Quine's Ideology, Metaphysical Dragons, Fidelity, Confabulations and other Stuff
Interview by Richard Marshall
'I wish Quine hadn’t chosen the word ‘ideology’. The word comes with way too much baggage. Once, during a presentation on the ideological differences between the A-theory of time and the B-theory of time, I had to explain that I wasn't giving a Marxist analysis of how conceptions of time are used by the ruling class to create a false consciousness for the purposes of dominating the working class. That sounds like an interesting idea -- but it’s definitely not what I’m talking about!'
'In metaphysics, there are various disputes that come down to a choice between rival theories. These rival theories have different ideologies that, presumably, are different with respect to how well they correspond to the world’s structure. So, we should base our choice at least in part on which theory we think is most likely to have the most accurate ideology. How do we determine which that is? I favor a virtue-driven methodology. Once a dispute reaches a mature state of stability -- in the sense that the main theoretical options are coherent, their consequences have been identified, and so on -- we can compare the rival theories with respect to various theoretical virtues, features of the theory that make it more likely to be correct.'
'There’s a conceptual difference between acquiring high-fidelity beliefs and avoiding low-fidelity beliefs. In theory, someone could care only about one or the other. But not so are victories over the structure of the world gained. The easiest way to avoid a gruesome ideology is to not employ any ideology whatsoever. That sort of ideological agnosticism seems to me to be psychologically impossible. It also seems to me to be epistemically criticizable. If fidelity has final epistemic value, then it is valuable to have high-fidelity beliefs. Someone who only tries to avoid low-fidelity beliefs will miss out on acquiring the high-fidelity ones.'
'Confabulation was first studied in clinical contexts with patients suffering from brain damage or psychological disorders. If someone takes those to be paradigm cases of confabulation, then it’s no surprise if they also think that confabulation is dysfunctional. But I think there are cases of confabulation occuring in completely ordinary contexts that undermine that presumption.'
'If we want a better understanding of sexual orientation, then we need to determine what behavior counts as an authentic expression of orientation. In practice, this means determining what its stimulus conditions are as well as the ways in which it can be masked or mimicked. Plausibly, the correct stimulus conditions require some degree of external freedom: there’s a reasonable diversity of potential partners, there’s no coercive force at play, stuff like that. It’s also plausible to think that the correct stimulus conditions require certain internal features, like the person being willing and able to have sex.'
Peter Finocchiaro works mainly in (analytic) metaphysics, with a focus on metametaphysics and methodology. He also works in areas of social philosophy, particularly in relation to social ontology, rationality, and bias. Here he discusses ideology in Quine's sense, ideology in metaphysics, ideological externalism and internalism, maximal, extreme, non-ontic ideological externalism, the epistemology of metaphysics (where the dragons live!), the relationship between fidelity and epistemic value, the Razor and the Comb, ideological parsimony and whether its an epistemic or a fidelic virtue, confabulation, how to decide whether ancient Greeks were bi-sexual, sexual orientation and masking, and, finally, just how bad is death?
3:16: What made you become a philosopher?
Peter Finocchiaro: To be honest, it kind of just happened. When I was younger, I had the personality of a philosopher. (Ask my mom about the bags-of-leaves story.) But I had never thought about being a philosopher in a professional capacity -- in fact, I didn't even know that was a thing that ordinary people could be. When I started undergrad at Syracuse University, I intended to major in Biology with vague plans to “be a scientist”. On a whim, I took Ethics and Value Theory with Laurence Thomas. That class was massive! There were so many students in attendance that they spilled onto the aisles of Grant Auditorium, which is supposed to have a seating capacity of over 350. I enjoyed the course enough that I decided to take two more Philosophy courses in the Spring. At the same time, my interest in “being a scientist” began to dissipate as I realized that “being a scientist” required doing science -- which at the time I very much did not enjoy doing. At any rate, I was enjoying my Philosophy classes, and so I asked one of my professors, Mark Heller, about majoring in Philosophy.
In the Fall semester of my sophomore year, I took Philosophy of Religion with Kris McDaniel. If anyone is responsible for me becoming a philosopher, it's Kris. I have this very distinct memory of him in class casually suggesting that dolphins might be persons. It was the first time that a philosophical idea slapped me across the face. I was hooked, and took as many courses with Kris as I could. I don't want to get too sappy, and I don't want to embarrass Kris too much, but he was basically my hero -- he kind of still is, actually.
Eventually, I guess, I decided I wanted to go to graduate school. I honestly don't remember when I got the idea or where it came from. I definitely didn't think it through and ask about what I was getting myself into. I was living in the moment. I enjoyed studying philosophy, and so I wanted to continue studying philosophy. It wasn't until a year or two into grad school (at the University of Notre Dame) that I realized I had embarked on a career. Luckily, it seems to have worked out so far.
I think that’s worth repeating. I'm lucky. I made it through 9 years of education without any major negative experiences. At every stage, my professors took care of me and treated me with respect. Far too many students aren't so lucky. That bothers me. A lot. Paying it forward -- doing what I can to ensure that my students are treated with the care and respect that my professors treated me -- is one way that I have tried to make peace with that luck.
3:16: You’re interested in a species of ideology – the ideology that Quine talks about. So first of all, can you sketch for us what ideology in Quine’s sense is?
PF: Sure! As I understand it, ideology is the means through which ideas are expressed and the ideology of a theory of metaphysics is the undefined terminology that is used to state the theory. So, let's contrast two theories of modality. One theory formulates claims about possibility and necessity using the sentential operators ‘◇’ and ‘□’. The other theory avoids using these operators; it formulates claims about possibility and necessity using quantification over possible worlds and a counterpart relation. These two theories have different ideologies. That's still true even if we can translate every sentence that uses modal operators into a sentence that uses counterpart relations and vice versa . What matters is what terminology the theory itself does not define.
Now, like many other technical terms, ‘ideology’ is used inconsistently. In contrast to what I just said, some philosophers take a theory's ideology to be the primitive concepts that are employed when stating the theory. Most of the time, the differences between these “interpretations” of ideology don't matter. Both my “linguistic” interpretation of ideology and the alternative “conceptual” interpretation can be fruitfully applied to philosophical issues. So… why do I go with the linguistic interpretation? If I'm being honest, it's mostly because talking about concepts makes me feel woozy. How do I determine if I'm employing one concept instead of another, similar, concept? How much can a concept change and still be the same concept? I don't know! Of course, similar issues come up when talking about language. But I feel somewhat more comfortable dealing with those.
That being said, the differences between the linguistic interpretation and the conceptual interpretation do matter in certain situations. I think syntactical differences in bits of language play an important role in our choice of ideology. Here's an example. Some philosophers claim that existence is not a property. If they're right, then maybe we should avoid theories that employ an existence predicate. Maybe we should instead adopt theories that employ existential quantification. This difference between existence-as-predicate and existence-as-quantifier makes sense to me. But I have a harder time seeing an analogous difference in concepts. And insofar as I can, I worry that it is just parasitic on the linguistic difference.
For what it’s worth, I wish Quine hadn’t chosen the word ‘ideology’. The word comes with way too much baggage. Once, during a presentation on the ideological differences between the A-theory of time and the B-theory of time, I had to explain that I wasn't giving a Marxist analysis of how conceptions of time are used by the ruling class to create a false consciousness for the purposes of dominating the working class. That sounds like an interesting idea -- but it’s definitely not what I’m talking about!
3:16: You’re interested in the role this ideology plays in metaphysics, and in particular the distinction between ideological externalism, on the one hand, and ideological internalism, on the other. So what is this distinction about? What are the two positions?
PF: When we’re choosing between two rival theories, we’re in part choosing between two rival ideologies. But what criteria should we use when making this choice? Some criteria are externalist, meaning that they judge the ideology on the basis of its relation to external features of the world. Here’s an example. Lots of philosophers think that some properties are more natural than others -- so, maybe the property being green is more natural than the property being grue (that is, let's say, the property of being either green and observed before 3000AD or blue and not observed before 3000AD). One externalist criterion evaluates the ideology of a theory on the extent to which it reflects these differences. So far as this one criterion is concerned, predicates that express natural properties are better than predicates that express unnatural properties. Other criteria are internalist, meaning that they judge the ideology of a theory on the basis of what historical, sociological, or psychological relations hold between it and those who use the ideology. Intelligibility is a good example of an internalist criterion. Here's another one: we should use predicates that we have a history of predictive success with. This internalist criterion would also judge that it would be better for us to use ’green’ rather than ‘grue’. But of course our success with ‘green’ is a matter of historical contingency. There might be aliens somewhere out there who have had just as much success using ‘grue’. This internalist criterion would respect that difference; each community should continue to use the predicates that they have had success with, even if that success is a matter of contingent luck.
Now, some characteristics might be instrumental toward some further end. For example, the fact that a predicate has a history of predictive success may suggest that the predicate tracks a natural property. But that instrumental connection is not what I'm talking about. Criteria that are genuinely internalist in the sense that I am interested in value these internal characteristics for their own sake.
That's the division of criteria. But there's also a division of balance. In some circumstances, the externalist criterion favoring natural properties and the internalist criterion favoring historically successful predicates make the same recommendation. But in some circumstances they don’t. When they don’t, how do we balance the criteria in a way that gives us determinate recommendations about what ideology to employ? That's a super tough question, and exploring its nuances has been an endless source of inspiration and despair.
Okay, coming back to the question you actually asked -- what are ideological externalism and ideological internalism? Broadly speaking, philosophers divide themselves into two camps. The philosophers who are in the camp that I call ideological externalism privilege externalist criteria in the sense that they predominately appeal to externalist criteria when they choose between rival ideologies and in the sense that they tend to weight externalist criteria more heavily than internalist criteria. This camp includes philosophers that would generally be called “realists”, like David Lewis and Ted Sider -- and me too, I guess. What about ideological internalism? No surprise here: the philosophers who are in this camp do the opposite, privileging internalist criteria. Again, the internalist camp includes those who would generally be called “anti-realists”. I think Quine belongs in this camp, despite his quasi-mythical placement as the “savior” of realist ontology.
It’s worth emphasizing that this division is not a matter of precise doctrinal disagreement. At the end of the day, the real work will come down to evaluating and balancing the full range of externalist and internalist criteria. But no one (myself included) has articulated a fully developed “meta-ideology” along these lines. Given that, the rough division between externalists and internalists is still pretty useful as a diagnostic tool. Lots of philosophers talk past one another because they don't recognize their implicit differences when it comes to how they think of ideology. My hope is that we will begin to uncover these differences and deal with them properly. My division between ideological externalism and ideological internalism is a first step toward realizing that hope.
3:16: You opt for a version of the externalist viewpoint don’t you? So what is this view, and why is this particular version of ideological externalism – non-ontic externalism - superior to “internalist” positions, and better than competing externalist views as well?
PF: So, let me start with the absurd terminological labels: I identify as a maximal, extreme, non-ontic ideological externalist. What that means is that I think that external criteria should apply to the full range of a theory’s ideology (maximality), the only criteria that should be relevant to judging ideology are external (extremeness), and the worldly features to which the ideology relates need not be things (non-ontic). Now, let me try to unpack all that labeling and (hopefully!) make it seem a bit less absurd.
I’m an ideological externalist because I’m a realist. At a minimum, I want a theory that accurately states what sorts of things exist -- I want a theory with the right ontology. But the ontology of a theory is expressed through its ideology. So, by changing the ideology of a theory, we can also change its ontology. Many philosophers are familiar with how this works for predicates: if someone wants to avoid an ontological commitment to chairs, then they can try to avoid talking about chairs and instead talk about simples arranged chair-wise. But the same basic idea applies to the full range of a theory’s ideology. Quine himself developed a formal system (“predicate functorese”) that eschewed quantification. If that system is adequate, and if a theory’s ontology is determined by the values of its bound variables, then any theory stated in that system has an empty ontology. So, if we are to have any hope of developing an accurate theory of the world, our choices about quantificational ideology ultimately have to be constrained by how the world is.
I’m a maximal externalist because this same line of reasoning can be extended to all other types of ideology. I don't have a general proof of this. But I think that all ideology can be “grue-ified” in some way or another. Imagine, for instance, deviant modal operators that behave like‘◇’and ‘□’ most of the time, but produce “unnatural” results in circumstances beyond our experience. Or imagine truth-functional connectives whose truth-tables change as the day of the week changes. We will always have ideological choices to make and these choices will result in different characterizations of the world. If you care about the differences in characterization provided by ‘green’ and ‘grue’, then you should also care about the differences in characterization provided by other types of ideology.
I’m an extreme externalist for similar reasons. If we let internalist criteria factor into our choice of ideology, then we lose the right to say that our resulting theory really is our best attempt at correctly characterizing the world. Here’s a small thought experiment. Suppose God descends from the heavens and hands us a book that contains the best characterization of the world -- best in the sense that everything it says is true and everything it says is stated in an ideology that matches the way the world is. Unfortunately, the book is unintelligible to us. Assuming that God is telling the truth, that we have reason to believe that God is telling the truth, and so on, I think we should accept that book as the definitive account of the world, even if we can’t read the book, even if we have no history with the concepts expressed in the book, and even if it disagrees with our current best book.
Now, I suppose I’m assuming a controversial view about what the final goal of metaphysics is. As I see it, the final goal is to get the most accurate characterization of the world that we can get. God’s book gives us that. But other philosophers might quite reasonably think that the final goal of metaphysics is something else -- like, to understand the world. God’s book contributes very little toward that goal. I don’t have an argument for what the final goal of metaphysics “should” be. In fact, I try to be a pluralist about this sort of thing. There are several importantly different projects that get labeled as “metaphysics”. Maybe some of these projects are more valuable or worth pursuing than others. But I'm not really interested in having that conversation. Let a hundred flowers (safely) bloom. The point is this: if what you care about is getting the most accurate characterization of the world that you can get, then you should be an extreme externalist.
Okay, the last label requires a bit more explaining. There's a way of seeing the world as nothing but individuated things and their qualities. Those things can be substances, or universals, or sets, or whatever -- the point is that the contents of the world could be written down on a numerated list. When we evaluate our ideology on the basis of externalist criteria, then, it's a matter of determining how well a given bit of ideology corresponds to something on that list. So, the name ‘Cicero’ corresponds to the line-item that is that ancient Roman guy, the name ‘Tully’ corresponds to the same line-item, the word ‘electron’ corresponds to the line-item property being an electron , the word ‘green’ better corresponds to some physical property that's on the list than does ‘grue’, and so on.
That sort of world view meshes well with ideological externalism when we limit ourselves to names and predicates. But what about quantifiers and truth-functional connectives? What line-item could these types of expressions correspond to? We could say that there are properties like existence and conjunction . But that doesn't seem right to me. (As I mentioned before, some philosophers already deny that existence is a property.) We could say that these expressions don't correspond to any line-item because they're not the kind of expression that corresponds to the world. I don’t like that answer because it abandons maximality. So that leaves us with one final option: reject the line-item world view.
What to put in its place is hard to articulate. Following Sider, I say that the world has metaphysical structure. But what metaphysical structure is is another question that I’m still thinking through. It isn’t a thing in the same way that a property is a thing. Structure cannot be divided into parts. The way I think about structure is similar to how we think about stuff -- in the metaphysically technical sense of ‘stuff’. There can be more or less coffee in my mug, but I cannot count the “coffees” in my mug. Similarly, the world can have more or less structure, but we cannot count the “structures” in the world. This way of thinking about structure suggests a holistic view about the correspondence between ideology and structure. I think that’s really interesting, but also really strange! What are the implications for this difference in ideological correspondence? Does it mean that, strictly speaking, the world cannot be divided into its physical structure, its logical structure, its modal structure, and so on? To be honest, I’m not sure, and I need to think more about it.
3:16: So if you think we should all endorse this view that maximal realism is true then how do we ensure that theories are using the right ideology?
PF: Yeah, so now we’re talking about the epistemology of metaphysics. Here be dragons!
I don’t think we can ensure that we’re using the right ideology. I also don’t think that we’re likely to discover the right ideology any time soon. (Geeze, I don't think I realized how much of a pessimist I am...) I do think, however, that we can make rationally motivated changes in our choice of ideology that are probabilistically more likely to be correct. That’s a kind of progress, I think, albeit a slow and often frustrating one.
Here’s how I think about it. In metaphysics, there are various disputes that come down to a choice between rival theories. These rival theories have different ideologies that, presumably, are different with respect to how well they correspond to the world’s structure. So, we should base our choice at least in part on which theory we think is most likely to have the most accurate ideology. How do we determine which that is? I favor a virtue-driven methodology. Once a dispute reaches a mature state of stability -- in the sense that the main theoretical options are coherent, their consequences have been identified, and so on -- we can compare the rival theories with respect to various theoretical virtues, features of the theory that make it more likely to be correct. For instance, one theory might be more simple than its rival, and might be for that reason more likely to be correct. The theory that exhibits the greatest amount of overall virtue is the theory that we should endorse. Of course, on this methodology there is no guarantee that the theory we identify as having the greatest virtue is the correct theory. For one, we might have missed another potential theory that would have had even greater virtue. But, even setting that aside, virtue is only conducive toward being correct. It’s entirely possible that the less virtuous theory turns out to be the correct one. Such is life.
That’s the basic idea. But that basic idea gets very complicated very quickly. First, there’s a cluster of issues (well-established by this point) in determining which theoretical features are genuine virtues as well as in determining when a theory has an advantage of virtue over its rivals. But there’s an importantly different cluster of issues that doesn't get as much attention. These issues are about potential conflicts in the value of virtue. Imagine that we have a choice between two theories and we are able to measure (somehow) the precise degree to which the theories differ with respect to two theoretical features that we -- in our imagination -- have definitely proved to be genuine virtues. But the virtues pull in opposite directions: Theory A is more virtuous than Theory B with respect to Virtue 1 to degree x, and Theory B is more virtuous than Theory A with respect to Virtue 2 to degree y. What do we do with that? How valuable are each of these differences? How do we decide which theory is overall more virtuous?
But even these questions presuppose something controversial: that there is a single value that these virtues are conducive toward. Conceptually, we can distinguish between the truth of a theory and the extent to which its ideology accurately corresponds to the structure of the world -- what I call the fidelity of a theory. Truth and fidelity are both ways that a theory can be “correct”, and so in that sense they are both epistemically valuable. But is this sort of epistemic value genuinely unified, and are truth and fidelity commensurable? I used to think so. Now I only hope so!
3:16: So what’s the relationship between fidelity and epistemic value? Why should we seek fidelity when avoiding the gruesome might be enough? Or is avoidance never enough in metaphysics?
PF: I think fidelity is epistemically valuable. Furthermore, I think it has final epistemic value -- it is valuable for its own sake, not merely because it is instrumental toward some other thing that is epistemically valuable. That’s a position that I initially expected philosophers to be open to. But I’ve gotten a surprising amount of pushback! Most commonly, philosophers respond by suggesting that the epistemic value of fidelity is ultimately reducible to the epistemic value of truth. So, for example, someone may say that fidelity is valuable only insofar as differences in fidelity track differences in the significance of the truths in question. That seems plausible, initially. After all, beliefs about fundamental physics seem to be more significant than beliefs about silly gerrymandered properties like grue . But significance amplifies value and amplifies disvalue: a true belief about something significant is much more valuable than a true belief about something trivial, and a false belief about something significant is much more dis valuable than a false belief about something trivial. So, if fidelity tracked differences in significance, then fidelity would also amplify value. But that seems wrong to me. It seems to me that high-fidelity false beliefs are more valuable than their low-fidelity counterparts. Yeah, they’re false, but at least they’re using the correct ideology. That’s better than nothing!
As for seeking fidelity versus avoiding the gruesome… let’s start with a more familiar idea. Everyone strikes a balance between seeking truth and avoiding falsehood. In theory, someone could choose to care only about one or the other. Then their life would be pretty simple! Imagine someone who was maximally cautious and cared only about avoiding false beliefs. They should just avoid believing anything. Mission accomplished. But, as William James said, “not so are victories either over enemies or over nature gained.” There’s something wrong with such an extremely cautious approach to the pursuit of truth.
The same can be said for the pursuit of fidelity. There’s a conceptual difference between acquiring high-fidelity beliefs and avoiding low-fidelity beliefs. In theory, someone could care only about one or the other. But not so are victories over the structure of the world gained. The easiest way to avoid a gruesome ideology is to not employ any ideology whatsoever. That sort of ideological agnosticism seems to me to be psychologically impossible. It also seems to me to be epistemically criticizable. If fidelity has final epistemic value, then it is valuable to have high-fidelity beliefs. Someone who only tries to avoid low-fidelity beliefs will miss out on acquiring the high-fidelity ones.
Now, all of that is only to say that the strategy of being extremely cautious with respect to avoiding the gruesome is bad. That leaves a wide range of other strategies for balancing the pursuits of fidelity, and I'm not sure which are permissible. Let’s return to the more familiar issue regarding truth. Some philosophers, inspired by William James, think that there is no single strategy for balancing the dual pursuits of truth -- acquiring true beliefs and avoiding false beliefs -- that everyone is rationally obligated to adopt. Rather, there’s a wide range of strategies that are permissible. Some people are more courageous than others, and some people are more cautious. This is not merely something to tolerate. It is a good thing. The world needs both kinds of people. But other philosophers worry that this value pluralism leads to problematic permissivism. If two people may rationally weight evidence differently, then perhaps in some situations they may rationally come to different conclusions despite being presented with the exact same evidence. All of these issues extend to strategies for balancing the dual pursuits of fidelity -- acquiring joint-carving beliefs and avoiding gruesome beliefs. I worry about this. I worry about this because it seems to suggest that irresolvable metaphysical disputes are inevitable. Imagine a dispute where Theory A is more “risky” than Theory B in the sense that Theory A is both more likely to carve the world at its joints and more likely to carve the world where its joints are not. A courageous metaphysician may for that reason prefer Theory A over Theory B. Similarly, a cautious metaphysician may prefer the “safer” Theory B over Theory A. If both of them are behaving rationally, then there’s simply nothing to be done about the dispute. Their differences in value have led them to a stalemate.
Actually, this issue about balance is another good example of what I mentioned before about implicit differences going unrecognized. Those of us who are maximal realists think that every bit of a theory’s ideology should be evaluated on the extent to which it accurately corresponds to the structure of the world. That includes a theory's logical connectives. But of course there’s a wide variety of equally expressively adequate sets of logical connectives that a theory may include: disjunction and negation, conjunction and negation, the Sheffer stroke, Quine's dagger, and so on. There seems to be little, if any, reason to choose one of these sets over the other. Thus, choosing one would be objectionably arbitrary. To avoid this issue, Sider has adopted an “egalitarian” position and endorses all of the relevant logical connectives. But some philosophers, like Michaela McSweeney, have argued against this kind of egalitarianism on the basis of its redundancy. She instead advocates for a sort of ideological agnosticism. It seems to me that which position someone should adopt ultimately depends on how they weight the two values of fidelity. Those who are cautious and primarily want to avoid carving the world where its joints are not should adopt agnosticism. But those who are courageous and want to carve the world where its joints are should adopt the comparatively more risky egalitarian position.
3:16: Occam’s Razor requires that we prefer ontological deserts to fecundity. Is your notion of the Comb doing to ideology what Occam does for ontology? Can you sketch for us what the Comb entails, what it brings to metaphysics and why you think that metaphysicians should use both Razor and Comb or neither?
PF: Yeah, that's right! The Razor has a long tradition in philosophy. And it plays an influential role in contemporary disputes in metaphysics. Dozens (hundreds?) of papers argue that such-and-such theory is better because it is more ontologically parsimonious than its rivals, or such-and-such theory is actually not as ontologically parsimonious as people think, and so on. Underlying these arguments is the virtue-driven methodology that I mentioned earlier. Ontological parsimony is a virtue -- it is conducive toward a theory being correct. Thus, we should use the Razor to select the more ontologically parsimonious theory. But there is some dispute as to how the Razor should be used. We could use it to minimize the total number of entities we posit. Or we could use it to minimize the number of kinds of entities we posit. Some philosophers think that only the second use is valuable. For example, there would be no ontological advantage to choosing a theory that posited 7 mathematical objects instead of 13; if someone is going to accept numbers into their ontology, they may as well accept an uncountably infinite amount of them.
But ontology isn’t the only thing we can groom. We can also compare theories with respect to how comparatively parsimonious their ideologies are. The Comb (Is the name kitschy? Yes. But is it cute? Also yes.) is used to select the more ideologically parsimonious theory. But, just as with the Razor, the Comb can be used in different ways. Should we minimize the total number of primitive terms we endorse? Or should we minimize the number of kinds of primitive terms we endorse? There are compelling reasons to resist the first usage. At a minimum, it would seem to entail that a theory that uses the existential quantifier is more likely to be correct than a theory that uses the existential quantifier and the universal quantifier. But it’s hard to see how that could be the case. We could use the Comb to minimize the number of ideological kinds we endorse. But, uh… what is an ideological kind, and how can we determine when two expressions are members of the same kind? I haven’t found a satisfying answer. At first pass, we could say that two expressions are members of the same kind if they are interdefinable. The existential quantifier and the universal quantifier are interdefinable, as are many other expressions that are intuitively members of the same kind. But ‘green’ and ‘grue’ are also interdefinable! Now, we could try complicating the relevant notion of interdefinability. But every attempt at doing so that I’ve seen -- including my own -- leads to some pretty counterintuitive results.
So how about we just don’t use the Comb? That’s not a good option, either. Without the Comb, the Razor is too dangerous. I mentioned before how we can change a theory's ontology by changing its ideology. So whenever we have a theory, we can try to make an ontologically more parsimonious version of it by messing around with its ideology and introducing all sorts of complex ideological devices that avoid ontological commitments. That’s cheating!
3:16: What’s the importance of ideological parsimony when trying to decide between competing metaphysical theories? Is it an epistemic virtue or a fidelic virtue?
PF: To prevent cheating, we should take ideological parsimony to be a theoretical virtue: a theory that is more ideologically parsimonious is more likely to be correct. But, as I keep stressing, a theory can be correct in the sense that it is true and a theory can be correct in the sense that it has fidelity. Consequently, if a theoretical feature is a virtue, it might be a virtue in the sense that it is conducive toward the truth of the theory or it might be a virtue in the sense that it is conducive toward the fidelity of the theory. (It’s also possible that it is conducive toward both.) So we can distinguish between epistemic virtues (or, more precisely, alethic virtues) and fidelic virtues.
Generally speaking, philosophers who take ideological parsimony to be a virtue seem to take it as an epistemic virtue. But I’m not so sure. The interaction argument I gave against cheating doesn’t establish whether ideological parsimony is an epistemic virtue or a fidelic virtue. Yet I think it’s better to take it as a fidelic virtue. First, differences in ideology often do not come with differences in truth. Remember, a grue-ified theory is every bit as true as its high-fidelity counterpart, and adding redundant ideology doesn’t risk adding falsehoods. Second, taking ideological parsimony to be a theoretical virtue coheres well with the general intuition that a simpler picture of the world is more likely to be correct. But ideological parsimony does not produce simplicity with respect to truth and falsity. Rather, ideological parsimony produces simplicity with respect to the proposed structure of the world.
3:16: What’s confabulation – why is it usually thought to be dysfunctional and why don’t you think it is always a bad thing?
PF: At first pass, confabulation is a behavior whereby someone sincerely asserts something even though what they assert is ill-grounded and even though they should know that what they assert is ill-grounded. Examples can be controversial, but think about V.S. Ramachandran’s stroke patient, who reported that they were pointing at the doctor’s nose even though their hand lay by their side entirely motionless. When the obvious was pointed (hehe) out to the patient, they moved on to some other ill-grounded, but shockingly sincere, assertion.
As the example suggests, confabulation was first studied in clinical contexts with patients suffering from brain damage or psychological disorders. If someone takes those to be paradigm cases of confabulation, then it’s no surprise if they also think that confabulation is dysfunctional. But I think there are cases of confabulation occuring in completely ordinary contexts that undermine that presumption. According to a well-known study by Richard Nisbett and Timothy Wilson, when people are given the choice between identical consumer goods, they will disproportionately choose the right-most option but assert that their choice was motivated by a perceived difference in quality. Many people, including my co-author, Sam Murray, and I, think that the Nisbett and Wilson study is a paradigm case of confabulation. These people provide what could be, in theory, genuinely justificatory reasons for their decisions; but clearly what’s really motivating them are positional biases.
To be fair, that behavior still seems dysfunctional when taken in isolation. But Sam and I argue that confabulation is not dysfunctional, in general, because confabulation plays a more general positive role in human psychology. What that positive role is, precisely, is a matter of controversy. Recently, Lisa Bortolotti has argued that confabulation is beneficial because it helps to preserve our sense of agency -- basically, without it, we'd have no choice but to confront the devastating reality that we are driven by an enigmatic web of unconscious influences. But that overlooks the essentially social nature of confabulation. Sam and I argue that the positive benefits of confabulation have more to do with the need to be a socially predictable and resilient agent. Take, for example, the fact that people vastly underestimate the odds that their marriages will lead to divorce. At some level, people are aware that they are underestimating the odds. But their marriages are important commitments, and that, we argue, leads them to non-intentionally but willfully ignore relevant information. Confabulation becomes dysfunctional when people are unable to effectively calibrate this willful ignorance, as is the case with Ramachandran’s stroke patient.
3:16: You’ve looked at the issue of Greek bisexuality in terms of whether to use the term with ancient Greeks is anachronistic, a mistaken projection of one culture’s cultural construct into another culture. You don’t think it is a mistake do you, although you do see epistemic problems with doing it don’t you? So why isn’t it a mistake, what epistemic problems does it face and can these problems be overcome?
PF: Here's something that happens every Pride Month with boring predictability. Some person will contrast our society’s narrow ideas about sexuality with the richly diverse history of sexuality across other times and cultures. So, for example, someone will celebrate bisexuality by talking about the way Alcibiades gushes over Socrates in Plato’s Symposium . (It’s an unfortunate example for them to highlight, maybe, given Alcibiades’s complicated reputation.) In doing so, this person will describe how sexual and romantic relationships between men were widespread and socially sanctioned in ancient Greece. So, their story goes, why should bi people today be stigmatized when so many people back then were also bi? Inevitably, someone else will respond to these comparisons by (rightly) mentioning that the ancient Greeks did not think of sexuality in the same ways that we do. So, their story goes, it is a mistake to say that anyone in ancient Greece was “really” bi.
This type of exchange is given theoretical weight by connecting it to the “essentialist” vs. “constructionist” debate. An essentialist thinks that there are objective categories of sexual orientation -- and probably also thinks that these categories are the categories we use now. Typically, though not always, an essentialist supports their essentialism by pointing to certain biological properties that can (they think) substantiate our current categories. So, a much-too-simplistic essentialist may say that there is a “gay gene” and whether or not someone is gay or not is fully determined by whether or not they inherited the gay gene. No reasonable essentialist would say that now. Simon LeVay (an important and much more reasonable essentialist) suggests that differences in sexual orientation emerge from differences in brain development, which is itself influenced by genes, hormones, and epigenetic processes.
In contrast, a constructionist thinks that there are no objective categories of sexual orientation and our current categories are created by society. Foucault is probably the most famous example (“The sodomite had been a temporary aberration; the homosexual was now a species.”), but Mary McIntosh said it first (“It is proposed that the homosexual should be seen as playing a social role rather than as having a condition.”) Constructionists may disagree with one another as to how, exactly, a society constructs sexuality. But they agree that there is no significant ahistorical and acultural thing for us to track. To some of them, society’s interest in sexual orientation seems akin to an interest in who prefers white meat and who prefers dark meat, or in who sleeps on their stomach and who sleeps on their back. Even if these are real differences that real people have, they are not the sorts of things that society should be structured around. (To put it in terms of the metametaphysics I was talking about earlier, someone could say that these differences are low-fidelity gerrymanders that fail to accurately correspond to the world’s structure. But that presupposes that social practices cannot create worldly structure.)
At first glance, it seems like constructionists should say that the bi advocate made a conceptual mistake. Since sexuality is socially constructed, and since ancient Greek society did not have a concept of bisexuality, it is a mistake to say that anyone in ancient Greece was bi. So, too, it is a conceptual mistake to say that Sappho was a lesbian, and it is a conceptual mistake to say that Emperor Ai was gay. In fact, our divisions between people who are either gay, bi, or straight are so specific to our own society that it is likely also a mistake to project these divisions onto people who are cotemporaneous but culturally distant.
I disagree. We can project our divisions onto culturally and historically distant people. To be clear, that’s not because our divisions are objectively correct in the sense that the essentialist says they are. Rather, even if the constructionist is right, projection is not a mistake -- or, more carefully, it's not a conceptual mistake.
I have a paper where I substantiate this claim by developing a model for characterizing claims about socially constructed features. To keep a long story short: the model (adapted from some neglected work by Iris Einheuser) complicates the notion of a possible world, defining it as an ordered pair of a set of conceptual practices and a “pre-conceptual” substratum. These elements can be mixed and matched. Any arbitrarily selected set of conceptual practices can be paired with any arbitrarily selected substratum, and the result is a possible world. In effect, then, what the bi advocate is doing is imposing their society’s conceptual practices onto the substratum of ancient Greek society and saying that, according to those conceptual practices, Alcibiades counts as being bi. Okay, so then what's going on with people who say that projection is a conceptual mistake? It seems to me that they limit modal space to only those worlds where the substratum supports the conceptual practices it is paired with. But such a limitation should be rejected; it entails that something as simple as “No one in ancient Greece spoke English” is also a conceptual mistake!
Now, as you mentioned, projection does face epistemic problems. Who someone sleeps with is undeniably shaped by their social context. So we cannot straightforwardly infer from their behavior what their underlying sexual orientation is -- if there even is such an underlying and “pre-social” sexual orientation. Similarly, when we project our socially constructed categories onto people who would find those categories alien, we may make inferences that are reliable in our context but not in theirs. For instance, most people today closely associate sexual orientation with “romantic orientation”. Even if this association holds for many people today, it certainly does not hold universally for all people of all cultures throughout history. So, there are epistemic problems with projection. But I think it's important to acknowledge that these problems aren’t unique to the constructionist position. Projection would face these epistemic problems even if essentialism is true. Thus, the essentialist/constructionist debate is irrelevant to issues of projection.
3:16: Another issue regarding sexual orientation is ‘masking’. So what is this, and what does it mean to say that sexual orientation can be intrinsically masked?
PF: Sexual orientation is a disposition. This disposition manifests in sexual behavior: sex, desires to have sex, stuff like that. Someone who is gay is disposed toward sexual behavior with people of the same gender, someone who is straight is disposed toward sexual behavior with people of the opposite gender, and so on. These dispositions are not identical to their manifestations, or lack thereof. Someone who takes a vow of celibacy doesn’t thereby become asexual and someone who lives their entire life on a desert island with no sexual contact and no opportunity for sexual exploration could still be straight, or gay, or whatever. Similarly, in some cases someone’s sexual behavior is not indicative of their underlying sexual orientation. Many such cases are tragic.
This basic idea gets messy and confusing when you start trying to develop it in greater detail. But we can get some help from the literature on the metaphysics of dispositions. A disposition manifests some characteristic property when it is in the appropriate stimulus conditions. So, for example, salt is water soluble because it will dissolve when put into water. Now, a disposition like water-solubility is straightforward; we know precisely what its characteristic property is -- dissolving -- and we know precisely what its appropriate stimulus conditions are -- being put into water. No analysis required! A disposition like fragility is less straightforward. A fragile coffee mug can shatter, but it can also crack or break in half. A fragile coffee mug can shatter when dropped, but it can also shatter when struck by a hammer. A disposition like sexual orientation is even less straightforward. Still, despite these complications, it can be fruitful to think about sexual orientation in these terms.
Here’s how. Something could fail to manifest a disposition even when it is in the appropriate stimulus conditions. Imagine, for example, if my coffee mug were covered in bubble wrap. If I were to drop it, the coffee mug would not break. That's because the bubble wrap masks the manifestation of the mug’s fragility. Now imagine I drop my coffee mug but, at the moment of impact, a wizard casts a spell of shattering on it. The coffee mug shatters, but not because it is fragile. That's a case of mimicking . (Of course, some philosophers would deny that masks and mimics are possible. They would say that the relevant stimulus conditions exclude situations where the object is covered in bubble wrap, or where there is a spell being cast. I think they’re wrong. But getting into why is a bit of a digression.) We can say something similar about sexual orientation. Sexual orientation is a disposition manifested through sexual behavior. But sometimes that behavior is masked, and sometimes that behavior is mimicked. That being the case, determining someone’s sexual orientation (including one’s own) is not as simple as observing how they behave. In fact, it’s not even as simple as observing how they behave across all possible worlds.
Why does all of this matter? Well, if we want a better understanding of sexual orientation, then we need to determine what behavior counts as an authentic expression of orientation. In practice, this means determining what its stimulus conditions are as well as the ways in which it can be masked or mimicked. Plausibly, the correct stimulus conditions require some degree of external freedom: there’s a reasonable diversity of potential partners, there’s no coercive force at play, stuff like that. It’s also plausible to think that the correct stimulus conditions require certain internal features, like the person being willing and able to have sex.
In fact, there’s a temptation to build all kinds of internal features into the stimulus conditions. If someone is in a situation where they don’t want to have sex, and the reason for why is based on some aspect of their psychology, then maybe they’re not in the relevant stimulus conditions. But that’s a temptation I think we should resist. Consider cases of internal conflict. For instance, “Side B” Christians deny that being gay is immoral but maintain that acting on their desires would be immoral, and therefore they decide to live a life of celibacy. Their behavior is determined by their deeply held moral belief. But I think it is a mistake to say their situation is one that falls outside the scope of the relevant stimulus conditions. Someone with less strongly held beliefs may abstain from sex less consistently. It would be objectionably arbitrary to include their situation but exclude that of Side B Christians -- where would we draw the line between beliefs that are sufficiently strongly held and those that are not? I think it is better to say that the manifestation of someone’s sexual orientation can be intrinsically masked by their moral beliefs. More generally, there are many ways that internal factors can prevent or enable sexual behavior. These sorts of internal conflicts should be characterized in terms of masks and mimics rather than in terms of what the relevant stimulus conditions are.
3:16: And why don’t you think death is bad?
PF: I do think death is bad! I just think it’s not distinctively bad. Most people think that there is a sort of asymmetry when it comes to our lives: it is bad to die sooner rather than later, but it is not bad to have been born later rather than sooner. I think (as does my co-author Meghan Sullivan) that such a position is untenable. Rationally speaking, people should be temporally neutral with respect to the length of time they live. In other words, one extra year of life in the future is, ceteris paribus , no more valuable than one extra year of life in the past. To prefer the former over the latter is to be irrational.
Of course, the ceteris are often not paribus . Sometimes, the precise time someone is born plays a profound role in how their life goes. Take my dad, for example. He was born on August 17. August 17 was number 154 in the 1969 conscription draft in the United States. So, if my dad had been born 6 years earlier, he would have been drafted to fight in the Vietnam War. August 14 was number 198, just barely past the 195 birthdays that were drafted. So, if my dad had been born 6 years and 3 days earlier, he would not have been drafted.
So, admittedly, the thing I'm saying is one of those philosopher’s abstractions removed from the real world. But this particular philosopher’s abstraction has a long history. Lucretius (a Roman poet and philosopher who lived in the 1st century BC) gave what was basically the same argument I gave. Importantly, though, Lucretius took the symmetry argument to show that death is not bad: because we do not prefer to have been born earlier, we should not prefer to die later. Meghan and I flip that: rationally, we should prefer more well-being to less whenever it is scheduled, future or past.
Lots of people deny the symmetry claim. Most influentially, John Martin Fischer and the late Anthony Brueckner argued that we should deny it because it is rational, in general, to treat the future and the past differently -- that it is rational to be future biased. Unsurprisingly, we disagree. We think future bias is irrational for reasons that are extensive enough to fill an entire book -- ask Meghan, she’s literally done exactly that!
3:16: And finally, are there five books you can recommend that will take the curious reader further into your philosophical world?
PF: 
1. Meghan Sullivan's Time Biases: A Theory of Rational Planning and Personal Persistence
2. Edward Stein's The Mismeasure of Desire: The Science, Theory, and Ethics of Sexual Orientation
3. Brian Epstein's The Ant Trap: Rebuilding the Foundations of the Social Sciences
4. Lisa Bortolotti's The Epistemic Innocence of Irrational Beliefs
5. C Thi Nguyen's Games: Agency as Art
3:16: As an option, I include a survey requiring one word answers (unless you want to expand).
PF: So, in the spirit of the question, here are my answers without any elaboration whatsoever!
a. A priori knowledge: yes or no? Yes
b. Abstract objects: Platonism or nominalism? Platonism
c. Aesthetic value: objective or subjective? Subjective
d. Analytic-synthetic distinction: yes or no? No
e. Epistemic justification: internalism or externalism? Externalism
f. External world: idealism, skepticism, or non-skeptical realism? Non-skeptical realism
g. Free will: compatibilism, libertarianism, or no free will? Compatibilism
h. God: theism or atheism? Atheism
i. Knowledge claims: contextualism, relativism, or invariantism? Contextualism
j. Knowledge: empiricism or rationalism? Empiricism
k. Laws of nature: Humean or non-Humean? Humean
l. Logic: classical or non-classical? Classical
m. Mental content: internalism or externalism? Externalism
n. Meta-ethics: moral realism or moral anti-realism? Moral realism
o. Metaphilosophy: naturalism or non-naturalism? Naturalism
p. Mind: physicalism or non-physicalism? Non-physicalism
q. Moral judgment: cognitivism or non-cognitivism? Cognitivism
r. Moral motivation: internalism or externalism? Externalism
s. Newcomb's problem: one box or two boxes? Two boxes
t. Normative ethics: deontology, consequentialism, or virtue ethics? Deontology
u.Perceptual experience: disjunctivism, qualia theory? Qualia theory
v. representationalism, or sense-datum theory? Representationalism
w. Personal identity: biological view, psychological view, or further-fact view? Biological
x. Politics: communitarianism, egalitarianism, or libertarianism? Communitarianism
y. Proper names: Fregean or Millian? Fregean
z. Science: scientific realism or scientific anti-realism? Science realism
Ai. Teletransporter (new matter): survival or death? Death
Aii. Time: A-theory or B-theory? B-theory
Aiii Trolley problem (five straight ahead, one on side track, turn requires switching, what ought one do?): switch or don't switch? Don't switch
Aiv. Truth: correspondence, deflationary, or epistemic? Correspondence
Av. Zombies: inconceivable, conceivable but not metaphysically possible, or metaphysically possible?Conceivable but not metaphysically possible.
ABOUT THE INTERVIEWER
Richard Marshall is biding his time.
Buy his second book here or his first book here to keep him biding!
End Time series: the themes
Huw Price's Flickering Shadows series.
Steven DeLay's Finding meaning series
Joseph Mitterer's The Beyond of Philosophy serialised
NEW: Reviews of new political books - A Hornbook of Democracy Book Reviews
NEW: How Dylan's Songs Work interview
NEW: Ernest Gellner Legacies series 1: Alan Macfarlane essay