(1) If one includes a purely contemporary physical ontology in with =
"physical explanation" (as most materialists and functionalists would =
want to do), then I think one has a good argument that HAL (or a person) =
is not conscious if that person is -fully- explainable in current =
physical terms.
(2) If one leaves out ontology, and just says HAL's/the person's =
physical side is -fully- explainable in -current- physical terms, then I =
think one has a good argument that (a) any consciousness would be =
epiphenomenal, and (b) there's no argument to believe he is conscious, =
aside from an argument from analogy from the premise of -our own- type-a =
consciousness, which argument isn't as strong as our confidence wrt =
other people's consciousness, and which argument is -very weak- for =
objects such as HAL with not only a different architecture (at any but =
the highest level), but different underlying hardware, the only =
similarity being behavior which is, as noted, completely explained =
without reference to type-a consciousness. (Might such a belief be =
properly basic, that is, held at least partially as a foundational =
belief without complete, independent evidential or argumentative =
justification? Wrt other people, I think that's very plausible [maybe =
even rationally necessary]; but for HAL, that seems very implausible. =
Hence, we rationally need a good argument, which we utterly lack.) =
However, you're very right in pointing out that this still appears =
-logically compossible- with even strong, type-a consciousness.
(3) You asked:
>How in the world can can one get at the ontology of the situation from =
the
>physical behavior *if* that behavior obeys the laws of nature? And if =
it
>doesn't, just how does the type-a consciousness act to violate those =
laws?
I think those are EXCELLENT questions: pithy, relevant, and tough. If =
we accept current physics as the last word on mind, it ain't easy AT ALL =
to see how type-a consciousness gets into the picture. This is roughly =
why I made the two arguments above, and also the corollary argument that =
if a being has type-a consciousness, it is not fully explainable =
(causally and ontologically) in current physical terms. The corollary =
to the corollary is that, unless one accepts eliminative materialism for =
people (scientifically, no problems, but intuitively rebarbative), we =
need either to accept that physics is limited wrt its explanation of =
people (or any type-a being), OR that we'll need a somewhat new physics =
to deal with type-a consciousness (at least ontologically, very possibly =
causally/predictively too). I'm in no position to say which option will =
be taken. If I were (at least if I were in convincing detail), then =
-I'd- be going to Sweden! :^> At this point (and dealing with your =
second question above), we are many orders of magnitude short of having =
the practical predictive power necessary to know if brains act wholly in =
accord with current physics (i.e., physics without recognition of type-a =
consciousness) or not. If SO, then there'd be an argument either for =
epiphenomenalism (consciousness having no causal powers) or for a new =
physics that's new "only" in its metaphysical interpretation of the =
physical equations (like a new rival to the Copenhagen and =
Hidden-Variable interpretations of quantum mechanics -- different =
assessment of deep reality/ontology, but the same empirically predictive =
equations). If NOT, then the new physics would need at least -some- new =
math, too. (My guess? The latter. But that's just a guess.)
(4) Why the emphasis on -current- physics? Because while there's no =
room (I argue, and I think you agree) for type-a consciousness in =
current physics taken as the complete explanation, conceivably there may =
be in future physics (as in (3)). But if that happens, you ask, don't =
we just have the same problem all over again, the problem of type-a =
consciousness effectively being eliminated in favor of purely physical =
entities? I don't think so. Let's imagine this "type-a physics", a =
physics that completely incorporates type-a consciousness into its =
causality and ontology. Is it true that type-a consciousness is then =
wholly physically explicable? Yes! But because of the new aspects of =
the type-a physics, my HALlish argument no longer works -- complete =
physical explicability no longer eliminates type-a consciousness, it =
just calls the very same thing "physical". (Perhaps there are new =
particles or fields: "nouons" [from Gk "nous", mind], or "soulons", =
or... :^> ) Indeed, presumably quite the contrary: now, given =
sufficient empirical knowledge about the object in question, we'd =
presumably be able to DERIVE the fact of consciousness (or not) FROM =
THIS KNOWLEDGE PLUS PHYSICAL LAWS, without needing introspective =
reports. (good thing, since maybe the object can't speak!) Cool. And =
bizarre. And utterly hypothetical, of course!
(5) I think we each feel like we've been talking past each other a bit, =
and that's all too common in difficult conceptual issues generally, and =
consciousness in particular. This is vague stuff. (The vagueness =
-doesn't- render logical argument impossible or unsound, it just passes =
the vagueness through to the conclusions.) And sometimes, because of =
our different intellectual backgrounds, we can be precise or imprecise =
in ways that don't appear clear to the other. So only patience and love =
will keep us from following in the footsteps of the more adversarial =
interlocutors on this mailing list. ;^> Let me know if I fail in =
either -- I'm only human. Not always verbally gentle like HAL. On the =
flip side, I don't kill astronauts. (This humor's getting bad -- maybe =
I should take a stress pill....)
--John
----------
From: David Bowman[SMTP:dbowman@tiger.gtc.georgetown.ky.us]
Sent: Wednesday, February 05, 1997 10:09 pm
To: evolution@ursa.calvin.edu
Cc: rylander@prolexia.com; dbowman@tiger.gtc.georgetown.ky.us
Subject: Re: Turing test
>The nutshell response is this: even as devil's advocate, you seem to be
>conceding -my- main point, which is that if we take HAL (or a person) =
to be
>completely explainable in physically reductionistic terms, then HAL (or =
a
>person) lacks type-a consciousness, despite the presence of type-b
>consciousness. This defeats contemporary materialism and also most =
popular
>versions of functionalism, in my view.
I'm sorry, but it seems we have a failure to communicate here. I did =
not
realize that this was your main point. I thought you were arguing for =
HAL
not having type-a consciousness based on the fact that a physical =
description
sufficient to understand the mechanisms of his physical operation didn't
require it. This is different than what you say above. My point did not
address materialism one way or the other. It is true that I'm happy to =
grant
your point above as long as it is understood that the rest of the =
description
involving the immaterial type-a aspect of HAL (or a person), although
necessary for a complete description for him, was not neccessary for
understanding his physical behavior. Thus, there is no =
objectively/physically
measurable way to tell if he has this type-a consciousness or not =
(unless his
behavior violates physical law in some way--in which case the =
extra-physical
part of his description *is* then essential for understanding his =
physical
behavior, and these violations of physical law then do definitely betray =
his
type-a consciousness).
>It appears that our differences wrt HAL lie in whether or not we deem =
HAL to
>be completely reductionistically explicable (which I didn't spend much =
of
>time addressing), but I don't think that's worth arguing about. I =
think he
>is; you seem not to think so.
Contrary to what I intimated as my namesake, my client is agnostic about
whether or not HAL is "completely reductionistically explicable" or not,
precisely because the difference is physically unobservable short of =
ongoing
miraculous behavior. It is the unobservablity of such physical effects =
that
led my namesake to conclude that HAL indeed had a type-a consciousness
because HAL's physical behavior passed the Turing test so well and his
actions were indistinguishable from a being with such a type-a =
consciousness
whether or not HAL actually had one. IOW, it was an easy mistake for =
David
Bowman, the astronaut, to make *if* it is found out that HAL didn't =
actually
have a type-a consciousness after all. This is because, short of =
ongoing
miraculous behavior, the behavior of a type-a being is indistingushable =
from
a type-b machine that passes the Turing test (by definition). David =
Bowman,
the astronaut, is convinced that HAL was a type-a (but can't prove it) =
and
you are convinced he is not (but can't prove it). My client is agnostic =
on
the matter. (I personally side with you that HAL is not a type-a, but I
admit that I have no argument other than my own personal bias to offer =
as
support for this belief).
> I'm willing to let the matter drop =
there. (I
>wouldn't be if, as you sometimes imply, my goal were to "convince" HAL =
that
>-I- am conscious; but I don't really care at all about that, partly =
because I
>don't think HAL is conscious.)
But is your goal to convince anyone else of your view by a sound =
argument?
>One note, to restate for emphasis: by "is completely explainable in =
current
>physically reductionistic terms", I don't mean merely "has an empirical
>component that entirely acts in accord with the current laws of =
physics,
>possibly with an immaterial component that is (for all we know) =
something
>else entirely." No sir. I mean this is the whole ball of wax, wrt
>causality/predictivity, AND ontology.
How in the world can can one get at the ontology of the situation from =
the
physical behavior *if* that behavior obeys the laws of nature? And if =
it
doesn't, just how does the type-a consciousness act to violate those =
laws?
> And I should re-re-emphasize =
"current"
>in this: I made quite explicit that seeing type-a consciousness as =
going
>beyond -current- physicalism doesn't entail "working miracles". =
can be
based on ambiguous definitions.
>Thanks for your comments, and sorry my arguments made you lose your =
good
>humor, Dave!
I'm sorry my words seemed to you like I lost my humor. I've been tired
lately so maybe I could have chosen some of them with more care. I'm =
really
not angry at all--just mystified by how we can keep talking past each =
other
without connecting. I think my apparent caustic style may reflect the =
fact
that I've read too many exchanges between Glenn and Jim (they set bad =
example
for me). :-) I suppose blaming others in one's apology is bad form =
though.
David Bowman
dbowman@gtc.georgetown.ky.us