[extropy-chat] Type vs. instance with respect to identity acrosstime (a reply to Jeffrey)

[Lee Corbin]

Slawomir and Jeffrey discuss identity.

> A B:
>> You say that it is a unique trajectory through space
>> and time that specifies an identity. (Correct me if
>> that is not a fair or correct representation).On the
>> face of it, I actually don't disagree with you.
>> Another way to describe that is that two different
>> trajectories are different precisely because they are
>> separated both in space and time (which are both
>> intertwined entities as far as I know). But this, in
>> addition to other "factors" is why I believe it is
>> appropriate to believe that no "self" continues
>> through time, and that we are in fact "dieing"
>> continuously (and it's not a big deal).

It is not a big deal, I guess you to mean, because
we are so accustomed to it. Almost anything that
happens to us each Planck interval can't be as
alarming as, say, Muslim fundamentalism. 

In fact, several people here, following Derek Parfit's
lead, do emphasize that survival is what counts. We
all of us *do* want to survive (except the poor souls
who are very depressed, for instance).

I choose to define "self" and what "I" am in a way
that is more consistent with the above daily facts
of life. That which does not change sufficiently from
ordinary day to ordinary day is what I happen to be.
Now sure, if tomorrow you put me in an unbelievably
intense Army existence, where I lopped off other
people's heads every day in some medieval setting,
and my best friends were crude barbarians who spoke
only an archaic version of Swedish, then I probably
would not stay me very long, no matter *how* 
continuous and uninterupted my process (a la Heartland).
Why isn't it better to change the meaning of these very,
very, very common words like "me" and "I" and "self"
to accomodate to this reality---so that our sentences
make sense legally and in daily life---than to kling to
[excuse me] idiotic redefinitions of what everyone
else means, so that I am a bit string that changes
every femto-second.  It's just a useless way of talking
in my NSHO.

Slawomir:

> Suppose we reduce a person to a string of bits (that is, "person=pattern of 
> bits"). Tomorrow or even a second later that person-representing string will get 
> longer or shorter (as new memories are being created and old ones evaporate) while 
> the bits that make up the string change also so that the pattern of bits is almost 
> certainly different at different times. If so, then each of these strings at 
> different times must be assigned different identity according to Leibniz's law and 
> if we still assume that persons are nothing but strings, we are forced to conclude 
> that people die constantly.

This is all philosophy.  It's impractical.  We should focus on what it
is about us that we treasure.  None of us is really interested in how
my bit string varies from moment to moment, or from year to year.
We are interested in survival.

Philsophy should be *prescriptive*.  What good is all this talk of
uploading if we can just wait around until it happens and then decide
if forking is okay, memory erasure is okay, and so on?  Good
philosophy should be about *decisions* that you would make.
Would you teleport?  Would you let your EEG go flat for a second
if a certain cancer cure demanded it?  And so on.

Lee


> However, we are more like minds and processes rather than brains and patterns. I 
> claim that "person=instance of process" is the only model that can accurately 
> reflect the dynamic nature of our minds in contrast to static patterns.
> 
> So how does this "person=instance of process" model performs with respect to 
> identity across time? Unlike patterns, processes are undefined in time intervals 
> equal to 0. It simply does not make sense to say that there is such a thing as 
> process-at-t3 and process-at-t789. Instead, any process is necessarily defined 
> across *all* ts in the interval equal to the duration of an instance of that 
> process. Because processes are defined across time, it is impossible to find a 
> single property y of an instance of the process at any t1 in the interval with a 
> value that is going to be different from a value of property y of this instance at 
> t2 (because these values will refer to the same exact instance).
> 
> Leibniz's law applied to "person=instance of process" model suggests that I'm still 
> the same person (instance of process) I was yesterday.
> ----
> 
> Even though the above analysis was done with respect to time, the same analysis 
> also applies to space as processes are necessarily defined across space. This is 
> why I still claim we don't die as we glide through time and space.
> 
> Slawomir