From this morning, there is a thread going on on Identity Commons mailing list (identity gangs) on “What is Identity?”.
The thread started off by quoting Kim Cameron’s definition of Digital Identity.
Digital Identity: the digital representation of a set of claims made by one digital subject about itself or another digital subject.
Then, Bob Blakly paraphrases American Heritage Dictionary in the same thread as:
the set of characteristics by which a thing is generally recognized or known
These two seems to be in a general agreement, but I would like to dig a little more because I have a bit of problem with American Heritage Dictionary definition.
When we talk about a term, it is always useful to get back to its root. The below is an excerpt from my lecture at Security Expo 2009 Tokyo.
The term “identity” first appeared in documents around 1570. It was a term that was derived from middle age french word identité, which was in tern formed from 5th century Latin Identitatem. Identitatem was a combined word of “idem et idem”, where “idem” is “same”. From this, it is apparent that the central notion of “identity” in fact is the “sameness”.
This is captured in the “Principle of Identity of Indiscernibles” by Gottfried Wilhelm Leibniz, (1646 ? 1716) .
Subject x and y are identical if any predicate possessed by x is also possessed by y and vice versa.
i.e.,
(1) 
This is rather controversial.
Clearly, The indiscernibility of identicals:
(2) 
holds, but not the identity of indiscernible.
To illustrate it, I have depicted the relationship between Subject and identity (partial identity) in the following figure.
In this figure, I have represented Subject as a molecular structure. It is there, but we cannot observe it directly because whenever we observe, it is merely a projection of it onto the cognitive surface or hyper-plane. Unfortunately, this mapping/projection is not one-to-one. Different Subject can map onto the same thing on a cognitive surface/hyper-plane. For example, if the cognitive hyperplane is degenerated to a “point”, then everything will have this same (partial) identity, i.e.,
Two objects having same (partial) identity may NOT be identical.
We sometimes call this “Partial Identity” a “Persona”.
The problem that I have with the American Heritage definition probably is that it is supposing there is a general cognitive plane (generally recognaized). There is no such thing in practice, unfortunately.
References:
Definition of identity?, http://lists.idcommons.net/lists/arc/community/2009-08/msg00000.html
Identity of Indiscernibles, http://en.wikipedia.org/wiki/Identity_of_indiscernibles