The Limits of Adverbialism about Intentionality (in Inquiry)

Inquiry An Interdisciplinary Journal of Philosophy ISSN: 0020-174X (Print) 1502-3923 (Online) Journal homepage: http://www.tandfonline.com/loi/sinq20 The Limits of Adverbialism about Intentionality Casey Woodling To cite this article: Casey Woodling (2016) The Limits of Adverbialism about Intentionality, Inquiry, 59:5, 488-512, DOI: 10.1080/0020174X.2016.1140071 To link to this article: http://dx.doi.org/10.1080/0020174X.2016.1140071 Published online: 26 Feb 2016. Submit your article to this journal Article views: 429 View related articles View Crossmark data Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=sinq20 Download by: [Casey Woodling] Date: 12 September 2016, At: 12:40 InquIry, 2016 VOL. 59, nO. 5, 488–512 http://dx.doi.org/10.1080/0020174X.2016.1140071 The Limits of Adverbialism about Intentionality Casey Woodling Coastal Carolina university, uSA ABSTRACT Kriegel has recently developed (but not fully endorsed) an adverbial account of intentionality, in part to solve the problem of how we can think of non-existents. The view has real virtues: it endorses a nonrelational (internalist) conception of intentionality and is ontologically conservative. Alas, the view ultimately cannot replace the act-object model of intentionality that it seeks to, because it depends on the act-object model for its intelligibility at key points. It thus fails as a revisionistic theory. I argue that the virtues of adverbialism can be had from within the act-object framework, provided we understand intentional objects correctly. I use Crane as a guide here, and build on his work on intentional objects. In the end, we can provide a suitable solution to the problem of thinking of non-existents within the act-object framework without adopting implausible ontological or metaphysical views. So, adverbialism is neither a possible stand-alone revisionary option nor a needed modiication of the common-sense act-object framework of intentionality. ARTICLE HISTORY received 9 June 2015; Accepted 5 January 2016 Philosophers have long kicked up the dust discussing thoughts about non-existents. These discussions typically track the following problem. We have two intuitively appealing claims that appear incompatible. First, it seems obvious that we can think of things that do not exist. For example, I can think of mythical creatures, such as Pegasus, even though mythical creatures do not exist. Second, it can seem equally obvious that when one thinks of a non-existent, one is thereby related to it in some sense and CONTACT Casey Woodling cwoodling@coastal.edu Present address: Casey Woodling, Coastal Carolina university, The Department of Philosophy and religious Studies, PO Box 261954, Conway, SC 29528, uSA § © 2016 Informa uK Limited, trading as Taylor & Francis Group InquIry 489 it must therefore exist in virtue of being a relatum. How can we accept both seemingly true ideas? If intentional objects must exist in virtue of being relata, it makes no sense to say that they are nonexistent. There is no obvious way to resolve this incompatibility. I shall follow uriah Kriegel, a recent contributor to this debate, in calling this ‘the problem of intentional inexistence’ (Kriegel 2007). Kriegel (2007, 2008, 2011) has shown how adverbialism about intentionality can be defended.1 Adverbialism as a model of intentionality rejects the idea that intentional objects have a constitutive role to play in intentional states. To say that intentional objects are constitutive for a given model of intentionality is to say that intentional objects are necessary components of intentional states according to that model. For an act-object theorist, there is no intentional state without an intentional object. not so for the adverbialist. For the adverbialist, conscious intentional states are not acts that relate thinkers to intentional objects; they are ways of thinking, instantiations of non-relational properties.2 Pegasus is not the object of my thought according to the adverbialist.3 In thinking of Pegasus, I am thinking Pegasus-wise. The adverbialist approach, then, says that we have genuine thoughts about non-existents, but in doing so we do not stand in any relationship to non-existents; there are no constitutive objects of thought.4 The instantiation of adverbial properties, such as thinking Pegasus-wise, is what makes a state intentional. no intentional object is needed.5 A substantial part of the appeal of adverbialism is thought to lie in its avoidance of commitment to intentional objects, entities many philosophers are wary of. Many philosophers share this desire to avoid 1 For a defence of adverbialism, see Kriegel (2011) reports a slight preference for a naturalistic theory of intentionality based on tracking properties over the adverbial theory he has developed. 2 On my own view, discussed in Section IV, intentional states also instantiate non-relational properties, though the story diverges importantly from adverbialism. 3 In this paper, I understand ‘intentional object’ and ‘object of thought’ synonymously. 4 The adverbialist can leave open that thoughts may have objects in the sense that thoughts could refer to or pick out objects, but this is a diferent commitment than holding that intentional states must have intentional objects in order to be genuine intentional states. 5 Adverbial properties are non-relational properties whose instantiation does not require the existence of any object in the individual’s environment. non-relational properties (or intrinsic properties as they are sometimes called) need not be adverbial properties, though. For example, it is possible that the property of having a thought about Barack Obama is both non-relational and non-adverbial. This is the view of the content internalist who is not an adverbialist, but believes that intentional properties are not ways in which we think but properties that we instantiate that make our thoughts about objects in the world. I argue for such a view in Section IV of the paper. 490 C. WoodlIng questionable ontological commitments. However, adverbialism cannot be the whole truth on intentionality. As a revisionistic theory, it promises to displace the common-sense act-object framework. It cannot do this, however, because it depends on the act-object framework for its very intelligibility at key points. Even though it cannot stand on its own as a theory of intentionality, adverbialism has some decided virtues: it construes intentional properties as non-relational and avoids questionable ontological and metaphysical commitments. However, those virtues can be had from within the act-object model of intentionality provided that we understand intentional objects correctly. I argue that Tim Crane has already articulated a notion of intentional objects that allows us to achieve the virtues of adverbialism from within act-object framework. The goal of this paper, then, is twofold. I aim to show that adverbialism depends on the act-object model of intentionality and that the virtues of adverbialism can be had within the act-object framework. In the end, we shall see that adverbialism, while not without its heuristic merits, cannot stand on its own as a theory of intentionality nor do we need it to for an ontologically and metaphysically conservative view of intentionality. Here is the plan. I explain adverbialism about intentionality in Section I. In Section II, I lay out a plausible desideratum that all revisionistic theories must meet. In Section III, I argue that adverbialism fails as a revisionistic theory because it fails to satisfy the desideratum discussed in Section II. In Section IV, I outline Tim Crane’s view of intentional objects and explain how it achieves the virtues of Kriegel’s account from within the act-object framework. In Section V, I end with a brief conclusion. I. Adverbialism about intentionality Before I say more about adverbialism, it will be useful to discuss Kriegel’s articulation of the problem of intentional inexistence as an inconsistent triad of claims. (a) one can represent non-existents. (b) one cannot bear a relation to non-existents. (c) representing something involves (constitutively) bearing a relation to it Kriegel (2007). Intuitively, we want to accept all these claims, but one of the claims must be rejected because they are jointly inconsistent: it InquIry 491 follows from them that we can bear relations to things to which we cannot bear relations.6 It may seem that (c) is a key commitment of the act-object model of intentionality, for intentional states are characterized as acts that are directed upon objects.7 Without an object of the intentional state, there is simply no intentional state for the act-object theorist. Crane (2001a) says that if we drop talk of intentional objects, then it is not clear what intentional states are about. If someone says that they are in an intentional state and there is no object, then it is hard to see what makes that state intentional, since it is about nothing. To have aboutness—to have intentionality—thoughts must be about something; they must have some object or other. Within this general framework, there is room for diferent views on the nature of intentional objects, but there are intentional objects of some sort. I take this to be the core commitment of the act-object model of intentionality. It is this core claim that the adverbialist denies. Kriegel’s solution to the problem is to deny (c). To what does this amount? Thoughts have content according to the adverbialist, but that content is not dependent on them having an intentional object. rather, it depends on the way in which the subject is thinking. Here is Kriegel. The alternative to the act-object theories are theories according to which being in an intentional act/state does not involve constitutively (though it may involve contingently) bearing a relation to an intentional object. on this view, thinking about something is never constituted by bearing a relation to that-which-one-thinks-about. I think of this view as a type of adverbialism about intentionality: to think about Vienna does not involve (constitutively) bearing an aboutness relation to Vienna, but rather engaging in the activity of thinking in a certain way—Vienna-wise. More generally, for any x, representing x does not involve constitutively bearing a representation relation to x; what it involves constitutively is representing x-wise. Kriegel (2008, 84). 6 It may seem that (c) is the least plausible claim in the set and thus this triad does not constitute a genuine philosophical puzzle because the claims do not carry equal plausibility. In thinking about intentionality, however, (c) is a claim that many are drawn to, so setting things up in this way captures how many thinkers irst come to see the puzzle involving representing non-existents. It has been suggested to me that the puzzle could be set up by noting that from (a) and (b) it follows that we do not represent non-existents by bearing a relation to them, a potentially puzzling claim in need of an explanation. That is a ine way to think of the puzzle, but I have chosen this particular set-up not only because it is the one Kriegel ofers but also because I think it does a good job of representing the problem of thinking of non-existents as most people irst ind it. 7 In Section IV, I explain why an act-object theorist is not necessarily committed to (c). 492 C. WoodlIng According to adverbialism even though thoughts do not always have objects, they can be true or false, satisied or unsatisied, and it is by means of these relationships—and not in virtue of having intentional objects—that they connect with the external world.8 An intentional state having an object contingently is possible for the adverbialist. All this means is that some intentional states actually refer to objects in the world. They need not have such objects to be intentional, though. Before moving to the next section, I need to explain the sense in which thoughts can have objects according to the adverbialist, and also bring out a crucial ambiguity that is present in discussions of intentional objects. As we saw, the adverbialist denies that intentional objects are needed for intentional states. The idea here is that thoughts can have intentional content without referring to an object, without having an intentional object. So, the adverbialist thinks of intentional objects as the referents of thoughts, which is a mistake. There is a rival understanding of intentional objects, one in which they have an independence from the referents of thoughts, which is crucial to the proper act-object view. I shall briely sketch this understanding of intentional objects so as to draw out a pervasive ambiguity in ‘intentional object’ in writings on intentionality before turning to Section II. I will say more about this ambiguity in Section IV, but it is helpful to discuss it briely now. on the view I defend, intentional objects are what subjects take their thoughts to be about. And what subjects take their thoughts to be about is not always what they refer to or pick out. If I take a thought to be about Pegasus, then it refers to nothing even though it has an intentional object, the object that I take it to be 8 Kriegel (2008) helpfully lays out the two main versions of adverbialism, inferentialist adverbialism and phenomenological adverbialism. While both reject that intentionality is a relation between a subject and an object, they difer on which non-relational properties of thinkers determine the content of thoughts. Suppose I am thinking of Pegasus. The inferentialist adverbialist says that my thought represents Pegasus in virtue of the fact that the thought plays a Pegasus-wise inferential role in my mental economy. The phenomenological adverbialist says that my thought represents Pegasus in virtue the phenomenal character of the thought; it has a Pegasus-wise phenomenal character. It is important to note that both versions appeal to internal factors to determine intentional content and do not require that the subject who has the thought stand in any relation to an object outside of the mind. InquIry 493 about.9 When we understand intentional objects in this way, we see that the intentional object of a thought is not always identical with the referent of a thought, since in the case of thought about non-existents, there are no referents. often times, of course, what we take a thought to be about is what the thought refers to. When I take my thought to be about President obama, obama is who is referred to. He is both intentional object and referent. However, there is a danger in discussions of intentional to slide illicitly between both uses of ‘intentional object’. As will become clear by the end of the paper, one of the keys to an ontologically and metaphysically conservative solution to the problem of intentional inexistence is seeing that intentional objects are the objects that subjects take their thoughts to be about; understanding intentional objects as constitutive of intentional states is perfectly sensible provided that we understand intentional objects in this way. If intentional objects are the objects subjects take their thoughts to be about, it is clear why all intentional states need intentional objects. If a subject takes his or her thought to be about nothing, it clearly has no intentionality. on my view ‘intentional object’ only has one true meaning—what a subject takes his or her thought to be about. It has been used to track the referents of thoughts, as in Kriegel’s defence of adverbialism, but this is problematic. At this point, we need to be mindful of the potential to use ‘intentional object’ in these two distinct ways. II. A desideratum for revisionistic theories It is not diicult to see that adverbialism is a revisionistic theory. our common thought and talk presupposes that we are related to objects. We must revise this, according to the adverbialist, in favour of seeing others and ourselves as having contentful thoughts in virtue of thinking in certain ways. In this section, using the example of the metaethical theory of expressivism, I want to establish a desideratum that any revisionistic theory must meet to be plausible. It is fair to say that expressivism is a revisionistic 9 This notion of taking can also be cashed out in terms of seeming. The object a subject takes his or her thought to be about is the object that the thought seems to be about from the subject’s point of view. And the seeming ixes the reality. There is no gap between what a subject takes a thought to be about and what the thought is about. One could also say, as has been said, that intentional objects are the objects that are presented to subjects in thought. 494 C. WoodlIng theory because if it is true, we must revise the common-sense idea that moral claims are expressions of cognitive attitudes, that is, we must revise the part of our shared conceptual scheme that holds that moral beliefs can be true or false. When one asserts, ‘Adultery is wrong’, in a normal context, we take that person to be expressing the belief that adultery is wrong, understanding that this belief is truth evaluable. Expressivism says that the standard interpretation is false. When one assents to the sentence about adultery, one merely expresses a non-cognitive attitude that is not properly evaluable for truth or falsity. In assenting to ‘Adultery is wrong’, one merely expresses a non-cognitive distaste for adultery according to the expressivist. It is clear that this forces us to revise common sense, which more or less assumes cognitivism is true.10 To be plausible, a revisionistic account must not presuppose or depend for its very intelligibility on the common-sense view it seeks to replace. Expressivism meets this desideratum. Expressivism is not parasitic on cognitivism (considered as a proxy for common sense) in any way that would prevent it from being a truly revisionary theory. Consider the following sentences. (1) Adultery is wrong. (2) don’t commit adultery! The expressivist’s claim is that when someone utters (1), they are really uttering (2). We do not have to depend on the conceptual framework of cognitivism to make sense of (2), that is, we do not have to irst understand the idea that adultery is wrong as a straightforward factual claim in order to understand the idea that someone may be disgusted by or have negative feelings about adultery. So, while expressivism forces us to revise common sense, we are doing it in a way that does not require the conceptual resources of cognitivism or our common-sense scheme.11 It should be clear why revisionistic theories must satisfy the desideratum. If revisionistic theories are in some way parasitic on 10 In saying this I do not mean to suggest that most people are aware of the thesis of cognitivism or conversant in metaethical debates. I simply mean that most people take ethical claims to express beliefs that can be evaluated for truth or falsity. In what follows, I take the basic commitments of cognitivism to articulate the common-sense view of the status of moral claims. 11 Strawson (1959) makes similar claims about various forms of skepticism. Skepticisms of various sorts fail if their coherence actually requires the portions of the common-sense conceptual scheme they attempt to revise to be true. Such skeptical challenges to common sense would not be candidates for completely new schemes, then, but merely fanciful ways of restating the claims of common sense. InquIry 495 the common-sense scheme they seek to replace, they do not truly represent an independent theoretical option. At most they amount to a fanciful way of reinterpreting the language of the theory they seek, but ultimately fail, to replace. III. Adverbialism and the desideratum The question I want to pursue is whether adverbialism meets the desideratum just discussed. If it fails to, then it fails as a stand-alone theory of intentionality. I argue that it fails for two reasons. First, it fails because the recipe Kriegel gives for understanding adverbial properties (directly introspecting them) relies on intentional objects (understood as the objects subjects take their thoughts to be about) and commitment to intentional objects is commitment to the act-object model of intentionality. Second, in order to truly understand novel adverbialist intentional state ascriptions (the language used to attribute intentional content as the adverbialist conceives of it), one must revert to the act-object model of intentionality.12 These two points show us that at best adverbialism cannot replace the act-object model of intentionality but amounts at most to a way of speaking from within the act-object framework or perhaps a heuristic that helps us see some important points about intentionality. III.i. Introspecting adverbial properties requires intentional objects To see one point where adverbialism relies on the act-object framework, let us begin with how adverbialism handles non-propositional intentional states (intentional states we would describe as ones in which a thinker is related to an object and not a proposition), which are easier for an adverbialist to analyse than propositional intentional states (intentional states we would describe as ones in which a thinker is related to a proposition). Consider the following intentional state report. (3) President obama is thinking about the budget deicit. 12 I take novel ascriptions to be ascriptions that an interpreter has not previously encountered. 496 C. WoodlIng obama does not stand in an aboutness relation to the budget deicit according to the adverbialist. He is merely thinking in a certain way. We should therefore translate (3) as follows. (4) President obama is thinking budget-deicit-wise. of course, the fact that we are already accustomed to using adverbs to modify mental states can give such adverbial intentional ascriptions the appearance of true intelligibility. We would not bat an eye if someone said, ‘She thinks through philosophical problems quickly’, or, ‘He hopes for the future unrealistically’. The adverbialist, though, needs to press adverbs into a diferent role. Instead of describing the way or method in which something is done, the adverbs must on their own express the entire content of intentional states. Is it coherent to talk about thinking in a budget-deicit manner? A irst pass may ind us attempting to parse the adverb in the familiar way as indicating some general pattern of mentality. Perhaps in uttering (4) the speaker is trying to say that obama’s thought about issues is growing progressively more troubling. This is not how we are supposed to understand the adverbial modiier, though. It is supposed to tell us how obama’s thought is directed, and it tells us that his thought is directed budget-deicit-wise. But what does this mean? Here is a helpful passage from Kriegel on how we can make sense of the idea that the content of a thought is to be cashed out in terms of thinking in some way or manner. What does it mean for my current visual experience to be intentionally directed laptop-wise, or for my current thought to be intentionally directed dragon-wise? Without an answer to this question the very intelligibility of the view is in question. The irst thing to note, however, is that experiential properties are amendable to grasping simply by ostension through something like direct introspective encounter. So given the relevant adverbial properties are experiential they may be grasped in this way. And indeed, when I relect on why adverbialism seems to me perfectly intelligible, and how I get a sense of grasping the property of being directed dragon-wise, it seems to me that it is simply introspective encounter with the property that afords me this grasp. By this I do not mean that introspection instructs me of the adverbial nature of exp-intentional properties; merely that it instructs me of what are in fact adverbial exp-intentional properties.13 Thus, 13 ‘Exp-intentional’ is shorthand for experiential-intentional. For our purposes, we can say that exp-intentional properties are conscious intentional properties. InquIry 497 just now I stopped writing and visualized (about a foot long) green dragon hovering motionless about a yard away from me directly in front of my eyes. When I attended introspectively to this visualizing experience, I encountered a certain property of the experience, a property we have theoretical reasons for construing as visualizing smallish-green-hovering-dragon-wise (Kriegel 2011, 155). one can get a good understanding of this by actually doing the introspecting. I imagine a small green dragon hovering in from of me. I can introspect my mental state and thereby grasp the adverbial property of being directed green-dragon-wise. Though it is not stated in the above quote, Kriegel also holds that after enough of such introspective episodes, one can gain an understanding of the general concept of being directed somehow. It is not obvious that the grasping of adverbial properties depends on the act-object model of intentionality. To make clear the tacit reliance on the act-object model, it is useful to consider propositional intentional states, intentional states that are properly expressed by reference to a proposition or a complete declarative sentence. As an exercise one should try to make sense of the following intentional state attribution without reference to intentional objects, which we need to do to satisfy the desideratum for revisionistic theories. (5) President obama believes that the budget deicit is the greatest threat to the prosperity of future generations of Americans.14 The adverbialist parses this as follows. (6) President obama believes budget-deicit-is-the-greatestthreat-to-the-prosperity-of-future-generations-of-Americans-wise. To grasp the adverbial property by way of introspection, you must put yourself in the same mental state as obama is claimed to have and then directly introspect the adverbial property of thinking budget-deicit-is-the-greatest-threat-to-the-prosperity-of-futuregenerations-of-Americans-wise. But the problem with this recipe for understanding is that I cannot put myself in this mental state without taking my thought to be about the budget deicit and the gqqqqreatest threats to future Americans, that is, I must take these things to be the objects of my thought in the process of 14 Kriegel (2007) holds that non-conscious, standing intentional states represent in virtue of their relation to corresponding conscious intentional states. 498 C. WoodlIng understanding obama’s thought. However, this is merely another way to talk about the intentional objects of a thought. So, it is not possible to say that you can do the instrospecting without intentional objects, since anytime subjects take their thoughts to be about some object, that object is the intentional object of their thought. once we introspect the target mental state, we can of course name the property anything we want. We can use adverbial modiiers, or we could call it property X, but the point is that it is impossible to get oneself into the proper introspective states needed for grasping these adverbial properties without implicitly making use of intentional objects. So, the introspective acts so crucial to the adverbialist story about how to understand adverbial properties in fact depend crucially on the to-be-revised framework, thus violating the desideratum. To this objection, the adverbialist might reply that it assumes a certain notion of an intentional object, the very notion that the adverbialist is rejecting. The very question is whether or not we need intentional objects, so it is not legitimate to assume that there are intentional objects in objecting to adverbialism. The full justiication for such a notion of intentional objects is given in Sections III.ii and IV. I do not in the end merely assume that my view of intentional objects is the correct one. Here is a sketch of the justiication. Adverbialism requires the act-object framework to be true, so it is not a stand-alone theory of intentionality. understanding intentional objects as what subjects take their thoughts to be about allows us to solve the problem of intentional inexistence from within the act-object framework. So, adverbialism is not needed to serve as an ontologically and metaphysically conservative view of intentionality. If intentional objects are the objects that subjects take their thoughts to be about, then we have an ontologically and metaphysically conservative solution to the problem of intentional inexistence from within the act-object framework. For now, it is enough to see that to put ourselves in the right introspective states, we really must take our thoughts to be about objects. I cannot put myself into the intentional state described by (6) unless I take my thought to be about obama, the budget deicit or future generations of Americans. We do not yet have to agree that the objects subjects take their thoughts to be about are intentional objects. The adverbialist may wish to reject this notion InquIry 499 of an intentional object in favor of thinking of intentional objects as the referents of thoughts. However, by the end of the paper, it will be seen that this is not a tenable position for the justiication just given. III.ii. Adverbialism and novel intentional ascriptions The second way in which adverbialism relies on the act-object framework can be seen when we consider whether adverbialism allows us to understand novel intentional ascriptions, the language used to attribute the intentional content of an intentional state to some thinker. It is widely recognized that in order for interpreters to understand novel phrases or expressions on their own and without additional contextual clues, those phrases and expressions must be compositional, that is, they must be decomposable into more discrete, meaningful syntactic units which interpreters are familiar with, so that the meanings of complete sentences can be understood based on the meanings and grammatical roles of the parts of the sentence. Kriegel (2007, 2008) himself admits that adverbialist intentional ascriptions lack compositionality because they are syntactically simple, though he thinks that there are workarounds. Before I talk about the problem adverbialism faces when it comes to compositionality and intentional content, it is useful to make these points at the level of language before they are made at the level of thought. In other words, I shall briely discuss how compositionality is related to linguistic content (or linguistic meaning) before I discuss how it is related to intentional content. Contrast the following two sentences. (7) The budget deicit is the greatest threat to the prosperity of future generations of Americans. (8) Islamic extremism is the greatest threat to the prosperity of future generations of Americans. It is obvious that the respective meanings of (7) and (8) are both similar and distinct. Compositionality is what allows us to see the similarities and diferences in meaning. If we had to treat the sentences as syntactically simple, then we would not be able to compare and contrast the linguistic content of these sentences because we could not break the linguistic content (or meaning) 500 C. WoodlIng down into the discrete parts that we need to in order to recognize the similarities and diferences in meaning. These points about linguistic content are related to points about intentional content. I understand linguistic content as the linguistic meaning of expressions, and I understand intentional content as the content of thoughts. obviously, these two notions of content are very closely related. To make intentional content public, we need to rely on linguistic content, but this is not to say that intentional content is the same as linguistic content. The distinction marks a vehicular diference in that each type of content is had by distinct types of things (linguistic elements and intentional states).15 There is no doubt much of interest to say about the relationship between the linguistic content of intentional ascriptions and the intentional content those ascriptions aim to describe. For my purposes, in what follows, I shall assume that to grasp the intentional content of a novel intentional state we must at a minimum grasp the linguistic content of the novel intentional ascription. This does not say that linguistic content of an intentional ascription always determines the intentional content of the respective state, but something much weaker: if the linguistic content of the ascription of a novel ascription is not understood, then we have no hope of understanding the content of the intentional state being described. It follows from the assumption just stated, and the lack of compositionality of adverbialist ascriptions, that we cannot understand the intentional content that is expressed by novel intentional ascriptions. Consider (6) again. (6) President obama believes budget-deicit-is-the-greatestthreat-to-the-prosperity-of-future-generations-of-Americans-wise. The only explanation, then, for the adverbialist for how we can understand the adverbial property expressed by the ascription in (6) is previous introspection of the property. of course, it seems obvious that we can understand the intentional content of (6) without actually having previously been in that very same intentional state ourselves. Adverbialism cannot explain this basic fact about our ability to understand novel intentional ascriptions. 15 This distinction allows that the core idea of content is the same in both types of content and what makes the diference is merely what has the content. InquIry 501 Another way to see the problem is to notice that each adverbial modiier must be learned on its own because each one is syntactically simple. Adverbial modiiers are like idiomatic expressions in that they cannot be grasped by grasping the meanings of the constituent words, but must be learned as an entire unit. Consider the idiomatic expression: ‘He isn’t playing with a full deck’. This has a certain literal meaning that can be ascertained by grasping the appropriate meanings of the terms that make up the sentence. However, it has a metaphorical meaning that cannot be grasped simply by knowing the appropriate meanings of the constituent terms. To learn a new idiomatic expression, such as ‘He isn’t playing with a full deck’, we cannot simply grasp the whole sentence meaning by grasping the meaning of the parts. The same is true for adverbialist intentional ascriptions. Kriegel discusses a related issue when he notes that the lack of compositionality might entail that a speaker may not learn all possible adverbialist intentional ascriptions; he sees this as no great concern. After all, the adverbialist intentional ascriptions can be learned, he thinks, one-by-one by way of the method of introspection discussed in Section III.i. However, as was discussed, this implicitly appeals to the notion of an intentional object, which should be understood as the object that the subject takes his or her thought to be about. And this is a notion of an intentional object that is part and parcel of the act-object framework. Even if this point can be addressed, problems remain for the adverbialist. Just consider (6), which is likely a novel intentional ascription for most readers; no one has diiculty in parsing this sentence, but if adverbialism were true, we could not parse it because it cannot be subjected to compositionality. Because we cannot parse it, we would have to learn it in the way we learn idiomatic expressions. But this is not required to understand novel intentional ascriptions. Except in exceptional cases, we do not learn them as units but are able to grasp their entire meanings based on the meanings of their parts and the grammatical roles of those parts. So, adverbialism is false because it cannot account for our ability to understand novel intentional ascriptions in the way we do.16 As we consider potential replies from the adverbialist, we 16 It may also be true that there are other implausible consequences for adverbialism when we grant that the adverbialist intentional ascriptions (or paraphrases as I sometimes call them) lack compositionality. I have kept my focus on its inability to explain how we can understand novel intentional ascriptions to streamline the discussion. 502 C. WoodlIng can see the point at which adverbialism relies on the act-object framework. An adverbialist may reply that we can in fact understand the compounds to which ‘-wise’ applies. To this, it must be noted that we can do so only by appealing to act-object framework. To understand, for example, what it means to apply the novel compound adverb ‘the-budget-deicit-is-the-greatest-threat-to-the-prosperity-of-future-generations-of-Americans-wise’ to an intentional state, we must not only be able to decompose the compound, but we also must be able to ask ourselves how the linguistic content of the compound adverb describes the intentional content in question. We must therefore understand the ascription not as a compound adverb, but as a syntactically complex expression which adverbialism does not allow. And, moreover, we must employ the notion of an intentional object to consider whether the linguistic content expressed by the syntactically complex expression correctly or incorrectly describes the intentional content in question. Asking oneself how accurately the linguistic content of the ascription describes a subject’s intentional content requires us to appeal to what the subject takes his or her thought to be about, to appeal to the subject’s perspective on the world. Additionally, the notion of an intentional object on ofer allows us to treat the constituents of the thought in question as discrete units and to ask how well each discrete bit of linguistic meaning describes each bit of intentional content. Consider an illustrative example that is familiar from discussions of content externalism (See Burge 1979). (9) Alf believes he-has-arthritis-in-his-thigh-wise. Interpreters encountering this ascription for the irst time must not only decompose it to understand it, but they must also ask themselves, having decomposed the sentence into smaller bits, how well those bits express Alf’s concepts. Successful interpretation requires an interpreter to consider how well certain constituents of the ascription, such the linguistic content expressed by ‘arthritis’ and ‘thigh’, capture the subject’s corresponding intentional content. In this case, we have enough contextual detail to know that the linguistic content fails to capture the idiosyncratic concept or content that the subject associates with the term ‘arthritis’, because Alf surely does not believe that he has an ailment that cannot occur in his thigh (but only in his joints) in his thigh. So, the notion of an InquIry 503 intentional object is required so that an interpreter can determine if the linguistic content properly expresses what the subjects take their thoughts to be about. We can see that adverbialism faces two insurmountable problems when it comes to interpreting intentional content. First, interpretation of intentional content requires compositionality at the levels of both linguistic and intentional content, which is ruled out by adverbialism. Second, the question of whether the linguistic content of the ascription properly expresses the intentional content requires the notion of an intentional object. of course, as was discussed earlier, Kriegel understands intentional objects differently, but his understanding confuses intentional objects with the referents of thoughts. once we understand intentional objects properly, then we will see that the process of understanding the adverbialist intentional ascriptions relies tacitly on the act-object framework. The lack of compositionality of adverbialist paraphrases is not a new worry. Jackson (1977) raises a compositionality-based objection to adverbialism about sense perception. The novelty of the point I am making is that the lack of compositionality not only blocks natural inferences one would make between the contents of intentional states (Jackson’s point), but that the lack of compositionality also shows that if we truly dispensed with the act-object framework, we could not understand novel intentional ascriptions nor could we measure how well the linguistic content of ascriptions describes intentional content. Kriegel is well aware of Jackson’s objection and responds to it by appealing to the determinable–determinate relationship (Kriegel, 2007, 2008). His response is made to Jackson’s objection that the lack of compositionality in the adverbialist paraphrases prevents natural inferences. Here is a sketch of the basic dialectic between Jackson and Kriegel. Jackson is worried about simple inferences from thought contents being blocked by adverbialism. For example, if a subject believes Islamic-extremism-is-thegreatest-threat-to-future-generation-wise, we cannot infer that the subject thereby believes Islamic-extremism-wise because we cannot decompose the compound adverb into simpler parts, and such decomposition is required to make the very natural inference here. on this point, Kriegel concedes that we cannot make 504 C. WoodlIng such an inference just using the adverbial modiiers; he ofers a workaround, however. We can make the inference by appealing to the determinable–determinate relationship. The property of believing Islamic-extremism-is-the-greatest-threat-to-futuregeneration-wise is a determinate of the determinable property of believing Islamic-extremism-wise. once we understand this, we can infer that when someone believes Islamic-extremism-is-thegreatest-threat-to-future-generation-wise, they thereby believe Islamic-extremism-wise just as when we know an object is blue, we can thereby know it is coloured since being blue is a determinate property of the determinable property of being coloured. Applying this move to my critique, the reasoning might be as follows. We can understand novel adverbialist ascriptions by knowing the determinate and determinable properties they express. This allows us to break up the entire adverbialist property into more discrete units, and thus it appears that such ascriptions are compositional when we think in terms of the determinable properties expressed by the entire determinate adverbial property. This response fails because it treats adverbial modiiers as if they were syntactically complex, but they are not according to the adverbialist. Just as compositionality was required to understand novel ascriptions involving adverbial modiiers, it is also required to decompose the determinate property an adverbial modiier expresses into more discrete determinable properties. Consider ‘Islamic-extremism-is-the-greatest-threat-to-future-generationwise’. If it is syntactically simple, then we cannot know that it embeds more discrete properties such as the property expressed by ‘Islamic-extremism-wise’. And we must know this for the move to be successful. The only move here for Kriegel would be to say that it is possible to know the determinable adverbial properties (e.g. Islamic-extremism-wise) that make up the entire, determinate adverbial property (e.g. Islamic-extremism-is-the-greatest-threatto-future-Americans-wise) by having previously introspected the entire determinate property. not only does this method of introspection depend on the notion of an intentional object, but it also provides no help for novel ascriptions, because novel ascriptions will nearly always express intentional content that we have not previously thought of and thus not previously introspected. InquIry 505 In short, the determinate–determinable property move does not help because it requires adverbial modiiers to be compositional and they are not. And the method of introspection cannot help because it does not provide assistance in understanding novel ascriptions but merely for those we have already encountered and introspected. So, the talk of determinate and determinable adverbial properties here amounts to at most a novel way of speaking from within the act-object framework and not a piece of a suitable theoretical replacement. To review, there are two points at which adverbialism relies on the act-object framework. First, we cannot put ourselves in the introspective mental states needed to grasp the adverbial properties that modify intentional states without relying on intentional objects as they ought to be understood. Second, if adverbialism is true, we cannot grasp the intentional content of intentional states expressed by novel intentional ascriptions. It may appear that we can understand novel ascriptions, but this understanding tacitly relies on the act-object framework. I should note that the fact that we must irst work from within our normal conceptual scheme to understand the adverbialist paraphrases is in and of itself no problem. That is not the objection being made here; starting with one’s normal conceptual scheme is a necessary step in any revisionistic theory. To understand the revision of some portion of our common-sense conceptual scheme, we must start with that portion of the normal scheme and then attempt to understand how the profered, revisionary change will displace the aspect of the normal scheme up for revision. For example, to understand the brain-in-a-vat hypothesis from discussions of skepticism, one must move from the common-sense conception of the world to the conception in which one’s thought and experience is caused by a computer program that is stimulating one’s disembodied brain electronically. What we must be able to do is maintain the coherence of the revisionary conceptual scheme once we have thrown away, as it were, the scheme being replaced. And this we cannot do for adverbialism. notice that this is not a point one could make against expressivism. It is not as if I must rely on common sense (for which we are understanding cognitivism to be a proxy) to make sense of the expressivist paraphrases of ethical claims. Expressivism may 506 C. WoodlIng indeed be false, but it satisies the desideratum for revisionistic theories. I can understand the expressivist paraphrases of moral claims without implicitly falling back on cognitivism. The moral appears to be that some revisionary theories can truly displace the portion of the common-sense conceptual framework that they seek to and some cannot. IV. An ontologically conservative solution within the actobject framework At this point, adverbialists may make a concessive move by admitting that even though their theory requires the truth of the act-object framework, adverbialism is still needed as a modiication to the act-object framework in order to avoid metaphysically and ontologically extravagant views of intentional object. This response can be answered by sketching a metaphysical and ontologically conservative view of intentional objects from within the act-object framework that does not require us to construe intentional properties as adverbial properties. I shall sketch such a view, appealing to the work of Crane (2001a, 2001b, 2013). Crucial to Crane’s view is the idea that intentional objects are schematic objects and not substantial objects. Substantial objects are objects that have speciic metaphysical natures. Physical objects, for example, are objects that are concrete and bound by the laws of physics. Knowing an object is a physical object allows one to know certain things of its metaphysical nature. However, not all objects are like this. Some objects, including intentional objects, are schematic. Knowing an object is a schematic object tells us nothing of its metaphysical nature. For instance, knowing that an object is thought of by some subject tells us nothing about what kind of object it is (or that it exists for that matter). To help make this point, Crane compares intentional objects to the use of ‘object’ in grammar where ‘object’ has a schematic sense. We talk, for example, of direct objects of sentences. Knowing that a term is a direct object in a sentence tells us nothing about the nature of the referent of the term that functions as the direct object. direct objects can be terms that stand for activities, abstracta and individuals, among other things. The same is true of intentional objects. Being an intentional object does not mean that the object of one’s InquIry 507 thought is an object of a certain kind. Contrast this with substantial objects. As we discussed, the notion of a physical object is the notion of a substantial object. Knowing an object is a physical object tells us about its nature. All physical objects are concrete and bound by the laws of physics. In contrast, there is nothing we can tell about the nature of an object just by knowing it is an intentional object of some subject’s thought. Also, importantly, intentional objects are objects only for subjects on Crane’s view. We cannot reify them and talk about them as if they had their own existence outside of some subject taking his or her thought to be about the object. To talk of intentional objects without reference to subjects and their perspectives on the world is to commit the reiication fallacy, the fallacy of divorcing intentional objects from subjects and then asking for the metaphysical and ontological status of intentional objects. This fallacy is at the heart of much confusion in discussion of intentionality and non-existents. The general form of the fallacy is as follows. It seems obvious that we can think of objects that do not exist (objects we ourselves make up, established mythical or ictional objects, or objects we mistakenly believe to exist but do not). What is the nature of these objects in and of themselves? They don’t exist, but perhaps they subsist. Perhaps they are mental objects, or perhaps they are abstract. All such conjectures make two errors. First, as Crane points out, they assume there is a metaphysical property that ties together all intentional objects into a kind, but there is no such property. only substantial objects have such metaphysical properties that tie them together into a kind. Second, the fallacy results from divorcing the target of the thought, the intentional object, from the subject who has the thought. So, in addition to saying that intentional objects are schematic objects and not substantial ones, we should also think of intentional objects as what subjects take their thoughts to be about, as I suggested earlier. Thinking of intentional objects in this way allows us to make the crucial distinction between the intentional object of a thought and the referent of a thought. All intentional states are presented to their subjects as having objects but not all states refer to things in the world. This allows one to say, along with Kriegel, that the mere having of a thought does not instantiate any extensional relations with anything in the world. Kriegel is right that our thoughts link 508 C. WoodlIng up to the world in terms of whether or not they are true or false or satisied or unsatisied, not in term of instantiating extrinsic properties. Crane’s view of schematic intentional objects allows us to accommodate this insight within the act-object framework. Here is how. Having a thought requires that subjects take themselves to be directed on an object, but the object need not be something that is actually in the world. Intentional objects are what subjects take their thoughts to be about from their points of view, and subjects sometimes take their thoughts to be about things that do not exist—both wittingly and unwittingly—so all thoughts have intentional objects in virtue of being intentional even if some thoughts (namely, thoughts about non-existents) fail to refer.17 Kriegel’s mistake is to understand intentional objects as the referents of thoughts and not as the objects that subjects take their thoughts to be about. This is an understanding of intentional objects that prevents a suitable solution to the problem of intentional inexistence from within the act-object framework. However, once we understand intentional objects to be the objects subjects take their thoughts to be about and not just the referents of thoughts, then we can see our way to an ontologically and metaphysically conservative solution. To summarize this section so far, the three key ideas to a metaphysically and ontologically conservative solution based in the act-object framework are: (i) There is distinction between the object of a thought (its intentional object) and the referent of that thought. (ii) Intentional objects are schematic and not substantial. (iii) Intentional objects are what subjects take their thoughts to be about and should not be thought of as independent of the subject whose thought is directed on them. Equipped with these distinctions, we can now reconsider our initial inconsistent triad. 17 For explanations of how intentional content can be cashed out without intentional objects, see Gorman (2006); Searle (1983). Truth conditions or satisfaction conditions, the thought runs, are enough to express the intentional content of all intentional states, including those about non-existents. I agree that we cannot make sense of the content of propositional intentional states unless we understand what their truth/satisfaction conditions are. But, I argue, in Woodling (forthcoming) that we cannot properly articulate the truth/satisfaction conditions of intentional states without talking about intentional objects because intentional objects are needed to properly ix the subject’s perspective and to get the truth/satisfaction conditions right. InquIry 509 (a) one can represent non-existents. (b) one cannot bear a relation to non-existents. (c) representing something involves (constitutively) bearing a relation to it. Keeping track of (i)–(iii) allows us to see that (c) of the triad is false. In representing an object we need not bear an extensional relation to it. So, ‘relation’ in (c) does not have an extensional sense, while ‘relation’ in (b) does have an extensional sense. This is a distinction that may seem similar to one that Searle (1983) makes regarding ‘about’. unlike Searle, I do not think that ‘about’ is ambiguous in intentional contexts. To say that a subject’s thought is about X is to say that the subject takes his or her thought to be about X. If there is no actual object in the world to which the thought refers, that does not impact the thought’s intentionality. one may think that there is a sense in which it is about nothing, but that sense of ‘about’ is not the proper sense of ‘about’ since it fails to capture the intentionality of the thought. My claim, then, is not that ‘about’ is ambiguous. My claim is that ‘relation’ is ambiguous and gets used in two distinct senses in the inconsistent triad, so the claims are not truly inconsistent. Here is a way to reformulate (c). (a) one can represent non-existents. (b) one cannot bear a relation to non-existents. (c’) representing something involves it being the intentional object of one’s thought. From (a), (b) and (c’), we cannot derive a contradiction. It does follow that we do not represent non-existent objects by bearing relations to them but in virtue of them being the intentional objects of our thoughts, which is a representation relation that does not instantiate any extrinsic properties. let me say a bit more here that may be helpful about this representation relation by appealing to the notion of a personal concept. X being the intentional object of a subject’s thought can be translated into talk about the way a subject’s concepts represent their referents or purported referents. If I take my thought to be about Pegasus, then a key constituent of my thought is my concept of Pegasus. This concept represents the purported referent, Pegasus, in a certain way. This talk of personal concepts matches up with the notion of an intentional object that I have been defending. The notion of an intentional object on ofer is a phenomenological one, which is to say that it is ixed by how 510 C. WoodlIng thoughts seem to subjects. The notion of a personal concept is how a subject thinks of an object in an individual or idiosyncratic way, so the intentional object of one’s thought on my understanding will always line up with a subject’s personal concept, since the way the thought seems to the subject will fully capture the individual and idiosyncratic details of his concept.18 Provided we keep track of (i), (ii) and (iii), we have a suitable solution to the problem of intentional inexistence within the act-object framework. The virtues of Kriegel’s adverbialism are had by the view that intentional objects are schematic objects for subjects. And this view does not require that we give a metaphysics of intentional objects since they have no unifying metaphysical nature nor does it require us to say that all intentional objects exist as referents on some plane of existence merely in virtue of being thought of. There is always some object that an intentional state is directed on, but the thought need not refer to something in the world solely based on the fact that it has an intentional object, because the intentional object is not ixed by an extensional relation between the subject and object but it is ixed by what a subject takes his thought to be about. The content of intentional states depends solely on monadic or intrinsic properties on the current view just as in Kriegel’s. Because we do not construe intentional objects as the referents of thoughts, but as what subjects take their thoughts to be about, the intentional content of thoughts does not depend on the relational or extrinsic properties of subjects. In short, we can achieve the virtues of adverbialism from within the act-object framework, which is good because adverbialism cannot stand on its own independently of the act-object model of intentionality. Even if it cannot stand on its own, the adverbialist may still try to argue that his or her view is needed as a necessary modiication of the act-object framework. Provided that we are careful in understanding intentional objects, though, we do not need adverbialism even as a modiication of the act-object framework. There is a notion of intentional objects that is metaphysically and ontologically conservative so it achieves the virtues that adverbialism has in these regards. In sum, the idea that adverbialism is needed 18 I should note that this talk of personal concepts (and the related notion of what a subject takes his thought to be about) are not contrivances for solving the problem of intentional inexistence. Subjects’ intentional content is always ixed this way. A defence of this amounts to a defence of content internalism, for which I do not have the space. InquIry 511 to displace or amendment to the act-object framework to avoid implausible metaphysical and ontological commitments is false. V. Conclusion Adverbialism promises an ontologically and metaphysically conservative solution to the problem of intentional inexistence. However, it is not the only solution that its this bill. Because adverbialism fails to stand on its own as a model of intentionality, philosophers looking for an ontologically and metaphysically conservative view of intentionality should adopt the account sketched in Section IV. More can no doubt be said about how representation is to be cashed out in terms of intentional objects. For the purposes of this paper, however, the necessary points have been made. Adverbialism is not a viable revisionistic theory but at best a heuristic device that allows us to see important points about intentionality. The notion of an intentional object as what subjects take their thoughts to be about is indispensable for a proper theory of intentionality; if we abandon such a notion, we leave the subject's perspective out of the picture and thereby fail to capture properly the intentional content of individuals' thoughts. While intentional objects have caused problems before in philosophical discussions, these problems were caused by a failure to keep in mind the important distinctions discussed in Section IV (see i–iii). If we are careful about things, we shall see that we have the resources within the act-object framework to deal with the problem of intentional inexistence. Acknowledgements A previous version of this paper was presented at the 2014 meeting of the Florida Philosophical Association in Tampa, Florida. Members of the audience, especially John Biro and gene Witmer, ofered comments that positively shaped my thinking about adverbialism and intentionality. An anonymous referee for this journal also deserves my thanks for perceptive feedback. Disclosure statement no potential conlict of interest was reported by the author. References Burge, Tyler. 1979. “Individualism and the Mental.” Midwest Studies in Philosophy 4: 73–121. Crane, Tim. 2001. Elements of Mind. new york: oxford uP. 512 C. WoodlIng Crane, Tim. 2001. “Intentional objects.” Ratio 14: 336–349. Crane, Tim. 2013. The Objects of Thought. new york: oxford uP. gorman, Michael. 2006. “Talking about Intentional objects.” Dialectica 60: 135–144. Jackson, Frank. 1977. Perception: A Representative Theory. Cambridge: Cambridge uP. Kriegel, uriah. 2007. “Intentional Inexistence and Phenomenal Intentionality.” Philosophical Perspectives 21: 307–340. Kriegel, uriah. 2008. “The dispensability of (Merely) Intentional objects.” Philosophical Studies 141: 79–95. Kriegel, uriah. 2011. The Sources of Intentionality. new york: oxford uP. Searle, John. 1983. Intentionality. Cambridge: Cambridge uP. Strawson, Peter. 1959. Individuals. new york: routledge. Woodling, Casey. Forthcoming. “The Indispensability and Irreducibility of Intentional objects.” The Journal of Philosophical Research.