understanding the importance and impact of anonymity and authentication in a networked society
navigation menu top border

.:home:.     .:project:.    .:people:.     .:research:.     .:blog:.     .:resources:.     .:media:.

navigation menu bottom border
main display area top border
« Technology and the law: Will reasonableness be the ruin of privacy? | Main | Ottawa to give police more power to snoop »

The Personal and the Empirical

posted by:David Matheson // 04:13 PM // August 17, 2005 // Core Concepts: language and labels

Sometimes when we describe information as "personal" we mean to cordon it off as information that others have no business knowing. This is often what's going on when one declines to answer a question on the grounds that "that's personal!" But here I want to talk about personal information in a broader sense. I'm interested in understanding the nature of personal information in the sense of the sort of information with respect to which one can, at least potentially, have privacy. I can have privacy with respect to information about what medications I might happen to be on, because that's personal information about me; but I can't have privacy with respect to information about the average annual rainfall in Ottawa, because that's obviously not personal information about me. This even though the information about my medications might quite properly be the business of others to know (e.g. my physician).

What is personal information in this broader sense? An adequate answer to that question is important for any robust theory of the individual's right to (informational) privacy, since if we want to understand what that right amounts to we'll also want to understand what the thing is – privacy -- to which the right entitles the individual. And since privacy is presumably a relation that holds between the individual's personal information and other people, we'll in turn want some understanding of what we're talking about when we speak of the individual's personal information.

Elsewhere I've argued that personal information is not necessarily sensitive information about the individual, on the grounds that even if the individual could care less about whether others know a bit of information about her (or even if most members of the society in which she lives could care less whether others know that sort of information about them) that's not enough to render the information in question non-personal; all it might show is that the individual could care less about whether personal information about her is known by others. I'm still inclined to say this. But this, of course, is not really to say anything about what personal information is; it's only to say what personal information isn't.

Once the personal and the sensitive are drawn apart in this way, it becomes tempting to suggest that personal information is just information specifically about an individual: personal information about me, for example, is just facts about me as opposed to other people (or no people at all). But there are serious problems with this liberal account of personal information. Consider the following fact (i.e. bit of information):

(1) David Matheson is David Matheson.

As Steven Davis has pointed out in a recent paper (see footnote 17 of his "The Epistemology and Normativity of Identifying and Identification"), this sort of trivial identity fact is, for all its triviality, a bit of information specifically about an individual, viz. me. No one else, after all, is identical to me except me! And so on the liberal account, (1) would count as a bit of personal information about me. But that's absurd: (1) is information specifically about me, to be sure, but not personal information about me.

Other counterexamples to the liberal account of personal information abound. Consider, for example, the following two facts about me:

(2) If David Matheson is a resident of Ottawa, Ontario, then he is a resident of Canada.

(3) David Matheson is not both married and a bachelor.

Again, (2) and (3) are both bits of information specifically about me. But it is quite implausible to say that they are bits of personal information about me.

So we're still in a fog. If personal information is to be equated neither with sensitive information about an individual, nor with just information specifically about an individual, what is it? My suggestion is this: personal information is empirical information specifically about an individual. To flesh this new account out a bit, I'm going to have to say something about the meaning of its key term, "empirical." And to do that, I have to say something about knowledge sources.

A knowledge source can be thought of, generally, as a cognitive process that begins with certain characteristic mental states (the source's "input") and -- provided all goes well -- results in knowledge (the source's "output"). Philosophers have distinguished six putative knowledge sources, so understood, distinguished from each other in terms of their distinct inputs and/or outputs:

Perception. This takes sensory appearances of things in the world (e.g. its seeming to me visually as if there is a dog in front of me) as input, and generates knowledge of things in the world (e.g. my knowing that there is a dog in front of me) as output.

Introspection. This takes appearances of one's own mental life (e.g. its seeming to me, when I direct my attention to my own mind, that I am annoyed with someone) as input, and generates knowledge of one's own mental life (e.g. my knowing that I am annoyed with that person) as output.

Inference ("Reason" in one sense). This takes knowledge of certain information (e.g. my knowing that Socrates is a man and that all men are mortal) as input, and yields knowledge of new information (e.g. my knowing that Socrates is mortal) as output.

Rational Intuition ("Reason" in another sense). This takes appearances of necessity (e.g. its seeming to me necessary that 2+2=4) as input, and generates knowledge of necessary truths (e.g. my knowing that 2+2=4) as output.

Memory. This takes past knowledge of information (e.g. my knowing yesterday that I had a latte in the morning) as input, and yields present knowledge of the very same information (e.g. my knowing today that I had a latte yesterday morning) as output.

Testimony. This takes knowledge that another person has claimed something (e.g. my knowing that the departmental chair has claimed that philosophy enrollments are up this year) as input, and yields knowledge of what she has claimed (e.g. my knowing that philosophy enrollments are up this year) as output.

Among these general sources, only Perception, Introspection and Rational Intuition seem to be basic knowledge sources, since only they generate knowledge from something other than knowledge. The other sources -- Inference, Memory, and Testimony -- only work if they've already got knowledge ultimately derived from the basic sources.

We can understand the notion of empirical information, however, just in terms of the basic sources. To say that a bit of information is empirical is to say that it can only ultimately be known through the operation of Perception and Introspection (through "experience", as philosophers are often wont to say). The contrast is with (as philosophers also often put it) "a priori" information. A priori information is information that can ultimately be known through the operation of Rational Intuition. (A little aside: you've probably heard about great debates in the history of Western philosophy between the "Rationalists" and the "Empiricists." The Rationalists think we can know a lot of significant a priori information, i.e. a lot ultimately just through Rational Intuition; that's why they're big on Reason. The Empiricists, by contrast, think that pretty much any significant information to be known is empirical, i.e. to be known only through Perception and Introspection; that's why they're big on Experience, or the Senses.)

So, in a nutshell, the claim that personal information is empirical information specifically about an individual amounts to this: personal information is information specifically about an individual that can ultimately only be known through the operation of Perception and Introspection. Rational intuition ("reason" in one historically important sense of the term) alone won't do the trick.

This account of personal information allows us to understand why the likes of (1)-(3) don't count as bits of personal information. Although they are bits of information specifically about an individual, they are not empirical bits of information specifically about that individual, since they can (I suggest) be known ultimately though Rational Intuition. They are bits of a priori information about me.

This account also makes it clear why personal information does not necessarily go hand in hand with sensitive information about individuals. For it's pretty obvious that not all empirical information specifically about an individual is sensitive information about the individual.

Comments

Not sure what you're having trouble with here, but "personal information" is information that's specifically about me (i.e., the entity that's making the claim or stating the fact).

In most contexts, your name is personal information so "David Matheson is David Matheson" is, indeed personal. "David Matheson is not both married and a bachelor" is not personal. It's merely an instance of a very general statement: "no one may be both married and a bachelor."

No need to bring philosophers into it at all.

Oh, BTW, I hope you meant that "the individual *couldn't* care less about " something.

Posted by: Dave Kearns at August 17, 2005 05:52 PM

Hi Dave. I think there's a misunderstanding here. I actually agree that my having the name I do is a bit of empirical information specifically about me. So it follows on the account I was offering that it's a bit of personal information about me. But notice that this is a different bit of information than the information that I am myself. The one is a contingent claim about what name I happen to bear, and may be quite significant in certain contexts; the other is a necessary truth about my being self-identical and is trivial in all contexts.

I think your comment about "David Matheson is not both married and a bachelor" is an interesting one, and points to something that fits quite well with my account. If "David Matheson is not both married and a bachelor" is not known directly through rational intuition, it may nonetheless be known ultimately through inference from something that is, viz. the general necessary truth "no one is both married and a bachelor." Of course, the same thing can be said of "David Matheson is David Matheson": if it's not known directly through rational intuition, it may nonetheless be known ultimately through inference from something that is, viz. the general necessary truth that everything (or everyone) is self-identical. The central point remains that neither "David Matheson is not both married and a bachelor" nor "David Matheson is David Matheson" counts as a bit of empirical information about me, because both are knowable (either directly or ultimately) through rational intuition. And so neither counts as a bit of personal information about me.

Posted by: David Matheson at August 17, 2005 10:56 PM

!!!Warning!!!!
this posting is by a philosopher-in-training.
(that was a joke and heads up for Mr. Kearns)

I find it strange David, that you call 1-3 'knowable' or information at all for that matter. I think that all three statements actually end up saying nothing really at all. They are almost void of meaning. To say x is x, x can't be both x and not x, etc. and so on is not really to say much. And these statements only seem to matter in this context in the world of theory and don't leave much room for the complexities of the real political world (i.e. someone experiencing being a little bit of both of the opposites of something, such as, man and woman (the hermaphrodite), gay and straight (the bi-sexual) and so on and so on). You see I think an argument can be made that you can be married [here I am asssuming the definition of marriage that you are using in your example, the dominant western conception of marriage] and be a bachelor, many husbands/wives act like bachelors/bachelorettes alot of the time. This does not negate their legal status as married but surely it also adds to their roster the identity of single. On paper they are married, but in life... not so much (the life context seems to me to be the important one in these situations). Logically the language does not make sense and we still call them married but really they are somewhere inbetween residing in the land of 'married bachelors'. I don't think that our language is always a reliable source for, or reflection of, our very complicated identities. Especially the language of logic, because we really are the most illogical of animals. I'm not sure how I got so far off topic.
Now, Getting back on track...
'David Matheson is David Matheson' is not personal information about you in this context, in this context David Matheson means nothing at all, of course. But if we transfer this theory into the active world and for instance, say you are wearing a uniform with your name on it, then that surely identifies you 'empirically' from those around you as in fact 'David Matheson', which makes that information personal if in fact the context in which it is experienced or used has effects on your personage (for instance, cases of mistaken identity). And it seems to me this is what is at issue in the world of privacy. So, I guess what I am curious about is if you are willing to allow context into your theory of personal information?
I should add for clarity that I am not a rationalist or an empiricist but have been diagnosed by a few "doctors" as an 'anti-foundationalist' (sometimes even a romantic nihilist??) so you and I are not occupying the same epistemic terrain on this particular theoretical concern, but I hope you respond anyhow. Also, I have not written as clearly as I could/should have, but the hour is late and the eyes are weary.

Posted by: krystal at August 22, 2005 02:06 AM

Hi Krystal. Thanks for the follow up.

I agree that the likes of (1)-(3) are not bits of personal information. But I don't think the way to account for this is to say that (1)-(3) are not knowable bits of information at all. I take knowable bits of information to be knowable true propositions (true claims, facts, etc.); I also take the likes of (1)-(3) to be knowable true propositions. That they are *necessarily* true propositions shouldn't prevent us from seeing them as knowable. (Compare math truths.)

You ask whether I'm willing to allow context into my theory of personal information. I think the answer is "yes." (But I'm a little worried that I may be misunderstanding your question; so let me know if I am.) In one sense, at least, my theory is all about contextualizing personal information. Information doesn't get to be personal, on my account, unless it's empirical; and one of the distinctive features of empirical information is that it tells us something about how things are in actual contexts (as contrasted with *a priori* information, which tells us how things must be in all contexts, actual or merely possible).

Posted by: David Matheson at August 23, 2005 11:49 PM

I think you are trying to be way too analytical and not enough historical or politcal in your search for an answer. To understand the debate on what is "personal information" you would probably need to understand and delve into the debates on private property, the history of the evolution of the term, the power structures that are built upon it, and so on... You can't start in a vacume and just analyse the term. It won't make sense.

It is an easy mistake to make, especially when one is brought up, as I was, in the tradition of analytic philosophy. I don't know how to answer your question, which is an important one, but I think your are not going to get far looking for the answer the way you are going about it.

Posted by: Henry Story at September 1, 2005 06:45 PM

Hi Henry. Were my question historical, etymological, or political in nature (e.g. How has 'personal' been used in the history of the English language? How is the management of personal information used to further various political aims?), then of course the methods you suggest would be the most appropriate. But my question is not any of these things; it is rather a conceptual question about the nature of personal information, understood as the sort of information about which one can have privacy. As such, it seems to me that the method I do employ in my attempt to answer it – essentially, just rational reflection -- is quite appropriate.

In saying that I'm being "too analytical," you may simply be expressing a lack of interest in the conceptual question. If so, fair enough. We all have our own special interests, and there's no need to claim that our interests are the only ones that should be of interest to others.

I would, however, take issue with the suggestion that the mere use of rational reflection to answer conceptual questions like the one I raise amounts to little more than working in a "vacuum." To my ear, this implies that the mere use of rational reflection to answer conceptual questions can never provide any good basis for answers to those questions. And that seems wrong. It is, for example, the use of rational reflection that allows us to see that personal information is neither necessarily sensitive information nor simply information specifically about an individual, but rather empirical information specifically about an individual.

You might of course object that my use of rational reflection to arrive at these conclusions is in fact a misuse -- that my arguments for them are flawed in some particular way or other. And while that is something I would welcome, it will require engaging in a little rational reflection of your own, on the relevant question.

Posted by: David Matheson at September 4, 2005 06:45 PM

main display area bottom border

.:privacy:. | .:contact:.


This is a SSHRC funded project:
Social Sciences and Humanities Research Council of Canada