[extropy-chat] Resend: Parfit's Reasons and Persons

[Jef Allbright]

[Very sorry, resending in plain text due to report that the previous
version was full of HTML garbage viewed on a Eudora 7 client.]



[Jef wrote:]
>>> I remember my disappointment with the tentativeness of his
>>> conclusions, but after so many years I don't remember the specifics.
>>> I just retrieved my copy of the book and upon perusing the table of
>>> contents I am surprised anew by its extent.  I'm also reminded that
>>> nearly any idea that feels original to me has been picked up or
>>> synthesized from other sources.

>On 4/17/07, Anders Sandberg <  asa at nada.kth.se> wrote:
>> Hmm, which implications would that be? If you have a list, I could try to
>> talk to him about them (he is literally one flight of stairs upstairs from
>> me).

> I'll review the book and organize my thoughts and get back to you on
> this as quickly as practical.  Who knows, I may find that I see it
> differently so many years later.

Anders, I spent a few hours last night reviewing _Reason's and
Persons_. The challenge to reassess my impression that Parfit stopped
short was a sufficient motive for me to review and type in all 154
topic headings and create the bookmark structure for my electronic
copy of the book.  On a possibly relevant note, was something of the
book's essence lost as I destructively scanned the sheets, or might it
be that the essence of the book was actually enhanced by being made
more accessible and interactive via the electronic medium?

While doing so I refreshed my recollection of the topics, subtopics
and arguments, and I found that my impression has not substantially
changed.  Very briefly it is this:

Parfit argues theories of personal identity and of morality. He does
so within the paradigm of classical analytical philosophy, uncovering
branches of reasoning and pruning those branches via the discovery of
inconsistencies at various levels.  This approach follows an esteemed
western philosophical tradition thousands of years old and still very
influential and active today, but such a paradigm, in its Aristotelian
rigor, transparently assigns the human reasoner a privileged position
as an effectively rational, objective observer of the world being
described.  This is quite acceptable when the topic being described is
within a sub-context that can be treated effectively objectively,
however the topics of personal identity and morality are subjective at
a fundamental level. This paradigm, strikingly to me, neglects the
subjective certainty, systems theory, evolutionary theory and
semiotics that (seem to me) essential to understanding not only our
world, but more importantly, our understanding of our world.

With regard to personal identity, Parfit shows very clearly and
correctly that there is no logically warranted basis for belief in a
discrete self.  He then proceeds to espouse what he calls the
Reductionist View, claiming that

  "...we cannot explain the unity of a person's life by claiming that
the experiences
  in this life are all had by this person . We can explain this unity
only by describing
  the various relations that hold between these different experiences,
and their relations
  to a particular brain . We could therefore describe a person's life
in an impersonal way,
  which does not claim that this person exists."

True to the spirit of Reductionism, it can't be faulted on its own
terms, but it misses the point that the meaning of personal identity
doesn't inhere in such detail, and stronger yet, such detail is to a
variable extent irrelevant and could be altered with no practical
effect on personal identity.  Personal identity has no ontological
status; it consists entirely in its role as a label.  Parfit
recognizes the logical inconsistency of the concept of a discrete
personal identity, but retreats into reductionism rather than taking
the conceptual leap to a more encompassing paradigm encompassing the
observer and the process of meaning-making.

I don't intend to join the debate on the strengths and weaknesses of
reductionism, but my main point is that explanations should be
evaluated according to their explanatory power which is necessarily
dependent on context.  I went down that same road many years ago and
found that there was no Self at the end.  This was significant,
because it meant I could stop looking for an ontic Self.  I then asked
myself, what is the *meaning* of self, and found a very effective
extensible operational description that has passed testing for several
years since.  Meanwhile, some people are still down near the end of
that road scratching around and looking for what *is* Self, and Parfit
appears to be standing at the very end of the reductionist road
shaking his head knowingly and saying "there is no self, but it leaves
tracks in terms of its relationships to other things."  There's still
a bit of unresolved dualism left in that view, resolvable by expanding
the paradigm.  People deal perfectly with personal identity every day,
without ever resorting to tracing the web of relations to a particular
brain.  Nature deals perfectly with soap bubbles without ever
computing the infinite expansion of the digits of pi.

Further, as I have argued in other posts, personal identity depends
not on any physical, functional or historical similarity whatsoever,
but rather on perceived agency with respect to an abstract entity. The
agency is perfectly knowable, while the entity is only indirectly
knowable, even if it's oneself. (There's only one mention of agency in
the book, and that's the "agency of hearing".)

With regard to moral theory, Parfit recognizes the moral problems of
narrow collective self-interest, but concludes that these imply the
need for a more *impersonal reductionist* approach.  He apparently
doesn't consider the possibility of a more coherent but typically
non-intuitive approach of *broadening* the context of
self-identification, in other words making decisions not impersonal,
but *more* personal, over larger context of decision-making.  This is
again due to operating within a  reductionist paradigm.  He describes
with great accuracy the pitfalls of consequentialist ethics, but does
not appear to consider anything like morality assessed as the extent
to which the values of an increasing context of decision-making are
expected in principle to be promoted over an increasing scope of
consequences.  He realizes that the discrete Self does not exist, but
does not follow the implications that a fully *effective Self* most
certainly does (else who makes decisions, and is assigned
responsibility for consequences?) and he does not consider that this
effective Self could effectively identify with an expanding sphere of
values much as a good mother identifies with her children, and further
in an expanding sphere of understanding of our causal and
consequential inter-relatedness which is in no sense arbitrary but
rather an increasingly probable outcome of evolutionary processes.

In the concluding chapter, Parfit mentions the Non-Identity problem as
yet unresolved, along with the Mere Addition "paradox" and the
resulting Repugnant Conclusion and Absurd Conclusion.  I have not
taken the considerable time that would be required to construct an
adequate reply, but these problems seem to be the natural result of
assuming a privileged status (both moral status and observer status)
for humans, rather than reasoning from a more realistic  Systems
Theoretical paradigm of computing agents reasoning within bounds,
evaluating choices relative to necessarily local sets of values, and
acting not as objectively rational goal seekers, but subjective values
promoters.

I recall that when reading the book several years ago, I was impressed
with the depth and breadth of logical rigor, but disappointed that the
work seemed transparently enmeshed in the classical paradigm of
analytic philosophy, with little or no consideration of the
implications of the semiotics of  subjective agents embedded in the
very reality being considered.

Parfit's work is a quite comprehensive and rigorous analysis of
concepts of personal identity and provides a valuable contribution to
the field, but remains confined within the limits of his paradigm.  In
the concluding chapter he acknowledges some weaknesses and paradox and
expresses his belief that "others could succeed."

A work as important as Derek Parfit's _Reasons and Persons_ deserves a
well researched and presented response and so I would not want to give
the impression of criticism of what stands, when my intent, time
allowing, would be one of extension. I apologize in advance for the
sloppy and incomplete presentation in this brief email.

Paradox is always a matter of insufficient context.  In the bigger
picture all the pieces must fit.

- Jef