Cheryl E

Cheryl E. Fitzgerald

Belief De Re: Causal Connections, Information, and Mistaken Identity Beliefs

Restricting myself to beliefs about concrete particulars and not abstract or conceptual objects[1], I declare that the distinguishing factor in determining whether a belief is de re or not is the origin or cause of the belief, not the nature of the belief itself—what is in the mind of the believer. While the nature of the belief itself is of philosophical interest, we cannot ignore the cause of the belief, for it is equally interesting, since beliefs are, in fact, caused and do not pop into our heads from nothing. It is not my intent to give any attention to the debate regarding whether beliefs are propositional or not, for my discussion is not about the belief itself. The meat of this essay is a response to “Self-Ascription and Belief De Re” by Neil Feit, in Philosophical Studies 98: 37-51, 2000, in which Feit presents what he believes to be a problem for the self-ascription theory and his solution to that problem.

The cause of a belief de re is the relation between the believer and the object believed about, a relation of acquaintance as it is historically called, but more clearly a relation of causal rapport. David Lewis made famous this explication of the relation—of which I need not quote this well-known passage—as a transmission of information from the object believed about to the believer, i.e., an external set of conditions.[2] Prephilosophically speaking, I have a belief about the candle on my desk because it is there in front of me and I am looking at it; I wouldn’t have a belief about it if it was never there, if I had never come in contact with it. My belief is not spontaneous, not even a choice: it is caused by an external state of affairs. This state of affairs is what defines my belief as de re; thus, this state of affairs is what calls out for our attention and analysis.

The information-theoretic approach describes the state of affairs: information is transmitted from an object to a conscious perceiver, where the information can be interpreted by the perceiver and a belief thus formed.[3] The event of a transmission of information is referred to as an information channel, the relation of acquaintance, causal rapport, all of which I use interchangeably with each other. A precise description of what takes place when the interpreted information causes a belief to be formed is left out of this essay, for it goes far beyond the scope here, and must be taken up in the philosophy of mind and cognitive science. “Information” is a technical term that refers to a “lawful dependence between signals.”[4] Light waves carry information about the objects they bounce off of, such as color and shape; sound waves carry information about the objects that made the disturbance, such as the type of instrument and how far away it is; the surface of an object carries information about its texture and perhaps material. Language also carries information in the same way, but one must be careful not to misunderstand the nature of this information: the information is carried via the sound waves and the specific collections of different sounds that we interpret as syllables, words, and sentences that act as signs for us. Even when someone is speaking French to me, and I do not understand French, information is still being transmitted to me as signals, I simply lack the correct knowledge to read most of them, but I can interpret some of the syllables and how they might be spelled out in letters, for example.

Causal rapport must not be defined too strictly: one need not be standing in direct perceptual acquaintance with an object. I can form beliefs de re about you if I am looking at a photograph of you, or if a friend of yours is telling me about you. De reness can be transferred and/or communicated from one person to another, from a book to a person, from the TV to a person, from the internet to a person, etc.[5] The point to understand regarding the transmission of de reness is that the information channel, or causal rapport, does not stop at the person in perceptual acquaintance with the object, but can be continued on to another person, or a letter, or a chat on the internet, etc.

Causal rapport is a necessary external condition for the existence of a belief de re; it is what we look to in order to find out what the belief is about, even when the believer is mistaken in what he thinks his belief is about.[6] The world dictates our beliefs: we do not choose to accept the information we receive, we do not even choose to interpret the information that catches our attention—we have been “built” by evolution to do so.[7] Our mental lives are a product of the continual ballet between our environment and our selves—however the ‘self’ is to be understood—nature v. nurture. When we desire to understand why a person behaves the way he does, we often look to his environment and how it affected him. The situation is no different here: what makes a belief, in fact, de re is that the believer stands in causal rapport with the object the belief is about.

Turning now to Feit’s article, I want to pay specific attention to the hypothetical scenario he claims presents a problem for the self-ascription theory. Frank has a crush on Fran, so he employs his friend Fred to write letters to Fran for him, signed as a Secret Admirer—Frank is not very good at romantic writing. Each letter expresses a bit of information about Frank, such as the fact that he is wealthy. Poor Fran has an honest and understandable mistaken identity belief: she believes that the writer of the letters and the secret admirer are identical. The question is: does her mistaken identity belief result in her forming beliefs de re about Fred regarding the information about Frank that she reads in the letters? Does she believe about Fred that he is wealthy?

Feit claims that she does not have a belief de re about Fred that he is wealthy. However, Feit claims that under the definition of belief de re according to Lewis’s self-ascription theory, it turns out she does. But, says Feit, this does not agree with the facts of the case.[8] He redefines belief de re as such:

S ascribes F to x if there is a relation of acquaintance R such that (i) S bears R uniquely to x; (ii) S self-ascribes the property of bearing R uniquely to a thing that has F; and (iii) if there is an object y, and a relation of acquaintance R¢, such that S’s self-ascription in (ii) depends on S’s having a belief b that mistakes x, under R, for y, under R¢, then S has b in virtue of thinking that the thing to which S bears R and the thing to which S bears R¢ are similar.[9]

The reformulation he presents is in his condition (iii), which states that if there is another object y that S is acquainted with, and S believes that because of the ways he is acquainted with x and y, they are the same object (the mistaken identity belief), S ascribes a property F, that in reality is exemplified by y, to x because of this mistaken identity belief, then S has a belief de re about x that it is/has F. It is best to draw upon the classic example of twin tennis pros to make this clear, and Feit uses this example. This particular case is also important in order to emphasize a distinction between it and the Secret Admirer case; Feit claims the distinction makes the latter a counterexample to the self-ascription theory.

The twin tennis pros case was first presented by Kent Bach[10]: it involves a mistaken identity between twins, one of whose existence I am unaware. I know Tim Gullikson, the tennis pro, but I don’t know that he has a twin brother Tom. The only noticeable difference between them is that Tim is right-handed and Tom is left-handed. I show up late to a tennis match to see that Gullikson is beating Glickstein, however, it’s Tom, not Tim. I notice that Gullikson is playing left-handed, and thus I ascribe the property of being able to also play right-handed to the person I see, Tom, the lefty, despite that the property of playing right-handed is exemplified by Tim, whom I am acquainted with in my memory. Feit points out one specific aspect of the Gullikson case, in opposition to the Secret Admirer case, that allows ascriptions made primarily to one individual “to pass through the [mistaken] identity belief, so to speak, and attach to the other one.”[11]

The difference, Feit claims, is that in the Gullikson case, I have the mistaken identity belief because of a “perceived similarity”[12] amongst the two individuals, the twins look remarkably similar in physical appearance, last name, as well as profession; in the Secret Admirer case, however, Fran does not mistake the writer of the letters for her secret admirer because she perceives similarities between them.

But Feit’s differentiation between a mistaken identity belief in virtue of a perceived similarity of two people and one that is not is an exercise of an arbitrary limit as to what counts as a proper relation of acquaintance, or perhaps worse, a too strict definition of acquaintance. There is no reason for us to deny that Fran has a belief de re about Fred that he is wealthy just because she does not perceive some similarity between Fred and Frank. And there is nothing wrong with this result: Fran might, indeed, have a mistaken belief about Fred that he is wealthy. Feit merely argues from his intuition that Fran would not have this belief de re, because he feels this to be a wrong answer to the question of whether or not she does. Perhaps Feit and I don’t share similar intuitions.

Essentially, Feit lacks the notion of causal rapport, and as such, his understanding of the relation of acquaintance is messy and fumbled. Let me alter the scenario just a bit. Fran hires a P.I. to find out who is writing her letters. Amidst everything the P.I. brings to her is a photograph of Fred about to drop an envelope in a public mailbox, and the envelope is addressed to Fran and lacking a return address. Fran can even recognize the handwriting on the envelope. Fran will most definitely believe about the man in the photograph that he is wealthy—despite his modest attire. Now I ask: can we really deny that Fran now has a belief de re about Fred that he is wealthy? It would be absurd to try, especially if we imagine having this conversation with Fran personally: she will point to the man in the photograph and insist that he is wealthy. What about this relation of acquaintance allows for the de reness of her belief, while the previous relation, her reading his letters, does not?

The original scenario is more complicated than it seems, not only because we must be clear about what defines the relation of acquaintance, but also because of Fran’s usage of her concept the writer of the letters. If Fran thinks about the writer of the letters, and she uses this phrase attributively and not merely referentially, and believes that he is wealthy, then she does have a belief de re about Fred that he is wealthy. However, if she uses the phrase the writer of the letters referentially to refer to her secret admirer, then she believes de re about Frank that he is wealthy, because her intent was to refer to her secret admirer. We can imagine her saying that the writer of the letters is wealthy; when we ask her, “Who?”, she clarifies, “Oh, my secret admirer!” Here, her belief is directed toward Frank, and not Fred, because she used the writer of the letters as a referential description to refer to her secret admirer.[13]

If we take Fran to be using the writer of the letters attributively, then her situation is no different from the one in which she is pointing at the man in the photograph, the causal rapport in each situation is enough to pump out a belief de re. She can be holding the letters and thinking to herself, The man who wrote these letters to me is wealthy, and there is no reason to deny that she refers to Fred, for she is standing in causal rapport with him by having read letters written by him. The only reason we can have for denying that her belief is de re is if the information channel is broken, as they sometimes get. A perfect example is the game telephone: by the end of the line the message is so far from being about what it was at the beginning, the information has been lost, the information channel broken. Now, it would be quite a different debate if one wants to claim that the information channel is not complete, and thus, Fran does not have a belief de re about Fred that he is wealthy. But that is not what Feit is arguing.

Feit would insist that the requirement of perceived similarity is a necessary distinction to make between the Gullikson case and the Secret Admirer case; however, I think his position is more clearly understood using the notion of causal rapport. In the Gullikson case, I am standing in causal rapport with Tim, by remembering him, and I am standing in causal rapport with Tom, by watching him, and I am receiving information about each man. The key point is that some of the information I am receiving from each man is similar, such as physical appearance, being a tennis pro with the name Gullikson, etc. In the Secret Admirer case, on the other hand, Feit would say there is a lack of similar information being received from each man, from each information channel. There could be such similar information if, for example, Fran notices by the smudges of ink on the paper and the handwriting that the person who wrote them is left-handed, and that the letter informs her specifically that her secret admirer is left-handed. Information of left-handedness would be transmitted via both information channels, and thus she could have a de re reason, so to speak, for believing them to be the same man. But this is not the case.

However, information has been overlooked, for there is, indeed, some information being transmitted via each information channel that is similar, and it is the property of being Fran’s secret admirer. Clearly, Fran’s causal rapport with Frank, i.e., the content of the letters, transmits the information that he is her secret admirer. But her causal rapport with Fred transmits the same information: the letters contain indexical words, such as “I”, “my”, being signed as “Your Secret Admirer”, etc., and because of this, Fred is conveying the information that the person writing the letters is her secret admirer. It happens to be false about him, but it is information nonetheless. And it is similar information Fran is receiving about Frank, that he is her secret admirer. Thus, Fran does have a de re reason to form the mistaken identity belief between Fred and Frank, for she is receiving similar information about them. We may not be able to call it a perceivable similarity, but in order not to make the issue too complicated by introducing such a complex topic as perception, we can leave it at the notion of similar information. After all, I can describe to you what my brother looks like, in as much detail as possible, and you may then see someone you take to be my brother because he seems to fit the description I gave you. But we cannot really call this a perceivable similarity in the sense that Feit uses it, since you have no visual perceptions of my brother. But I do, and I am transmitting that de reness to you.[14] Essentially, there is no reason to require that the similar information received is perceptual—but then again, that will depend on how one defines perception, and that issue is beyond the scope of this essay.

If one is suspicious of my claim that the information Fran is receiving about Fred and Frank is similar in a significant way, let me draw an analogous situation that I think makes it more clear. You are supposed to meet with a woman with whom you have been set up on a blind date. She calls you, and in the midst of your telephone conversation, she gives you a description of herself and tells you that she does modeling. Now, in talking directly to her, you have a belief de re about her that she looks like [whatever description she gave you] and that she does modeling. At the designated place and time, you wait for her to arrive. You see a woman coming towards you who fits exactly the description you were given, and you even think to yourself that she does look like a model—her extravagant appearance, the way she carries herself, etc. You go on about your date and have a wonderful time.

What you didn’t know is that this woman is not the woman you spoke to on the phone, in fact, this person is not even a woman, but a man. The information you were given was about him, the description and the modeling—he models as a woman, so convincing that almost no one can ever guess he is really a man. You undeniably have beliefs de re about his appearance, as well as a belief de re about him that she models, even though it never came up in conversation, for the information you were told on the phone you ascribe to the person you met because of a mistaken identity belief. You were in causal rapport with two people you spoke to and interacted with directly, i.e., one on the phone and one in person, so we cannot deny that you have beliefs de re about them. You received similar information about each of them, and thus believe them to be the same person, even though you did not have any perceivable information about the woman with whom you spoke on the phone. In fact, the woman with whom you spoke on the phone does not even remotely fit the description she gave you, nor does she model, but you still have these beliefs de re about her. You are deceived, not because you misread the information, but because you interpreted it just as you were supposed to, for the information you received about the two people was similar, even though it was not similar because you saw two people that have similar qualities. Thus, we have no reason to require that the mistaken identity belief is caused by one’s perception of similar qualities in two people; all that is needed is that some kind of significantly similar information is received about each individual via causal rapport.

The result is that Feit’s reformulated definition of belief de re can apply to the Secret Admirer case: Fran can have a belief de re about Fred that he is wealthy. The reason is, of course, that her mistaken identity belief is a result of her mistaking one individual, under one relation of acquaintance, for another individual, under another relation of acquaintance. Feit claims this does not occur because Fran mistakes them first before having any reason to—she assumes that her secret admirer would write his own letters. However, she can have good reason to believe the writer of the letters and her secret admirer are the same person, for she is receiving similar information about each: the letters utilize indexical words and are all signed as “Your Secret Admirer.” Feit is wrong to say that Fran comes to have the mistaken identity belief because of an assumption. If we asked her what caused her to believe that her secret admirer wrote the letters, she would most likely tell us that the fact that the letters are signed the way they are, the presence of such sentences as, “I have an unusually large bank account.”, etc. are the reasons she came to believe the letters were written by the one who admires her. She would admit that if they were written in the third person, she’d have no reason to believe the writer is her admirer. But indexical words such as “I” refer to the one who utters them, or writes them, and not to someone else.[15] Obviously, this is a special use of the word “I”, such as if I were dressed up in costume and told you, “I am Napoleon”; there is something unuttered but implied. What I am really expressing to you is, “I am dressed up to look like Napoleon,” but because of the social situation, I need not utter this sentence, only the former one, in order for you to understand. But Fred has not given any clue that his is a special use of the word “I”, and so Fran lacks the needed information about this particular situation that would prevent her from having the mistaken identity belief.

Feit would claim that the way Fred, qua writer of the letters, is conveying information about himself, namely, posing as her secret admirer, is not enough to count as similar to Frank to cause Fran’s belief that the writer of the letters and the secret admirer are identical. He requires that the qualities be actual and perceivable similarities. However, not only do I find this requirement to be unnecessarily strict, Feit does not even make a real argument for it. The most he says about it is thus:

If…the subject has the identity-belief because she thinks that there is a [perceivable] similarity between the thing to which she bears R [a unique relation of acquaintance] and the thing to which she bears R¢ [another unique relation of acquaintance], then de re belief is possible. But if the subject has the identity-belief for some other reason, as in the case of the secret admirer, then there can be no de re belief.[16]

Feit believes that Fran has the mistaken identity belief because of a de dicto assumption, namely, that secret admirers write their own letters. Despite that I think Fran does have a de re reason to believe the writer and admirer are identical, i.e., because she receives significantly similar information about them, Feit gives no argument for his claim that a de dicto assumption about the identity of the individuals prevents ascriptions passing through the identity belief: he just asserts it as if it were inherently and clearly true. Even worse, Feit claims that the “facts of the case” do not result in Fran having a belief de re about Fred that he is wealthy[17], but he never clearly states what these facts are and why they are facts that she does not have such a belief. It feels as though Feit merely has this intuition, and again I have to state that I have differing intuitions than he does.

But if one is going to claim that Fran does not believe de re about Fred that he is wealthy, then we must argue that Fran does not use the description the writer of the letters attributively, but rather referentially, to refer to her secret admirer. We can say that she casually attributes the property of having written the letters only for the purpose of referring to the one who is her secret admirer. The attribution comes out false, but the reference is successful nonetheless. And this is the referential use of descriptions. Arthur Falk refers to this as the dominance theory:

[T]he believer can have intentions that her own conceptions of the things she has beliefs about can be overridden, dominated in her attitudes by the truth, if her conceptions of them contain error, with the effect that her attitudes still succeed in being about them, despite herself.[18]

This follows as a consequence of the information-theoretic approach, that the world, in an important way, dictates our beliefs; we “wrest control from the world of the ensuring of [our] attitudes’ referential success.”[19] We can see the dominance theory in effect commonly everyday. You and I are in a store and we pass by two foreigners; later I make a comment about them, referring to them as the two men who were speaking Spanish. Knowing exactly whom I was talking about, you reply, “You mean, the two men who were speaking Portuguese.” You don’t ask, you state it, because my referential use of the description was successful; the truth, that I am unaware of, dominates over my beliefs so that my beliefs really are about those two men, despite that my conception of them was mistaken. It is in this way that we can say that Fran does not believe de re about Fred that he is wealthy, because she uses the description the man who wrote me these letters referentially.

If we think about my earlier addition to the scenario to include Fran’s P.I., we can ask ourselves what might happen if Fran confronts the man who did in fact write the letters, Fred. Let’s imagine her introduction to him in two steps. First, at introduction, all her beliefs about her secret admirer will be ascribed to Fred, i.e., she will think to herself, Ah, here is the man whom I have been thinking about. But then she finds out that he is not wealthy. She would probably be angry that she was lied to, because she will still think that Fred, qua writer of the letters, is her secret admirer. Second, she finds out that he is not her secret admirer, but just the author of the letters. Most likely her reaction would involve a moment of pause, followed by an, “Oh.”, and all of the beliefs she had about Fred, as she mistook him for her secret admirer, would be retracted. She might ask, “Well, then where is my secret admirer, the man who is wealthy?” This is a clear indication that her conception of the writer of the letters was purely referential, not attributive, and that she intended her beliefs to be about her secret admirer. Despite that her conception of her secret admirer contains an error, that he wrote the letters, her belief about the writer of the letters that he is wealthy successfully points towards Frank instead of Fred because her belief is not really about the actual writer of the letters.

In order to show how this is the case, a strict regimentation of belief reports is absolutely necessary. I use the regimentation that Arthur Falk uses in his book Desire and Belief[20], as I have found it to be the most clear and correct. In the simplest presentation, a belief de re is reported as such: a subject S believes about some object x that it is (has the property) F. The “about-position” contains the actual object the belief is about, while the “that-clause” contains what the subject believes of that object. If the subject is mistaken as to what her belief is about, then the truth of her mistake can be revealed, and set straight without contradiction, in the “about-position” of the regimented belief report. Since Fran is using the writer of the letters referentially, and not attributively, then her belief is not about the writer of the letters, Fred, it is about her secret admirer, for that is who she intends to pick out. Regimented, Fran believes about her secret admirer, namely Frank, that he wrote the letters and is wealthy. When she believes, The writer of these letters is wealthy, she uses the phrase referentially to pick out her secret admirer—whom her belief is correctly about—but then mistakenly attributes to him the property of being the writer of the letters. To say that she uses the phrase referentially, even though she does attribute the property to her secret admirer, is only to say that she does not use the phrase to pick out the individual of whom it is true, but rather to pick out the individual she intends to pick out, to whom she mistakenly attributes this property.

One might ask how this can fit in with my earlier argument, that because Fran receives similar information about both men, she can have a belief de re about Fred that he is wealthy. Essentially, each argument is a different perspective on the scenario, for Fran can mentally use the phrase the writer of the letters either attributively or referentially in her belief, and the situation will look different for each use. But Feit does not specify which use he takes Fran to be using, and thus the scenario is more complex than he makes it. In my first argument, Fran uses the phrase attributively: causal rapport can fix the reference of a description for a believer.[21] Fran can recognize the causal rapport she stands in relation to Fred and use that to fix the reference of her use of the writer of the letters; the reference would be something like: the man with whom I stand in the information channel that begins with the writing of these letters. But in my second argument, Fran does not fix the reference of the phrase this way: she does not recognize the causal rapport she stands in relation to Fred as a reference fixer. Rather, she immediately thinks of her secret admirer as the one who wrote the letters, and thus uses the phrase referentially to refer to Frank. She is not really interested in the writer of the letters; she is interested in her secret admirer. And despite that there is information conveyed in the letters that the writer is the admirer, i.e., that similar information Fran should be receiving about Frank and Fred, if Fran is using the writer of the letters referentially and not attributively, then the information channel between Fred and Fran that should convey that similar information is broken, i.e., Fran does not receive and process that information. It would seem that it is the second explication of the scenario that Feit has in mind.

As I have shown, the reason that Fran does not have a belief de re about Fred that he is wealthy has nothing to do with lacking any perceived similarity between the two mistaken individuals. The self-ascription theory—fleshed out much more fully than the simple definition of belief de re that Lewis gave, which is the only depth that Feit gives the theory—does not fail to give the “right” answer, according to what Feit believes that to be. It can give the right answer, as long as one is willing to take a fleshed out version of the self-ascription theory, namely, one that includes the notions of causal rapport and information, as well as a clear understanding of the distinction of two possible conceptions of a description, referential use and attributive use. Thus, it would appear that Feit’s understanding of the self-ascription theory is merely lacking a full body. The Secret Admirer case is not a counterexample to the self-ascription theory: it can be tackled, without Feit’s reformulation of the definition of belief de re. What is needed is not any sort of reformulation of the definition, but rather a full fleshing out of what that definition entails, and all of the factors involved. A simple definition can never be taken as the extent of a theory. Thus, I accuse Feit of jumping the gun before he has taken in the whole scene.