Is so-called Phenomenal Intentionality Real Intentionality?

Abstract

This paper addresses the title question and provides an argument for the conclusion that so-called phenomenal intentionality, in both its relational and non-relational construals, cannot be identified with intentionality meant as the property for a mental state to be about something. A main premise of the argument presented in support of that conclusion is that a necessary requirement for a property to be identified with intentionality is that it satisfy the features taken to be definitory of it, namely: the possible non-existence of the intentional object (the fact that an intentional state may be directed towards something that does not exist) and aspectuality (the fact that what is intended is always intended in some way, under some specific aspect, from a particular perspective). By taking this premise on board, I attempt to show that phenomenal intentionality cannot be identified with intentionality because, appearances notwithstanding, it ultimately satisfies neither of the two above mentioned features.

1 Introduction

My main aim in this paper is to provide an answer to the title question—an answer that, as it turns out, will be negative. The argumentative path I follow to reach this conclusion can be summarized in the following steps: (i) a necessary requirement for a property to be identified with intentionality is that it satisfy the features taken to be definitory of it; (ii) there is a general consensus that two such features are the possible non-existence of the intentional object (the fact that an intentional state may be directed towards something that does not exist) and aspectuality (the fact that what is intended is always intended in some way, under some specific aspect, from a particular perspective); (iii) appearances notwithstanding, so-called phenomenal intentionality ultimately satisfies neither of the above mentioned features; (iv) therefore, phenomenal intentionality cannot be identified with intentionality.¹

In my paper I shall take as my critical target what in my view counts as the core thesis behind the phenomenal intentionality theory. Such a thesis can actually be seen as the conjunction of two sub-theses, namely:

• (i) There is an experiential property that is determined by phenomenality, and that we tend to describe using terms like ‘of’ or ‘about’, and

• (ii) Such an experiential property is intentionality (or at least the basic form of it).

My criticism concerns (ii), not (i). I do agree with the advocates of the phenomenal intentionality theory that what they label ‘phenomenal intentionality’ is a real property having an experiential nature that makes it introspectively accessible. What I contest is that such a property is to be identified with intentionality, notwithstanding the fact that we tend to describe it using terms like ‘of’ or ‘about’. My reason for so claiming is that in my view such a property does not ultimately satisfy what everyone (and the advocates of the phenomenal intentionality theory are no exception in this regard) takes as the defining features of intentionality.

The structure of the paper is the following: Sect. 2 provides an overview of the main theses to which the advocates of the phenomenal intentionality theory subscribe, as regards both intentionality and its relationship with phenomenal consciousness. The next two sections present my criticisms of the idea that so-called phenomenal intentionality can be equated with intentionality. The criticism is presented with regard to versions of the theory that endorse a relational analysis of the metaphysical structure of the property of intentionality (Sect. 3) as well as to versions that opt for a non-relational analysis (Sect. 4).

2 The core idea behind the ‘Phenomenal intentionality research program’

That phenomenal intentionality exists is the core thesis of a research program within analytic philosophy of mind that has recently taken the stage as a radical theoretical alternative to the allegedly competing research program that (starting from the ‘70s with the works of Fodor, Dretske and Millikan among others) dominated the philosophical landscape in this area up to the last decades of the previous century.² For brevity’s sake, in what follows I shall use ‘PIT’ and ‘PITers’ to refer to the thesis of the existence of phenomenal intentionality and its proponents respectively.³ PIT is a thesis about the source of intentionality. What PITers claim is that intentionality—by which they intend to refer to the property primarily possessed by mental states to be about something,⁴ or to have a content that makes the state semantically evaluable as true/false, satisfied/not satisfied, accurate/inaccurate—has its source in phenomenality or, more precisely, in the phenomenal character of conscious states which is, in turn, characterized as that feature which accounts for their what-it-is-likeness.⁵

In conferring pride of place to phenomenality in a theoretical account of intentionality, PITers take a radically critical stance both towards those approaches that deny any kind of entailment between intentionality and phenomenality, and towards those other approaches that, while acknowledging some kind of dependence between the two properties, endorse the claim that intentionality takes priority over phenomenality. Thus, as far as the relationship between the two properties is concerned, PITers not only oppose the ‘separatist position’ (the so-called two-separate-realms conception of the mind that conceived the mind as an inhomogeneous field devoid of any unifying features),⁶ but also the ‘reductive representationalist position’, which, while acknowledging that phenomenality cannot be separated from intentionality, claims that the latter is independent (both metaphysically and explanatorily) of the former.⁷ Regardless of how different these two positions are from each other, they share the assumption that phenomenality is irrelevant to intentionality. Such an assumption, according to PITers, albeit supposedly functional to the naturalization project, cannot but generate a radically misleading account, one that leaves something essential for intentionality out of the picture. Intentionality, they claim, cannot be accounted for merely in terms of externalist-tracking relations between mind and world, because there is a subjective aspect to it—an aspect they cash out in phenomenological terms—that can be accounted for only by adopting a first-person perspective which, for principled reasons, cannot be captured within the third-personal, objective methodology that standard tracking accounts adopt. This point is important in clarifying the exact import of the criticism that advocates of the Phenomenal intentionality research program raise against the Naturalist externalist research program. What the former claim is not that the latter provides an incomplete account of intentionality, an account that should be supplemented by a parallel account of the subjective/phenomenal aspects of intentionality. Rather, they claim that such an account is wrong because no intentionality can actually arise from mere externalist-tracking relations. In their view, the real source of intentionality is phenomenality. It is the latter that provides the ‘spark’ from which intentionality arises.⁸

It should be stressed that the Phenomenal intentionality research program is not a monolithic research project, but rather one that comes in many varieties.⁹ In my paper I shall take as my critical target what in my view counts as the core thesis behind the phenomenal intentionality theory. Such a thesis can actually be seen as the conjunction of two sub-theses, namely:

• (i) There is an experiential property that is determined by phenomenality, and that we tend to describe using terms like ‘of’ or ‘about’, and

• (ii) Such an experiential property is intentionality (or at least the basic form of it).

My criticism concerns (ii), not (i).¹⁰ I do agree with PITers that what they label ‘phenomenal intentionality’ is a real property having an experiential nature that makes it introspectively accessible. What I contest is that such a property is to be identified with intentionality, notwithstanding the fact that we tend to describe it using terms like ‘of’ or ‘about’.¹¹ My reason for so claiming is that in my view such a property does not ultimately satisfy what everyone (and PITers, or so it seems to me, are no exception in this regard) takes as the defining features of intentionality.

Before I present my argument, some clarifications are in order. In my critical discussion I shall confine my attention to intentionality meant as the property of mental states of being about something. In Brentano’s famous formulation,¹² this characterization is coupled with another that Brentano presents as equivalent, according to which intentionality is the property of mental states (or of psychic phenomena, to use his terminology) to have a (truth-evaluable) content. After Brentano, these two characterizations have generally been distinguished. In the now well-known terminology introduced by Kim (1996), they are labelled respectively ‘reference intentionality’ and ‘content intentionality’. Along with many others in the intentionality debate (see, e.g. Crane 2013), I shall here take the former to be the most basic one in the following sense: a mental state can be referentially intentional without possessing a propositional content (one can desire something, an ice cream say, without desiring that the ice cream be stored in the fridge, for example), yet no mental state can possess content intentionality without also possessing reference intentionality (one cannot think that so-and-so is the case without also thinking about something to the effect that it is thus-and-so). I shall therefore assume that a necessary requirement for a mental state to be intentional is that it possess reference intentionality.¹³ Accordingly, I shall address the question as to whether so-called phenomenal intentionality is identical with intentionality by considering whether it is identical with reference intentionality.¹⁴¹⁵

That being said, I shall proceed as follows: No property can qualify as the property of intentionality unless it satisfies its defining features. As far as intentionality is concerned, two such features are the possible non-existence of the intentional object and aspectuality.¹⁶ Let me illustrate them briefly.¹⁷ The first feature concerns the fact that for a state to be about something it is not required that there actually exist an object which the state is about.¹⁸ In this regard, intentionality differs radically from any ordinary relational property. Think for example of the relational property of kicking a ball. As a matter of fact, one cannot instantiate such a property unless there exists an actual, concrete ball one is kicking. Not so as far as intentionality is concerned. Consider for example of Le Verrier’s thinking of Vulcan (the supposed planet orbiting around Mercury and the Sun). Although Vulcan does not exist, Le Verrier’s thought exists and it instantiates the property of being about something (a putative object).¹⁹

Let us now consider the feature of aspectuality. What this feature is meant to capture is the fact that the entity a given intentional state is directed at is always intended/presented in a particular way. Hammurabi’s thoughts that Hesperus is bright and that Phosphorus is bright are both of one and the same object, the planet Venus, and yet they are different thoughts. What makes them different is precisely the aspect under which Venus is intended/presented in the two cases. This feature, aspectuality, distinguishes intentionality from any ordinary relation one can stand in to objects. As far as any ordinary relation R is concerned, if a subject S stands in relation R to an object O, his/her standing in such a relation is independent of the way in which the object is intended/presented. If S kisses Whoopi Goldberg, he/she therefore kisses Caryn Elaine Johnson, that is, the famous actress who played the role of Sister Mary Clarence in Sister Act, no matter how Whoopi is intended in the two cases. Not so with intentionality.

Well, does so-called phenomenal intentionality satisfy these two features? In order to start addressing this question, we first have to pin down the property in question. Let us thus consider what PITers are talking about when they talk about phenomenal intentionality. According to them, the property in question is an experiential property that we tend to describe using terms like ‘of’ or ‘about’, and that we can literally encounter in our mental life by directing our attention inward. This point is very clear in Mendelovici, who takes phenomenal intentionality to be a feature that, in paradigmatic cases (a present perceptual experience, a thought we are currently having), we “notice introspectively in ourselves and are tempted to describe using representational terms like ‘of’ or ‘about’” (Mendelovici 2018, p. 4).²⁰ Analogously, Kriegel (2011, p. 46) claims that phenomenal intentionality (or experiential intentionality as he labels it) is simply introspectively manifest. Well, the question is: is it true that whenever we entertain a paradigmatically conscious intentional mental state there is an experiential property we can introspectively notice that we are tempted to describe using terms like ‘of’ or ‘about’? Moreover, and more importantly, are PITers right in claiming that such a property is identical with intentionality?

As a start, it is hardly disputable that one cannot enjoy a conscious intentional mental state without being conscious of/being aware of something. What one is conscious of/aware of is presented to one. Thus, there is undoubtedly a presentational element in any conscious intentional mental state and this element is something that we experience. But this is true not only of paradigmatic intentional mental states, such as occurrent thoughts and perceptions, but also of other conscious mental states, such as emotions, bodily feelings and moods, whose intentionalist treatment can be disputed. Whenever one undergoes a conscious mental state, one is not only conscious, but also conscious of something (no matter how incomplete, partial, obscure, confused such an awareness may be). Is this introspectively manifest fact sufficient to establish PIT? Well, even granting that it suffices to show the plausibility of PIT first sub-thesis, it is hardly sufficient to establish the second sub-thesis. ²¹ What is required to that end is to assess whether the ‘ofness’ in question—let us label it ‘phenomenal ofness’ (PO),²² the property of being (experientially) presented with something and that we tend to describe by using terms like ‘of’ or ‘about’—is the ‘ofness’ of intentionality, that is: the mind’s intentional direction upon something. ²³ I can find no better way of handling this question than to consider whether PO possesses the two main defining features of intentionality, namely: the possible non-existence of the intentional object, and aspectuality. I shall address this issue by considering two possible accounts of the metaphysical structure of PO: the relational and the non-relational. According to the former, a mental state’s property of being of something is a two-place relation, that is, something that involves two relata, one of which is the intentional object the mental state is directed at. In the latter rendering, being of something does not (constitutively) involve bearing some relation to something; rather, it consists in instantiating some kind of monadic property.

3 The relational account of phenomenal ofness

Let me start by providing a preview of the main points in this section. After a brief overview of the relational account, I shall introduce what I regard as a mandatory constraint of explanatory adequacy that the adoption of the thesis of the phenomenal nature of intentionality imposes on PITers. I shall then consider whether there are (or could be) relational accounts able to comply with the requirements the constraint imposes. I shall address this issue by taking into account three possible varieties of relationalism which differ from each other mainly as regards the way the intentional object is conceived. I shall claim that two of those varieties are unable to conform to the requirements the explanatory adequacy constraint imposes. I shall then present what I take to be the only relational account of PO that could fit the bill, namely the account that takes intentional objects to be items having a phenomenal nature. Finally, I shall consider whether PO so construed satisfies the two defining features of intentionality. The verdict will be negative. On this ground I shall conclude that the prospect of providing a relational account of PO that satisfies both the constraint imposed by the thesis of the phenomenal nature of intentionality and the two defining features of intentionality seems doomed to fail.

According to the relational account of intentionality, to entertain an intentional mental state is for its subject or for the mental state she is in to stand in some kind of relation with something. That intentionality has (at least) a dyadic structure is considered by most people the common-sense view, and the one that best conforms to our pre-theoretic intuitions as regards the nature of intentionality. Such a view, its advocates claim, seems to be confirmed not only by our introspective access to our own mental states, but also by the kind of verbal reports we provide of them, and by the kind of logical inferences that those reports license. As regards the first point, it seems undisputable, they say, that our occurrent intentional mental states strike us as relational: our thoughts introspectively present themselves to us as relations to what we are thinking about. Moreover, the kind of language we ordinarily use to report our intentional mental states exhibits a relational structure: we normally report them by using nouns and adjectives (such as: ‘I am seeing a red apple’, ‘I am thinking of my cat Virgola’), rather than adverbs, for example, and such reports license the kind of logical inferences we frequently draw (such as: there is something I am seeing/thinking about). It is true that it is, in principle, possible to gloss such ascriptions by using non-relational phrases, but such renderings, the relationist observes, strike us as artificial and unfaithful to the phenomenology of our intentional experiences.

Of course, these considerations are not by themselves sufficient to demonstrate that intentionality is relational. As regards the first point, one could claim that introspection, far from revealing to us the true nature of our intentional mental states, systematically misleads us by presenting as relational what is in fact a monadic property. Along these lines, an opponent of the relational account could object that the alleged relationality that according to the relationist we experience when we entertain an intentional mental state (say, when thinking about x), is just an effect of the way in which we inevitably describe our thinking when we reflect on what any such thinking is and that therefore introspectively-based considerations demonstrate nothing. Moreover, as to the second point, one could object that the relational form of the linguistic reports we make of our mental states is wholly compatible with those states’ having a monadic structure, because the way we speak is no sure guide to the metaphysical structure of reality. Be that as it may, the fact remains that the relational account is considered by most scholars as the one that best captures our pre-theoretic intuitions as regards the structure of intentionality.

That being said, it is worth stressing that such an account, while being the prevailing one outside the PIT camp, does not find many supporters within it.²⁴ Most PITers actually endorse a non-relational account, because in their view it possesses all the virtues of the relational one, minus its defects.²⁵ A major problem often raised against relationalism concerns its ontological commitment towards entities we apparently have no independent reason to think exist, apart from the fact that they are needed to cover the intentional object role in cases in which the intentional state does not refer to anything existent. Accounting for such cases is actually one of the thorniest problem for all variants of relationalism. It is true that its advocates can avail themselves of a plethora of possible solutions,²⁶ but the fact remains that they cannot avoid crowding the ontology with ‘exotic’ (Sainsbury 2010) or ‘peculiar’ (Schellenberg 2011) entities whose postulation appears to be ad hoc and not independently motivated.

If the ontological problem afflicts relationalists of any stripe, the way it afflicts those who also defend PIT is even more severe. This is so, in my view, because the phenomenological account endorsed (in particular, the claim that intentionality is an experiential, phenomenal property) imposes a constraint of explanatory adequacy that narrows down considerably the logical space of the available options. What such a constraint demands is that the account provided of the property of intentionality conform as closely as possible with the way its instantiation in a subject’s occurrent mental state strikes her, from her first-person point of view. Well, how does the instantiation of the property of intentionality strike us from the first-person point of view? First of all, it strikes us as a property that relates us to items that present themselves to us as mind-independent and, in many cases, as external to our mind and concrete. Moreover, what it seems that we are being presented with remains constant in all phenomenally indistinguishable states such as for example a veridical perception of a yellow lemon and a hallucination as of a yellow lemon qualitatively identical to the real lemon one is seeing.²⁷ On the ground of these considerations we can lay down the two following phenomenological requirements that the explanatory adequacy constraint imposes upon PIT²⁸:

• (a) conformity requirement (which prescribes an account capable of capturing the way the intentional object presents itself to the subject, namely: as mind-independent, and in many cases as external and concrete)²⁹;

• (b) uniformity requirement (which prescribes a uniform treatment for all cases that present phenomenal indistinguishability).

Let us take these two requirements on board and consider whether there are (or there could be) relational accounts of (reference) intentionality within the PIT camp suitable to comply with them.³⁰ Three main possible varieties of the relational account are distinguishable on the basis of the way in which the nature of the intentional object (i.e. the right-hand side relatum of the intentional relation) is conceived, namely: (i) concretism (the intentional object is a concrete item), (ii) abstractivism (the intentional object is an abstract item), (iii) phenomenism (the intentional object is an item having a phenomenal nature). I shall now consider them in this order.

Concretism is the view according to which what we are related to in our intentional mental states are ordinary concrete items, that is mind-independent entities endowed with a spatio-temporal dimension. Notwithstanding its initial appeal, concretism fails as a general theory of the nature of the intentional object, because, quite simply, very few of the ‘entities’ our intentional mental states are directed at actually consist of ordinary concreta. This is certainly so as far as mental states intentionally directed towards non-existent entities (putative entities that do not exist) are concerned: whatever the intentional object of Le Verrier’ thought that Vulcan is a planet, say, turns out to be, it certainly cannot be equated with an ordinary concretum. A concretist could handle the problem raised by mental states intentionally directed towards non-existent entities, and more generally by mental states directed towards non-concrete entities, by narrowing down the application domain of her proposal, and then providing a different treatment to all the outlying cases. Although a viable solution, it is not one that a relationalist who endorses PIT could avail herself of—not, at least, if she cares (as she should) to comply with the requirements that the constraint of explanatory adequacy imposes on her account of intentionality. Such a move would in fact violate the uniformity requirement which, as already stated, prescribes a uniform treatment for all the cases that present phenomenal indistinguishability. As a case in point let us consider a veridical perceptual experience of a yellow lemon and a phenomenally indistinguishable hallucination. Although the two experiences differ radically as regards the fact that the experiencing subject turns out, in the former case but not the latter, to be in causal contact with a real, concrete yellow lemon, there is no difference from a phenomenological point of view in their what-it-is-likeness. In both cases, it seems to the subject as if there is a real, concrete yellow lemon in front of her. Compliance with the uniformity requirement demands that one treat the two cases in the same way and, in so far as one cannot say that in the hallucinatory case one is related with an ordinary concrete entity, the same must be said as regards the veridical case. In order to comply with such a requirement, the relationalist must opt for a relatum that is available regardless of whether there actually is a real, concrete entity that the subject happens to be related to.³¹ A possible solution is to turn to abstract entities. Let us thus consider the second variety of relationalism, which actually represents the favoured position among advocates of the relational account of PIT.³²

According to abstractivism, a phenomenal intentional state of thinking or, more generally, experiencing something consists in standing in a relation to some abstract entity whose existence is not tied to that of the mental state.³³ As the advocates of abstractivism claim (see e.g. Bourget 2020), such a relation can, in turn, be understood in different ways: it can be considered for example as a similarity relation, or as an irreducible mental relation such as an awareness, an acquaintance or a grasping relation. Moreover, it can be taken to be the same across all phenomenal intentional states, or to vary from one kind of state to the other.

Notwithstanding its undisputable capacity to provide a treatment of the nature of the intentional object able to cover also cases of states directed towards apparently non-existent entities, in full conformity with the uniformity requirement, abstractivism has met with several criticisms both inside and outside the PIT camp. One can group them into four main types on the basis of the nature of the problem raised, namely: ontological, explanatory, epistemological and phenomenological. Let us consider them briefly, in this order.

As regards the ontological problem, it must be acknowledged that while the entities that abstractivism commits itself to are undoubtedly less ‘exotic’ and ‘spooky’ than those postulated by other kinds of relationalism, it nonetheless remains that even abstractivism commits itself to a rather cumbersome ontology.³⁴

The problem that the second kind of criticism raises is how abstracta could play some role (either causal or constitutive) in the explanations of our occurrent conscious mental states given that the latter are concrete phenomena. According to Kriegel’s articulation of this problem, to adopt one such kind of explanatory strategy seems to contravene the widespread accepted principle of the explanatory closure of the realm of concreta according to which every aspect of the domain of concreta can be fully explained by appealing to what goes on inside that realm.³⁵

A third kind of criticism has to do with epistemology. The worry raised in this connection is that if abstractivism is endorsed, then unacceptable epistemological consequences ensue. This criticism has been explicitly developed by Kriegel (2011), and although it specifically targets a particular variant of this position, namely Platonic abstractivism, it can actually be extended to cover all the other variants as well.³⁶ The problem that Kriegel raises is that abstractivism implies a model of the epistemic relation between perceptual experiences and perceptual beliefs that is utterly wrong when applied to our knowledge of ordinary physical objects. According to such a model, our perceptual beliefs about such objects could never be immediately justified on the ground of our perceptual experiences, by directly ‘endorsing’ them (as would be demanded by a correct model). For that would require that our perceptual experiences make us directly aware of concrete ordinary objects, whereas, according to abstractivism, what our perceptual experiences are immediately directed at are abstract entities.³⁷

Let me now move to the phenomenological objection. I consider this objection the most serious one for an advocate of PIT, because, as I have already stressed, such an advocate should comply as far as possible with the requirements of phenomenological adequacy. And such requirements demand, among other things, that the account provided of the property of intentionality conform as closely as possible with the way its instantiation in a subject’s occurrent mental state strikes its subject from her first-person point of view. Considering the case of thoughts about non-existent entities (i.e. putative entities that do not exist), Kriegel (2007, p. 311) observes that “phenomenologically, the entities we are aware of when we think of dragons and parrots present themselves to us, from the first-person perspective, as external concreta, not as abstracta or mental concreta”. The severity of this objection reveals its true force as soon as one applies it beyond the restricted domain of states apparently directed towards non-existent entities, and considers ordinary cases of veridical perceptual experience.³⁸ As has already been stressed, a relationalist PITer who wants to conform to the uniformity requirement cannot say that in such cases the relatum of the intentional state is an ordinary concrete object. Rather, she has to say that such a relatum is always an abstract entity, and this seems to contravene the conformity requirement.³⁹ Bourget (2020) discusses this objection and concludes that in his view there is a sense in which what we are aware of when we introspect our current visual experiences is a way the world could be (a possible state of affairs), and that therefore it does not seem implausible to say, in line with abstractivism, that what we are aware of are actually abstract entities. In my view, such a proposal could sound plausible if it also explained how an awareness of something abstract can nonetheless present itself to us as of concreta (and, in particular, as of concrete particular entities). Actually, the fact that our awareness has this feature, at least in many cases (in particular, and paradigmatically, in the case of ordinary perceptual experiences), is a phenomenological datum that any advocate of PIT should explain, and it seems that abstractivism has yet to properly address it. Thus, all in all, as things currently stand, although abstractivism can satisfy the uniformity requirement, it does not satisfy the remaining requirement. Moreover, it is far from obvious that any future refined versions of it could satisfy that requirement without betraying the spirit of the abstractivist position.

As it turns out there exists a sort of tension between the two phenomenological requirements. For, on the one hand, the uniformity requirement imposes compliance with the claim that what we are aware of can never be concrete, external entities. On the other hand, the conformity requirement demands an account of the introspective datum that our experiences present themselves to us in many cases as of concrete, external entities. The challenge facing the relational account of PO lies in explaining how these two aspects could be conjunctly satisfied. Is there no way that a relational account could comply with the two requirements and yet not commit itself to the idea that what we are aware of in entertaining our occurrent intentional mental states are concrete, external entities? In short, could something different from a concrete, external entity present itself to us in our introspection as a concrete, external entity after all?

I think that a possible positive answer to such a question is available, and in the remaining part of this section I shall present and discuss what I take to be the only relational version of PIT able to comply with the phenomenological requirements. After having introduced it, I shall try to assess whether so called phenomenal intentionality, construed along the lines of that relationalist version, can be equated with intentionality by considering whether it satisfies its two defining features.

Let us now consider what kind of relatum, different from both an ordinary, concrete, external entity and an abstract entity, could be congenial to the relationalist. The kind of relatum I want to consider can be characterized as the something a subject is phenomenally presented with whenever she is entertaining a conscious intentional state. No matter whether the state in question is veridical, illusory or even hallucinatory, a subject is always presented with something. Were this not so, it would make no sense to say that the subject is having a conscious experience. Whenever a subject has a conscious experience there is always something that presents itself to her. What thus presents itself is what the subject takes her experience to be of, namely what the experience seems to be of from the subject’s point of view.⁴⁰ On this ground I consider it appropriate to qualify such a relatum by using the label ‘phenomenal entity’.⁴¹ The position I am envisaging is therefore one according to which intentional objects are items having a phenomenal nature. Let me try to expand on this point in order to better fix our grip on the notion. To begin with, let me say that a phenomenal entity must be something able to phenomenally manifest itself to a subject. This, however, does not seem sufficient to qualify something as a phenomenal entity. For even an ordinary, worldly object satisfies this requirement, and yet it would be inappropriate to qualify it as a phenomenal entity (without adopting some radically subjective form of idealism). I deem it definitory of a phenomenal entity not only that it (phenomenally) manifest itself/appear to someone, but also that it be a kind of entity that coincides with its manifestation/appearance, that is, a kind of entity whose being and nature are wholly exhausted by its appearance. As for its being, a necessary and sufficient condition in order for a phenomenal entity to exist is that it appear (to someone, at a given time): a phenomenal entity exists only if it appears and if it appears to someone. Ditto for its nature. There is no distinction between appearance and reality as far as phenomenal entities are concerned: the reality of a phenomenal entity coincides with its appearance (as is the case, by the way, with anything having a phenomenal nature). Whereas an ordinary object can appear in different ways without these differences in its ways of appearing affecting its identity (the coin appearing now round and then elliptical in consequence of my changing position relative to it remains one and the same coin), in the case of phenomenal entities any change in ways of appearing amounts to a change of identity, a change in what appears. As it will turn out, both aspects of phenomenal entities (that concerning their being and that concerning their nature) are crucial in assessing the claim as to whether phenomenal ofness (PO) satisfies the two defining features of intentionality.

Before considering this point, let me address a question I have thus far left unanswered, namely: in what sense can phenomenism be taken to satisfy the phenomenological requirements—or, at a minimum, to satisfy them better than any other proposals within the relational PIT camp? The uniformity requirement is of course satisfied in so far as the suggested relatum, a phenomenal item, is available whether or not the subject is actually related with a real, concrete object in the world. Both in a case of veridical perception and in one of a phenomenally indistinguishable hallucination, the subject is confronted with qualitatively identical kinds of objects of awareness.⁴² But what about the other requirement? Isn’t it true that phenomenal entities, qua entities having a phenomenal nature, are neither mind nor subject-independent? My answer is that even though this is true, what the conformity requirement actually demands is not that the relatum be subject/mind-independent, but that it be a kind of item that presents itself as such. As things stand, phenomenal entities do so present themselves, and therefore satisfy the requirement in a way no other item does.

What we have to assess at this point is whether PO understood as a relation whose right-hand side relatum is a phenomenal entity can be identified with intentionality, that is, with the property whose defining features are the possible non-existence of the intentional object and aspectuality. Let us start from the first feature and consider whether it is possible for a given mental state to be directed towards something that does not exist, under the assumption that the kind of entities that play the relatum role are phenomenal entities. In other words, let us consider whether a phenomenal entity could fail to exist if the state directed towards it exists. In principle, a necessary condition for this possibility to obtain is that the relatum does not present any kind of ontological dependence on the state that is directed upon it. When the relata are taken to be either concrete or abstract entities this condition is satisfied, because the existence of such entities, being mind-independent, is not necessitated by the existence of the mental state directed upon them. But how do things stand as regards phenomenal entities? Of course, any existing phenomenal entity could have failed to exist if the state targeting it had not been accomplished. This is obviously a possibility in so far as phenomenal entities are not necessary entities. But if the state exists, then so does the phenomenal entity that the state targets, because the latter ontologically depends on the former and, as a consequence, its existence is necessitated by the existence of the state. It therefore follows that the above stated necessary condition cannot be satisfied as far as phenomenal entities are concerned. The conclusion to be drawn is that PO understood as a relation whose right-hand side relatum is a phenomenal entity cannot be taken to satisfy the possible non-existence feature.

Let us now move to aspectuality and consider whether PO, conceived along the previously suggested lines, presents this other defining feature of intentionality. I shall start by saying that in order for PO to present this feature it should be possible for two conscious intentional mental states of a subject to be of the same phenomenal entity while presenting it in different ways, so as to make it possible for her not to realize that it is one and the same entity that is being presented in different ways.⁴³ That is: it should be possible for a Frege case to arise. But can a Frege case arise as far as PO, so conceived, is concerned? If what I have said above concerning the being and nature of phenomenal entities is right, I think that the correct answer is negative: if the ways of presentation are different, what the experience is phenomenally of is different as well and, as a consequence, no Frege case can arise in this area. In other words: if you change the way in which the presentatum (e.g. a phenomenal entity) is presented, you change the presentatum itself. This is so because, in this case, a change in ways of presentation amounts to a change in ways of appearing—and a change in ways of appearing, in this case, is a change in what phenomenally appears.⁴⁴ We can sum up this point by saying that as far as phenomenal entities are concerned, there is no distinction between the ‘how’ and the ‘what’, because the two coincide (simply because the reality of a phenomenal entity—what it ultimately is -, coincides with its appearance—with the way in which it phenomenally appears). But without this distinction, no Frege case can arise. My rationale for this claim hinges on the metaphysical nature of phenomenal entities: a phenomenal entity has its own mode of presentation (i.e. its way of appearing) ‘built into it’, in the sense that it is individuated by the very way in which it presents/manifests itself to the subject’s awareness. It is, so to say, an entity-cum-perspective, i.e. a perspectival entity, and a perspectival entity cannot be given from different perspectives because, by its very nature, it is not independent from a perspective. In particular, it is dependent on the perspective that, by individuating it, makes it the very entity that it is. I therefore conclude that in so far as relational PO, in what amounts to its most phenomenologically plausible version, does not satisfy either of the two defining features of intentionality, it cannot be taken to be identical with it.⁴⁵

One could reject this conclusion and claim that the relationalist could actually account for the two features by resorting to the notion of reference taken as an extrinsic relation different from the relation of phenomenal ofness. By mobilizing that notion, the objection goes, the relationalist could say that all that is required in order for the two features to be satisfied is (i) (as for the feature of the possible non-existence of the intentional object) that a mental state directed towards a phenomenal entity may fail to refer (because there is no entity in the world that corresponds to/matches what the subject takes her state to be about) and (ii) (as for aspectuality) that two mental states directed towards two distinct phenomenal entities may actually refer to one and the same item. Is this a way out of the problems we have raised, or rather an acknowledgment of the impossibility of accounting for the two features as features of intentionality within a framework that takes it to be identical with the property of phenomenal ofness? I shall postpone my critical assessment of the manoeuvre underlying this objection to the next section, where I shall discuss a variant of it made by the non-relationalists.

4 The non-relational account of phenomenal ofness

The non-relational account of intentionality is the prevailing position within the PIT camp.⁴⁶ According to it, intentionality is a property whose instantiation does not require any relatum playing the role of that-which-one-thinks-about. There are actually different versions of non-relationalism. Bourget (2020) uses the label ‘aspect view’ to group them under a single heading. Within this view, as the name suggests, that-which-one-thinks-about (an object, a property or a combination thereof) is taken to be an aspect of the mental state itself, where the aspect is something that does not exist separately and distinctly from our representation of it.⁴⁷ It is important to stress that the non-relational account (just as the relational one) aims at providing an explanation of the nature of intentionality, i.e. an answer to the question of what intentionality (or at least basic intentionality) really is, au fond. The nature of intentionality, according to the aspect view, turns out to be very different from how the common-sense view takes it to be; in particular, it turns out not to involve constitutively any relation with intentional objects. Of course, in so far as non-relationalism dispenses with intentional objects, it does not run into the ontological problems that afflict the rival account. For there is no need within the non-relational framework to introduce ‘peculiar’ or ‘exotic’ entities that play the relatum role in cases where the state is (putatively) directed towards something that does not exist. Regardless of whether what one thinks about exists or not, to think about something is always for a given state to have an intrinsic aspect, one whose existence is independent of the existence of anything outside the thinking subject’s head. Apart from the obvious advantage of avoiding (or keeping to the minimum) cumbersome ontological and metaphysical commitments, non-relationalism is claimed to present further important advantages as compared with the relational account. First and foremost, it is claimed to account for what is called the ‘psychological involvement’ of the contents of our mental states, i.e. the fact that what we think about shows up in our mental life as something that we entertain, and that enjoys a psycho-phenomenological reality that we can introspectively access.⁴⁸ Conversely, according to advocates of the aspect view, no relational account seems able to explain this crucial feature of mental content.⁴⁹ Granting that these are real virtues of this position, it is important for us to consider here, firstly, whether the non-relational account of PO satisfies the phenomenological adequacy requirements (which, as we have said, any account within PIT must satisfy) any better than the rival relational account and, secondly and more importantly, whether the non-relationalist explanation of PO allows the two defining features of intentionality to be ascribed to it.

In what follows I shall address these two issues by focusing on Kriegel’s position, which amounts to a form of adverbialism in the theory of intentionality patterned on the account originally put forward for sense perception by people like Ducasse and Chisholm.⁵⁰ According to adverbialism, ‘thinking about something’ (say, a lemon), for S, has to be analyzed as ‘S is thinking lemon-wise’ (or S is thinking lemon-ly), where, for S to think lemon-wise/lemon-ly is for S to engage in the activity of thinking in a certain way.⁵¹

Well, what about the phenomenological adequacy of this proposal? Let me consider this point before moving to my main criticism, aimed at showing that non-relational PO is not intentionality because it does not satisfy its two defining features. The critical remarks I am going to make are meant to set the stage for my further step by calling into question the identification of intentionality with an intrinsic aspect of mental states (be it an adverbial modification or whatever).

As regards the uniformity requirement, there is no problem with satisfying it within the non-relational framework in so far as the account provided of intentional states is entirely internalistic, and therefore indifferent as to whether there actually exists something in the world that the state targets. Let us then consider the conformity requirement. As a matter of fact, our conscious intentional mental states present themselves to us as directed towards something (that is as having relational directedness), and what they are directed towards presents itself to us as mind-independent and in many cases as concrete, and external. The apparent relationality of the directedness of our intentional mental states and the apparent mind-independence, concreteness and externality of that which they are targeted at are all traits that characterize the way in which intentionality strikes us phenomenologically. One would expect them to be preserved within an account that treats intentionality as an experiential property having a phenomenal nature. And yet, this is not so within the account we are considering. But how can a position that rejects these traits count as adequate from a phenomenological point of view given the fact that it takes intentionality to have a phenomenological nature? As a start, it must be stressed that the characterization of intentionality as an intrinsic, non-relational property of the mind seems to conflict with the role that people have traditionally ascribed to it, namely that of accounting for our mind’s openness to the world. If one sticks to this traditional way of characterizing the role of the property, the notion of intrinsic directedness appears to be a contradictio in adiecto. How can directedness be intrinsic if directedness has to connect us to the world? In other words, do intrinsicness and directedness not pull in opposite directions?

The very fact that the idea of intrinsic directedness is hard to swallow is a point many PITers acknowledge. Kriegel (2011), for one, admits that there is something perplexing, and somewhat mystifying, about the notion itself. And yet, that such a property exists can, in his view, be introspectively ascertained, as is the case for any other experiential property.⁵² In his view, the tension between intrinsicness and directedness is handled by getting rid of the claim that it is intentionality that connects us to the world. He admits that many of our phenomenal intentional states do connect us to the world, but what makes them phenomenally intentional and what makes them connect us to the world are, in his view, two different properties. A thought, Kriegel claims, connects us to the world in virtue of being true (or accurate or veridical), not in virtue of being a thought. In this way, according to Kriegel, the tension is resolved by de-powering the ‘explanatory burden’ placed on the notion of intentionality. This should explain the apparent relationality of the directedness of our intentional mental episodes.

A similar move is made as regards the other trait that characterizes the way intentionality strikes us phenomenologically. It seems to us that intentionality connects us, in most cases at least, to mind-independent, concrete, and external objects. But this is not really so, according to the non-relationist, because intrinsic directedness is never object-involving and does not constitutively involve any concrete particular whatsoever. And yet, the non-relationalist maintains, it is possible to do justice to the sense in which our conscious experiences are often as of mind-independent, external, concrete particular objects by accounting for this in phenomenological terms. The general idea is that the phenomenal character of our conscious experiences involves phenomenal properties that account for the feeling of being directed ‘mind-independent-external-concrete-particular-wise’.⁵³

Whether the explanations provided by the non-relationalist actually succeed in removing the impression of the phenomenological inadequacy of the proposal very much depends on one’s philosophical sensibility. As for me, while acknowledging the astuteness of the attempts made to account for the mismatch between how intentionality strikes us introspectively and how it really is, I do find it surprising that a position that defends the phenomenological/experiential nature of intentionality could accept the idea that there is such a huge mismatch between appearance and reality as regards intentionality itself. If intentionality were not claimed to have a phenomenological nature, there would be no problem in claiming that the way in which it presents itself to us might not reflect its real nature. But if intentionality is claimed to have a phenomenological nature, such a move is, in my view, hard to swallow.

One could retort that the claim that intentionality is an experiential/phenomenal property does not imply that its true nature should be revealed to us by introspection, and conclude on this basis that PIT, while being deeply revisionistic as regards the nature and structure of intentionality, is actually the true theory on the subject.⁵⁴ That PIT is revisionistic (and that this is so in particular as regards the non-relational versions of it) is a point that is acknowledged and even stressed by most of its advocates. As a matter of fact, that a theoretical proposal on a given phenomenon is revisionistic is not by itself a false-indicating feature. It may actually turn out that many of the traits we pre-theoretically ascribe to a given phenomenon are in fact not possessed by it, and that other traits, which do not figure in our pre-theoretical understanding of it, actually do characterize the true nature of the phenomenon. This is hardly a questionable point. And yet, any revisionistic proposal has to respect some basic requirements if it aims at truly counting as a new account of a given subject rather than a change of subject. Well, are those basic requirements satisfied in our case? If one agrees that no account can actually amount to an account of intentionality unless what is labelled ‘intentionality’ satisfies the defining features of intentionality, then, as I shall try to show in the following, there are good reasons to doubt that non-relational PO is identical with intentionality.⁵⁵

Let us consider whether PO in its non-relational construal satisfies the two defining features of intentionality then. At first glance, one could have the impression that non-relationalism has no chance of accounting for them, because both features are actually framed in relational terms. In fact, it is standard to formulate the two features by saying that the object a given state is about may not exist (as regards the possible non-existence feature) and, analogously, that the object a given state is about may be presented in different ways (as regards the aspectuality feature). Does this prevent non-relationalism from accounting for them? The typical move made by the non-relationalist in this regard is to claim that the notion of aboutness featured in such formulations is actually ambiguous between two different readings, one referential and one intentional, and that whereas the former requires a relational treatment, the latter does not. Non-relational or intrinsic aboutness is meant to account for the states’ intentional directedness, which is an essential property of any intentional state. By contrast, relational or extrinsic aboutness is meant to account for what amounts to a contingent property of an intentional state, namely the property of being related to something in the external world. According to the non-relationalist, if one draws this distinction and reformulates the relational aboutness-talk into non-relational aboutness-talk, one can account for the two above mentioned features of intentionality within a non-relational framework. Is she right in this regard? Does the suggested move actually do the trick? Kriegel, for one, is very optimistic about this and claims it to be a virtue of the adverbial approach that it succeeds in accounting for the two defining features of intentionality in a very elegant and ontologically parsimonious way.

At first glance, Kriegel seems to be right in so claiming. As regards the feature of the possible non-existence of the intentional object, one can say that it is perfectly possible for an intrinsic property of a mental state to be instantiated even though there is no entity to which the state is related. For within this framework, as has already been said, being related to something is an extrinsic property of the mental state and, consequently, there is no need that this property be instantiated in order for the state’s intrinsic property of being-directed-somehow to be instantiated. It is therefore possible for a mental state (a thinking episode, say) to be directed Pegasus-wise, for example, even though Pegasus does not exist, and therefore even though the state turns out to be related to nothing in the real world.

Much the same holds, mutatis mutandis, for the other defining feature of intentionality, namely aspectuality. Also in this case nothing seems to prevent the possibility of two mental states that exemplify different intrinsic properties (say, being directed Hesperus-wise and being directed Phosphorus-wise) being externally related to the same object (the planet Venus). For, as in the previous case, there is no necessary connection between the instantiation of those intrinsic properties and the instantiation of the extrinsic property. So, does non-relationalism, and in particular, adverbialism, ultimately succeed in coping with the problem I have raised?

In what follows I shall express my reservations regarding Kriegel’s optimistic assessment. At first glance, the non-relationalist seems able to acknowledge within her framework the two defining features of intentionality. But the point to consider is whether she succeeds in accounting for them as features of what she takes intentionality to be, that is, as features of intrinsic directedness, features that an intentional mental state possesses purely in virtue of instantiating the non-relational property of being directed-somehow. In order for this to obtain, it ought to be possible to reduce the property of extrinsic, referential aboutness to the property of intrinsic, intentional aboutness in such a way as not to make any reference whatsoever to the former in the explanation provided. Well, can this be sensibly done? Let us consider the two following (adverbialist) rephrasings of our two features:

• i. A state directed-somehow (say, Pegasus-wise) may not be referentially about anything (i.e. may fail to refer)

• ii. Two states directed-somehow in different ways (say, one directed Hesperus-wise and the other Phosphorus-wise) may be referentially about (i.e. may refer to) the same object (the planet Venus).

As things stand, both (i) and (ii) exploit the notion of referential aboutness. Can any such reference be avoided? Well, let us consider what the result would be if one got rid of that notion by completely reducing the relational referential-talk to purely non-relational adverbial-talk. What one would obtain is something along these lines:

• i.* A state directed-somehow may fail to be directed-somehow (!);

• ii.* Two states directed-somehow in different ways may be directed-somehow in the same way (!).

As it turns out neither (i*) nor (ii*) return the intended meaning, provided of course that they return something sensible and not merely gibberish.

What do these considerations tell us? First, they tell us that there is a sense of the notion of aboutness that seems not to be captured by a purely non-relational account, and such an irreducible sense seems to be indispensable for providing a sensible formulation of the two features. A second and more substantive point is the following: if a mental state’s having those two features depends not only on that state’s intrinsic property of being directed-somehow, but also on that state’s contingent property of being or not being related with something, then it follows that those two features cannot constitutively belong to any intrinsic property of the state; a fortiori, they cannot constitutively belong to the state’s intrinsic property of being directed-somehow. If what I have said is correct, it follows that the chances of the non-relational versions of PIT of providing an account in which PO turns out to be identical with intentionality are very slim indeed after all.

5 Conclusion

If what I have said in the two previous sections is on the right track, it follows that so-called phenomenal intentionality is not intentionality. This is so in my view because it cannot be identified with that basic form of the property, reference intentionality, that any other form presupposes.

An advocate of PIT could actually contest this conclusion by rejecting the underlying assumption that a mental state cannot be intentional unless it exemplifies reference intentionality, and claim instead that all that is required for a mental state to count as intentional is to possesses some content that makes it assessable for accuracy.⁵⁶ A thorough discussion of this paper’s title question would therefore require a further section that addressed the issue of whether phenomenal intentionality can be identified, if not with reference intentionality, at least with content intentionality after all.

Reasons of space prevent me from discussing this further issue here. In the absence of such a further section, what the present paper actually licenses is a weaker conclusion than the one previously drawn. Namely, either the conditional conclusion that if it is true that reference intentionality is presupposed by content intentionality, then so-called phenomenal intentionality cannot be identified with intentionality; or the non-conditional conclusion that so-called phenomenal intentionality cannot be identified with reference intentionality.

Ideally, a thorough discussion of the present topic ought also to include a section aimed at saying what so-called phenomenal intentionality ultimately is. Of course, such a section would not be needed if one endorsed an eliminativist stance towards it—but this is not the position I am inclined to. As I said in Sect. 2, I am sympathetic towards PIT’s claim that there is an experiential property that depends on phenomenality, and that we tend to describe using terms like ‘of’ or ‘about’. Moreover, and more importantly, I am strongly inclined to think that such a property has a crucial role to play within a general theory of intentionality. As previously clarified, what my criticism aims at targeting is the claim that such an experiential property is intentionality. Well, granted that PO is a real property, three questions immediately arise. First: What, then, is PO, ultimately? Second: Does PO, although not identical with intentionality, have any role to play within a complete account of intentionality? Third: In the positive case, how would such an account relate to the competing accounts in the philosophical literature? In particular, would it be a competitor to both the Naturalist externalist research program and the Phenomenal intentionality research program, or would it rather be a sort of synthesis of those two supposedly competing approaches aimed at overcoming their respective limitations and partialities in order to promote a picture of intentionality able to do justice to its actual complexity?

A detailed treatment of these issues would obviously require a separate and perhaps more ambitious paper. My aim here has been confined to a critical assessment of the claim that identifies intentionality with a phenomenal, experiential property. If what I have said is correct, such an identification is ultimately unconvincing for the reasons stated. Even though the overall import of the paper is almost entirely negative, I think it can provide a positive service in improving our understanding of a territory that, albeit widely explored, is still poorly understood in all its complexity and richness.