January 12, 1972

Professor Annette Baier
100 Maple Heights Road
Pittsburgh, Pa. 15232

Dear Annette:

Your letter is such a rich tapestry of issues and assumptions that it is with more than a little trepidation that I undertake a reply. Yet the issues are so central and the assumptions, or at least some of them, so plausible, that even a partially successful reply would lay the groundwork for a far more effective exchange of ideas on intentionality than our tentative efforts of last November.

Since the specific thrust of some of your questions seems to me based on a isapprehension of my views, I shall not take them up in the order in which you raise them. Rather I shall attempt to delvelop a framework in terms of which I can reinterpret your questions in a way which preserves their general objective and their challenges while enabling me to give something like answers.

I shall begin with some remarks which, though they may seem to suggest that our intuitions are so different that the prospect of agreement is almost beyond hope, may, nevertheless, help locate the roadblocks which stand in our way. My first remark concerns your claim (pp. 2-3) that whereas

(a) Jones fears Smith, as an examiner

admits of the substitution of "co-referential expressions for 'Smith,' " i.e., in your terminology, constitutes an 'S-extensional' context for 'Smith.'

(c) Jones fears Smith
(d) Jones fears the examiner

on the other hand do not. Your intuitions seem to tell you it would be incorrect to substitute "the secretary of the orchid fanciers' club" for "Smith" in (c) and for "the examiner" in (e), even though

-2-

we are given that

Smith = the examiner = the secretary of the orchid fanciers' club.

I think that you are just wrong about this. You write,

I suggest that we can explain the apparent S-intensionality of (c) and (e) [I note in passing that you originally characterize this S-intensionality as "straightforward"] if we suppose that a perspicuous rendering of any intentional sentence will contain an as-phrase, and that when it is omitted, one assumes that the noun in the omitted as-phrase would duplicate the noun which is the direct object of the intentional verb. (P. 2.)

I shall neglect the sweeping reference to "all intentional sentences" and focus attention on sentences like (c) and (e). What, then, are you claiming? Surely that, "perspicuously rendered,"

Jones fears the secretary of the orchid fanciers' club

becomes

Jones fears the secretary of the orchid fanciers' club qua secretary of the orchid fanciers' club

and that

(e) Jones fears the examiner

becomes

(e') Jones fears the examiner qua examiner.

You even suggest that, perspicuously rendered,

(c) Jones fears Smith

becomes

(c') Jones fears Smith qua Smith.

I am amazed that the absurdity of this last examples which you so bravely swallow with the help of Kaplan sauces¹ did not open your

---

¹I am not sure that I understand what it would be to fear Smith qua Smith. I can understand the contrast between fearing Smith on the whole and fearing Smith qua examiner -- of this more later. But the former is surely not to be identified with fearing Smith qua Smith.

---

-3-

eyes to the fact that in demanding that "perspicuous renderings" include qua-phrasess you are demanding that the explanation of a person's having an attitude be included in the very statement which ascribes it to him.

I don't think I can be exaggerating, for as you use 'perspicuously render' it seems to be equivalent to 'make explicit' (see 1.9 from bottom on p. 2). There might be some merit to arguing that if I say "Jones fears the examiners" I 'imply' in something like Grice's sense and, depending on the context, to a greater or lesser degrees that Jones fears the examiner because he (the examiner) is the examiner. But there seems to me to be no merit whatever to the claim that a reference to a ground of Jones' fear of x, in the sense of 'ground' in which, for example, the fact that Smith is an examiner is a ground of Jones' fears is part of the very meaning of 'Jones fears x.'

In other words I think that it is just as wrong to suppose that a qua-phrase is implicit in the meaning of

Jones fears the secretary

as it is to imagine that there is an implicit because-clause, and for the same reason: qua-phrases and because-clauses are means of adding an explanatory component to the context in which they occur.

Yet qua-phrases, though providing an 'explanatory component', are not simply another way of saying "because." To draw on an example from the last paragraph of your letter, I can admire Smith in several respects. I can also scorn him in other respects, and be indifferent to him in still others. We can therefore distinguish admiring Scott all things considered, from admiring Scott in one particular respect or group of respects. Of courses not all uses of qua-phrases involve

-4-

distinctions corresponding to that between admiring someone all things considered and admiring him in a certain respect. Sometimes the character in question is simply either present or absent, and the qua-phrase picks out the respect which is the ground of its presence, thus,

John chose the cube qua die

Here we are obviously in the neighborhood of

Jones chose the cube because it was a die

and the more guarded

Jones chose the cube because he believed it to be a die.

It might be helpful here to pause for a moment and take into account the occurrence of qua-phrases in contexts which are not mentalistic. Consider

(1) The fire was caused by Mrs. O'Leary's cow.

Surely any co-referential expression (e.g. "Bossy") can be substituted salva veritate for "Mrs, O'Leary's cow." The fact that a qua-phrase can be attached, thus

(2) the fire was caused by Mrs. O'Leary's cow qua that which overturned the lantern

would scarcely tempt us to suppose that (1) is misused unless it could be "perspicuously rendered" or "made explicit" as

(3) The fire was caused by Mrs. O'Leary's cow qua Mrs. O'Leary's' cow.

Notice, also, that (2) is in the neighborhood of

(2') The fire was caused by Mrs. O'Leary's cow because (or, by virtue of the fact that) it kicked over-the lantern.

-5-

This enables me to make the point that although statements of the form

(4) x was caused by y

are conceptually tied to statements of the form

(5) x was caused by y qua φ-ing

(things are causes by virtue of something they do.) This does not mean that unless we add a qua-phrase to a statement of form (4) we would be asserting that which would be perspicuously rendered by

(6) x was caused by y qua y.

Rather it means that when (4) is made explicit, it has the form

(4') (Ef) x was caused by y, qua f-ing (i.e. x was caused by y, qua doing something)

.

The parallel point, which, I believes was at the back of your mind when you wrote the last paragraph on p. 2. is that although the form

(7) x fears y

is conceptually tied to

(8) x fears y qua φ

(Things are feared by virtue of some character they are believed to have. This does not mean that unless we add a qua-phrase to a statement having form (7), we are asserting that which would be perspicuously rendered by

x fears y qua y.

Rather it means, at most, that made explicit, (7) has the form

(7') (Ef) x fears y qua an (or the) f (i.e. x fears y qua being of a certain character).

-6-

But I am getting ahead of my story. My primary aim, up to this point, has been to give some plausibility to the idea that qua-phrases are explicable -- though not in any simple way -- in terms of because-clauses. In the special case of psychological attitudess the because-clauses characteristically concern the subject's beliefs. Thiss however, does not mean that psychological attitudes are reducible to states of belief accompanied by bodily sensation and propensities to behave, even if 'accompanied by' is construed as involving a causal relationship between the beliefs and the sensations-cum-behavior. Attitudes involve not only beliefs but intentions in that broad sense which includes appetites, desires and aversions (roughly, intention candidates). Thus, in first approximation, fear involves the belief that something threatens harm. But it does not consist in this belief together with unpleasant bodily sensations and, for example a propensity to shudder. It also involves something like the thought 'Would that I avoid harm.' In other words, when I say

Jones fears x because x is about to examine him

I attribute to Jones something like the reasoning

x is about to examine me
One who is examined is likely to suffer
Would that I avoided suffering
So. would that I escape from this situation.

This thought-nexus, accompanied by unpleasant sensations and, perhaps, shudders, consists of more than Jones' belief that x is about to examine him, which belief gives its content to the because clause. For the reference to the believed possibility of harm or suffering is implicit in the term 'fear,' and the because-clause picks out the distinctive feature of Jones' other beliefs which explains why Jones finds his situation fraught with that possibility. Thus my statement implicitly attributes to Jones the premise

One who is examined is likely to suffer.

-7-

I must emphasize, then, that at no stage in our discussion was I concerned to reduce the intentionality of attitudes to the intentionality of beliefs. My concern was rather to stress the propositional character of intentionality -- where 'proposition' covers both theoretical and practical thoughts -- by showing how the use of qua-phrases in characterizing psychological states can be explicated in propositional terms. That, for reasons indicated above, I found the point of qua-phrases to consist in picking out beliefs by no means involved in a commitment on my part to the idea, which you attribute to me in your first paragraph, that the intentionality of a psychological state "is concentrated in the implied belief-that," and that the remainder consists of non-intentional items such as bodily sensations and behavior clustering around the belief.

These remarks may explain why I am unhappy about your interpretation of my views in terms of your (a_s), and why I pass over your first comment on (a_s) at the bottom of p. 3. On the other hand, since non-intentional elements (which, although they involve sensations and feelings, may be represented, for our purposes, by the term 'shuddering') are characteristic of fear, I can deal separately with your second comment (top p. 4). There you raise the problem of cross-reference:

You must quantify over individual concepts to express [characterize?] the beliefs, but you need cross reference not to the concept Smith, but to Smith himself, if 'shudders at' is a behavioral relation. Will you construe it also as involving the concept Smith. If so, there will be no need to transform fears into beliefs, since fearing behavior has now become intentional.

To this, the answer is the quantification over intensions (e.g. individual concepts) by no means has the consequence you ascribe to it. Thus consider

-8-

(8) Smith is tall and Jones believes that Smith is the tallest man in Pittsburgh

To capture both references to Smith by a quantified variable we need only make use of the fact that

Smith is tall

is logically equivalent to

that Smith is tall is true

in which 'Smith' stands for the individual concept Smith. The latter clause is, in turn, logically equivalent to

that he is tall is true of Smith

where, again, 'Smith' stands for the individual concept Smith. Thus (8) above is equivalent to

That Smith is tall is true and Jones believes that Smith is the tallest man in Pittsburgh

and to

That he is tall is true of Smith and Jones believes that Smith is the tallest man in Pittsburgh.

the relevant quantified counterpart of these are, respectively,

(Ei) That i is tall is true and Jones believes that i is the tallest man in Pittsburgh.
(Ei) That he is tall is true of i and Jones believes that i is the tallest man in Pittsburgh.

Thus, if we look ahead to transparent belief, the following is available

(Ei) i ME Smith and that he is tall is true of i and ^oB_j i is the tallest man in Pittsburgh.

None of this involves construing tallness as an "intentional" state or attribute. The same considerations hold of the following example:

Jones shudders at Smith and believes that Smith threatens Jones

-9-

This becomes

(Ei) That Jones shudders at i is true and ^oB_j i threatens Jones.

or, if we wish to construe the belief as transparent, the following are available:

(Ei) That Jones shudders at i is true and i ME Smith and ^oB_j i threatens Jones.
(Ei) That Jones shudders at him is true of i and i ME Smith and ^oB_j i threatens Jones.

This no more requires that shuddering be intentional than the previous example requires that tallness be intentional. Needless to say, an adequate theory of reference must explain the sense in which the two occurrences of 'Socrates' in

Socrates is wise and it is true that Socrates is snub-nosed

are 'co-referential,' even though, strictly speaking, the first occurrence refers to Socrates and the second to the individual concept Socrates. Clearly we have here an extended sense of 'co-referential' which is to be justified in terms of semantic theory. The problem is closely related to the traditional problem of explaining the sense in which

snow is white

and

that snow is white is true

give us "the same information." But that is a story for another occasion.

 

I postponed taking up the third problem you raise on p. 4 save insofar as it hinges on the assumption that the reference to Jones' belief that Smith is an examiner comes in via conjunction, rather than via the context

. . . because Jones believes that Smith is an examiner

-10-

yet an introductory comment is in order. You argue that (a_s), which you attribute to me as the analysis of 'Jones fears Smith qua examiner,' involves transparent belief. "It seems to meet your definition of transparent belief, in that Jones does ^tbelieve that Smith is dangerous, since he believes that Smith as examiner is dangerous and Smith as examiner is materially equivalent to Smith." (p. 4.) I was puzzled by this, since (a_s) does not contain

Jones believes that Smith as examiner is dangerous

(I would scarcely end my analysis of qua-phrases in general by tracing qua-fears to qua-beliefs.) It then occurred to me that you are tacitly construing me to hold-that

Jones believes that Smith is dangerous and Jones believes that Smith is an examiner

to be equivalent to

Jones believes that Smith as examiner is dangerous.

I most certainly do not. And this brings me to the comment on your "thirdly" which I wish to make at this time.

I begin by noting that in the paragraph which follows the one introduced by "thirdly," you suddenly realize that I might have something stronger in mind than conjunction and/or temporal coincidence, and you consider the possibility that I might be relying on a counterfactual. Having come this close, however, you dismiss the idea that

(1) Jones fears Smith as examiner

can be analyzed as

(1') Jones fears Smith and Jones would not fear Smith unless he believed that Jones is an examiner

on the ground that "Jones fearing Smith in both capacities is compatible with 'Jones believes Smith is secretary and fears him as examiner.'" I am not clear exactly what the argument is, but the

-11-

crux of it seems to be that (1') implies that if Jones did not believe that Smith is an examiner, he would not fear him, whereas we know that Jones might fear Smith even though he didn't believe him to be an examiner, because he might also fear Smith as secretary. The thrust of this argument (if I have it right) is that the attempt to explicate qua-phrases in terms of counterfactual conditionals about beliefs cannot cope with the aspectual character of attitudes and emotions. The latter is a recurring and central theme in your letter and it is time that I dealt with it directly.

Suppose I were to claim that

(3) Jones admires Scott qua author of Marmion

can be explicated as

(3') Jones admires Scott because he believes that he is the author of Marmion [which he believes to be an excellent novel].

You might begin by pointing out that (3) is compatible with

(4) Jones admires Scott qua conversationalist

whereas (3') is incompatible with

(4') Jones admires Scott because he believes that Scott is an excellent conversationalist

To this you might add that both (3) and (4) are compatible with

(5) Jones does not admire Scott all things considered (or on the whole or period).

After all, the following makes sense,

(6) Jones does not admire Scott (sans phrase), though he does admire him qua author of Marmion and qua conversationalist.

But all of this, while true, relies heavily on the subtle logical grammar of verbs expressing attitudes towards things or persons. I will not attempt to sketch a theory of this grammar,

-12-

but will limit myself to some provisional formulations which point in the direction of my general thesis, i.e. the thesis that the primary form of intentional objects is propositional.² The general

---

²I have also consistently neglected the niceties which are required by taking into account propositional nominalizations other that-clauses.

---

sort of thing I want to say can be put as the claim that the form

(A) x fears y

where y is a thing or person, is ultimately to be understood in terms of the form

(B) x fears that y will Z (and, by Z-ing, W him, which would cause him harm].

A limiting case of (B) is

(C) x fears that y will do something [which will somehow cause him harm].

I said that (A) is 'ultimately' based on (B) because, as we shall see, there is a wide range of locutions which mediates between them.

Now I take it as evident that x's fear that y will Z involves the thought (on x's part) that y will Z. Similarly, to add an additional element of realism, x's fear that y might Z, involves the thought the y might Z. (I have already noted that the practical thought that harm is to be avoided is also involved, as well as that fear involves characteristic feelings and behavioral tendencies.) These considerations enable us to explain why, on a given occasion, x fears that y might Z, by a reference to what he believes.

-13-

x fears that y might Z because he believes that y might Z and that a Z-ing by y would (directly or indirectly) threaten him harm.

How do we get from 'x fears that y might Z' to 'x fears y'? We might begin by introducing the locution (which is not a pure invention)

y has a fear of y

interpreting this to mean

(EZ) x fears that y might Z.

This would be an initial step in interpreting 'fear of' in terms of 'fear that.' The next step would consist in noting that a person can have several fears-that which pertain to the same thing or person. He can also have both hopes and fears with respect to the same thing or person. If it makes sense to speak of 'summing' fears, 'summing' hopes, and 'balancing' hopes against fears, we could say that

x fears y on the whole as

the sense of

x's fears of y outweigh his hopes of y.

Clearly all this is a (rather makeshift) program, and not a result. Let me throw one more item into the hopper in addition to the verb 'to Z' we need the noun 'Z-er' (one who, or that which, can be expected to Z when the appropriate circumstances obtain). This would enable us to relate

x fears y qua Z-er

to

x fears that y will (or might) Z [and by Z-ing directly or indirectly harm him].

-14-

Thus

x fears y qua Z-er

would entail

x has a fear of y

but not an unqualified

x fears y.

It would, on the above analysis, be a mistake to explicate

x fears y qua Z-er

as

x fears y because he believes that y is a Z-er.

But it would not be obviously wrong to explicate it as

x has a fear of y because he believes that y is a Z-er.

My aim in these last paragraphs has been to defuse that part of your argument for the indispensability of qua-phrases which hinges on the fact that one can admire Scott in one respect and despise him in another. Notice that I am not claiming that what is conveyed by qua-phrases is dispensable. It is, indeed indispensable. What I am claiming is, rather, that this indispensable information is ultimately to be understood in terms of because-clauses, i.e. in propositional terms.

In our discussions my basic point about your use of mythological examples was the methodological one: First become clear about the role of qua-phrases in contexts which do not introduce a mythological or fictional universe of discourse. When that has been dones one, one

-15-

should be in a position to see what, if any, interesting features of qua-phrases emerge in fictional contexts. Otherwise one runs the danger of starting too many hares at once. There are, of course, interesting puzzles about reference to fictitious entities. I do not, however, see that

Jones admires Ahab qua sailor but not qua monomaniac

or

Ordinary readers despise Uriah Heep as a hypercrite, but politicians despise him for his clumsiness.

raise any problems about qua-phrases qua qua-phrases than your example about Scott. But let's take up the whole question of fictional reference in a separate discussion.

I return now to your "thirdly" paragraph on p. 4. You argue that if

Jones believes of Smith that he is dangerous

is true, then so are

Jones believes of Smith as examiner to be that he is dangerous

and

Jones believes of Smith as secretary that he is dangerous.

You then look back at the view, which you attribute to me, that

Jones fears Smith qua examiner = Jones shudders at Smith and/when he believes that Smith is dangerous and/when he believes that Smith is examiner (a_s).

and argue that if we interpret the beliefs in (a_s) as transparent, it follows that "Jones fears the secretary of the orchid fanciers' club and what is more fears him as such."

Now the form of the argument is not clear. Part of what you need to show is that there is a way, to the legitimacy of which I

-16-

am committed, of transforming (a_s) into a counterpart which ends in

. . . and/when Jones believes Smith is secretary.

(From now on I simplify by dropping out the 'when'). As I reconstruct the situation, it occurred to you at this point, that if you find me committed to the move from

Jones believes of Smith that he is dangerous (i.e. ^tbelieves that Smith is dangerous).

to

Jones believes of Smith as secretary that he is dangerous

then you are home free for the latter surely entails

Jones believes Smith is dangerous and Jones believes that Smith is secretary.

It is for this reason that you appeal to the (supposed) co-referentiality of 'Smith,' 'Smith as examiner,' and 'Smith as secretary'. For the co-referentiality of 'Smith' with the weaker phrases 'the examiner' and 'the secretary' clearly cannot justify the move from

. . . Jones believes Smith is dangerous and Jones believes that Smith is examiner

to

. . . Jones believes that Smith is dangerous and Jones believes that Smith is secretary.

That you are (rightly) uneasy about this use of 'Smith as examiner' emerges on p. 5, where you ask "was it correct of me to say that Smith as examiner ME Smith?" To fix our ideas, let us consider

(1) Jones believes of Smith that he is dangerous.

Given that Smith is the examiner, it follows that

(2) Jones believes of the examiner that he is dangerous.

-17-

Does it also follow that

(3) Jones believes of Smith as examiner that he is dangerous?

Clearly not. For it doesn't follow from (1) together with the above assumption that

(4) Jones believes that Smith is the examiner

but (3) surely entails (4). Indeed (3) entails that Jones' belief that Smith is dangerous is (at least in part) grounded in his belief that Smith is the examiner. This strongly suggests that qua-phrases cannot simply be put in a box with names and descriptions, but play a much more complicated role which must be contextually explicated. The modifier 'as examiner' doesn't simply build another referring expression out of 'Smith' as does, for example,

Smith, the examiner

Let us take another look at qua-phrases in non-mentalistic contexts, It is not implausible to hold that

The cube as a piece of celluloid is inflammable

is, roughly, equivalent to

The cube is inflammable because it is a piece of celluloid.

Suppose that the cube is a die. We can derive

The die as a piece of celluloid is inflammable

but not

The cube as a die is inflammable

for this would entail

The cube is inflammable because it is a die.

which is, of course, false.

-18-

Notice that a fuller discussion would require recognition of the fact that things can have causal properties by virtue of several facts about them (in which case we have parallels to admiration and fear), Thus,

Smog is unhealthy both qua irritant and qua allogen.

Would you not agree that an account of qua-phrases in such contexts can be given in which they are analyzed in terms of the explanatory use of that-clauses?

I come now, as I near the end, to your beginning. You write

In Science and Metaphysics you say that a complex like "this white thing" is essentially incomplete, a sentential fragment, and its complexity is derivative of the sentence conplexity or predicative character of "this is white."

You infer that I would give this as a reason for claiming that

He fears this white thing

is to be analyzed in such a way that the analysand contains the sentence 'This is white' in an intentional context such as 'He fears . . .' or 'He believes . . .' The answer is that I would do nothing of the sort. I have already argued that

He fears this white thing

in a context which permits the substitution of any co-referential expression. Hence it is by no means equivalent to

He fears this white thing qua white.

The most that the SM passage commits me to is that

He fears this white thing

is related to

This thing is white and he fears it

as

The table is larger than this white thing

-19-

is related to

This thing is white and the table is larger than it.

[I use this circumlocution, because, for Donellan-like considerations, the referential use of phrases of the form 'this φ' cannot be simply dissolved away into the referential role of 'this'.] Obviously the phrase which identifies the object of a persons fear need have nothing to do with the respect in which he fears it. Curiously enough, it is you who seem to hold that

He fears this white thing

when made fully "explicit" or "perspicuously rendered" becomes

He fears this white thing as white

and, hence, to finding yourself surrounded by leukiphobes.

I come now to your "last query." You write:

The fundamental intentional context for you is not belief, nor intention, nor any other propositional attitude, but the semantic attitude of taking 'red' as a red, [I assume you mean 'a ·red·'].

You point out that taking something as something is a paradigm of q-intentionality, and so are puzzled

. . . why you want all the derivative cases of intentionality to take a propositional form, when the parent case does not?

The answer is that the assumption of your query is false. There is, indeed, such a thing, as taking a linguistic episode to be of a certain kind. Thus it makes sense to say

Jones heard the episode as a ·red· (or a ·Tom is tall· or a ·2 + 2 = 4·)

But these examples have the same form as

Jones saw the object as a bear (or pawn, or executioner)

-20-

and are to be given the same type of analysis. An initial or crude attempt would take the following form

Jones responded visually to the object with the thought ·this is a bear· (the thought that this is a bear)

Correspondingly, we would have

Jones responded aurally to the episode with the thought 'this is a ·red·' (the thought that this is a ·red·)

If so, then you would not have found a counter example to my analysis of qua-phrases in terms of that-clauses.

But I suspect fhat what is misleading you is the fact that for me what is fundamental about intentionality³ is the ability of people

---

³Which is by no means the same as 'what is the fundamental intentional context.'

---

to emit verbal behavior which conforms to the criteria for being classified as ·red·s, &middit;Tom is tall·s, ·2 + 2 = 4·s, etc. It is simply wrong to speak of this ability as the ability to take various behaviors as ·red·s, etc. To run these together would be, as I see it, a paradigm case of confusing (what is fundamental to) language as thought with (what is fundamental to) language as communication.

Well, on re-reading what I have written, I am depressed by its high density and promissory-note-ishness. I doubt if it achieves one-tenth of what I hoped to achieve when I began it. But at least it places our discussion in a larger context and it may just happen that one or two of the things I have said will turn some ducks into rabbits.

As ever,

Wilfrid