From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!neat.cs.toronto.edu!cbo Mon May 25 14:05:47 EDT 1992
Article 5694 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!neat.cs.toronto.edu!cbo
Newsgroups: comp.ai.philosophy
>From: cbo@cs.toronto.edu (Calvin Bruce Ostrum)
Subject: Re: AI failures
Message-ID: <92May16.003923edt.48037@neat.cs.toronto.edu>
Organization: Department of Computer Science, University of Toronto
References: <1992May13.173251.17396@organpipe.uug.arizona.edu> <1992May13.234419.17060@news.media.mit.edu> <92May14.134243edt.47895@neat.cs.toronto.edu> <1992May14.234328.12094@news.media.mit.edu>
Date: 16 May 92 04:40:08 GMT
Lines: 254


Marvin Minsky asks for explanation of what he apparently takes to be
   my fit-induced ranting:
| > ...  Even if I am wrong, his sentiment
| >and Marvin's boosting of it help to promote a discouraging, and perhaps even
| >a dangerous, view, which is that there is no point to moral reasoning since
| >"it's all circular". 
| 
| 1. Well, I don't know what you mean by discouraging.  Discouragement is
| in the mind of the listener, not in the nature of the assertion.

I believe I am using "discouraging" in more or less the normal, everyday 
sense.

As for the second sentence above, this is a nice turn of phrase, but I 
don't see how it affects my point.  While it may be roughly correct in
what I take to be its implication that different people may be discouraged 
by different things, this does not change the fact that these things that
are discouraging are not always (or, perhaps, even usually) in the mind, 
but are often in the outside world.  I also object to the narrowing of 
discouragment to the *mind*.  It is usually a complete person who is 
discouraged.  The expression "it's in his mind" is not the most usual way 
of talking, and could be unfortunate for two reasons.  First, the expression
"it's in his mind" is often used to disparage the reality of something, 
or, even if the reality is acknowledged, to place the responsibility 
for the trouble it may cause on the person whose mind it is.  (This is not 
to say that I am necessarily accusing Marvin of doing these things). Second, 
the expression may make us tend to forget that the mind, and the person, 
are affected in more ways than just by receiving cognitive input. 
Marvin aboves limits this receiving even further to "listening" to the
expression of a few, which is okay when considering the direct effects
a view has, but which is not okay when considering its indirect effects.

Now, to the point that different people are discouraged by different things:
I agree this is true, and I did not say that *everyone* would be discouraged
by the view that there is no point to moral reasoning.  Hans Moravec, for 
example, in his megalomaniacal urge to solve "wonderful problems" and 
build all sorts of new persons to play with at his whim, would probably
be uplifted by the assertion that "there is no point to moral reasoning".
Moral reasoning, and the windbag notion of morality generally, would really
just get in the way of his projects, or so he seems to think. And, sadly,
there do appear to be quite a few others that share his sentiments. 

However, I should hope I would not have to explain that many people feel
that it is important to do the right thing, and they feel that reflecting
on what to do (by engaging in moral reasoning) is an effective way to
increase their ability to succeed at this.  I think it would clearly
be discouraging to many of them if they thought many people thought
this activity was pointless, and if no one else would engage in it with
them, but, say, just shrug and ignore them (or worse yet, beat them up).

| 2. Dangerous to whom, and how.  The danger of faithfully adopting some
| unsupported answers to questions is hard to assess.  Please elaborate.

Note, by the way, that I did not say that "faithfully adopting 
unsupported answers" was dangerous.  What I did claim was the view 
that 'there is no point to moral reasoning since "it's all circular"' 
was perhaps dangerous. Marvin is either misreading me, or jumping way too
far ahead of me. 

There are two distinct ways this view could become dangerous.  One is that
people might, if convinced of its truth, stop relying upon moral reasoning,
and rely primarily upon their moral intuitions: if it feels right, do it.
This is a bit stronger than "faithfully adopting some unsupported answers", 
because unsupported answers may still be used in moral reasoning (I've 
been using this phrase "moral reasoning" quite a lot, without really saying
what it is. I just mean regular reasoning, as far as possible, but reasoning
with and about moral concepts).  But if people think moral reasoning is 
pointless, they might have less tendency to ever do it.  They may end up
paying an inordinate amount of attention to their initial moral impulses. 
I doubt the results of this are be to be recommended. I don't intend to
argue for this claim at this time, however. Particularly not in
this newsgroup, where it, like the rest of this post and nearly all of
this thread, does not really belong. (pace Bill Skaggs).

The second way that this view could become dangerous is not related to
adopting unsupported answers about what is morally right at all.  It seems
reasonable to believe that the view that moral reasoning is pointless would
tend to undermine the entire institution of morality.  If reasoning can't
be used to arrive at new moral truths, like it can in other domains, people
will start to emphasise a sociological view, and say, more and more, that 
"morality is just what we were taught", "morality is just what you can 
get away with", "morality is for suckers", "morality and expedience 
are more or less the same thing", "morality is for humanist windbags", etc.

They may end up inverting the order of complexity, and think that science
can easily provide both the questions and the answers that should truly
matter to us -- acting on, and pushing for social policy based on, 
views such as the following (courtesy of Hans Moravec):

| To me it's utterly clear that such a utilitarian, mathematical
| approach to human (and animal, plant and robot) behavior is light-years
| ahead of one based on ancient humanistic concepts like morals.
| At best the humanistic concepts are oversimplified, distorted,
| cartoons of the underlying utility mechanisms.

and offering as an example of their sophisticated "scientific" approach the
following sort of analyses, meant also to "justify" what they plan to do, 
given the opportunity (again, from Hans):

| There are many situations (self defense, war, capital punshment, 
| abortion) where present society condones elimination of individuals 
| that are judged to have a negative  net worth.

The only way such people would fail to be dangerous, in my opinion, is
if they are prevented from exerting a great influence.  Fortunately, I
believe that serious attempts *will* be made to prevent them from exerting 
an undue amount of influence, should they ever start to acquire the 
potential to do so.  But they remain a potential danger, nevertheless.
I almost forgot: To whom?  We moral folk of course.

I admit that I have an optimistic view of the power of moral deliberation,
and believe that people's employment of it, both as private individuals 
and as influential public figures, has historically produced some good 
results. I am aware that many do not think so, and I also admit that I 
have not proven this to my own satisfaction yet. But even if some 
"scientific" and completely deflationary Marxist type explanation such 
as that pointed at by Hans Moravec is correct, I remind myself and others
of Marx's own dictum that the point is not to understand reality, but to
change it.  I don't see that we should cease talking in terms of right
and wrong just because they turn out to be superfluous concepts in 
explaining human history.

| 3. In your fit of annoyance, perhaps you didn't actually read the
| whole page. I didn't say there was no point; onm the contrary I
| suggested that there were powerful and perhaps productive social
| consequences:

I certainly do not like Marvin's claim that my annoyance (if that is what 
it was) took the form of a "fit". Even worse is the suggestion, which appears
to be being made, that this (non-existent) fit coloured my reasoning to the 
extent that I was unable rationally to evaluate his posting (to the best of 
my extremely limited but nevertheless sober ability).  I would like someone 
to point out the "fit-like" behavior that I exhibited in that posting, that 
may have lead Marvin to his conclusion.

Marvin seems to think I am attributing to him *personally* the view that 
there is no point to moral reasoning because it is circular.  In fact, I 
did not say anything of the kind.  All that I said was that he "boosted" 
a statement (recall the "bravo"?) which helped to promote this view.  
I even acknowledged that the statement boosted may not entail the view,
but may only promote it (i.e., help to make it more common. It can do that
without entailing it).

Once he has attributed to me a claim that I never made, Marvin attempts
to show that the claim is false by citing other parts of his posting, and
suggesting that I made this false claim because I never read these parts.
But I did read them, and in fact they are exactly what made me able to say
the following, which Marvin really does seem to have missed:

| But as Marvin points out in the above quote, reasoning is circular in ALL 
| domains, when carried on at a sufficiently abstract level.  Nor is it 
| obvious that all such circularity is vicious, as opposed to virtuous, 
| another point which Marvin makes clear.

While it is true that I toned it down a bit, ("nor is it obvious") it
is clear that it was Marvin's stronger statement later in his posting 
that enabled me to attribute this somewhat weaker view to him.  And it's
clear I acknowledge Marvin's view that circularity need not imply
pointlessness.

I agree with Marvin that circular reasoning can be virtuous, although 
it is not clear to me exactly how (please, everybody, no one-line
solutions unless they are informed by intimate familiarity of some part of 
the large and serious philosophical literature on the subject. Otherwise, 
I suspect I will merely be offered crude reinventions of some already 
existing wheel). That is why I objected to his cheering of the original 
statement, which in its completeness, was (courtesy of Bill Skaggs):

| >  The problem with all this theorizing about morality is that it's
| >inescapably circular.  This is immediately obvious once you
| >realize that the fundamental question is:  "What is the right
| >system of morality?"

To me, the gist of this statement is that there is something *wrong*
with moral reasoning, and that this is exactly because of its "inescapable
circularity".  This may not be what Bill actually meant.  He may have
meant that only the few specific lines he sighted, and perhaps referred 
to with "this", consituted the "theorizing".  But since there seemed 
to be nothing distinguished about those lines, I assumed that the "this" 
meant "all this kind of 'morality' stuff".  Alternatively,  Bill may have 
meant "problem" as "challenge": i.e., not that there was something wrong 
with moral reasoning (rendering the whole enterprise suspect) but rather 
that it was difficult and challenging (making it hard but not impossible 
to achieve significant results).  However, I do not believe that Bill 
intended either of these possible interpretations. And, I claimed (see 
above) that even if I was wrong about the view I attributed to Bill, 
others still maintained a similar view, and that this kind of statement 
nevertheless tends to support that view.

I believe that Marvin's statement "bravo" is properly taken under usual
rules of English implicature as referring to the gist of this statement.
But his lengthy quotation from Society of Mind did *not* support the notion
that there is something wrong with moral reasoning, compared with any other
kind of reasoning.  That was the main point of what I said in my brief 
posting.  Probably Marvin intended his approval to extend only to the 
assertion of circularity, removed completely from its important surrounding 
context.  But that's akin to saying "bravo" to something like:

 "The problem with all these black people is that they have [some feature 
  or other that Marvin finds very interesting, but which is possessed by
  all people, including of course, all white people]. This can be seen 
  by [asking yourself some particular question]."

I wanted to point out that it's probably not a good idea to say "bravo"
to such a passage.

Finally, Marvin appears exasperated at being unable to figure out what
I was trying to say.  He makes a guess, which I think is an extremely
wild one:

| Are you saying that we
| should support moral reasoning under all circumstances, even it if has
| circular foundations, and no matter where it might lead?  My goodness,
| what _are_ you trying to say?

I thought it was quite clear what I was trying to say (in terms of what
we should support, that is).  We should support the view that there is a 
point to moral reasoning, regardless of any arguments that claim "it's all 
circular".  I certainly don't think I ever said anything resembling the 
above suggestion!  Nevertheless, I will briefly comment on it.

The section from Society of Mind that Marvin quoted did not really
distinguish moral reasoning from any other kind of reasoning. (In fact,
the index entry for "morality" does not even refer to this section!) Hence,
I would answer Marvin tentatively by saying that we should support moral
reasoning in the same kind of way we support reasoning in general. In fact,
it is not even really clear where one kind of reasoning leaves off and 
another begins.

Marvin's own question, in fact, is a moral question.  If I were to answer
it with "yes", he might very well answer "why?".  In other words, he is
looking for a reason.  He is expecting some moral reasoning.  For sure, at 
some point, and in specific circumstances, sitting around reasoning would 
not appear to be the right thing to do.  It might lead to being eaten by 
the enemy, for example, and that might be something that earlier moral 
reasoning, or intuition, or even maybe observation, had decided was wrong.  
So of course moral reasoning is not to be supported under all 
circumstances: it seems clear that there are at least some circumstances 
where it would be the wrong thing to do.  What is less clear is what those 
circumstances are.  Pencils (ahem, workstations) ready, Hans?  Let's 
calculate!

---------------------------------------------------------------------------
Calvin Ostrum                                            cbo@cs.toronto.edu  
---------------------------------------------------------------------------
   If I tell you that to let no day pass without discussing goodness
   and all the other subjects about which you hear me talking 
   and examining both myself and others is really the very best thing that
   a man can do, and that life without this sort of examination is not
   worth living, you will be even less inclined to believe me.
   Nevertheless, that is how it is.
   	-- Socrates, according to Plato, "Apology"
---------------------------------------------------------------------------


