Angela Mendelovici

Research

My research is on intentionality, consciousness, and the relationship between the two. Much of my work argues for a view on which all intentional states are phenomenal states. This is a version of the phenomenal intentionality theory, the view that all original intentionality arises from phenomenal consciousness alone. On the view I defend, the intentional mind is limited to the phenomenal mind, making intentionality more limited than is conventionally thought. While perceptual experience might represent fairly rich contents, thought content is fairly impoverished, and nonconscious states, like standing beliefs and the nonconscious states posited by cognitive science don't have genuine intentional properties at all, though thoughts, and standing beliefs and desires may have rich derived representational properties on a more permissive notion of representation.

My research contributions can be organized by themes and topics that appear in multiple works:

  1. Fixing reference on intentionality and our theory-independent access to intentionality
  2. The phenomenal intentionality theory
  3. Representationalism
  4. Superficial character and deep nature
  5. Reliable misrepresentation and the mismatch problem
  6. Moods
  7. Olfactory experiences
  8. Self-ascriptivism and the contents of thought
  9. Non-propositional contents
  10. Naturalism about intentionality
  11. Mental combination

Click here for a list of publications with abstracts.

1. Fixing reference on intentionality and our theory-independent access to intentionality

The central project that my research focuses on is that of providing a theory of intentionality, roughly understood as the "aboutness" of mental states and other items. Before offering a theory of intentionality, it is important to be clear on what a theory of intentionality is a theory of. I argue that many existing definitions of intentionality, including definitions in terms of aboutness, truth conditions, and intentionality's purported role in psychological explanation, do not get at the core notion we are interested in when we are interested in intentionality. I propose an alternative ostensive definition of intentionality based on introspectively accessible paradigm cases of intentional states. On this way of fixing reference on intentionality, intentionality is a phenomenon to be explained, rather than a posit in a theory explaining something else.

This way of fixing reference on intentionality suggests two theory-independent ways of finding out about intentionality: the first is through introspection. I argue that, in limited circumstances, introspection can at least partly inform us of the specific contents of specific intentional states. For example, introspection can tell us that we are currently representing <grass is green> rather than <snow is white>. The second way of finding out about our intentional states is through their psychological roles. Intentional states tend to cause related states and behaviors, so observing these related states and behaviors can provide insight on their contents.

Relevant works:

2. The phenomenal intentionality theory

The phenomenal intentionality theory (PIT) is the view there is a kind of intentionality, phenomenal intentionality, that is nothing over and above phenomenal consciousness, and from which all other kinds of intentionality derive. Versions of PIT have been defended by various authors, including Uriah Kriegel, David Pitt, Katalin Farkas, Terence Horgan, John Tienson, George Graham, Brian Loar, and David Bourget. In my work, I have carved out the space of possible versions of PIT, argued for PIT as a general approach to intentionality, and defended a specific version of PIT, which I call "strong identity PIT".

Relevant works:

My specific version of PIT, strong identity PIT is the view that all intentionality is phenomenal intentionality (strong PIT) and that the precise relationship between phenomenal intentionality and phenomenal consciousness is one of identity (identity PIT).

Strong PIT faces challenges in accounting for allegedly intentional states that are not phenomenal, such as standing propositional attitudes (e.g., beliefs we are not currently entertaining), non-conscious states posited by cognitive science (e.g., non-conscious states involved in early vision), and thoughts whose contents outstrip their phenomenal characters (e.g., thoughts with wide, object-involving, or abstract contents). I argue for a largely eliminativist strategy with respect to these kinds of states and contents: The contents of thoughts that outstrip their phenomenal characters and the contents of standing propositional attitudes are merely derivatively represented in much the same way that conventional representations like stop signs are taken to represent their contents. Given what "intentionality" ends up picking out (see above), this sort of derived representation is not a species of intentionality. The non-conscious states posited by cognitive science are also not intentional, though they might be informational, computational, or track various items in the environment. Again, the sort of representation involved here is not a species of intentionality.

Representative works defending strong identity PIT:

3. Representationalism

When combined with the view that all phenomenal states give rise to intentionality, strong identity PIT entails representationalism (or intentionalism), the view that phenomenal consciousness is nothing over and above intentionality, perhaps that meets some further conditions. (Indeed, it entails pure representationalism, the view that phenomenal consciousness is nothing over and above intentionality alone.) So, strong identity PIT faces some of the same challenges as representationalism, such as those of accounting for mental states that seem to be phenomenal but not intentional, such as moods. I have argued for pure representationalism and for a representationalist treatment of moods.

Relevant works:

4. Superficial character and deep nature

In several works, I introduce and make use of the distinction between the superficial character and the deep nature of intentional states and contents. When we ask of a particular intentional state, "What is this state's content?" we might be asking two different questions. First, we might be asking what particular contents the state represents, where some possible answers are that it represents <grass is green> or that it represents <the cat is on the mat>. Second, we might be asking what is the state's deep metaphysical nature, what it really is, where some possible answers are that it is a set of possible worlds, a structured proposition, or an adverbial modification of the representing subject. The first kind of question concerns what I call the superficial characters of intentional contents, while the second kind of question concerns their deep natures.

The distinction between superficial character and deep nature also applies to intentional states. Specifying an intentional state's superficial character involves characterizing its content in a way that distinguishes it from other contents and specifying any non-content features of the intentional state, e.g., whether it is a belief, a desire, or a visual experience. Specifying an intentional state's deep nature involves specifying its content's deep nature and the relationship it bears to that content, e.g., whether it is a primitive relation to set of possible worlds or a set of causal relations to the components of a structured proposition.

The distinction between the superficial character and deep nature of intentional states and contents is important for several reasons. One is that it provides a way of isolating what the theory-independent ways of finding out about the contents of our intentional states mentioned above really tell us. Introspection and considerations of psychological role, in the first instance, only tell us the superficial character of intentional states, e.g., whether a particular intentional state represents <the cup is on the table> rather than <an octopus is wrapped around my arm>. These theory-independent considerations arguably do not reveal the deep nature of intentional states and contents, e.g., whether my thought that there is a cup on the table is a relation to set of possible worlds, a relation to a structured proposition, or something else.

Another reason the distinction between superficial character and deep nature is important is that it clears the way for potentially promising positions on debates concerning the kinds of contents that can be represented by experience. For example, propositionalism is sometimes taken to be the view that all mental states have propositional contents, which is taken to imply that mental states have contents all mental contents have a propositional superficial character (they represent that such-and-such is the case) and that all mental contents are relations to propositions or proposition-like entities (such as sets of possible worlds or structured propositions). But there are two separate issues here, one concerning whether all contents have propositional superficial characters, and another concerning whether all contents have propositional deep natures, and different types of considerations bear on the two issues. Distinguishing between the deep and superficial questions allows us to clearly see space for a view on which all contents have a propositional superficial character but non-propositional deep natures. Along similar lines, I have argued that some perceptual experiences, such as olfactory experiences, have singular superficial contents without having object-involving deep natures.

Relevant works:

5. Reliable misrepresentation and the mismatch problem

I have argued that tracking theories of intentionality, which aim to account for intentionality in terms of causal, informational, or teleological tracking relations between mental states or mental representations and items in the environment, face a fairly blatant problem of empirical adequacy: they fail to attribute contents correctly in certain paradigm cases of intentionality, such as in the case of color experience. More precisely, in the case of the problematic states, tracking theories attribute contents that differ in their superficial characters (see above) from the contents we have theory-independent reason to ascribe. I say that contents that are dissimilar in superficial character fail to match, and, so, the tracking theory faces a mismatch problem.

Many mental states that pose a mismatch problem for tracking theories systematically misrepresent the world around us. For example, visual color experiences track surface reflection profiles or the like but represent primitive colors, which are arguably uninstantiated. But since such mental states do a good job of tracking various worldly items, they reliably misrepresent: they misrepresent in similar ways in similar circumstances. I have argued that reliably misrepresenting mental states can be as useful for various tasks as veridically representing mental states. I have also argued that tracking theories cannot accommodate non-contrived ("clean") cases of reliable misrepresentation, and that this is a problem for them even if there are in fact no such cases.

Relevant works:

6. Moods

Moods are sometimes thought to pose a challenge to representationalism, the view that phenomenal consciousness is nothing over and above intentionality, perhaps in combination with some further ingredients. I argue that moods and emotions are both intentional, representing the same affective properties, such as elatedness, sadness, and anxiousness (see above). However, while emotions represent these affective properties as bound to particular represented objects, moods represent them entirely unbound. In other words, while emotions represent propositional contents attributing affective properties to represented objects, moods represent mere proprietal contents: the affective properties themselves. For example, an emotion experience might represent a particular event as sad, while a mood might represent mere sadness. Note that the distinction between propositional and proprietal contents is one concerning the superficial character, rather than the deep nature, of content (see above), so my view is not committed to the claim that actual affective properties enter into the contents of moods.

Relevant works:

7. Olfactory experiences

While it is widely agreed that olfactory experiences have contents, there is disagreement on the precise objects of olfactory experiences, i.e., on what olfactory experiences represent as having olfactory properties. I argue that some olfactory experiences seem to present everyday objects, such as basil leaves, cakes, and garbage bins, as having smells, while other olfactory experiences seem to present smells as taking on a life of their own, wafting from their sources and leaving them behind. I argue that the objects represented by the first kind of olfactory experience are everyday objects, like basil leaves, cakes, and garbage bins, while the second kind of olfactory experience represent what I call "bare olfactory objects", which are ad hoc objects only represented as having olfactory properties and spatial properties. I also argue that the properties represented by bare olfactory experiences are primitive olfactory properties which happen to be uninstantiated, and that olfactory experiences misrepresent everyday objects or bare olfactory objects as having primitive olfactory properties. Since olfactory experiences do a good job of tracking molecular properties of everyday objects and odors, they reliably misrepresent (see above), which helps explain why they are so useful and how they can help us acquire justified true beliefs about putatively smelly things. (Note that since bare olfactory objects are not represented as having or lacking any further properties, bare olfactory objects might end up referring to odors or portions of air, so olfactory experiences representing bare olfactory objects do not misrepresent simply in virtue of the objects they represent.)

Relevant works:

8. Self-ascriptivism and the contents of thought

We intuitively take our thoughts to represent more than just the contents at the forefronts of our minds when we have them. For example, we might take the thought we would express with the words "Modal realism is false" to have a content involving a characterization of modal realism as the view that possible worlds exist in the same way that the actual world exists, even though no such content is in the forefront of our minds when we think the thought. We might also take our thoughts to represent wide contents, including object-involving contents. For example, we might take Oscar's water-related thoughts to represent H2O, even though he knows nothing of the chemical constitution of water, and we might take a thought about Eleni to involve Eleni herself as a constituent. Accommodating such alleged contents is particularly challenging for the phenomenal intentionality theory (PIT), the view there is a kind of intentionality, phenomenal intentionality, that is nothing over and above phenomenal consciousness, and from which all other kinds of intentionality derive (see above). The problem for PIT is that the above-mentioned alleged contents of thoughts are not phenomenal contents, and it's not clear how they might derive from phenomenal contents.

I argue that while many of our thoughts' alleged contents are not in the forefront of our minds, there nonetheless is a type of content that is. I call this content immediate content, since it is the content we are immediately aware of, or that is before our mind's eye. These are the contents that we "grasp" (see Bourget 2015). There are different views on the superficial characters of immediate contents: they might be or involve words, imagery, or gisty, partial, or schematic understandings of the corresponding alleged contents. What matters for my purposes is that the immediate contents of thoughts are distinct from their alleged contents.

My view of the immediate and the alleged contents of thoughts is that immediate contents are phenomenally represented, while derived contents are merely derivatively represented in much the same way that stop signs derivatively represent <stop!> and the word "cat" derivatively represents <cat>. I propose a self-ascriptivist view of how we derivatively represent the alleged contents of thoughts. According to self-ascriptivism, we derivatively represent a content by ascribing it to our mental states or ourselves. The way that we ascribe thoughts' alleged contents to ourselves is by being disposed to accept certain cashings out of the contents of concepts involved in thoughts. For example, we are disposed to think <by <modal realism>, I mean <the view that possible worlds exist in the same way that the actual world exists>>. In virtue of this, we count as ascribing <the view that possible worlds exist in the same way that the actual world exists> to <modal realism>, and <modal realism> comes to derivatively represent <the view that possible worlds exist in the same way that the actual world exists>. In turn, the thought that we express with the words "Modal realism is false" comes to derivatively represent a content involving the complex descriptive characterization of modal realism.

As the above example shows, when we take our accepting that our immediate contents cash out into complex descriptive contents, these contents derivatively represent the complex descriptive contents. We can similarly ascribe object-involving or externalist contents to ourselves or our mental states. This happens when we are disposed to think thoughts like <by <water>, I mean the clear watery stuff around here>. In this case, the complex descriptive content <the clear watery stuff around here> is used, rather than mentioned, so the content we end up self-ascribing is its referent, i.e., H2O, rather than the descriptive content itself.

Given the way I fix reference on intentionality (see above), self-ascribed content is not intentionally represented, though it is represented on a more permissive use of "represent". So, my self-ascriptivist picture is compatible with strong PIT, on which all intentionality is phenomenal intentionality (see above), though there is a nearby fall-back position that does take self-ascribed contents to be intentionally represented, which is not compatible with strong PIT (but is compatible with what David Bourget and I call "moderate PIT" here and here).

Relevant works:

9. Non-propositional contents

Propositionalism is often taken to be the view that all mental states represent propositional contents. But, I argue, there are different ways of understanding this characterization. On one way, propositionalism ends up being a view about the deep nature of intentional contents on which all intentional contents are proposition-like entities, such as sets of possible worlds or structured propositions consisting of ordered pairs of objects and properties; this is deep propositionalism. Another way of understanding propositionalism is as the view that all contents have a propositional superficial character in that they represent that such-and-such is the case; this is shallow propositionalism. I have argued that the two views are often conflated and that it is important to distinguish between them, particularly since different considerations support either view.

I have also argued against both deep and shallow propositionalism. I've argue that deep propositionalism is false because contents are aspects of intentional states rather than distinctly existing entities like sets of possible worlds or structured contents. In my work on moods, I've argued that shallow propositionalism is false because there are proprietal contents, contents representing mere ways things are, such as <cat> or <happy>. (This case against shallow propositionalism is complementary to the case that is often made based on intentional states with objectual contents, like <John Oliver> or <Eleni>.) In my works on olfactory experiences, I argue that although a proprietal view of certain kinds of olfactory experiences might be tempting, olfactory experiences in fact have propositional contents. The differences between moods and olfactory experiences help us better understand when proprietal content attributions are warranted.

Relevant works:

10. Naturalism about intentionality

Much work on intentionality has focused on the project of "naturalizing" intentionality, which is usually taken to involve showing how it can be constructed out of non-mental ingredients. Naturalism is often taken to be a constraint on a theory of intentionality, one that rules out theories such as PIT, which appeal to mental ingredients, such as consciousness. David Bourget and I have argued that naturalism cannot be used to argue against PIT. Briefly, the only appropriate constraint on a theory of intentionality in the area is that it appeal to entities that exist. Since consciousness exists, it is not inappropriate for a theory of intentionality to appeal to it, regardless of whether it is a naturalistic ingredient. Bourget and I also argue that there are alternative ways of understanding naturalism as a methodological constraint that includes, perhaps among other methodological principles, a commitment to empirical adequacy. On this way of understanding naturalism, PIT arguably ends up being more naturalistic than alternative tracking theories, which arguably cannot account for all paradigm cases of intentionality (see the mismatch problem above).

Relevant works:

11. Mental combination

I have recently become interested in the question of how mental items - intentional states and contents, phenomenal states and phenomenal characters, or subjects - combine to form more than their mere aggregates. On my favored view of intentionality (see above), intentionality is non-relational. This view bears some similarity to the adverbial view of perception, which is thought to have trouble accommodating for the structure of perceptual states and contents (this is Frank Jackson's many properties problem). I have argued that every view of intentionality faces a challenge when it comes to explaining intentional structure. This has led me to consider the combination problem for panpsychism, which concerns panpsychism's alleged inability to explain how various microphenomenal features can combine to give rise to macrophenomenal features, such as our own. I have argued that many aspects of panpsychism's combination problem are special cases of more general problems of mental combination, such as the problem of mental unity and the problem of accounting for intentional structure. I have tentatively suggested that, although our minds might be able to understand physical modes of combination, they are not be equipped to understand mental combination.

Relevant works:

accepted by OUP