[extropy-chat] Re: Re: (Ethics/Epistemology) Arrow of Morality [Was:The statement that

[john-c-wright at sff.net]

Dear Mr. Allbright, My apologies for the delay in penning this reply, but other
matters have whelmed me. You have given me much to think about, and,
unfortunately, much to say, so I apologize for the length of this letter. 

You will see in the letter that my understanding is limited and weak. Your
conception comes from something alien to my rationalistic tradition, so perhaps
you can explain your conception to me in simple terms. Please do not interpret
my disagreement as any sign of disrespect.

Now, to the matter:

To the best of my limited understanding, your conception of an arrow of morality
has three shortcomings: first, it is useless to any who do not accept mere
survival as the ground of morality; second, it is mute to determine what objects
should be included or excluded from the moral order, some of which are already
universal in any case; third, morality by its nature must be treated as if it
were an absolute by its partisans, or else it has no ability to act as a moral
code. 

1. I asked for a description of what is meant by your idea that morality is
“what works”, especially were a given system of moral rules or reciprocities
works better for a larger group than for a smaller included group. Identifying
“what works” in an empirical science is easy: when the results as witnessed by
our eyes are as predicted, the theory giving rise to the prediction is said to
have “worked.” If the results are other than predicted, no matter what good the
theory may be on other grounds, it does not work as an empirical predictive theory.

I then asked how to apply this to a normative science, such as ethics, where we
are not dealing with theories predicting what will happen, but, rather, with
maxims of what men ought or ought not to do.  

You say: "what works" means a structure that will tend to survive and grow,
regardless of whether it is fully comprehended by any observer system.

I submit, however, that there is no ground to say one thing “works” and another
“does not work”, without a normative axiom beforehand to define what works. 

Mere survival is insufficient for this end: if the human race, for example, were
promised mankind would enjoy a population level on average of two hundred
millions, guaranteed to survive at least two hundred thousand years, if only we
were absorbed into a Borg cube, and lost our souls; or if we were, given the
alternative, offered a population only of one hundred millions and a span of one
hundred thousand years if we are members of the United Federation of Planets, I
would select the Federation over the Borg for myself and my children. 

My point is that only if I accepted the normative axiom that mere survival at
any price were the supreme governing moral principle, am I would obligated to
accept the offer of the Borg to be assimilated.  They offer twice as many
survivors to last twice as long. But this is not a axiom I accept: there are
times when it is better for the nation to perish than that one innocent man
should die. Thank goodness, those times are rare, but the mere existence of
normative values no lower than mere survival makes me chary of accepting your
formulation without some additional argument to support it.

2. The second basic problem with the “arrow of morality” formulation, is that it
cannot be used to tell in what direction the arrow of morality should grow, and,
hence, cannot tell a man how he should act. This is a specific application of a
general philosophical error when dealing with evolving or changing standards: a
standard, by definition, if it changes, cannot be used as a standard. 

A moral code is a specific formulation of the universal morality; the main point
of difference from culture to culture or age to age is the scale of the moral
code: whether the moral order protects and commands one’s neighbor’s only, one’s
tribes or nation, or all mankind. This seems to be the arrow of morality of
which you speak, the motion from a parochial to a cosmopolitan moral code. 

The Stoic, the Christian, the Buddhist and the Mohammedan each embrace a code,
which is universal and cosmopolitan. Pagan codes of honor (with all due respect
to my pagan ancestors) are parochial; pagan gods were meant to be the tribal
gods of a given tribe, and their rules were never claimed to protect or to bind
strangers from the antipodes. But the monotheistic religions made the assertion
of universality. These systems claim to apply to every living soul. The
parochialism of the previous tribal gods was rejected by the Roman and absorbed
by the Hindu: antislavery societies, believe it or not, existed among the
Imperial Romans and during the Christian Dark Ages. Likewise, the followers of
the Prophet were forbidden to enslave any of their fellows who submitted to the
Will of Allah the Compassionate, the Merciful. The size of the group to be
considered covered by the moral order increased from the local tribe to all
mankind with these cosmopolitan religions.  

Communism and Nazism, of course, reverse this. These aberrations spring from a
higher culture and reject it, restricting the moral order no longer to mankind,
or even to Christendom, but only to members of a favored race (in the case of
the Nazi, the so-called Aryan)  or to members of a favored economic class (in
the case of the Communist, the so-called Proletarian). Their savagery exceeded
that of the ancient pagans, perhaps in part because the knowledge that they were
betraying a conception something finer and higher than their own tormented them. 

Oddly enough, in the modern era, two factions among us are attempting to
increase the scope of morality in two opposite directions. Some would insist
that the moral order protect animals; others would insist that the moral order
protect unborn babies. The first would outlaw carnivores as cannibals, the
second would outlaw abortionists as infanticides. 

Now, is there any way to predict or prefer which way the arrow or morality will
go in the future? If we grant human rights to beasts, they might increase in
survival and growth (unless animal population numbers drop once they are no
longer domesticated for food); if we grant human rights to fetuses, they
personally will increase in survival, and families who otherwise would go
childless will grow. So which way is the arrow of morality supposed to grow?
Your formulation of “what works” seems as ambiguous as a Delphic oracle. 

3. No matter what the viewpoint from the objective observer as to the actuarial
benefit of adopting or rejecting specific innovations to a given moral code,
from the viewpoint within a moral code, the moral code itself will contain
reasons to explain and support itself. 

The observer outside the moral code talking to the partisan within the moral
code may say anything he likes about the “growth and survival” benefits of a
particular innovation; but, unless he speaks to the specific reasons why the
partisan adheres to a particular moral code, the information is of no value to
the partisan. 

An example might make this clear. Suppose we have two men, both of whom agree on
the basics of a moral system. Let us say one is a Franciscan Monk, the other is
a member of the Military Order of the Knights of the Temple of Jerusalem. Both
are Christians, but one has taken a vow of pacifism, the other, a vow to recover
Jerusalem from the Paynims by forces of arms. 

An objective observer shows them Game Theory. He explains to our Monk and our
Knight a simple game, called the Prisoner’s Dilemma, where if two players each
cooperate with the other, they break even; if one cooperates and one betrays,
the betrayer wins; if they both betray, both lose. Our objective observer
convinces them that one and only one strategy is favorable over the long term:
the strategy of simple reciprocity. Namely, a player who is willing to cooperate
with other players until betrayed, to betray once each time he is betrayed, but
to forgive and cooperate again next trial.  Our observer might urge, for
example, that the pacifist retaliate upon certain occasions in order to deter
further attacks; or he might urge the Crusader to fight only defensively, and to
cooperate with the Turk whenever possible.  

No matter what these calculations are, Christianity is fundamentally
otherworldly, so that their survival rate on Earth could not have been the prime
concern to our Monk and our Knight when they took their vows. 

Even if the Monk and the Knight are carefully convinced by our observer, his
arguments can have no effect on them, because they do not share the normative
axiom that survival and growth are paramount concerns. 

In other words, a basic problem with the “arrow of morality” approach is that it
is in fact not objective, merely one philosophy like any other, apparently a
form of utilitarianism. It will not convince anyone who does not already share
the axioms of utilitarianism (albeit, it might be useful as a predictor of which
philosophy will be the most popular or longest-lasting.) 

As a related thought, let me submit that humanity has no need of changing
standards in morality, nor any ability to change those standards even if it
wanted to. Morality, in human experience, is one unified structure, and always
has been: a set of maxims or imperatives commanding human action. 

I propose that the emphasis on certain maxims or imperatives, their rank or
priority, might differ from one man to the next or from one school of thought to
the name, but the general moral order of the universe known to all men through
their natural reason. What is amazing about the various ages and races of man,
is not that we see, here and there, customs of particular cruelty or degeneracy,
such as temple prostitution or human sacrifice, but that we see nearly universal
agreement on the basics: the Eightfold Path of the Buddha and the Ten
Commandments of Moses cover the same points, as do the utterings of Lapland
witches and the staves of Norse prophecy. Even vicious beasts like the
Communists can only justify their shocking inhumanity, their brutal
mass-murders, mass-lies, mass-robberies, and so on, by reference to a moral
maxim (charity to the poor) which has at least the same pedigree as the moral
maxims condemning murder, lying, and robbery. 

Hence, the universality of moral maxims suggests that moral systems cannot
differ in their fundamentals. They differ in the arguments used to support the
maxims, and they differ in the different weight given the moral maxims compared
one with another. (The Chinaman and the Jew, for example, will both acknowledge
the moral maxim of respecting one’s parents: but Chinese tradition has a much
more elaborate and demanding system of ranks within each family than the Jewish.) 

We can look at morality only one of two ways: from the inside, or from the
outside. From the inside, we, as moral beings, can weigh in our consciences the
wisdom of changing the emphasis or scope of a moral rule to which we all defer
as authoritative. The arrow of morality can grow only in the direction already
implied, but not yet come to flower, in the maxims we already accept. From the
outside, we, as purely rational beings, can look at some aspect of morality or a
particular moral code in non-moralistic terms, such as, for example, looking at
the incentives which cause certain formulations of local moral codes to flourish
or gain partisans, while other diminish. The difficulty with the arrow of
morality formulation (as I understand it) is that it cannot bridge this gap
between inside and outside. 

The man inside it does not need it: Christendom already preaches and practices
toleration of dissent. The man outside cannot use it: knowing a certain code
will reach more people or “works better” than another code provides no
particular motive to amend a moral code. It might or it might not, depending on
what he thinks the moral stature of “working better” is.  

JCW


--- Original Message ---
From: Jef Allbright <jef at jefallbright.net>
To: ExI chat list <extropy-chat at lists.extropy.org>
CC: john-c-wright at sff.net
Date: Wed, 13 Apr 2005 09:13:54 -0700
Subject: Re: (Ethics/Epistemology) Arrow of Morality [Was:The statement that

> john-c-wright at sff.net wrote:
> 
> >My apologies for the length of this post, but as a humble disciple of
> >philosophy, I find such specualtions sweeter than wine, despite that the eternal
> >questions have been debated, well, eternally.
> >  
> >
> That some questions are debated eternally is a strong hint to widen back 
> and view such  issues from a broader perspective. 
> 
> I too enjoy philosophical speculation but available time is quite 
> limited and after three iterations I seem to have made almost no 
> progress conveying even the first of my points - the significance of 
> nested scopes of context.
> 
> >Below, A is for Allbright and W is for Wright. 
> >
> >A: "I'm afraid my main point was lost -- again due in no small part to my
> >tersity -- the point being that *all* subjective input should be considered, but 
> >then weighted according to its [ultimately subjective] grounding in 
> >empirically verifiable "reality"."
> >
> >W: Your point was not lost on me, I hope. What I was trying to do was argue the
> >opposite side of this very question. My argument was twofold
> >
> >(1) The statement is self-contradictory. Those who argue that truth (what you
> >call "input") is ultimately subjective, argue as if truth is objective. Those
> >who put the word "reality" in scare quotes argue as if they are talking about
> >reality, that is, real reality without any scare quotes. 
> >  
> >
> As I've stated three times now in this discussion, I do believe in 
> ultimate reality, highlighted by my statements that from a god's eye 
> point of view, all is objective.
> 
> Certainly I do argue from assumptions of truth and consistency, as has 
> been agreed and doesn't require pointing out yet again.  I suspect that 
> a relatively superficial interpretation of my statements on subjectivity 
> and incompleteness may lead one to conclude that I am arguing from an 
> existentialist, nihilist, or (forbid) postmodernist point of view, but I 
> have tried to point out that (1) I assume there is an ultimate reality 
> (but I don't say it's static), (2) I point out that our knowledge of 
> reality is necessarily incomplete and subjective, and (3) I argue that 
> there is progress toward increasingly accurate approximations of 
> reality, measured in terms of what works.
> 
> It would be foolish to argue the validity of your three self-evident 
> truth statements, except (as I tried to do) to point out that the system 
> doing the truth evaluation is itself limited to a subset of reality 
> (truth) and is therefore inherently limited in its ability to certify 
> absolutely the truth of any proposition.  No matter how obvious a 
> proposition may appear, it may be superseded by a more encompassing 
> interpretation.  This is a pragmatic approach to truth, but does not 
> deny the correctness of your examples within the domains in which they 
> were intended.
> 
> >(2) The statement rests on the assumption that moral maxims can be supported or
> >denied by means of reference to statements of observation, what you call the
> >naturallistic fallacy. 
> >
> >For example, comparisons of the statistics of the crime rates and the use of
> >torture might tell you whether or not torture has a deterent effect on crime,
> >but this reveals only whether it is efficient, not whether it is morally upright
> >to use torture as an instrument of law-enforcement. 
> >
> >The statement that torture is efficient is a contingent statement: the statement
> >is true if the statistics support it, false if not, and in any case is dependent
> >on the accuracy of the demographic data. The statement that torture is barbaric
> >is an absolute statement. The statement may be true of false, but, no matter
> >what, statistics will not show whether the statement is true or false because
> >"barbarism" is a moral condemnation, not an thing that can be measured by a
census. 
> >
> >  
> >
> Yes, I hope we can agree that the Naturalistic Fallacy of attempting to 
> derive "ought from is" is in error because value judgments are 
> necessarily subjective.
> 
> >A: "My theory doesn't provide absolute moral answers, but it claims that 
> >there is a rational basis for finding increasingly moral answers."
> >
> >W: Do you agree that the idea of increasingly accurate measurements only makes
> >sense if there is some real thing being measured? 
> >  
> >
> 
> I've said that I assume (point of faith) an ultimate reality against 
> which our measurements are necessarily approximate and incomplete.  Can 
> we get beyond this starting point?
> 
> >We cannot get ever-more-precise measurements of the speed of light in a vacuum
> >unless the speed of light actually exists. Likewise, we cannot get increasingly
> >ever more objective and increasingly ever more correct maxims of morality unless
> >there actually is a moral order to the universe. 
> >  
> >
> 
> John, I'm breaking protocol here, but I would like to draw your personal 
> attention to this statement, which is causing me some frustration.  I've 
> repeatedly made this point.  In fact, during our last exchange I showed 
> that you had inverted this point and now you are challenging me to make 
> it again!  So here it is, copied verbatim from the previous exchange:
> 
>     On the contrary, I argue that from the God's eye view, morality is
>     in fact objective. However this ultimate view is only approachable,
>     but not obtainable.
> 
> Please remember, I started by saying my intention is to make three 
> points relating to the following:  (1) nested scopes of context, (2) 
> subjectivity and nature of Self, (3) what we call "good" and thus moral 
> is measured by what works, and what works over a larger context is 
> (necessarily subjectively) considered better.
> 
> >A: (Works better with what end in mind?) Anything that subjectively promotes
Self. 
> >
> >W: This is a subtle thought, and I am sorry you have no time to write it out
> >more clearly. 
> >
> >If you do get a chance at some later date to expound on this principle, I, at
> >least, would be interested in the disquisition. 
> >
> I have no hope of currently being able to convey this, which I consider 
> a more subtle point, considering the difficulty we're still having with 
> the previous.  You may wish to refer to Derek Parfit, _Reasons and 
> Persons_, for a rigorous philosophical analysis of a large portion.  
> Buddhist thinking and discoveries in cognitive science fill out the concept.
> 
> >My main question would be how to
> >reconcile that three examples I gave of the heroic Achilles, the saintly John
> >the Baptist and the wise Socrates with this principle of self-growth. It seems
> >to me that  the hero, the saint, and the philosopher all value something greater
> >than himself (glory, God, or truth) for which he is willing to lay down his
life. 
> >  
> >
> Yes, the key is that while all moral choice is necessarily subjective 
> (from the viewpoint of self), the Self with which one identifies can be 
> (and generally is) greater than the conventional concept of self limited 
> to one's physical body.  We identify with our goals which extend outward 
> and into the future; we identify with the society within which we are 
> enmeshed, and we identify (at a less than conscious level) with kin and 
> other members of our in-group.
> 
> As I stated in an earlier exchange, and quoted a few paragraphs above, I 
> do not propose to provide absolute moral answers, but I claim that we 
> can develop a rational basis for finding increasingly moral answers.
> 
> In the cases of Achilles, John the Baptist, and Socrates, their actions 
> can be considered moral to the extent that they were performed with the 
> "greater good" (with which they must necessarily identify) as their 
> objective.
> 
> >I would be interested to see how self-sacrifice can be reconciled a philosophy
> >which takes self-growth as its foundation. 
> >  
> >
> 
> I hope this was sufficiently elucidated above.
> 
> >A: (Works better for whom?) Works better for Self.  "Better" is inherently
> >subjective [meaning dependent on the observer].  Self means that with which one
> >identifies.  
> >
> >Q: This sounds like a formal description rather than a moral maxim. I suppose
> >one could define "self" broadly enough to include the divinity or the community
> >so as to explain the self-sacrifice of saints and heroes. (In other words,
> >Socrates considers his "Self" to be the laws of Athens, and loyalty to their
> >precepts, even when the laws are in the wrong, justifies his drinking hemlock.)
> >  
> >
> 
> Yes, as described above.
> 
> >But, by the same token, one could define the "self" and the "growth" of
> >Raskolnikov to include that he must kill an innocent old crone and her
> >halfwitted half-sister. 
> >  
> >
> Raskolnikov, in his dysfunctional state of mind, believed he had moral 
> justification for his actions.  He believed, necessarily subjectively, 
> and within his limited context of awareness, that his actions were for 
> the greater good.  Along the same lines, political assassinations have 
> been performed, atrocities have been committed in the name of religion, 
> and preemptive war has been carried out, all for ostensibly moral reasons.
> 
> In all these examples, we can see that moral goodness is subjective and 
> limited in terms of context of awareness.  No absolute moral maxim helps 
> here, because such moral absolutes can and have been used 
> interchangeably by either side.  My point is that we can not know 
> absolute moral certainty, but we can recognize that as the context is 
> broadened, in terms of the number of actors, the variety of 
> interactions, and the time over which interactions occur, we can 
> evaluate, with increasing agreement, the relative morality of an action 
> in terms of how well it (subjectively) works.
> 
>  From this realization, we can proceed to develop a science of 
> principles of effective interaction.  Our understanding in this area is 
> growing with studies in complex systems theory (and other areas) which 
> can be applied at all scales from the cosmic, through inorganic 
> chemistry, through biological evolution and on to human culture.
> 
> - Jef
> http://www.jefallbright.net