p. 142).9 Kripke's externalism about meaning entails that in a different situation the corresponding qualitative statement 'Water is not composed of H2O' would mean something different than the English sentence 'Water is not composed of H2O'. Hence when we imagine an epistemic situation qualitatively identical to our own, we imagine a situation in which an utterance of 'Water is not composed of H2O' is true. The word 'water' means something other than what it actually does. In that situation there is no water around though there is stuff that is called 'water' by us. When Kripke writes that we can imagine "having discovered" or "finding out" that water is not H2O, he is speaking loosely; we cannot really imagine it turning out that water is not H2O.10 This is still the Error Theory. Let me describe the error in a way that hints at why I think the view is implausible. You believe that water is H2O, and probably also that water is necessarily H2O. However, you acknowledge that you—the actual you, not the imagined you—might be wrong. For all your evidence, it seems that you can imagine being mistaken, imagine really discovering that water is not H2O, imagine it really turning out that water is not H2O. Kripke disagrees. Insofar as we take ourselves to be imagining discovering that water is really XYZ, not H2O, we are mistaken. That is not what we imagine. We imagine instead a situation in which 'water' refers to something other than water. Thus whether the situation imagined is an epistemic possibility, whether imagined "as actual" or imagined "as counterfactual," Kripke insists that we do not imagine the impossible situation we take ourselves to imagine.11 I conclude there is ample evidence that Kripke endorses Unimagined and Confusion, thus subscribing to the Error Theory about imagining. Is the Error Theory correct? Set aside modal truths and modal epistemology for a moment. On their own, how plausible are Unimagined and Confusion? How plausible is it that we cannot imagine certain impossible situations, and that we make mistakes about what we imagine when we try to? I contend: not very. Leaving aside modal considerations, Kripkean claims about imagination should strike us as implausible.12 On our commonsense understanding of imagination, the answer to (Q1) is yes. I will present and explain a range of cases that demonstrate this. Of course for my cases to be compelling, they will have to involve situations that are generally acknowledged to be impossible. Hence these cases, if compelling, will also lay the groundwork for our discussion of (Q2). The Error Theory and a view that answers "yes" to (Q2)—what I will call a doxastic view—differ in how they diagnose our past episodes of imagining. Error Theorists insist that, when we think back on previous imaginings, we never imagined impossible situations, even back when we were ignorant of their impossibility, and even if we previously (and mistakenly) thought the situations were possible. Doxastic theorists disagree; they offer a model for past modal error, which we examine in section three. Let's begin with an example that shows we are not usually wrong about what we imagine. You are picking up Michael, who you've never met, at the airport and you imagine him having brown hair. When he arrives and turns out to be redheaded, you do not conclude: "Oh, I didn't really imagine him at all. It seemed to me that I was imagining him, but this guy has red hair, so I must have been imagining someone else." That is a misguided reaction. You were indeed imagining Michael, just not as he really is. There is no mistake about that. The Error Theory entails that in scenarios very much like the first airport case, you are making a mistake.13 Suppose that Michael turns out to be a woman (there are some women named "Michael") and further, suppose that biological sex is necessary. The Error Theorist will contend that now you should retract: "Oh, I didn't really imagine that person at all. It seemed to me that I was imagining Michael was a man, but Michael is a woman, so I must not have been imagining Michael after all." That too seems like a misguided reaction. You were indeed imagining Michael, just not as she really is. (You might laugh and tell her, "I imagined that you were a man!") Why think there is mistake about that? And notice that according to the Error Theorist whether or not you are making a mistake hinges on the truth of our supposition about the necessity of biological sex. That too seems implausible. The only reason to hold such a view is a prior commitment to imagining as something like a telescope through which we survey genuine possibilities. There is a better alternative. My favored model pictures imagination as constructing scenarios out of existing materials rather than as a telescope through which we survey genuine possibilities. When we think about imagination as constructive rather than telescopic, it becomes clearer how we can imaginatively construct impossible scenarios: we combine ingredients that together amount to an impossible situation. But the fact that the resulting scenario is impossible does not impede the construction. Further, given that we understand our construction materials, the suggestion that we are confused about what we imagine should sound peculiar. For example, I visually imagine my great-great-grandmother even though I have no idea what she looks like; maybe I form an image of a woman who resembles Madame Chiang Kai-Shek. That does not make it Madame Chiang Kai-Shek that I'm imagining. I'm imagining that the woman who looks this way is my great-great-grandmother. The question of how I know this—how I know that it is my great-great-grandmother I am imagining, rather than some other woman—doesn't really make sense. Wittgenstein makes this point with his King's College example (1965, p. 39). When one imagines King's College on fire, there is just no doubting that one is imagining King's College, and not something else, e.g., a similar-looking part of UCLA, or a miniature replica of the College. As a general rule, when we imagine something there is just no doubting that we have imagined that something. The construction model respects commonsense about imagination's flexibility and power, including our power to imagine the impossible. Lewis Carroll's stories should be ample enough evidence, but here is a more recent example. Imagine that in the present a teenager's father is "initially" an ineffectual loser with a crappy job who is pushed around by his boss. Imagine the teenager travels back to 1955 in a DeLorean and, through a series of mistaken-identity-fueled madcap adventures, changes his father into confident leader. As we visualize a scene between the teenage son and his adolescent dad we imagine that the scene is taking place in 1955 for "the second time"—the "first time" through 1955 the son wasn't even born yet! To makes sense of the story we have to imagine something impossible, that thanks to the teenager's intervention, 1955 happens differently the "second time" through. Ordinary constructive imagining is not concerned with metaphysical possibility. When we reflect on Back to the Future time travel we might realize that the situation is impossible: 1955 could not happen "twice." But unless we are fixated on imagining (only) metaphysical possibilities, that will not prevent us from constructing the time travel scenario in imagination. We might realize that the situation we are imagining is impossible, but we still happily follow along with the story and imagine it. Reading philosophy of time travel might spoil the story because we realize that, sadly, in reality there is nothing we can do to change the past, but it does not spoil the story by rendering it unimaginable. I conclude that when we reflect on what and how we imagine "in everyday life," as when we imagine Michael turning out to be a woman, or past-changing time travel, it is plausible that we can imagine the impossible. Let me now offer several more examples to solidify the reader's intuitions. I imagine myself receiving the Fields medal for proving Goldbach's conjecture. Imagine renowned mathematicians marvel; given my limited background, they reckon my discovery to be the most startling since Ramanujan's. It is clear that I imagine (and I suggest that you have imagined too) that I really have proved it. I imagine that my Fields medal-winning journal article contains the proof. I am not imagining myself as some kind of charlatan; my imagining would have quite a different character if I were. I can also engage in a similar imaginative project: I can imagine disproving Goldbach's conjecture. Now maybe if this were to actually happen it would be a more stunning feat, because most mathematicians believe the conjecture to be true. That is irrelevant.14 My imaginings do not contain any mathematical detail. I do not imagine any steps in my prize-winning proofs; I cannot emerge from my reverie and snap my fingers, say "That's it!" and start writing. I'm simply imagining some heretofore undiscovered, yet, as far as my imagining goes, unspecified, mathematical details that I have miraculously managed to uncover. In my view, we cannot imagine worlds in which there are naturally purple cows, time machines, transparent iron, a moon made of green cheese, or pure phenomenal colors in addition to those we know. Anyone who attempts to do so will either fail to imagine a world or else will imagine a world that only seems to have the property of being a world in which the thing in question exists. Can we imagine a world in which there is transparent iron? Not unless our imaginings take place at a level of structural detail comparable to that of the imaginings of condensed-matter physicists who are trying to explain, say, the phenomenon of superconductivity. (p. 79) Van Inwagen appears to suggest that imagining requires a level of specific detail that would, of course, have to be absent in the Goldbach's conjecture cases. Let me begin with an exegetical note. Van Inwagen sketches these requirements on imagination in the context of discussing whether we have any reason to believe that such things (time machines, a moon made of green cheese, transparent iron) are possible. He notes that possible situations are embedded in entire possible worlds; for the situation to be possible, there must be possible laws of nature and the possible initial conditions that permit it. But philosophers advancing modal claims rarely ever attempt to fill in details about the possible laws of nature or possible initial conditions, and van Inwagen thinks this is to their discredit. Hence to charitably interpret van Inwagen, we should read him as suggesting that imagination-insofar-as-it-provides-evidence-for-possibility demands far more detail than we can generally supply.17 I am specifically interested in how plausible Unimaginable is in its own right, setting modal considerations aside. A view like van Inwagen's, outside of this modal context, holds little plausibility. The suggestion that to imagine any situation whatever we need to imagine key microstructural details would be far too restrictive; almost nothing would count as imaginable by that criterion. We could not imagine Hillary Clinton winning the 2008 Democratic primary unless our imagining descended to the level of voter neurochemistry, nor could we imagine a table two feet to the left of its present location unless we knew enough physics to imagine a) the table's microstructural features, b) the initial conditions and c) the laws of nature that would permit the table to be in the new position. It seems clear that we can imagine situations without having to imagining the rest of the possible world; it's that very feature of imagining that van Inwagen thinks has gotten philosophers into trouble. We can imagine a wizard casting a magical spell without bothering to imagine how magic works. I conclude that we should take the two Goldbach conjecture cases at face value, and admit that both are imaginable in the ordinary sense. My problem with the more technical notion of imagination-insofar-as-it-presents-evidence-for-possibility is that it's obscure. This more technical imagination could either be a second kind of imagination, which is hard to believe, or it could be a restriction on ordinary imagination. But it is hard to see how to develop the restriction without allowing prior modal convictions to drive our opinions of what we imagine. ("Well, that's impossible, so that can't be what I imagine in this more technical sense.") There is nothing wrong with inventing this more technical sense, but it would be useless as evidence for possibility. We would be reading our prior modal convictions into imagination, rather than using imagination to help us discover modal conclusions. I conclude that, setting modal considerations aside, Unimagined and Confusion are implausible. The construction model offers a better way to understand everyday imagining than the telescopic model. And in everyday imagining there is nothing puzzling about constructing situations that turn out to be impossible. (In section four I will give a theoretical account of the construction model that will explain how we construct impossible scenarios in imagination.) Imaginings are contentful states of mind and we are generally not confused about the contents of our own mind. Insofar as the Kripkean strategy rests on Unimagined and Confusion, it too is implausible. We should see whether a better explanation is available. A natural thought is that Kripke's remarks can serve as the foundation for a better explanation. Kripke's remarks were informal, and neither intended to lay out a detailed theory of imagination nor a comprehensive modal epistemology. Yablo (1990, 1993, 2006) admits that blunt assertions of Unimagined and Confusion are implausible, but that is because they are overly broad. Kripke's point should be that given enough information we cannot imagine certain impossible situations. Our ability to imagine what turns out to be an impossible situation is explained by our ignorance of that information. To give a name to contrast with the telescopic and construction models, we might call this the doxastic model: once you believe the relevant facts, what was once imaginable becomes unimaginable. In this section I will outline Yablo's analysis of several cases and, building on the examples from section two, argue that he too must make implausible claims about imagination. that I find p imaginable is explained by [my unawareness|my denial] that (a), and/or [my unawareness|my denial] that (b).19 In some cases your imagining p is explained by your unawareness or denial of a defeater. Ignorance facilitates imagining impossible situations. Yablo proposes that when we do not know the relevant q, or the relevant modal conditional, we can imagine that p, but such imaginings are fragile in the sense that once our ignorance of (a) and of (b) disappears, we can no longer imagine the impossibility. Let's see how Yablo applies his model to the cases we have already discussed. The ancients might very well have imagined Hesperus distinct from Phosphorus. But they could do so only because they were ignorant of the fact that Hesperus is Phosphorus.20 "But we could imagine veridically believing them to be distinct, back when we thought they were distinct." True but irrelevant; it remains that Hesperus ≠ Phosphorus is now epistemically possible, but not now [imaginable]. (1993, p. 23n48) Once we learn that Hesperus = Phosphorus, that renders Hesperus ≠ Phosphorus unimaginable—provided we know the relevant modal conditionals. Pre-Kripke most (all?) philosophers and ordinary folk could imagine Hesperus distinct from Phosphorus, though they knew them to be identical. In this case, they were ignorant of the modal conditional that Kripke proved: if H = P, then □(H = P). Once apprised of this modal conditional, Hesperus ≠ Phosphorus becomes unimaginable (provided that learning the modal conditional doesn't cause us to reject the non-modal fact). Post-Kripke nothing changes for ordinary folk. But philosophers aware of Kripke's writings may still imagine Hesperus distinct from Phosphorus if they deny the above modal conditional. If they know both the modal conditional and that Hesperus = Phosphorus, then Error Theory delivers the right result: we cannot imagine Hesperus ≠ Phosphorus. We imagine some similar situation that we mistake for one in which Hesperus ≠ Phosphorus. Similarly, Oedipus can imagine the impossible situation of himself being king without Jocasta ever existing only because i) he falsely believes that Jocasta is not his mother; and/or ii) he is unaware of cogent Kripkean arguments for the necessity of origins.21 Yablo's position on Goldbach's Conjecture (GC) is that neither it nor its negation is imaginable. The best we can do is imagine a situation in which mathematicians erroneously hail the discovery of a counterexample to Goldbach's Conjecture. Symmetry between GC and not-GC considerations prevents us from imagining that the number is in fact a counterexample. Yablo concludes that not-GC is not imaginable; but neither is it unimaginable; Yablo calls not-GC undecidable.22 I examine undecidability in detail in §3.4. Some scientific identities, like water = H2O, are to be treated like "ordinary" identities. We can imagine them only in ignorance either of the fact that water = H2O or of the fact that if water = H2O, then □(water = H2O). When we seem to imagine discovering that water is XYZ, we are imagining a similar situation and mistaking it for one in which water ≠ H2O. This is a good place to point out another aspect of Yablo's account, his gloss on epistemic possibility (1993, §ix). When impossible e seems imaginable, that is because you could have thought something true with the "same e thought"—roughly, the same internal mental act. Had you been in a different situation, say one in which XYZ had fallen from the sky and filled the rivers and lakes, then you could have thought something true with your water-is-not-H2O-thought. But the "same internal mental act" in the counterfactual situation would not have expressed the proposition that water is not H2O, so you are not literally imagining that water is not H2O. Recently Yablo has developed a more complex treatment of some scientific identities, like heat is mean kinetic molecular energy. The details do not matter for our purposes, so I set them aside.23 Next I will argue that Yablo's analysis of each one of these cases is incorrect. His doxastic model conflates what is believable with what is imaginable. You might not find something believable or possible, yet you can imagine it nonetheless. Does Yablo's model for modal error fit the cases I described in section two? In particular, does c) of Yablo's model accurately describe what happens when we try to imagine the cases from section two? At first gloss, it appears not. I explain the prima facie problem in this subsection; I will then look into the details the next two subsections, which will increase our confidence that Yablo's model cannot be salvaged. If you found the arguments in section two persuasive, then despite the fact that you know the relevant non-modal q and the relevant modal conditional if q then □not-p, you were still able to imagine Twain punching Clemens, water = XYZ, past-changing time travel, and so on. Yablo's c) is incorrect: our ability to imagine these impossible situations is not explained by our ignorance of a non-modal fact and/or a modal conditional. In some ways this should not be at all surprising. In each case we imagine that a certain non-modal fact, q, fails to obtain. But, after all, even though we know the relevant non-modal qs, the whole point of imagining is to imagine that certain actual world facts do not obtain. So far there is no reason to accept Yablo's suggestion that adding knowledge of a non-modal fact—even coupled with knowledge of the relevant modal conditional—will change what we find imaginable. This aspect of the doxastic model fails to accurately describe imagination. How might Yablo reply? Yablo's text suggests two different strategies. The first we might call the "modal conditionals" strategy; I will address that strategy in the next subsection. The second strategy is a version of the "modal conclusions first" reasoning we briefly discussed in connection with van Inwagen; I will cov
Referência(s)