Re: The Copy Paradox

Nick Bostrom (bostrom@mail.ndirect.co.uk)
Sat, 15 Nov 1997 01:04:55 +0000


Lee Daniel Crocker wrote:

> >>> Absolutely, but "you" are not a few hundred pounds of protoplasm,
> >>> you are the way matter reacts when it is organized in a specific
> >>> complex way. You are not a noun, you are an adjective.
> >>
> >> I'd say I'm a verb.
> >
> > You are both making category mistakes here. The sort of things that
> > are verbs or adjectives or nouns are words and perhaps concepts. But
> > human beings are not words or concepts, although there are words that
> > denote human beings and at least one concept of human being. To say
> > that you are a verb is to imply that you are not alive, because
> > neither words nor concepts are living things.

> I think the distinction he was trying to make is not that of
> "noun/verb" but of "medium/message" or "substrate/pattern", which
> is a useful one.

Sure, it's just that as a philosopher I am allergic to signs of
conceptual confusion.

> Noun/verb/adjective is an artificial distinction
> based on conventions of language (natural languages, that is;
> Lojban has no such concept), but "hardware/software" is a real,
> fundamental distinction. Most of the talk here about identity is
> just arguing definitions for the natural-language "I", and there
> are two axes of argument: one, does it refer to hardware or to
> software (to which I think most here will agree upon the latter),
> and the second axis, does "I" refer to the particular instantiation
> of the pattern (in Lojban "le sevzi") or to the abstraction of the
> pattern independent of particular instances ("le ka sevzi"). The
> answer to that is again just a matter of language convention, not
> the subject of rational debate.

That is controversial claim (which I personally happen to be
somewhat sympathetic towards). It is controversial because to many
people it seems as if the concept "I" is not arbitrary; even though
one may be hard put to define necessary and sufficient criteria for
personal identity, it seems that the notion has important relations
to values that are not shared by other notions we could form. That
is to say: if we stipulate that the word "I#1" should mean this
thing, and "I#2" should mean that, then there still seems to be some
thinking to be done about whether one would be willing to upload in
circumstances where identity is concerved in sense #1, but not in
sense #2, etc.

[Derek Parfit's "Reasons and Persons" (1984) is an important
philosophy book about the problem of personal identity. Max's PhD
dissertation (on his web page) is also on this topic. I haven't read
it yet but plan to do so when I get time.]

Nick Bostrom
http://www.hedweb.com/nickb