Intentional Objects
Comments on Tim Crane's "Intentional Objects"
Tim Crane, “Intentional Objects,” Ratio, Vol. 14, No. 4 (December 2001), pp. 336-349.
Intentional states and acts can take objects even if those objects do not exist, and this is so whether our intentional states and acts are about things, events, properties, or anything else. Tim Crane develops this idea by considering some claims made about intentional objects by John Searle; if we put these together with the recognition that intentional objects need not exist, we get an aporetic triad:
Intentional objects are ordinary existing entities.
Intentional objects are what intentional states are about.
Some intentional states are about things that do not exist.
If we put these together, we get the claim that there are ordinary existing entities that do not exist.
Proposition (2) is just the definition of ‘intentional object’; if we get rid of it, we just change the subject, even if we keep the same words. Proposition (3) seems to be recognized quite widely. An atheist would generally say that the theist is talking about something that does not exist. (Indeed, most atheists, of the kind who make arguments for atheism, at least, take themselves to be building their arguments by thinking carefully about something that they are sure does not exist. Are all such atheists being irrational in thinking that they are thinking about something that does not exist?) If someone hallucinates something that is not really there, surely they can later think and talk about what they hallucinated. People say that Pegasus does not exist, and still talk about it; they say that the planet Vulcan turned out not to exist, and still talk about it; we experience optical illusions and after-images, and the like, and can think and talk about them. Proposition (3), in fact, seems very secure.
Therefore it seems that Proposition (1) is the culprit. We might try to moderate (1), removing ‘existing’; then we get the claim that there are ordinary entities that do not exist, which at least seems odd, since it’s unclear what ‘ordinary’ means if we are applying it to entities that do not exist. It just does not seem to be the case that intentional objects are ordinary existing entities. So we need a theory of what they are instead.
We can rule out the idea that they are mental objects. When we look for something in the world that turns out not to exist, it makes no sense to say that we were looking for something in our minds. Ponce de Leon was looking for the Fountain of Youth; he was not looking for a mental idea or image of the Fountain of Youth. He presumably already had that.
We have a dilemma, then: either Proposition (3) is false or there are nonexistent objects. Therefore, since one of the key issues that requires us not to identify intentional objects with ‘ordinary existing objects’ is that intentional objects need not exist, we need an account that lets us say that there are things that do not exist. This is, of course, directly opposed to broadly Quinean accounts of what we can say there are, which for a significant period was the dominant view in analytic philosophy. It’s not surprising that it was, of course. There is a convenience with assuming that the existential quantifier, or the logical domain, is automatically concerned with actual existence. (Then you can think that you don’t need an account of what the quantifier or the domain is actually doing and shove off all the work on reality rather than doing it yourself.) How do two people think about the same non-existent object? How can objects be indeterminate in the way intentional objects would have to be?
Crane argues that we can resolve the dilemma by rejecting an assumption shared by both horns of the dilemma, namely, that intentional objects, considered as such, are things of a certain kind. Sometimes when we call something an ‘object’, of course, we are just using it as a sort of subset of ‘thing’. In other cases, we certainly contrast ‘object’ with ‘thing’; if we call something the object of our attention, we do not mean that it is the thing of our attention. ‘Intentional object’, Crane argues, has to be more like the latter than like the former.
To say that something is a thing of a certain kind means that it has to be subject to particular forms of general conditions that make it of that kind – things like having a particular location in space and time, or involving intentionality, or the like. When we talk about physical or mental things, Crane suggests, we can have an ontological account of it, what Crane calls a ‘substantial’ conception of it. He wants to argue, though, that intentional objects differ in not being susceptible of such an ontological account. On Crane’s diagnosis, both horns of the dilemma make the mistake of assuming that there can be an ontological account of intentional objects. There is nothing that anything has to be, in itself, in order to be an intentional object. We call them ‘intentional objects’ relationally, of course, as being the objects of the intentional states or acts of a subject. That is all there is to them, though.
When two people hear the same piece of music, we can also often say that they do not hear the same things. One may have a more musical ear than the other, and therefore hear things about the music that the other cannot. This does not affect the nature of the music at all. Rather, it’s more as if the world for each listener is different, although, of course, these distinct ‘worlds’ are linked by the actual music.
As Crane notes, this is a route to something like what Husserl and phenomenologists mean when they talk about intentional objects. Analytic philosophers have tried to avoid this in various ways. For instance, some would agree with Anscombe that intentional objects can be given a grammatical account, as the direct objects of transitive intentional verbs. Crane allows that there is a sort of parallel here, but it seems clear that taking it actually be an account is equivocating on the term ‘object’ again. Other attempts to think of intentional objects as ‘shadows of language’ or ‘shadows of grammar’ seem to run into the same problem. And they seem, moreover, to get the order of explanation wrong; the grammar is not even the right kind of thing to explain the phenomena of intentionality.
Other attempts to evade the positing of intentional objects involve representation. On this view, intentional states have representational content, i.e., a way the world is represented as being, and intentionality is regarded as not involving a relation to an object at all. Crane argues that in fact one cannot explain intentionality in terms of representation. One reason for thinking this is that intentional states can be about the same thing but present it in different ways; likewise, intentional states can be about different things but present them in the same way.
Further, our most common examples of representation – sentences, pictures, and the like – quite clearly don’t represent simply in and of themselves. They are so derivatively, which in practice means that we have to explain why they represent, and why they represent this or that, in terms of things like thoughts and intentions. Sentences only get their meanings from how they are used by thinking subjects. But how do they do this? A natural way would be to say that they gain the relation of representation. If, however, you are trying to use representation to reject the view that intentionality involves a relation to nonexistent objects, you can’t do so by invoking a kind of representationality that involves a relation to nonexistent objects. Similar problems arise when we try to account for representation of nonexistent using causation. Ultimately, any account of intentionality by appeal to representation or causation seems to require recognizing something very like an intentional object. If we are going to say that A is about B, there needs to be a way that B can be an object with respect to A.
When we look at this overall argument, we see that there are actually two distinct things that Crane wants to go together. We have here, first, a positive argument that intentional objects must be posited and are not eliminable, for any adequate account of any context involving what we often call ‘intentionality’ and ‘aboutness’. We are agreed with Crane on this positive affirmation of intentional objects, and think that not only do his arguments to this conclusion work, you can develop almost endless variations on them, all for the same conclusion.
In addition to this positive argument, Crane has a deflationary argument, in which he argues that, in positing intentional objects, there is not really any more to be said. That is, if we posit intentional objects, they come with no further baggage; indeed, they can carry no further baggage. We can admit intentional objects without having any ontological account of them, because they have to be admitted but there is no ontological account of them. This we think a harder sell, and we don’t think Crane is successful in arguing it. A major problem with this attempt is that we need to be able to distinguish intentional objects; it seems that this would have to be in terms of intentional states or acts. Intentional states or acts also have to be distinguishable, and we have to be able to identify when we are in one intentional state rather than another (e.g., desire rather than belief). This means that intentional states and acts have to be things of a certain kind, that is, things that meet certain general conditions to be what they are, and it seems some of those general conditions have to be general conditions met by the intentional objects. This seems to make them ‘things’ in Crane’s sense.
Nonetheless, we think that Crane is in the vicinity of something right here. A way we can salvage something like Crane’s view is by taking intentional states or acts to be things that involve certain ways being related. The foundational term of this being-related is something about the subject. However, as we have known since Aristotle, every being-related is a to-something, which requires that there be another term of the being-related. For intentional states or acts this other term is what we call an ‘intentional object’.
What Crane is getting right is that, for intentional states or acts, the being-related requires no more foundation than what is in the subject; that is, to have the being-related, we don’t need there to be a further thing as foundation. This is not because the intentional object meets no general conditions, nor because there can be no ontological account for it; it is because the intentional state or act is itself sufficient foundation for taking an object, so sufficient, in fact, that it does not require that the object actually exist. Pace Crane, we have to know something about how the object is given as the term to the intentional state or act. (Our own view is that this would have to be along the general lines of formal signification, or objectual causation as a kind of formal causation, or constitution, or purely relative being, or some combination of these. Crane would almost certainly reject these because they yield accounts in which the object is in some way immanent in the state or act, and he wants to avoid this.) We do need an ontological account of intentional objects. But, while we think the line is not quite where he thinks it is, Crane is right that there is a sense of ‘thing’ in which we need to contrast ‘thing’ and ‘object’. Everything he says as a negative characterization of intentional objects seems right. But we do have some hope of going beyond this and giving a positive characterization.
(It should be said, of course, that “Intentional Objects” is fairly early. Crane has a later work, The Objects of Thought, concerned with this topic, which provides some refinement and development, and looks more closely at how this integrates with Crane’s own physicalism. We hope at some point to discuss this work directly, after we have looked at a few more articles on this. But it’s good to look at what seems workable and what seems problematic in the original conception.)
