[joho] JOHO - April 15, 2004

  • From: "David Weinberger" <self@xxxxxxxxxxx>
  • To: <joho@xxxxxxxxxxxxx>
  • Date: Thu, 15 Apr 2004 18:47:31 -0400

***** JOHO Interval ******
**************************
     April 15, 2004
**************************
Editor: David Weinberger (self@xxxxxxxxxxx)
Web Version (Color! Fonts! Links!):
http://www.hyperorg.com/backissues/joho-apr15-04.html
*********************************
To view this issue correctly, please use a
monospaced font such as Courier.
*********************************

        A NEW TYPE OF JOHO

        I'm continuing to try out new flavors of JOHO.
        This one is a long-ish article, unsuited to a
        blog, without any ancillary material. Still
        feeling my way...

        And thanks for all the suggestions and comments
        about what to do with and about JOHO.

---------------------------------
THERE'S NO "I" IN "IDENTITY"
OR: WHY CLARK KENT ISN'T SUPERMAN'S SECRET IDENTITY 


        A Digital Identity is the representation of a human
        identity that is used in a distributed network
        interaction with other machines or people...

        A Digital Identity consists of two parts:

        1. Who one is (identity)

        2. The credentials that one holds (attributes of
        that identity).

        From: "What Is Digital Identity"
        DigitalID World
        http://www.digitalidworld.com/local.php?op=iew&file=aboutdid


The term "identity" was confusing enough in the real
world. Its meaning in the digital world is even more
ambiguous. Since momentous decisions about the
nature of online life hang on this verbal ambiguity,
we need to be extra-special careful about the real-
world assumptions that are guiding our expectations
about digital identity.

So, here's my program. Let's start with the real
world meaning of identity. Then let's see if we can
use that to clarify identity's digital meaning. 

Now, here are my hidden aims:

1. To suggest that we not confuse self and identity.

2. To argue against the idea that since in the real
world we go around identifying people all day,
identifying people ought to be the default online.

WARNING: I'm going to go the loooong way around to
say these obvious things.

IDENTITY IN THE REAL WORLD

Ordinary Language philosophy arose as a way out of
some vexing problems. For example, we've banged our
heads against the wall for thousands of years trying
to figure out what "reality" is. What makes
something "real"? Is it because it has matter? Is it
because it exists independent of our awareness? If
so, how could we tell? And is that where headwaiters
come from? Many thousands of bored freshmen (and one
Woody-Allen-ish gag) later, we're no closer to
understanding what makes reality real.

Along come the witty Ordinary Language philosophers.
Stop with all the pondering of those special words
in philosophy, they say. Instead, they recommend,
look at how we use them words in casual
conversation, for that's where words get their
meaning. For example, "reality" shows up in phrases
such as "In reality,..." in which it functions like
the word "However." We don't spend thousands of
years trying to figure out what "However" is because
we know it's just a way of tell listeners that we're
about to deny what we just said. Only philosophers
make the mistake of thinking that "reality" is the
name of something. In short: Ordinary Language
analysis subverts the attempt to figure out meanings
in abstraction from how they are used.

We can argue about whether ordinary language
exhausts the meaning of words, yet there is a wisdom
in ordinary language that I think can be useful.

So, how do we use "identity" in ordinary language? A
proper analysis would gather many examples. And it
would carefully differentiate between the full
cluster of words: "identity," "ID" "identify,"
"identification," etc. But I'm a lazy sot so I'll
just shoot from the lip.

Like "reality," when we abstractly think about
identity, we often treat it as a noun denoting some
stable object: My identity is "who I am." Yet, in
ordinary language, we generally only use the word
when we are going from ignorance to knowledge about
someone. "The police did not release the witness's
identity," for instance, is different from "The
police did not release the witness's dog." A dog is
a thing. An identity is the information sufficient
for people to figure out which "who" is the witness,
typically connecting the "who" with a name and
address, but sometimes other information: "Police
identified her as Mary Smith of Elmville, an
engineer at a local software company."

So, maybe an identity isn't a thing after all. Maybe
it's just a way the police talk when they either
don't know your name and address, or when they've
just figured out your name and address.

Further, just because you have an identity when the
police are protecting you doesn't mean that you have
an identity when there's no question about who you
are; once your identity has become public knowledge
and taken for granted, people aren't going to talk
about it any more. If you do talk about your
"identity," it's likely that you are maintaining
multiples because you are a spy or a gangster, and
even then you're distinguishing between who you
"really are" and your false identity. That's why
Superman is Clark Kent's secret identity but Clark
Kent isn't Superman's secret identity. Language
isn't that neat.

Likewise, we only rarely identify someone. Again,
the word only applies (in ordinary language) when
there's some doubt about you: I link you to some
relevant information about you. The information is
relevant to some action that follows from having
identified you: I'm going to charge you with a
crime, apologize for my rude treatment on the bus,
etc.

Finally, notice that "ID" is not simply an
abbreviation of "identity." We use the word "ID" to
refer to the tokens by which we authenticate that
some information about us (name, credit card number,
age) is in fact about us and is true.

Conclusions:

1. We use the language of identity -- "identity,"
"identify," "ID," etc. -- perfectly clearly in
ordinary conversations. But, when we then try to 
think about what constitutes "identity" in the
abstract, we start to get lost because we assume
that nouns have to name things. Nah. Maybe we don't
really have identities in that sense.

2. We talk about identities generally when we are
trying to go from doubt to knowledge, connecting a
thing or person to information about that thing or
person. We make that connection for the sake of its
consequences.

IDENTITY IN THE ONLINE WORLD

There's tons to be thought and said about what it
means to be a person on line. I don't mean to
dismiss such talk as a mistake that ordinary
language analysis could solve. Not at all. I'm just
not going to talk about this issue here.

When we hear "digital identity," we naturally expect
it to have some relationship to "identity" in the
real world. But, "identity" doesn't have a "natural"
meaning in the digital world because language hasn't
had time to get ordinary in cyberspace. So, we
thrash about, trying to get a precise definition of
a word that simply doesn't have one yet. We bring
the word to explicitness and fall into the same trap
as philosophers considering the term: We tend to
think that identities are things: My online identity
is my "who."

Rather than importing the philosopher's sense of the
term, why not be guided by its ordinary language
meaning in the real world? In the real world, we use
"identity" when we're trying to (1) go from doubt to
knowledge (2) by connecting something at hand to
other information (3) in order to accomplish
something. Isn't that what digital identity is? Or
am I confused again?

There is a disanalogy, of course. In the real world,
identification connects a person to further
information. But within the digital world, there's
only information, so digital ID connects one bundle
of information with another; e.g., the information
on this form can be trusted because it accords with
information from some trusted source. Nevertheless,
with digital ID, the real world person is still
there, just indirectly: It's my credit card number
or my address to which the goods will be shipped.
So, while the initial information is information,
it's still tied to the person.

        [In fact, it may be the case that the equation is
        switched with digital identity. In the real
        world, I identify a person by connecting her to
        some information about her. In the digital world,
        I identify some information by connecting it
        (indirectly) to the person. But I just thought of
        this before hitting the "send" button so I'm not
        confident there's anything to it.]

So, what does the ordinary language meaning of
"identity" in the real world suggest about digital
identity?

1. In the real world, we don't identify everyone. We
only identify those about whom we have doubts that
we have to resolve for some purpose. Identifying is
not the default in the real world. Nor, IMO, should
it be online.

2. Real world identifying is the connecting of the
thing/person at hand with information relevant to
our purpose. There is nothing in this process about
a "real self" that has "properties." In the same
way, digital identification is about connecting
what's in hand with other information we need for
some purpose. That's the sense in which there's no
"I" in "identity."

3. The cluster of information that gets connected to
what's at hand should be limited to what's needed to
accomplish the purpose of the identification. But
the purposes of, say, a merchant and customer are at
odds in this regard: merchants like to know lots
more about customers than is required to complete a
sale because merchants want a relationship, not a
mere transaction. That's not what we mean by
"identification," though. That's more like
"investigation." Digital ID ought to be used to
refer to connecting what's at hand to the minimum
set of information.

4. In the real world, an ID is a unique token that
is evidence that some other information about a
person is true: A driver's license connects you to
permission to drive and to a birth date, etc. It'd
help people like me not be so confused if "digital
ID" meant the same thing.

5. In the real world, not every case of going from
doubt to certainty by connecting myself to other
information is a matter of identification. For
example, I can get out of the parking lot for free
by showing a ticket that's been stamped by one of
the merchants. That connects me to other information
(that I made a purchase) but doesn't identify me. We
should strive to keep the same range of options
online; I should have to identify myself only when
there is some legitimate reason to.

We have evolved a careful, subtle set of usages for
the terms of identification. We've done so because
they serve important social purposes. Let's hope
that in bringing "identity" to the digital world,
we're guided by the nuances shown in ordinary
language, not by the ham-fisted assumptions we bandy
about in our real world thinking.

                       ----

APPENDIX

1. This essay arose in response to a correspondence
with Timothy Bouma. In one of his comments on a late
draft, he says: "...identity is required when two or
more things need to be considered as being different
from each other, and these different things need to
be treated differently." Good point. That's why I
don't have to show personal ID to get the parking
ticket discount: I am a member of the class of
people who bought something in the mall, and that's
the only relevant distinction. This may be a useful
principle for digital life: We should only have to
identify ourselves to the level of distinction that
justifies a difference in treatment. (Shall we title
this "Bouma's Law"?)

2. I blogged the question that is this piece's
subtitle and, of course, found out that the question
has been considered in depth by Superman fanboys.
See Seth Finkelstein's comment on my blog's comment
board. [1] Is there no question that goes unanswered on
the Web?

[1]
http://www.hyperorg.com/blogger/mtarchive/002581.html#comments

--------------------------------------------------
EDITORIAL LINT

To unsubscribe, send an email to 
joho-request@xxxxxxxxxxxxx with "unsubscribe" in the
subject line


Make sure you send it from the email address you
want unsubscribed. There's more information about
subscribing, changing your address, etc., at
www.hyperorg.com/forms/admin. Topica manages our
mailing list: www.topica.com/lists/joho. In case of
confusion, you can always send mail to me at
self@xxxxxxxxxxx

Any email sent to JOHO may be published in JOHO and
snarkily commented on unless the email explicitly
states that it's not for publication. ----

The Journal of the Hyperlinked Organization is a
publication of Evident Marketing, Inc. "Hyperlinked
Organization" is a trademark of Open Text. For
information about trademarks owned by Evident
Marketing, Inc., please see our Preemptive
TrademarksTT page at
http://www.hyperorg.com/misc/trademarks.html



Other related posts:

  • » [joho] JOHO - April 15, 2004