1. Introduction
1.1. Personal and Impersonal Knowledge of Persons
We regularly mark a distinction in everyday life between knowing someone in a distant or detached manner, on the one hand, and knowing someone in a more intimate fashion, on the other. The former is frequently captured by talk of merely knowing of or about someone, whereas the latter is often signalled by talk of knowing her. Few, if any, for example, would be willing to grant that I know that famous academic with whom I have never interacted, even though no one would want to deny that I, like most other members of the profession, know of her. She would in all likelihood be amused, perhaps even offended, to learn that I claim to know her. I know a lot about Michelin-starred chef Heston Blumenthal, but it would be a deceptive act of self-aggrandizing to let on to my friends that I know the man himself. The hiring committee member who pores over a job candidate’s dossier might get to know quite a bit about her on that basis, but it would seem quite wrong of him, for all that, to give others the impression that he knows her. The truth of an affirmative answer to another hiring committee member’s question “Do you know this candidate?” does not follow simply from the fact that the respondent has carefully read and absorbed the details of the candidate’s dossier.
Let us call the detached knowledge of persons that is typically marked by talk of knowing of or about someone impersonal knowledge of persons; and we may call the intimate knowledge of persons that is usually claimed in talk of knowing the person herself personal knowledge of persons. It might be suggested that the distinction is not really an epistemologically significant one, on the grounds that it is simply a function of whether the knowledge of persons in question is accompanied by affective attitudes. That is, one might think that knowing someone personally involves bearing just the same sort of general epistemic relation toward her as knowing her impersonally, the only difference being that the one also involves such things as liking her, disliking her, loving her, hating her, being friendly toward her, being hostile toward her, and so on, whereas the other does not.
But this cannot be right. It may be true — indeed I will argue below that it is — that knowing someone personally typically plays an important role in our affective lives that knowing someone impersonally does not, but this does not imply that the distinction between the two is to be understood merely in affective terms. I know various people personally with respect to whom I bear no particular affective attitude at all: I neither like nor dislike them, I neither love nor hate them, I am neither friendly nor hostile toward them. And it is clear that the impersonal knowledge of persons can come enmeshed in strong affective attitudes. I might not know that famous academic personally, but I might at the same time admire her a great deal as a result of all the knowledge about her that I happen to have acquired. I am positively enthusiastic about Blumenthal given what I know of him, but I do not know him personally. The hiring committee member might come strongly to dislike the job candidate on the strength of what he reads about her in the dossier, despite not knowing her personally. Indeed, that the distinction between personal knowledge of persons and impersonal knowledge of persons does not rest solely on a difference in affective stance is illustrated by the extreme emotions that fans often direct at their favoured celebrities, whom they know impersonally but not personally, however much they may feel otherwise.
Could the distinction be grounded in a simple sort of epistemic reciprocity? Perhaps the only difference between your knowing someone impersonally and your knowing her personally is that, in the former case, you know things about her but she does not know things about you, whereas in the latter case both of you know things about each other.
Upon reflection, however, this suggestion seems mistaken since epistemic reciprocity in this sense seems clearly possible even in cases of impersonal knowledge of persons. The famous academic might just happen to have heard things about me, with the result that we both know each other impersonally despite not knowing each other personally: we both know of or about each other, but neither of us knows the other in any sort of intimate fashion. Two political candidates might not know each other personally at all, and yet know a lot about each other (perhaps even admiring or detesting each other) on the strength of having read each other’s publicly available biographies.
The distinction between impersonal and personal knowledge of persons seems rather to rest on a difference concerning the agency, or voluntary participation, of those known. With personal knowledge of persons we go beyond impersonal knowledge of persons because of a unique participation on the part of those known: if I know you personally, you have participated in my knowing in a particular way, made the right sort of active contribution to it; you have not merely played a passive role in being known. For all the impersonal knowledge of her that the hiring committee member acquires of the job candidate by poring over her dossier, she may yet not have contributed to his knowledge in the active manner required to ground personal knowledge of her. I do not know that famous academic personally because she has not participated in the right sort of way in my acquisition of knowledge about her. The same is true when it comes to my knowledge of Blumenthal: it is impersonal knowledge of him because it lacks the sort of active participation from him that would be required to know him personally.
Note that I am not here claiming that a general distinction between detached and intimate knowledge of objects has always to do with a difference in the agency of the objects known. We do, after all, sometimes draw this distinction with respect to non-personal objects. Do I know Astana? Well, no; but I do know of it. I can know about Mycenaean history without really knowing Mycenaean history. I do not merely know of or about my car; I know it. Clearly, in such cases, the difference between detached and intimate knowledge cannot rest on a difference in the agency of the objects known, for the objects in question — cities, domains of inquiry, automobiles, etc. — have no agency at all. But my point specifically about detached as opposed to intimate knowledge of persons is compatible with this. It is also compatible, I should note, with the suggestion that even in the cases where we draw a distinction between detached and intimate knowledge of non-personal objects, the distinction nonetheless has something important to do with a difference in agency — in the agency of the knower as opposed to the object known. It might well turn out to be the case that I merely know of Astana, and don’t know Astana, because I have not played the right sort of active role in my acquisition of knowledge about it.
1.2. Personal Knowledge of Persons: Why the Lack of Attention?
In what follows my concern will be with the knowledge of persons in the intimate sense I have articulated — with the personal knowledge of persons. The topic deserves significant philosophical exploration, I think, for among the varieties of our epistemic contact with reality, the knowledge of persons in this sense plays a distinctly meaningful role in everyday life. You can add stack upon stack of know-of- ss or know-about- ss, but you will not get rid of your loneliness until you add some know- ss. The same can be said in general of propositional and procedural knowledge: however many know-that- ps and know-how-to-φs you acquire, your loneliness will remain in the absence of some new know- ss. It is exciting to learn that someone you know has recently gone on to do great things; learning that someone you merely know things about has made such strides does not carry the same thrill. In many contexts, it really is who you know, not what you know, that counts.
It is surprising, then, that the knowledge of persons, so understood, has attracted as little attention as it has in contemporary analytic epistemology. The usual approach is to mention it briefly, along with other putative instances of the knowledge of non-propositional objects, as a sort of propaedeutic to the study of more popular epistemological concerns. Zagzebski (Reference Zagzebski, Greco and Sosa.1999, 92–3) suggests two explanations for this lack of attention. First, the knowledge of persons, as one ostensible form of non-propositional knowledge, is not thought to be as straightforwardly transferable from one individual to another as are other forms of knowledge. If I know that a colleague is ill, for example, it is easy to see how I can pass along this knowledge to you: I have just got to tell you she is ill. It is more difficult to see how I can cause you to know my colleague simply by telling you things about her. The idea, presumably, is that this limited transferability of the knowledge of persons has made it seem to epistemologists to be too subjective (or at any rate not intersubjective enough) an epistemic phenomenon to be the topic of any profitable, objective analysis.
Second, there is a widespread conviction in analytic epistemology that reality is fundamentally propositional in character. Any cognitive contact with reality that does not take something propositional as its object, therefore, is assumed to be a less complete connection than one that does. Since the knowledge of persons is assumed not to take propositions, but rather persons, by way of object, it has been relegated to a lesser status in the domain of epistemological interest.
1.3. Desiderata for an Account of Personal Knowledge of Persons
The central aim of this paper is to provide an account of the knowledge of persons (again, in the relevant personal sense) that satisfies four desiderata: (a) it sheds light on how it is that the knowledge of persons, as a unique form of knowledge, involves the specially active participation of those known, (b) it blocks the inference of the first of Zagzebski’s explanations from limited transferability to unmanageable subjectivity, (c) it debunks the assumption of the second of Zagzebski’s explanations that the knowledge of persons is non-propositional, and (d) it helps to explain the distinctive value of the knowledge of persons.
Three accounts of such knowledge can be gleaned from the existing literature, scant though it is. None of them, I will argue in the next section, survives refutation by clear counter-example. In section 3, I will then lay out a novel account that avoids the counter-examples discussed, before going on in section 4 to show how my account satisfies the desiderata (a)–(d) and illuminates both the propositional and testimonial character of the knowledge of persons. I will conclude in section 5 by considering what might seem to be a puzzling consequence of the account when it comes to our knowledge of ourselves.
2. Existing Accounts
2.1. The Acquaintance Account
Feldman (Reference Feldman2003, 11) offers an account of the knowledge of persons according to which for one person to know another is for her to be acquainted with the other. And to be acquainted with someone, he says, is both to have met her and (“possibly”) to remember having done so.Footnote 1
This account is at once too broad and too narrow. Craig suggests that how well one person knows another is not correlated with the frequency of their meeting (nor, we might add, with the robustness of their remembering it): “I may have spent hours with that fellow-commuter, yet scarcely know him at all. On the other hand there are people in whose proximity I haven’t spent anything like so long and yet know (clearly because of the manner of contact with them) quite well” (1990, 144). But I think we can go further. If I am at a bar and exchange only a few brief words with a memorable stranger about how crowded the place is — I do not even get so much as her name — I will have met her and remember it, but I will not know her at all. Hence meeting someone and remembering it is not, contrary to the acquaintance account, sufficient for knowing her to any degree whatsoever. Nor is it even necessary. Suppose that you and I come across each other’s profile on an on-line dating service and, as a result, over time share many details about our lives by e-mail. Having shared enough of the right sorts of details, we would quite properly regard ourselves as knowing each other. Even so, a reasonable question that might arise is whether we should take our relationship to the next level and actually meet. Granted, it is not uncommon to hear talk of people “meeting on-line.” But I take that to indicate a form of virtual meeting distinct from the direct, experiential form required by the acquaintance account, i.e., from meeting in person. “We met on-line last year, but didn’t actually meet until just last month,” after all, seems quite natural and consistent.
2.2. The Recognition Account
Another account of the knowledge of persons, raised and rejected by Craig, holds knowing someone to consist of the ability to recognize her. Craig points to one way in which this account fails: “I don’t know Queen Elizabeth II,” he writes, “but I would be pretty good at recognizing her; probably some people who do know her would be no better” (1990, 147). The account also fails because the ability to recognize someone is not necessary for knowing her. Imagine that a dear friend from the past, with whom I have been out of contact for some time, has a radical physical makeover and loses a very large amount of weight. As I approach her on the street, I have no idea who she is. I just cannot recognize her until she stops me and explains the situation. I then realize that this is my dear friend, whom I have known for years.
2.3. The Congeniality Account
Craig’s own suggestion is that the knowledge of persons is at its core about a capacity for congenial interaction: to know someone is to have the capacity to interact with her “more smoothly and successfully than is generally the case when two more or less randomly selected persons come into contact with each other” (1990, 147).
I believe that this congeniality account also misses the mark. To illustrate, suppose that two total strangers, A and B, are both exceptionally friendly, sensitive, perceptive, and intelligent. As such, were A and B to meet they would appear to onlookers to be the best of friends: their interaction would be quite congenial — well beyond the norm in terms of its smoothness and success. Yet, for all that, A and B do not know each other; they are total strangers. So having the capacity for congenial interaction with another is not a sufficient condition for knowing her.
Moreover consider another pair, A and B, who used to be business partners. A falling out over certain financial matters has made it psychologically impossible for these two ever to work with, indeed speak to, each other again. Their past business required them to reveal various things of a personal nature about themselves to each other. But they never did so with any pleasure, because both of them were, as they continue to be, exceptionally private by nature. A and B were also, as they are now, both extremely terse and sour in their dispositions, having neither the capacity nor care for any form of friendship beyond the most austere utilitarian sort. Now, whenever they happen to meet, A and B appear to onlookers to be total strangers, and their interaction is — as it will always be — no more smooth or successful than one would expect of two more or less randomly selected people who happen to cross each other’s path. Given their history and character, A and B no longer have the capacity for congenial interaction with each other. Yet they both know each other. Hence, having the capacity for congenial interaction with someone is not even a necessary condition on knowing her.
3. The Communication Account
3.1. A First-Pass Formulation
In an early discussion, Aaron remarks in passing that with the knowledge of persons “the object known communicates information” (1944, 75). I think this inchoate observation contains an important insight that my own account of the knowledge of persons may be understood as building upon: one individual knows another, I propose by way of a first-pass formulation of what I will call the “communication account” of the knowledge of persons, just in case the first individual knows personal information about the second as a result of the second’s having communicated the information specifically to her.Footnote 2
Personal information is the sort concerning which an individual can have privacy. No one, obviously, can have privacy with respect to information that is not about persons at all — that, say, water is H20, that 12 × 13 = 156, or that lightning is a rapid discharge of electrons.Footnote 3 Nor does the fact that information is about persons, even particular persons, suffice to render it open to privacy. The information that humans are animals, that the United States has a population in excess of 300 million, that I am myself, and that my rate of free fall is that same as that of a small pebble, is all information about persons, but none of it the sort concerning which individuals can have privacy. To be able to have privacy about some piece of information about you, that information has to reveal something rather intimate about you, something that cannot be inferred about you simply on the basis of the logical, statistical, or scientific laws under which you fall.
When I say that one person has communicated information to another, I mean that she has intentionally revealed information to the other. If unbeknownst to me you rifle through my desk and discover information I have stored in it about the organization for which I work, I cannot properly be said to have communicated this information to you, since I had no intention of revealing it to you. Similarly, if you surreptitiously eavesdrop on what I believe to be a private conversation between me and my friend, you may well learn all sorts of interesting, personal information about me and my friend, but not because I have communicated it to you. I did certain things that caused you to learn the information, but not intentionally.
Moreover, it is possible for an individual to communicate personal information about herself to another without communicating it specifically to her. By purposely exposing facts about my intimate life to the public at large, I may well be communicating personal information about myself to you as a member of the public, in the relevant sense of intentionally revealing personal information about myself to whomever happens to be a member of the public. But that does not mean that I will be communicating the personal information about myself specifically to you, since I may have no intention to reveal the information to you in particular.
Note that the specificity condition does not require that the individual who communicates the information know the individual communicated to by name. There are various ways by which an individual can communicate to another specifically without knowing her name, e.g., by noticing her in a crowd and by dropping a line or two into a general address to the crowd with the intention that that individual learn something personal about the speaker.
It should also be stressed that communication in the relevant sense needn’t be verbal, for there are a myriad of non-verbal ways in which we, every day, intentionally reveal personal information about ourselves specifically to others. Suppose, for example, that my co-worker has just this morning learned that I was denied a promotion for which I had recently applied. She drops by my corner of the office to find me acting in an uncharacteristically grumpy manner. I need utter no words at all in order thereby intentionally to reveal such personal information about myself specifically to her as that I am presently upset at the denied promotion, that this is the sort of thing that can really get under my skin, etc.; my intentionally effected demeanour and body language can easily do that.
3.2. Counter-Examples Avoided
The communication account, so put, successfully avoids the counter-examples to which the alternative accounts discussed in section 2 are subject. The meeting with the memorable stranger at a bar does not involve coming to know her because the information exchanged in the meeting was not personal.Footnote 4 Our getting to know each other through the on-line dating service does not require our meeting because it comes through the electronic revelation of personal information about ourselves specifically to each other. Like Craig, I would no doubt be pretty good at recognizing Queen Elizabeth II, but I do not know her because she has never communicated personal information about herself specifically to me. Despite the inability to recognize my radically altered friend from the past, I continue to know her because the fact that I continue to know personal information about her that she has revealed specifically to me remains. However smooth and successful their initial interaction would be, the total strangers have never revealed personal information about themselves to each other in particular, so they do not know each other. The one-time business associates continue to know each other, despite their inability to get along with any above-average degree of smoothness and success, because of the personal information they know about each other as a result of their past exchanges.
3.3. A Sounder Formulation
There is, however, a potential wrinkle in the communication account as I have formulated it above: It is possible for one individual to communicate personal information about herself specifically to another without the latter recognizing it, and we probably would not want to say that the latter knows the former in such a case. A notices B, a stranger, across the crowded club they both happen to be in. Wanting to know whether B would consider the possibility of a more intimate encounter with someone of her sort, but insecure enough not to come right out and ask, A persuades some friends to relay the information to B that she finds her very attractive. But A insists that the friends not reveal to B the further piece of information that she has asked for their help. B gets the word as the night goes on, and sizes A up, knowing now that A is interested but not that she has intentionally revealed this bit of personal information specifically to her by means of the willing friends. On the first-pass formulation of the communication account that I gave above, B would thus know A, but this seems wrong, since B is unaware that A has communicated the personal information specifically to her. Thus, unless we are willing to allow for the possibility that one individual can know another without knowing that she knows her, it seems that we must add some kind of manifestation, or higher-order awareness condition, to the first-pass formulation of the communication account.
A second, sounder formulation of the communication account could thus be put as follows: One individual knows another just in case the first knows personal information about the second as a result of the second’s having communicated the information specifically to her, and she — the first individual — is aware of this. Or, more succinctly: One individual knows another just in case the first knows personal information about the second as a result of the second’s having communicated the information specifically and manifestly to her.Footnote 5
None of this is to say that the recipient of the communication (i.e., the knowing person) have only true beliefs about the identity of the agent (i.e., the person known). If as your secret admirer I firmly resolve to retain that status, I may cause you falsely to believe that I am not the one who keeps sending you those pleasant, anonymous gifts. I may also open up to you, quite honestly, about all sorts of other pieces of personal information about myself. In such case, on the communication account, you can properly be said to know me because you know personal information about me that I have specifically and manifestly communicated to you, despite the fact that you falsely believe that I am not your secret admirer.
3.4. Knowing or Merely Believing?
The issue of false beliefs concerning, or misleading communication from, the person known leads to a more general worry, however, about the communication account presented above: Might it not be overly strong to insist that the knowledge of persons requires knowledge of personal information about them, channelled through the relevant sort of communication? Knowledge of the information requires not only that the knower believe the information, but also that her belief is true and justified (or warranted in some suitably externalist fashion). Why is mere belief not enough?
This worry effectively proposes an alternative communicative account for our consideration, along the following lines: One individual knows another just in case the first believes personal information about the second as a result of the second’s having communicated the information specifically and manifestly to her.
One might argue for a distinction without a difference here, on the grounds that communication entails knowledge: for one individual to communicate information to another is for the first intentionally to cause the other to know that information. In that case, one individual’s believing personal information about a second individual as a result of the second’s having communicated it to her in the relevant sort of way would entail that the first individual knows the information. But perhaps that is an unsatisfactorily quick way of dealing with the worry. To present the worry in its strongest form, we may grant that for one individual to communicate information to another, all that is required is that the first individual intentionally cause the second to believe the information. We may even grant that there is a reasonable sense in which information can be about an individual even though it is false. Why then prefer my “knowledge-based” communication account to the weaker, “mere belief-based” account highlighted by the worry under consideration?
The answer is that there seem to be clear cases in which mere belief, channelled in the relevant sort of communicative way, is not enough. Consider, for example, a situation in which one individual, A, comes across the profile of another, B, in an on-line dating service. A makes e-mail contact with B, and B responds over time by communicating via e-mail a great deal of intimate information about herself specifically and manifestly to A. But all of this information is false — right down to the information conveyed by the fake digital pictures of herself that she forwards to A. Moreover, A’s only doxastic contact with B is through this communication.
Does A know B? The answer seems clearly to be that she does not. To be sure, she may think she knows B, but she does not really know her. Indeed, part of the “fun” that B might be having lies in the fact that she is causing A to believe lots of false things about her, thereby causing A to think that she knows her, even though she does not.
Why does A not know B? Plausibly, because she fails a truth condition on knowing B: In order for A to know B, A must not merely believe personal information about B as a result of B’s having communicated it specifically and manifestly to her; A must also truly believe the information in this way. The mere belief-based communicative account, unlike the knowledge-based communicative account I have presented, includes no such condition, which is why it cannot handle such cases.
Consider now another case. B is notoriously deceptive with respect to those she meets for the first time: she loves to regale them with almost entirely false stories about herself and her grand life-accomplishments. She knows from the witness of friends, however, that A is highly susceptible to accepting the word of others who assert their claims confidently, even in cases where there is good reason to doubt what they say. Upon meeting A for the first time (up to this point, A has never so much as heard of B), B begins by telling her, very confidently, that she has just moved to the city in which they are presently located. She then goes on, equally confidently, to explain to A why she has just moved: because she had been so successful in her previous investment banking career, she was made a job-offer to work in the city by the most prestigious of investment firms, the offer was much too good to refuse, and so on. B’s claim about having just moved to the city as it happens is true, but everything else she says to A is made of whole cloth. Indeed, she even gives A a false name. Moreover, there are many indications available to A that B is not being honest, such as that the story B tells does not fit well with other things that A knows about the city and investment firms, and that, despite her confidence, B’s body language and manner are strongly suggestive of a disregard for the truth. Still, because of A’s lamentable susceptibility to blind belief in highly confident assertion, she ends up believing everything B says.
A thus ends up getting a true belief in personal information about B as a result of B’s having communicated that information to her specifically and manifestly. Does she know B on this basis? Intuitively not, even though she may be convinced that she does. And why does A not know B? Again, plausibly, because she fails a condition that my knowledge-based communication account includes but the mere belief-based communication account does not — this time a justification condition. A has acquired a true belief in personal information about B — that she has just moved to the city — as the result of B’s having communicated the information specifically and manifestly to her; but A does not thereby know B because, clearly, she lacks justification for that belief: the various indications of B’s dishonesty make the belief quite unjustified.
4. Desiderata Satisfied
4.1. Participation
Let us now return to the four desiderata of my overall aim articulated at the end of section 1. The first of these desiderata, (a), is nicely satisfied. On the communication account I have presented, if I know you (in the relevant personal sense), you specially participate in this knowledge because you intentionally single me out as the specific recipient of your communication. You are no mere passive object that is known; you play a specially active part in the construction of my knowledge of you.
By contrast, neither the acquaintance account, the recognition account, nor the congeniality account seems capable of satisfying this desideratum. It is clearly possible to have met (and to remember having met) someone, and hence know her on the acquaintance account, without her actively participating in the process at all. Suppose, for example, that you introduce me by name to A at a party, but (knowing that my name is associated with a certain political movement to which she objects) she refuses to acknowledge my presence or return my consequential greeting. On the acquaintance account, I thus know A, but she has not played an active role in the construction of my knowledge of her. It is also possible to recognize another individual without her making any active contribution to whatever epistemic states supervene on this recognitional ability. I might be able to recognize Elizabeth II simply by virtue of the photographs and descriptions I happen to have come across of her in tabloids. Although the recognition account implies that I thus know the queen, whatever knowledge might be involved seems to be of the wrong sort, since the queen has obviously not participated in its active construction. Moreover, whatever knowledge might be carried by the capacity for congenial interaction shared by two total strangers (e.g., procedural knowledge of how to engage each other with an above-average degree of smoothness and success), it will not be of a sort such that the one has it only by virtue of the active contribution of the other. So the congeniality account also leaves us without an answer to the question of why the knowledge of persons involves the specially active participation of those known.
4.2. Limited Transferability
If I have argued successfully for my communication account, desideratum (b) is clearly satisfied as well: My defence of the account shows that we can give a profitable, objective analysis of the knowledge of persons, and that such knowledge is not an overly obscure epistemic phenomenon, even though it has the feature of limited transferability that Zagzebski highlights. Moreover, the communication account affords us insight into why the knowledge of persons should have this feature: I cannot cause you to share my knowledge of a colleague simply by telling you things about her because the propositional knowledge thereby exchanged does not carry an intention on my colleague’s part to reveal personal information about herself specifically and manifestly to you. Or, more strictly, it need not: it is conceivable that my passing along personal information about the colleague to you constitutes the means whereby she does communicate personal information about herself specifically and manifestly to you. My colleague tells me something intimate about herself, and asks that I share it with you along with the information that she made the request of me; I do, and she thereby reveals personal information about herself specifically and manifestly to you, with my help. On my communication account, you thus know the colleague. But that is all to the good, since it is intuitive that you do know her in such a circumstance.
4.3. The Propositional Character of Personal Knowledge of Persons
The third desideratum, (c), is satisfied by virtue of the fact that on the communication account I have presented, the knowledge of persons reduces to a kind of propositional knowledge. On that account, Feldman turns out to be wrong when he claims that the distinctive features of the knowledge of persons entail that “not all knowing is propositional knowing” (2003, 11). The knowledge of persons is a kind of propositional knowing, albeit a unique kind.Footnote 6 For on my communication account, coming to know someone consists of coming to know propositions about her of a particular sort — personal information about her — even if only by means of her specifically and manifestly revealing the propositions to a particular recipient. An appropriate analogy would be to source-specific propositional knowledge. The perceptual knowledge that there is a green coffee cup before me is of course propositional, but it is perceptual in light of the particular way in (i.e., source from) which it was acquired. The rational knowledge that all squares have four right angles is a distinct form of propositional knowledge because of its derivation from the faculty of rational intuition.
4.4. Personal Knowledge of Persons as Testimonial Knowledge
Indeed, there may be more than just an analogy here: on my communication account, the knowledge of persons looks like a subspecies of one sort of source-specific propositional knowledge, viz. testimonial knowledge. To see this, consider the core notion of testimonial knowledge that p: knowledge that p based on someone’s testifying that p, where someone’s testifying that p amounts to offering p as evidence (on some matter that is taken to be as of yet unresolved) by stating that p.Footnote 7 It would seem that every case in which an individual communicates personal information about herself to another is one in which she offers that information as evidence by, effectively, stating that information — at least on the reasonable assumption that stating something need not always be done via expressing it in language. Hence every case in which one individual comes to know something about another as a result of the other’s communicating it to her will be one in which she knows the information on the basis of the other’s testimony, which in turn means that on the communication account the knowledge of persons is a particular form of testimonial knowledge. If I know an acquaintance by virtue of her having communicated specifically and manifestly to me the information that, say, she frequently reads pulp fiction before drifting off to sleep at night, then I know her on the basis of her testimony about herself: she has offered up the information that she frequently reads pulp fiction before bedtime as evidence (on a matter that she took to be unresolved in my mind at the time — on what her reading habits are like) by stating that information to me, specifically and manifestly. A friend knows me because I have specifically and manifestly communicated to her such information about myself as that I love big breakfasts and wonder about the literal truth of intentional-state ascriptions to non-human animals. My friend’s knowledge of me is in these respects testimonial, since by communicating such information to her in this way, I have offered the information to her as evidence (on what I took to be unresolved questions she might have had about me, such as my eating habits or interests in the philosophy of mind) by stating it to her.
4.5. The Distinctive Value of Personal Knowledge of Persons
We come finally to the last desideratum, (d). How does my communication account help us to explain the distinctive value of the knowledge of persons in the relevant sense?
I think the answer is to be found by considering that, on that account, the knowledge of persons involves a recognition of the personal status of the knower by what is known that is typically not found in other forms of knowledge. From the fact that you know me it follows on the communication account that I have taken pains to single you out in particular as a being worthy of my self-revelation — as a person who can appreciate what it is that I reveal about myself. Few, if any, other forms of knowledge carry this implication. I can merely know of or about other persons, know that 12 × 13 = 156, know that a colleague is ill, know how to build better bridges, or even know how to live a moral life until I am blue in the face, but none of it will bring me another’s recognition of my personal status until I come to know her. And an awareness of this recognition is deeply valuable to us, as anyone who has wrestled with loneliness or sought out new friends and lovers can attest.
5. A Concluding Worry: Self-Knowledge
5.1. Self-Communication?
I have presented the knowledge of persons as a relation that obtains between two (or more) distinct persons. There is a reason for that: on the communication account proposed, the knowledge of persons is not something that one typically has with respect to oneself. To know oneself, on this account, would require communicating personal information about oneself specifically and manifestly to oneself; and we are unlikely to find many instances in which this sort of self-communication occurs. I say “many” and not “any” (and “typically” rather than “ever”) because it seems to me that there may in fact be a very limited number of realistic instances to be found, particularly within the general phenomenon of “reminding oneself.” If I have an especially wide-ranging discussion with my psychotherapist, for example, I might find it important to record some of the more illuminating things I have learned about myself in my private journal. Upon later reviewing the journal I may well remind myself of these insights in such a way that counts as having, at t1, communicated personal information about myself specifically and manifestly to myself at t2. The journal transcription seems to be an intentional revelation of my past self to my future self, and the information so revealed is personal. It is moreover directed specifically at my future self, who recognizes that specific intention.
5.2. Limited Self-Knowledge
Still, it is a consequence of my communication account that we rarely ever know ourselves, and this may be puzzling because it seems to run flat against a long-standing philosophical tradition that privileges self-knowledge over knowledge of others. It has been only very recently, for example, that analytic epistemologists have begun to worry that there are special difficulties involved in understanding our knowledge of ourselves (and even then the problems seem more semantic than epistemological in nature),Footnote 8 and the dominant conviction since at least Descartes has been that whereas there is a big problem when it comes to our knowledge of other minds and elements of the external world, no such problem holds with respect to our knowledge of ourselves.
Does this consequence stand as a significant problem for my communication account? I think not. For consider, first, that there is a very intuitive sense — the aforementioned epistemological tradition notwithstanding — in which claims of self-knowledge are strikingly odd. Imagine a colleague inquiring about whom I know as a result of my previous tenure at another academic institution. If, in addition to claiming to know A, B, and C (all of whom still work at the other institution), I go on to add “And of course let’s not forget that I know myself,” my colleague could be excused for thinking that I am making a bad epistemologist’s joke, because there is one very good sense in which my claim to self-knowledge in this context comes close to making no sense. Given the intuitively bizarre nature of many such claims to self-knowledge, it should be a virtue of an account of the knowledge of persons that helps us to see why they are bizarre. And the consequence of the communication account that we typically know (and come to know) others but not ourselves helps us do just that. This might give the communication account a (further) theoretical leg up on the alternative accounts considered in section 2, but I am not willing to stake much on it. Whether it does will depend on answers to such questions as whether we are typically (and in the relevant sense) acquainted with ourselves, recognize ourselves, or have the capacity to interact in a congenial manner with ourselves, and I do not at present have much to say about such answers.
Secondly, it should not be forgotten that my communication account is an account of the personal knowledge of persons, not an account of the impersonal knowledge of persons. It is thus silent about the extent to which we know ourselves impersonally. I think this is worth bearing in mind partly because the more famous injunctions to self-knowledge seem aimed at the impersonal knowledge of ourselves — as things that we come to discover ethically important (say) information about but do not in any interesting sense reveal to ourselves. Reflection on one’s past behaviour, the use of reason to draw out one’s status vis-à-vis one’s perceived human nature or proper function, and the acquisition of certain techniques for improving one’s moral condition, for example, would seem to be characteristic ways in which the well-known Delphic advice γνωθι σεαυτον is to be followed; but these are much more like the ways in which we might know others without their making any specially active contribution to that knowledge than they are to the ways in which we might know them as active contributors. Thus, those impressed with the importance of self-knowledge for concerns that go beyond the epistemological realm need not find anything particularly troubling, it seems to me, with the communication account’s consequence that we rarely ever know ourselves, for that is consistent with the possession of a good deal of impersonal knowledge about ourselves.Footnote 9
The silence of my communication account about the extent to which we know ourselves impersonally is also worth bearing in mind because it helps preclude a concern one might have about the account’s standing at odds with the ubiquity of first-personal knowledge of ourselves. I take it to be an obvious fact that we know a great deal about ourselves from the first-person perspective: every time I knowingly utter a true sentence of the form “I am F,” for example, I express knowledge of myself from the first-person perspective. But the consequence of the communication account that we rarely ever know ourselves in the relevant sense is also consistent with the point that we frequently have first-personal knowledge of ourselves, for there is nothing about knowing ourselves from the first-person perspective that entails knowing ourselves in such a way that we play a specially active role in the acquisition of the knowledge. We can — indeed, it seems that we typically do — know ourselves from the first-person perspective in a very passive way. Consider Perry’s famous example of the acquisition of new first-personal knowledge:
I once followed a trail of sugar on a supermarket floor, pushing my cart down the aisle on one side of a tall counter and back up the aisle on the other, seeking the shopper with the torn sack to tell him he was making a mess. With each trip around the counter, the trail became thicker. Finally it dawned on me. I was the shopper I was trying to catch.
[…] At first characterizing the change seems easy. My beliefs changed, didn’t they, in that I came to have a new one, namely, that I am making a mess? (1979, 3)
Note the passive language in terms of which the acquisition of the first-personal knowledge is expressed: “it dawned on me,” not “I made myself realize”; “my beliefs changed,” not “I changed my beliefs”; “I came to have a new one,” not “I informed myself.” All of this seems quite right: first-personal knowledge of ourselves is not usually something to which we make any specially active contribution; we just find ourselves possessing or acquiring it. That suggests that first-personal knowledge of ourselves is usually of the impersonal, not personal, sort.
