2. Pessimism in education

I will start with a passage from moral philosophy.

I assume that ‘vulgar relativism’ is what one often finds in the undergraduate classroom. Williams places the command to be tolerant and refrain from condemning other cultures as the conclusion drawn from the truth of Relativism. But it is clear, I think, that an aversion to condemnation and a concern for toleration are often the driving forces behind the acceptance of vulgar relativism. The commitment to Relativism is undertaken in the first place because it is seen as the proper expression of one’s concern for tolerance.

Here are a few observations. First, like many philosophers, I think that Williams’ quick way with the popular and vulgar form of Relativism is masterful and spot on. Second, it is masterful because it identifies what made that view seem appealing and reveals that this is a bad reason to accept the view. That is to say that it offers a diagnosis of the state of mind of someone who sees their way into this view. Third, in my experience, this brilliant bit of prose proves almost spectacularly ineffective in the classroom. Those sympathetic to ‘vulgar relativism’ are not much moved by it. Now, surely, there are plenty of reasons why this might be, not least of which is my own pedagogy. There is also the uncomfortable issue of how effective we could ever reasonably expect ourselves to be in this respect. And surely Williams’ dismissive language doesn’t help.

My aim here isn’t to check the impact of my own teaching style or Williams’ tone, nor to savour the cold comfort of student apathy. I want to explore the idea that this passage is ineffective because it is diagnostic and in particular that it diagnoses the motives of those who hold a view. That is to say, it is ineffective because, for those who hold the view, it purports to tell them what they really care about or what really underlies their commitment to a particular view. My thought is that the Williams passage and my experience teaching it reveal both the importance of achieving self-knowledge in education and the fact that there are barriers to transmitting such knowledge by testimony. In order to make progress in one’s education one needs to know one’s own concerns and motivations for thinking as one does. But this form of knowledge cannot be transmitted by testimony.

Let’s start with the importance of self-knowledge. I take it as obvious, and will refer to it as ‘the platitude.’

THE PLATITUDE: Self-knowledge is often required in education.

We are not blank slates. We enter educational contexts with a set of commitments, however inchoate, a body of evidence, however inconsistent, a set of biases, however invisible to introspection, and so on. One reason why teaching isn’t telling is that students need to integrate new information into their existing set of concepts, beliefs, values, and motivations. Sometimes this process is seamless and automatic.Footnote ² But other times it is not. In our case, the undergraduate needs to grasp that her own commitment to vulgar relativism is underwritten by a concern for tolerance. It is only by doing so that she can consider whether that concern is in fact at odds with what follows from Relativism. And being self-conscious about what really matters to her would, one might hope, weaken the hold of implausible claims that matter less, and help the student to understand her own commitments. Our example involves values, but the point isn’t restricted to such cases.

Let’s operate with a schematic example. I am teaching Williams on Relativism. My student, Riva fancies herself a Relativist. Readers of Williams would label her a vulgar or unregenerate relativist. Now, as an experienced teacher, I know both that Vulgar Relativism is an incoherent doctrine and that those who endorse it are typically moved by a, no doubt laudatory, concern for toleration. But, inevitably, when they discuss it, they see this as a consequence of Relativism, which is supposed to be more or less obviously true. Furthermore, Riva could see her way through the issues better, and be able to understand Williams’ point, if she could self-consciously understand her concern for tolerance as leading her to adopt the vulgar view, rather than as a consequence of it. It is because she assumes that rejecting Relativism goes hand in hand with an intolerant outlook that she is attracted to the Relativist position. So, Riva needs self-knowledge and I have the knowledge of her mind that she needs.

EXPERTISE: A teacher’s experience can equip her with the knowledge of a student’s mind that is required in education.

This knowledge might be arrived at in different ways. As suggested above, it might be general, since it might draw on one’s experience with students over the years. One might justifiably believe that Riva really cares about tolerance because students are often like that. Or the knowledge might be more specific, responding to idiosyncrasies about an individual student. All that matters is that the expert possess it.

Given the Platitude and Expertise, we should expect that the teacher can help here by simply transmitting knowledge about the student’s mind. I have claimed that the Williams example, as it unfolds, suggests otherwise. The student cannot come to possess the self-knowledge required for education merely by being told what she thinks or cares about. The teacher’s knowledge is such that, quite generally, there are barriers to transmitting it by testimony.

THE OBSTACLE: Sometimes the knowledge of the student’s mind possessed by the teacher, thanks to her expertise, is such that it cannot be transmitted to the student by testimony, given the aims of the student.

The aim of the student, in the context, is first-person knowledge. In a different essay Williams gets at the phenomenon as follows: ‘If someone feels… this way, there is never much point, it must be said, in telling him that his feelings involve a muddle: the feelings probably come from some place which that comment will not reach.’ (Reference Williams2008, 137) This predicament, of not being able to reach the place where another’s feelings and thoughts come from by simply identifying it, is common in education, and indeed, common in our lives more generally. We want to explain what gives rise to it.

Notice that our puzzle is in some respects narrow. We want to understand what it is about self-knowledge, and the expert’s knowledge of another’s mind, that explains why there is a barrier to transmitting knowledge of another’s mind by testimony in some cases. It shouldn’t be thought that we are after an explanation of every case where there are barriers to transmitting knowledge of another’s mind by testimony, since in some cases it will be accidental that the knowledge has that object. For example, suppose my student confuses numbers with numerals. I realize that they are confused in this way, but, because they do not have the relevant understanding, I cannot help them simply by telling them of their confusion. They need instruction on the difference in order to understand that they are confused. It doesn’t seem to me that the obstacle here depends on facts about self-knowledge. It is just an instance of the fact that transmission of knowledge by testimony requires levels of understanding that another might not possess. What our examples reveal, and what Freud reveals, is that there is something special about self-knowledge and something distinctive about the expert’s knowledge that explains barriers to testimony unique to that case.Footnote ³

3. What the expert knows

My claim is that the Obstacle arises when one’s knowledge of another’s mind has a special character: it is proleptic. I set this idea out in this section.

The idea of prolepsis is familiar. Having its source in grammar, it involves representing something as being in a condition that is not yet actual. The cliché example: ‘he was dead as soon as he walked into town.’ Call this a ‘proleptic claim’: it asserts that the subject has a property in advance of his actually having it. ‘Proleptic knowledge’ is knowledge that would be expressed in such a claim. In child development, prolepsis refers to the act of treating a child as if she were in a more advanced stage, for example, as if she could speak, as a means of bootstrapping her into that stage (Little Reference Little2008; Bakhurst Reference Bakhurst2011). Now a proleptic claim, including in the context of development, is not made true or appropriate merely by the truth of some prediction. I am going to die. But it is not true of me, I hope, that I was dead as soon as I came into town. That’s because there isn’t the right kind of relation between my current state and the future one. What is the right kind of relation? It’s not clear that we can give necessary and sufficient conditions that will apply to all cases. But roughly: there is an internal relation between the earlier and later stages which is identified in the proleptic claim. The earlier stage ‘contains the seeds’ of the later one; the later stage is a natural development or articulation of the earlier. (Organic metaphors come readily to mind when making proleptic claims.) In some sense, one was ‘really’ or ‘all along’ in the later stage by being in the earlier one. There is a process underway of which the earlier and later stages are parts.

Consider a talent scout who travels around the country looking to recruit young athletes. The scout’s expertise consists in being able to tell what the athlete will become, if she has the right training regimen and so on. If there is such an ability, it consists in seeing the potential as manifest in current practice. Even though what the athlete does, here and now, may be clunky or not quite graceful, the scout can see in the performance the manifestation of a competence to be developed. She can see now where things are going. Unsurprisingly, then, she will see things that others cannot and she will make claims about the athlete, proleptic claims, that others are not in a position to make or confirm.

My main claim is that what’s distinctive of the teacher’s knowledge of the student’s mind is that it is proleptic, and that means: it would be expressed in a proleptic claim. For example: ‘what matters most to Riva is tolerance.’ To make a proleptic claim is to commit to two things. First, she now holds an attitude, though inchoately or in a primitive way, and second, were she to think a certain way, it would take a more articulate form, the one identified in the proleptic claim. I think this idea is familiar; it comes up when we talk of what another ‘really’ thinks or feels. To make such a claim is to make a claim about how she would think and feel, clearly, explicitly and consciously, were she to reflect on matters in a particular way, given her current outlook. Again, a claim about what one really thinks or feels is not a mere prediction about how one would under some other circumstances, for one aspect of it is how one does think and feel now. The claim ‘Riva really cares about tolerance’ isn’t made true by the fact that, were you to brainwash her, she would have that concern. It is made true only if, were her thinking to continue along the lines implicit in it, she would grasp this.

What would count as thinking along these lines? It is helpful here to put the point in terms of Williams’ idea of an internal reason (Reference Williams1981). Internalism is a thesis about what is required for a fact to count as a reason for a subject to act. Internalists hold that whether a fact provides a reason for a subject to perform an action depends on whether the agent has some end or motivational state which would be served by his so acting, in light of the relevant fact. That is, we cannot have reasons to act that have no nontrivial connection with our existing motivations and concerns. As Williams notes, any plausible Internalism will have to idealize over both an agent’s beliefs – correcting for false beliefs – and her motivations – ironing out inconsistency. Thus, on Williams’ view, the fact that P is a reason for A to phi if and only if there is a sound deliberative route from A’s current beliefs, her current motivations, and the belief that P, to the desire to phi.

I don’t want to debate the truth of Internalism. Instead I want to focus on the idea of a sound deliberative route. On so-called Humean views sound deliberation is restricted to instrumental reasoning (Williams Reference Williams1981, 102). But Williams’ conception is more capacious. For example, he grants that imagination can ‘create new possibilities and new desires’ (105). As he puts it: ‘practical reasoning is a heuristic process, and an imaginative one, and there are no fixed boundaries on the continuum from rational thought to inspiration and conversion’ (110). Even without fixed boundaries, what this makes room for, and is meant to make room for, is an internal connection between a subject’s current stock of beliefs, desires, and feelings, and those she would hold by reflecting, deliberating, imagining and otherwise articulating things in a particular way. This means that there are true internal reasons claims that specify a reason an agent, at the time, might not accept, because she has not undertaken the sound deliberative route and thereby understood the light in which the fact reveals the action, given her existing outlook. Such claims are, in the sense given above, proleptic. They are made true by facts about what a subject’s motivations would be if she reflected or deliberated from her current stock of mental states, that is to say: if she were to articulate her current outlook, perhaps using imagination, and render its elements consistent. Thus, both aspects of a proleptic claim – its future orientation and rootedness in one’s present condition – are found here.Footnote ⁴

Return to our case. The idea of prolepsis is most appropriate in developmental contexts, and education and therapy are, of course, such contexts. What is proleptically true of one is what is true of one now, though only in a dim, muddled, or incomplete way, and which would be true in a more robust and clear way were one to pursue a particular developmental trajectory. The teacher’s experience equips her with proleptic knowledge of another’s mind. That means that she understands the vagaries of the student’s current thinking in terms of that toward which it is reaching, the state it is aspiring to be, as it were.Footnote ⁵ It is plausible to suppose that the same can be said of the therapist: she can see the subject’s behaviour as an expression of a state of mind that is realized here and now only inchoately.Footnote ⁶

The idea that what is known proleptically is present in the subject’s mind here and now, though only in an inchoate or incomplete form is important. The object of the expert’s knowledge is the subject’s mind, not simply as it will be, but as it is now in light of where it would be, were things to progress a certain way.Footnote ⁷ I leave it an open question how best to theorize this idea of a mental state existing in an inchoate or primitive form. Consider here a helpful remark from Richard Moran about what therapy achieves. He identifies a

Attitudes that are inchoate are those which have not been internalized in this sense and so are not recognizable to the subject. That’s a minimal requirement of the idea. This might mean that the object or content of the attitude is not determinate (whom are you angry at?) or the attitude type is not determinate (is it anger or something else you are feeling?) Or perhaps both are determinate, though the attitude is isolated from one’s other mental states and so cannot be recognized. Any of these views is consistent with the point made here, and perhaps we don’t need to choose one: perhaps different cases work different ways. We need only rely on what I take to be the familiar idea that our mental states can take a peripheral or inchoate form, as when one is in a muddle about what one thinks on some matter or how one feels about another, and that, in such a case, being in such a muddle isn’t merely a matter of ignorance, but is a feature of the mental state itself.Footnote ⁸ We needn’t suppose, and perhaps we shouldn’t, that being in this inchoate form is the same thing as being unconscious.Footnote ⁹ Our conscious feelings can be inchoate, incomplete, undeveloped (Harcourt Reference Harcourt2017). I have focused on two features of this phenomenon: that even when inchoate, these attitudes are the truthmakers of proleptic claims and that there is an internal, developmental relation between the earlier and later stages of the attitude. As mentioned, on its own this is plausible. But, arguably, we must accept something along these lines to make sense of the practice of therapy. As Edward Harcourt puts it: ‘If there is to be any sense to the idea that the analyst renders a mental item “digestible” for the analysand, it must surely be the very same thing that is as it were dragged from the category of the perforce only somatizable to the category of the thinkable’ (4).

My claim is that people, paradigmatically but not exclusively therapists and teachers, can possess proleptic knowledge of another’s mind. But one shouldn’t get carried away with this. To say that teachers and therapists possess this kind of knowledge isn’t to claim that they always do. Nor is it to claim that they should principally aim at bringing the student to the place where they think they are heading. Taking oneself to have proleptic knowledge might lead to a paternalistic attitude and an overly restrictive sense of what the interaction can lead to. My point is only that there is this kind of knowledge and that it is not accidental to their status as experts that therapists and teachers possess it. Part of their expertise consists in their being able to possess this sort of knowledge and act on it so as to guide the subject. What is puzzling is that the guidance must take some form other than testimony.

4. Transmitting proleptic knowledge

There is no in principle barrier to transmitting proleptic knowledge in other cases. Talent scouts can report what they see and others, who don’t see it, can take their word for it. Indeed, the scout can transmit this knowledge to the athlete herself. And two people can share proleptic knowledge of some third person’s mind. So if there is a problem with testimony it is a problem with transmitting proleptic knowledge of the subject’s mind to the subject herself. This section explains the problem. Roughly, the subject seeks first-person knowledge. But proleptic knowledge of one’s own mind could not be first-person knowledge.

First-person knowledge has a number of distinctive features, but here is one.Footnote ¹⁰ Knowing that someone else believes that P leaves entirely open, from my point of view, the question whether P. Believing that Roger believes that P doesn’t commit me, in any sense, to it’s being the case that P. Likewise with the other attitudes. I know that Roger wants to run away from it all; but I see nothing good about that. And if I know my own attitudes in a third-personal way, for example, by drawing some conclusions upon observing my behaviour on film, then the commitment doesn’t carry over. If I notice that I have the very look of someone who is angry with Roger, I don’t thereby take any stand on the question whether Roger’s behaviour merits anger. But things are different from the first-person point of view. There, in some sense, taking a stand on what attitudes you have involves taking a stand on the objects of those attitudes or the questions to which those attitudes serve as answers. Many in the self-knowledge literature have a finger on this. Andre Gallois argues that we can only possess first-person knowledge of beliefs that are subjectively justified (Reference Gallois1999). Richard Moran claims that ‘the expression of one’s belief carries a commitment to its truth’ (Moran Reference Moran2001, 92). Matthew Boyle, following Moran, claims that the first-person perspective is not ‘alienated’ from the attitudes known (Boyle Reference Boyle2011). Jordi Fernandez makes the observation that ‘we feel pressured to occupy the mental states that we attribute to ourselves’ (Reference Fernandez2013, 14). This feeling is a phenomenological feature of self-ascriptions that he calls ‘assertiveness.’ Christopher Peacocke argues that second-order beliefs ‘ratify’ first-order judgments (Peacocke Reference Peacocke, Wright, Smith and MacDonald1998). Likewise, putting the idea in a handy slogan, Tyler Burge writes that, ‘the first- and second-order perspectives are the same point of view’ (Burge Reference Burge1996, emphasis in original).

It is a matter of controversy how best to characterize this normative dimension of first-person knowledge. The dominant view has it that the first-person perspective involves reflectively endorsing an attitude (Moran Reference Moran2001). But perhaps that is too strong. We might operate with a weaker idea. From the first-person perspective, the attitudes known make sense to one, and they make sense in light of the reasons one takes oneself to possess. For example, it makes sense to you that you believe that P in light of what you would recognize as the reasons for believing that P. It makes sense to you that you are angry at Roger, in light of the reasons you take there to be for anger. One might think this is too intellectually demanding, unless we recognize that the experience of conscious judgment or conscious emotion can themselves render self-ascriptions intelligible in the required way. Thus, plausibly there are two ways one might find the attitude intelligible. First, you might be in a position to offer a rationalizing explanation of it, appealing to the reasons on the basis of which it is held. This doesn’t require reflective endorsement, since one can know one’s reasons without endorsing them. Second, the first-person perspective might involve a self-conscious experience, like the experience of fear or judgment, in light of which the self-ascription is intelligible. Here again, one ‘feels the force’ of the attitude known though one might not, on reflection, endorse it.Footnote ¹¹

Let’s label this phenomenon the ‘normativity of the first-person perspective’, NORMATIVITY for short.

NORMATIVITY: First-person knowledge of a reason-responsive attitude M constitutively involves knowing that I am in M and taking it that M is, from my point of view, to some degree intelligible or justified.Footnote ¹²

Suppose that this is right. There is a normative perspective built into the first-person perspective. Occupying it involves finding the attitude known intelligible in light of one’s current conscious outlook on one’s reasons. Although we are in want of a more developed account of what this comes to, the crude idea is sufficient to solve our puzzle. The problem with proleptic knowledge is that it refers to a condition that, by definition, you are not yet in a position to make sense of. If you could then you would be in it, and it would no longer be proleptic. If Normativity is true, then the only attitudes that can be known first-personally are those which we can make sense of, either by self-consciously and attentively experiencing them, or by being able to offer a rationalizing explanation of them. But if an attitude is the object of proleptic knowledge, then it is not true, here and now, that you can make sense of it in that way. For if the attitude reflected one’s current sense of the reasons for holding it, then, if you were rational, you would, in fact, hold the attitude for those reasons. But then the claim that one holds the attitude would not be proleptic.

Consider Riva again. She values tolerance, but is unable to see that this has distorted her view of Relativism. That means that, from her current vantage point, it would not make sense to say that she cares most about tolerance. Making sense means seeing it as, to some degree, appropriate or intelligible in light of the reasons she has. But for things to make sense in that way is the same as occupying that vantage point on the world, to embrace the fact that the concern for tolerance was what mattered and not Relativism per se. Yet, ex hypothesi, seeing things that way is not yet actual. Put simply: proleptic knowledge concerns how the subject would see things and not strictly how she does. First-person knowledge is restricted to how she does see them. So proleptic knowledge cannot be first-person knowledge. The subject needs to come to articulate her attitudes in such a way that they make sense to her in this sort of way, that they conform to her sense of the reasons she has.

A natural worry is that the truth of NORMATIVITY alone explains why it is inappropriate for the expert to offer testimony; proleptic knowledge is a red herring. That’s because accepting another’s testimony is not sufficient to equip the subject with the normative perspective on her mind. Testimony from an expert is no different from other third-personal ways of knowing one’s mind, like the case of observation mentioned earlier.

I think this is right, as far as it goes. Deferring to another about one’s mind is not enough to understand from the inside what one believes, that is, for the attitude known to make sense to one. But, at best, this would explain a defeasible presumption against testimony in some cases. That is because it is possible that, as things stand here and now, the subject can make sense of the attitude attributed to her by another; upon being told what attitude she holds, she might immediately recognize it in the first-person way. As it were, the normative perspective might be provided by the subject herself, even if it isn’t transmitted from the speaker. In such a case, the presumption against testimony – that it cannot lead to first-person knowledge – is defeated.

But this presumption cannot be defeated in the case where the expert possesses proleptic knowledge. For it is built into the very idea of such knowledge that the subject cannot, just now, make sense of the attitude ascribed to her, at least by her own lights. Thus, while NORMATIVITY might explain PESSIMISM, it is not enough to explain why it is quite generally inappropriate for the therapist or the teacher to put her expertise to use by offering testimony. We need the idea of proleptic knowledge of another’s mind for that.

Here is another worry. A proleptic claim picks out an attitude that a subject could self-consciously hold were she to articulate her beliefs, emotions, and so on a certain way. I have claimed that the expertise of teachers and therapists (though it isn’t restricted to them) equips them with proleptic knowledge. That might give one the impression that any knowledge we have of another’s mind would specify a condition which we could expect the subject to come to terms with it herself. And that might seem extremely implausible. Suppose that I have Oedipal desires.Footnote ¹³ It is plausible that, if you tell me this, I won’t come thereby to possess a proper understanding of it. But is it plausible to suppose that I must even in principle be able to come to terms with this? Must we assume that I can by something like a sound deliberative route?

Four points on this. First, on the weak reading of NORMATIVITY I offered, ‘coming to terms’ with an attitude needn’t involve anything as robust as reflectively endorsing it or integrating it into one’s self-conception. Attitudes which, on reflection, we would not endorse can be known in the first-personal way. For example, I might feel a strong urge to smother a screaming infant, and know this in the first-personal way. The desire makes sense to me, from my point of view, in some minimal way, say because it seems to me that the screaming MUST STOP! But that doesn’t mean that I endorse the desire. Second, our puzzle is restricted to attitudes that can be known in the first-personal way. There are many attitudes, perhaps including Oedipal desires, Kleinian phantasies, and implicit biases, that aren’t like that. In those cases, we needn’t suppose that, if another knows that we hold the attitude, then we could come to know of them in the first-personal way. Third, our claim is restricted to knowledge of the subject’s current attitudes. There are many claims of the form: ‘S would be in M, were she to do X’ that aren’t proleptic claims. Suppose you are trying to convince someone not to go to graduate school. You might say something like this: ‘when you are thirty, things that seem unimportant now – owning a house, saving for retirement, starting a family – will matter a lot. And you may regret that you do not have them.’ This might be true and yet it needn’t be proleptic knowledge, since it needn’t be the case that, here and now, the subject has these desires, though only inchoately. And in this case, as with the subpersonal attitude case, there may be barriers to testimony.Footnote ¹⁴ But since the claim doesn’t refer to the subject’s current condition, it isn’t an instance of the puzzle we are addressing, which concerns cases where a subject goes to another in search of knowledge of her current attitudes. Fourth, that the subject can articulate the attitude in question is consistent with the fact that she won’t and that it would be unreasonable to expect her to. Some people are too entrenched in self-deception to self-consciously grasp what they truly want or believe. Consider the analogous point for Williams’ idea of an internal reason. There are cases where a subject has a reason to do A, because her motivational states are such that, were they ironed out for consistency, she would have desires which would be served by her doing A, but where the subject can’t be expected to appreciate this, because there are barriers to her rendering her attitudes consistent in the required way. That doesn’t mean it isn’t true that she has that reason.

5. Further considerations

Let’s recap. The teacher and therapist’s expertise consists, in part, in being able to understand the subject’s mind in the sense of seeing where things are going and seeing these nascent possibilities in her current behaviour and speech. That’s what equips her with proleptic knowledge. Knowledge of this form is ineffective when transmitted to the subject herself because qua proleptic it cannot be first-person knowledge. While the thesis I called NORMATIVITY might explain PESSIMISM, understood as a defeasible presumption against deferring to others about one’s mind, it cannot explain THE OBSTACLE. NORMATIVITY and the claim that the relevant expertise consists in an ability to acquire proleptic knowledge can explain this. Thus, the idea of proleptic knowledge and the thesis NORMATIVITY can make sense of what might otherwise seem puzzling features of the practices of education and therapy.

I want to conclude by considering how to understand the interaction between the subject and the expert in light of this. I’ll consider two related issues. First, why do we go to people with proleptic knowledge in our attempts to gain self-knowledge? Second, how do they help?

It surely isn’t an accident that teachers and therapists often, by dint of their expertise, possess proleptic knowledge. So their possession of this knowledge must have something to do with why we go to them in the first place. The answer that naturally recommends itself is as follows. The subject wants first-person knowledge, and the expert possesses knowledge of what she really believes, wants, or feels. So presumably the subject wants first-person knowledge of the attitude known by the other. But she cannot possess first-person knowledge unless she is able to make sense of these attitudes in the way captured by NORMATIVITY. That must mean that what she aims at isn’t simply knowing her mind, as it stands, but articulating and developing it to become what is described in the expert’s proleptic claim. That is, the subject doesn’t just want to know her mind, but wants to understand it in a way that is inseparable from articulating it. In articulating the attitude to the point where she can make sense of it herself, the subject is, as a number of authors have emphasised, taking responsibility for the attitude and owning it.Footnote ¹⁵

That the first-person perspective involves a sense of responsibility and ownership for one’s attitudes connects to the original intuition behind Pessimism. There is something odd about simply deferring to another about one’s mind. There is something annoying about having another tell one what one really thinks or feels. It’s odd because doing so forecloses the possibility of making sense of oneself. If this is not just odd or suboptimal, but objectionable, that might be because it involves abandoning responsibility for one’s attitudes. And it is annoying because it is a case of another person telling us our business. Now it is a matter of considerable controversy how precisely to capture the idea of agency and responsibility here.Footnote ¹⁶ This much is clear: our attitudes are not voluntary. If they were, then there would be no point in seeking help from others in knowing them. It is perhaps rude or annoying for another to tell you that you are angry or what you believe, but this is quite different from them telling you what you are going to do tonight. This is presumably because of the in between status of our attitudes: they are up to us, but only in a sense. The idea of the proleptic is helpful here. We do not always self-consciously form our attitudes on the basis of reasons, inferentially or noninferentially. Many exist in an inchoate way, outside the periphery of our awareness. They are things for which we are responsible, but, often, they are things for which we need to take responsibility, rather than things that are, from the get go, our deeds. And to take responsibility in this way, another’s testimony, when they possess proleptic knowledge, is not enough. Of course, this is consistent with the fact that another’s testimony can play a role: it can be the impetus for the subject to reflect and articulate things for herself. The parallel with moral testimony is helpful: others make moral statements and these can lead us to figure things out for ourselves. What is objectionable isn’t taking what another says into consideration, but simply taking them at their word.Footnote ¹⁷

The subject’s interest in knowing her mind is not easily separable from an interest in seeing that her attitudes conform to her sense of reasons. This is why, when one is in a state of ignorance, more than information is required; one needs to work through the issues oneself and come to understand the attitude in light of one’s reasons. This provides us with the resources to answer the second question. How do teachers and therapists (and friends, for that matter) help us to know our minds if not by testimony? The answer, presumably, is: by helping us work through the issues so that we might ourselves take responsibility for how we think and feel.