(A guest post by Peter Lupu. Minor edits and comments in blue by BV.)
There are at least two ways in which the relational character and object-directedness of intentional states such as beliefs, wants, desires, seekings, etc., is motivated:
A. The individuation of intentional states;
B. Aristotle’s belief-desire model of explaining actions.
I. Motivation (A). Consider the following:
1) Jake seeks the golden mountain;
2) Jake seeks the keys to his car.
Clearly, (1) and (2) express two different intentional states of the same individual. But, in virtue of what do (1) and (2) express two different intentional states? It appears that the best and only explanation for the difference is that the two cases relate Jake to two different objects: i.e., in (1) Jake seeks the golden mountain; in (2) Jake seeks his car keys; and, of course, the golden mountain and Jake’s car keys are two distinct objects.
The point, though correct, needs to be made with a bit more exactitude. Presumably, the intentional states are numerically distinct in virtue of occurring at different times. If so, someone could reply that what makes the states different is their occurring at different times. The question, however, is not what makes the states token-distinct, but what makes them type-distinct. And the answer can only be that it is distinctness of intentional object that explains type-distinctness of states.
So far so good: we succeed in individuating different intentional states in terms of their respective intentional objects; i.e., the objects towards which the intentional states are directed. But now we face a problem. Unlike Jake’s car keys which certainly exist, the golden mountain does not. How can a non-existent object serve to distinguish one intentional state from another? How can the golden mountain, which fails to exist, be identical with or different from Jake’s car keys, which certainly exist? It seems it makes no sense to say that the golden mountain is identical to or different from Jake’s car keys. But if this makes no sense, then how can we use the golden mountain to distinguish (1) from (2)? This is one way to highlight the dilemma of intentionality. On the one hand, intentional objects are often the only way we can distinguish evidently different intentional states. On the other, when the intentional state is directed towards a non-existent object, then individuation appears to collapse. Yet we still distinguish (1) from (2) even though we know very well that the golden mountain does not exist. How do we do it?
I see it differently. Every intentional state in virtue of being object-directed has a content, but it is not the case that for every such state there exists something independent of the state that satisfies the content. The intentional states underlying (1) and (2) are type-distinct in virtue of their different contents regardless of whether one of the contents is satisfied and the other is not, or both are satisfied, or neither are satisfied. It is the content that gives the intentional state its object-directedness, not the externally existing object.
Here is a crude analogy. Think of an arrow, the 'arrow of intentionality.' It has a shaft and a head. The shaft is the mental act which is a state of a mind. The head is the content. Every arrow has a head which points in the direction of the arrow's flight, but not every arrow hits something. The object, which may or may not exist, is that to which the arrow head points.
The phrase 'intentional object' invites confusion. It could be used to refer to the entity, if there is one, which satisfies the content which is inherent in every intentional state and which makes it object-directed. Peter appears to be using 'intentional object' is this way. Accordingly, only some intentional states have intentional objects. On the other way of understanding 'intentional object,' every intentional state has an object, whether or not said object exists. On this second understanding, the object would seem to be the same as the content.
I don't think we will make any progress with this topic without a content-object distinction. (The Brentano student Twardowski saw the necessity of it long ago in his 1894 monograph Zur Lehre vom Inhalt und Gegenstand der Vorstellungen.) If we make the content-object distinction we have an easy solution to Peter's dilemma. We don't need existing intentional objects to distinguish Jake's two mental states. They are distinguished by their different contents.
II. Motivation (B). Consider Peter’s case:
(3) Peter believes that the Superstition Mountains exist somewhere in Arizona;
(4) Peter believes that visiting the Superstition Mountains will enrich him spiritually;
(5) Peter wants to enrich himself by visiting the Superstition Mountains;
(6) Peter decides to embark on a journey to find and visit the Superstition Mountains;
So,
(7) Peter sets out on a journey to Arizona in order to find and visit the Superstition Mountains.
The best explanation we have for the action described in (7) is to appeal to the intentional states expressed in (3)-(6). But how exactly does this explanation work? Well, something like this is going on: Peter’s action of embarking upon a journey to Arizona in order to find and visit the Superstition Mountains corresponds to the intentional object towards which his intentional states described by (3)-(6) are directed; namely, the Superstition Mountains. And since Peter believes that the Superstition Mountains exist somewhere in Arizona rather than in Texas, he sets out to journey to Arizona rather than to Texas in order to find the Superstition Mountains.
(Note: I simplify a bit: (3) and (4) are beliefs and, hence, their intentional objects are truth-value bearing entities such as propositions. The simplification, so far as I can see, does not alter the central point.)
So far so good: the relational character and the object-directedness of intentional states play a crucial role in explaining actions. And this form of explanation works because the object towards which the action is directed is *the very same object* as the object towards which the intentional states are directed. Hence, we can use the later to explain the former. I cannot see how else we could explain actions such as these.
But now consider Jake’s case:
(8) Jake believes that the Golden Mountain exists;
(9) Jake believes that finding the Golden Mountain will make him rich;
(10) Jake wants to get rich by finding the Golden Mountain;
(11) Jake decides to embark upon a journey to find the Golden Mountain;
So,
(12) Jake sets out on a journey to find the Golden Mountain.
Since (3)-(7) and (8)-(12) are exactly analogous, the explanation should work in exactly the same way. And I suppose that if Jake were to really set out on a journey to find the Golden Mountain, then his family and friends will undoubtedly use (8)-(11) in order to explain his actions. And so if someone were to inquire as to how exactly the explanation works in Jake’s case, we would be tempted to repeat the same account we have offered in Peter’s case. The explanation works because the object towards which Jake’s action (described by (12)) is directed is *the very same object* as the object towards which his intentional states expressed by (8)-(11) are directed. Hence, the later explain the former.
But now we face a problem, for the Golden Mountain does not exist. And since the Golden Mountain does not exist, then there seems not to be any particular object towards which Jake’s actions are directed. And if there is no particular object towards which Jake’s action can be directed, then there is no object that can be the same as, or different from, the object towards which Jake’s intentional states are directed. And of course we might add that since the Golden Mountain does not exist, Jake’s intentional states also fail to be directed towards any existing object. So our neat account that worked so well in Peter’s case seems to break down in Jake’s case. Yet, clearly, Jake embarks on a journey to find the Golden Mountain *because* he has the intentional states described in (8)-(11). What else could cause him to set out on a journey under the circumstances described?
III. Suppose we restate (1) as follows:
(1*) Jake seeks *a* mountain that is made out of gold.
Notice that unlike the original (1), (1*) is no longer a *singular intentional state*. Replacing the definite phrase ‘the golden mountain’ or the name ‘Golden Mountain’ with the indefinite ‘a mountain that is made out of gold’ changes the character of the mental state from a singular thought to a general thought. Therefore, unlike in the case of (1), (1*) no longer requires us to relate Jake to an individual mountain. We can think of Jake’s seeking a mountain that is made out of gold as an attitude towards any mountain that has the property of being made out of gold.
But surely any mountain made of gold must be an individual mountain. That's one point. More importantly, whether Jakes seeks the golden mountain or a golden mountain, the content of his thought is general, not singular. It is true that 'the golden mountain' is a definite description while 'a golden mountain' is an indefinite description. What that means is that at any given time, and in any given world, there is at most one x such that x is golden and a mountain. But at different times in the same world, or in different worlds at the same time, different things can satisfy the definite description 'the golden muntain.' Therefore, a thought directed to the golden mountain is a general thought, not a singular one.
The following example may help. Suppose Tom admires the president of the USA. 'The president of the USA' is a definite description. It is not a name. Different people satisfy the description at different times in a given world, and at the same time in different worlds. For example, in the actual world, in January of 2011, Obama satisfies the description. But there is a possible world in which MCain satisfies the description in January of 2011. So the content of Tom's thought is general, not singular. It is not directed to a specific individual such as Obama, but to any individual that satisfies the definite description. But even if Tom admires Obama, it is only via some definite description that he can access him, e.g. 'the first mulatto president of the USA.'
What about the fact that no mountain that is made out of gold exists? Wouldn’t this fact, just as before, deny us an object towards which Jake’s mental state is directed? Sure! But, unlike previously, the absence of a suitable individual or particular object is no longer an impediment, for general thoughts are never directed towards a particular object.
That can't be right. Jake's mental state is object-directed whether or not there is an external object. This is the crucial fact that must not be lost sight of. To avoid this mistake, we must distinguish content and object. The content is what gives the state object-directedness, and every intentional state has a content. I don't know what to make of the claim that general thoughts are never directed to a particular object. Suppose I want a used Harley-Davidson Sportster for less than $4,000. Many bikes fit the description and any one of them will satisfy my desire. Suppose there are ten of them in my area. My thought is not directed to any one of these ten to the exclusion of the others, but it is directed to a particular hunk of American iron, for only a particular such hunk can satisfy me. I am not hankering after some abstract object such as a set or conjunction of properties. You can't want a motorcycle without wanting a heavy metallic spatiotemporal particular. You can't want a goldmine without wanting a particular goldmine. The fact that you don't want a particular particular goldmine does not alter the fact that you want a particular goldmine.
So, then, what are general thoughts directed to, if anything at all? I suggest the following rough answer. General thoughts are a fusion of two mental states. The first of these mental state is a first-level mental state that is directed toward one property or the union of several *properties*. The other mental state is a second-level state and it is directed toward a *proposition*; namely, the proposition that the property or the union of several properties towards which the first-level mental state is directed is instantiated in the actual world.
(IV) The current proposal preserves the relational feature of intentional states, including the two considerations that motivate the relational view (A) and (B) above. And we have supplied surrogate relata towards which the intentional states are directed, albeit the surrogates are not particulars.
(i) Consider motivation (A) about individuation.
We can still maintain that (1*) and (2) are different because they relate Jake to two different entities. In (2) Jake is related to his car keys; in (1*) Jake is related simultaneously to a proposition (i.e., the proposition *the union of the properties ~is a mountain~ & ~is made out of gold~ is instantiated*) and to a union of these properties. Clearly, Jake’s car keys are distinct from either one of these entities. Therefore, (1*) and (2) are distinct intentional states.
The mistake here, I believe, is to think that the content of an intentional state depends on the existence or nonexistence of the external object which would satisfy the content were the object to exist. But it doesn't. Suppose I am seeking my car keys and unbeknownst to me the person who finds them destroys them. The seeking is not at all affected in its nature as an intentional state by the passing out of existence of the keys. And if I am seeking something such as a gold mine which turns our not to exist, that is not to say that I was seeking all along an an abstract object, a proposition.
(ii) Consider motivation (B) about explanation of action.
Peter’s case remains as before because since the Superstition Mountains exist, they can serve as the object of Peter’s intentional states. However, Jake’s case (i.e., clauses (8)-(11)) needs to be recast along the lines of the above proposal. I will refrain from the laborious task of doing so except noting that since we now view some of Jake’s relevant intentional states as general thoughts, they are simultaneously directed towards a suitable union of properties (i.e., the property of being a mountain & the property of being made out of gold) and a suitable propositions (i.e., the proposition that the union of these properties is instantiated). So recast, these intentional states can still explain his action of embarking upon a journey to find a mountain that is made out of gold.
Such an explanation will proceed roughly as follows. Since Jake believes that a mountain made out of gold exists and since he decides to find something that simultaneously exemplifies the properties ~is a mountain~ and ~is made out of gold~, he sets out on a journey to find this object. Moreover, this account explains why if Jake is sufficiently persistent, his journey could last a lifetime.
(V) Existential Exportation.
The above sketch of this account also explains why existential exportation is invalid in the kind of intentional contexts under consideration. First, notice that when it comes to general thoughts, such exportation is never valid and for obvious reasons. It simply does not follow that there exists a particular woman Jake wishes to wed from the fact that Jake wishes to wed a woman.
Second: according to the above account, some intentional states that seem to have the form of singular thoughts are indeed singular thoughts; however, other intentional states that seem to have the form of singular thoughts turn out to be general thoughts. The case of Jake seeking the golden mountain is an example. While in the former cases the statement that expresses the existential exportation of the original intentional object is true, if the original thought is true; in the later cases such a statement is false, even if the original thought is true. Thus, we cannot authorize such inferences generally because we do not know in any given case whether what appears to be a singular though is indeed a singular thought, or whether it is merely a general thought disguised as a singular thought. Since the difference between genuine singular thoughts versus those who are general thoughts disguised as singular thoughts is not merely a matter of logical form, but depends on the actual state of the world (e.g., whether a mountain made out of gold exists), we cannot in general certify such inferences as a matter of logic alone.
I would say that what you are calling Existential Exportation is never valid, and I would also question whether there are any truly singular thoughts.
(VI) Concluding Remarks: The above proposal offers a gesture towards EO’s insistence that the surface structure of sentences about intentional states may not always reveal their deep structure. As I have argued above, this is exactly what happens when the putative intentional object does not exist. Then what appears on surface to be a singular thought is in fact a general thought with a totally different structure. However, agreement ceases at this juncture.
As I have insisted above, the difference between cases illustrated by Peter’s case and cases illustrated by Jake’s case cannot be discerned based simply on matters of logical form or linguistic analysis. The world has something to say about these matters. And it is for this reason that certain types of inferences fail in these contexts. Second: intentional states remain relational. However, the nature of the entities towards which intentional states are directed may vary radically depending upon whether the intentional state is a genuine singular thought versus cases where what appears to be a singular thought turns out to be a general thought instead. Suitable concrete, particular, physical objects can normally serve as the intentional objects of intentional states in the former cases (unless, of course, the intentional state is about abstract objects such as mathematical entities). Not so in the later cases, where entities such as propositions and unions of properties must be reckoned as the intentional objects of intentional states.
Consider
(A1) Jake says ‘I am looking for my car’ (having forgotten where he parked it after getting ‘trolleyed’ and having to take a taxi home)
(B1) Jake says ‘I am looking for a car’ (i.e. he is looking for to buy one, but with no actual car ‘in mind’).
(C1) Kate says ‘Jake says he is looking for a car’.
Some points. First, note that what Kate says is true both of what Jake says in both (A1) and (B1). Second, she is not describing a mental state, but rather something accessible to any competent user of English, namely what Jake is saying. Therefore, if there is any problem about this example, it is not a problem just about individual or personal or private mental states. Thirdly, the fact that Kate’s statement is true of both (A1) and (B1) suggests that there is some extra content (let’s say a ‘singular content’) that is included in (A1) but not in (B1), and Kate’s statement is true because it is more general, i.e. omits specifying the extra singular content.
With this agreed (hopefully), consider
(A2) Jake says ‘I am reading about Bilbo Baggins’ (he is reading The Hobbit)
(B2) Jake says ‘I am reading about a hobbit’
(C2) Kate says ‘Jake says he is reading about a hobbit’.
The first point also applies here. What Kate says is true of both (A2) and (B2). So does the second point – Kate is describing what Jake says, not what he thinks. If any problem occurring here is parallel to any problem about mental states, it follows that the problem is equally a problem of language, not (individual) psychology. (Perhaps it may be a problem about group psychology, if we agree that semantics is group psychology – that’s a separate issue).
The third point also appears to apply. The difference between (A2) and (B2) appears to involve generality. Specifically, what Jake says he is reading about appears to involve a singular meaning in (A2), but a general meaning in (B2).
Peter: >> The above proposal offers a gesture towards EO’s insistence that the surface structure of sentences about intentional states may not always reveal their deep structure.
It’s a nice gesture but I don’t believe the distinction you are drawing between ‘singular’ and ‘general’ thoughts is the right one. I believe that there can be genuinely singular thoughts which have no existing object (which I think you deny). This is not the same as the distinction we need to draw, which is between propositions (sentences) which have existential import, and propositions of a similar surface structure which do not. I have briefly tried to justify this claim above. The difference between (A) and (B) above appears to be between singular and general content. But the second (i.e. A2 and B2) involve non-existent objects (hobbits), so the distinction in question cannot be to do with how the world is.
This is closely related to the points I made in earlier posts about the common object of ‘worshipped Zeus’. The identity of the ‘intentional object’ seems to be numerical identity. Where you have numerical identity you automatically have other issues like explaining singularity, individuation, haecceity and singular concepts and so on and so on.
My POV on this (very briefly) is (1) that there are singular and general meanings, and that this distinction underlies the distinction between singular and general thoughts. I.e. we have to describe a singular thought using a term that has singular meaning. (2) Singular thoughts can be empty just as general thoughts can. (3) the distinction is linguistic/logical, rooted in language and semantics, and nothing actual, such as an ‘intentional object’. The description of certain thoughts requires an accusative, but the accusative is logical/semantic only. (4) Singular meanings do not require ‘haecceities’. We do not explain individuation in terms of features of the world, but rather in terms of logic or language or semantics.
Posted by: Edward Ockham | Thursday, January 06, 2011 at 04:02 AM
Peter concludes >> The above proposal offers a gesture towards EO’s insistence that the surface structure of sentences about intentional states may not always reveal their deep structure. As I have argued above, this is exactly what happens when the putative intentional object does not exist. Then what appears on surface to be a singular thought is in fact a general thought with a totally different structure. However, agreement ceases at this juncture.
I find it hard to believe that the 'deep structure' of a sentence (surely a very abstract thing?) depends in any way on the existence or otherwise of the object(s) to which it seemingly refers. This for the reason Bill gives, that the world may change the status of the objects without reference to the sentence. Likewise, the deep structure of an intentional state cannot depend on the status of its object. We could place GM-seeking Jake down in a possible world in which the GM did exist and his behaviour (initially) would be just the same.
I don't agree with Bill when he says of Tom's admiring Obama that the content of Tom's thought is general rather than singular. Surely Tom knows the difference between admiring an individual and admiring, say, good public speakers. But it does seem right to say that it's only via something akin to a definite description that Tom can access Obama. The fact that we can't substitute a verbal description of Obama into sentences expressing intentional states directed towards him without change of meaning doesn't imply that Tom's accessor for Obama can't be expressed in words. It just means that it's not made of words.
Bill also says he would question whether there are any truly singular thoughts. When someone says 'x is not a true F' or 'there are no true Fs' I suspect that there is a conflation of nearby concepts occurring, or a conflation of points of view. In this case, from the first person pov there most definitely are singular thoughts, directed, say, towards Obama. The previous sentence induces one. On the other hand, if whatever it is by which I access Obama is some kind of description, then from a third person pov, it may well appear general. For when I use it in the world as it is then it seems to latch on to an individual. But when I express it to you as best I can in words it will be a description, ie, general.
Posted by: David Brightly | Thursday, January 06, 2011 at 04:47 AM