From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!usc!cs.utexas.edu!uunet!psinntp!scylla!daryl Thu Jan  9 10:34:03 EST 1992
Article 2552 of comp.ai.philosophy:
Newsgroups: comp.ai.philosophy
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!usc!cs.utexas.edu!uunet!psinntp!scylla!daryl
>From: daryl@oracorp.com
Subject: Re: Penrose on Man vs. Machine
Message-ID: <1992Jan8.133305.22630@oracorp.com>
Organization: ORA Corporation
Date: Wed, 8 Jan 1992 13:33:05 GMT

Mikhail Zeleny writes:

> Since the notions of a program halting on a given input, or a theory
> being consistent are fundamentally second-order, i.e. non-recursive,
> our ability to understand them is sufficient evidence of our ability
> to perform non-algorithmic tasks.

If your basic assumption is that understanding is non-algorithmic,
then Penrose' argument is unnecessary, and certainly doesn't add any
additional plausibility to your assumption. The fact that you can talk
about second-order notions doesn't imply that you are using second-order
reasoning; ZFC can talk about second-order properties, as well.

> Indeed, it is arguably true that all understanding is fundamentally
> non-algorithmic;

The question of this thread is whether the Penrose argument is
evidence in favor of this claim. Bringing in additional arguments that
our understanding must be non-algorithmic doesn't help to show that
the Penrose argument is correct; for an argument to be correct it is
not sufficient for the conclusion to be true, as you know.

> however, in view of our past disagreements, I shan't repeat an
> argument to that effect, limiting myself to the claim that it is
> intuitively obvious to me that I am capable of understanding.

The argument is not whether you understand, but whether machines are
also capable of understanding.

Daryl McCullough
ORA Corp.
301A Harris B. Dates Dr.
Ithaca, NY 14850-1313



