Re: future language

From: Scott Chase (
Date: Fri May 10 2002 - 23:57:17 BST

  • Next message: Scott Chase: "Re: Fw: future language"

    Received: by id AAA01237 (8.6.9/5.3[ref] for from; Sat, 11 May 2002 00:03:39 +0100
    X-Originating-IP: []
    From: "Scott Chase" <>
    Subject: Re: future language
    Date: Fri, 10 May 2002 18:57:17 -0400
    Content-Type: text/plain; format=flowed
    Message-ID: <>
    X-OriginalArrivalTime: 10 May 2002 22:57:17.0546 (UTC) FILETIME=[087F14A0:01C1F876]
    Precedence: bulk

    >From: "Douglas P. Wilson" <>
    >Subject: Re: future language
    >Date: Fri, 10 May 2002 09:31:29 -0700
    > > And I guess the guy who killed him thought he was saving the world from
    > > another fascist, but may instead have created a martyr whose party may
    > > do better in the election as a result.
    >Oh, for goodness' sake, one certainly doesn't need to kill fascists to
    >weaken, or eliminate fascism. Fascism exists all over the place, and is
    >seen in the sunny, and intellectually enlightened, most pleasant,
    >British Columbia, Canada, where I live -- we've had a surprisingly large
    >of Nazi's here, though the overwhelming mass of people opposes them
    >and sometimes tries to stamp them out with fascist tactics like denying
    >free speech.
    >But fascism is demonstrably a mathematical mistake, and being the most
    >politically naive person I know, -- not entirely an accident, since I
    >fond of Descartes' naive realism, though actually a slave to pragmatic
    >pragmatism myself -- sorry, rambling again -- being the most politically
    >naive person I know, as I started to say, I delude myself with the delusion
    >that once this mathematical error is pointed out to the fascists they will
    >quietly disband and go about their separate ways, no longer any threat to
    >anyone or anything.
    >The mathematical error is a neglect of the importance of error-covariance.
    >is explained on my homepage, (all too inadequately, sad to say), with a
    >stock-market prediction example (roughly, don't hire the triplets to
    >the market, they won't correct each other's mistakes but will reinforce
    >and also then mistakenly believe their consensus bodes well for accuracy),
    >instead hire three very different people to do it, so they are only likely
    >agree when all three are right about something. On my home page I take
    >more words to say this, and have only an artificial example, but there is a
    >telling little table, so it's better than nothing. If tolerant but
    >try visiting it at www.SocialTechnology.Org/dpwilson.html.
    >'Fascism', for those whose formal education glossed over this stuff, is an
    >exact literal translation of 'solidarity', and means "like people standing
    >like", similar people, e.g. all "racially pure" Germans, for example,
    >by to support one another, while casting out all those who differ, Jews,
    >Jehovah's Witnesses, homosexuals, etc. -- the ancient Roman fasces, or
    >of (very similar) sticks, (ancestor of the bassoon, the "ill wind that
    >blows good"), is the appropriate imagery here, as the name implies.
    >But in any system more complicated than a dead stick, similar people,
    >processing units, vitamin pills, and so on, make similar mistakes. The
    >mathematical term for 'making similar mistakes' is 'error covariance'. A
    >simple example should suffice.
    >I take two or three small multivitamin pills a day. But not all of the
    >brand and composition -- three different brands and compositions, in
    >More bottles to purchase, open, and remember to keep straight, so a
    >but clearly a safer way to go.
    >Biochemistry is not an exact science, and this fact sometimes creeps over
    >the technological correlative of biochemistry, the similarly named branch
    >chemical engineering that makes all these vitamin pills. So occasionally
    >manufacturers screw up a batch and instead of taking a few milligrams of
    >niacinamide, for example, you get a few milligrams of nicotine or some
    >very similar, extremely toxic pyridine derivative (did I remember that
    >correctly? I once did study a little biology and a little chemistry, but
    >was in some previous century or other).
    >So if you do what I don't, and take one large or three little vitamin pills
    >the exact same brand, composition and batch number, you get three times as
    >nicotine and still no niacinamide. The smokers on the list might see that
    >as a
    >good thing, neglecting as they do every day the fact that nicotine is quite
    >poisonous. What I do, on the other hand, is much safer and better and
    >in all respects (perhaps the only area of human life in which I display
    >virtues), since I will continue to get the vital daily dose of niacinamide
    >two of the three different brands of pills, and instead of three times the
    >batch-screw-up dose of nicotine, as the other person gets, I will get only
    >one. (Actually I think niacinamide, while nice, is not quite so vital as I
    >pretended above).
    >The three-different-pill regimen has demonstrably lower level of the
    >substance nicotine, but it has this lower level precisely because it has a
    >demonstrably lower level of the dangerous property error-covariance, which
    >much more poisonous than nicotine.
    >This is not an artificial example, merely one dimly remembered -- but there
    >real examples (somewhere, I've seen them) of people getting very sick,
    >eventually, or even dying (rare) because they took a single large pill from
    >source of chemicals, when they would have been quite OK had they taken
    >small pills from several unrelated sources.
    >Say, that's not a bad example, I should add it to my ever-changing,
    >rarely-very-good, homepage, where it might help to delude people into
    >I really do have the answer to all of society's problems, through careful
    >error-covariance minimisation and combinatorial matching based on the
    >There now, that's a sentence just as long and incomprehensible as most of
    >ones on that never-adequate homepage, which I shall immediately go and
    >and improve the overall health of it, with a dose of the most essential
    >vitamins and minerals, if you ladies and gentlemen will excuse me for a
    > dpw http://www.SocialTechnology.Org/dpwilson.html
    Niacin used to make me flush out (meaning my skin would turn red and I'd
    start feeling a little warm and fuzzy for a couple seconds). Not exactly my
    ball of wax.

    Your "error co-variance" idea, as I said in reply to Physics Phil Jonkers
    sounds a little like a concept I've encountered a long time ago called
    "groupthink" I found a definition and elaboration in a couple social
    psychology texts I've got lying around.

    Robert Baron and Donn Byrne's 9th edition of _Social Psychology_ (2000.
    Allyn and Bacon. Boston, p. 516) define "groupthink" as follows:

       (bq) "The tendency of the members of highly cohesive groups to assume
    that their decision can't be wrong, that all members must support the
    group's decision strongly, and that information contrary to it should be
    ignored" (eq)

    David Myers also discusses "groupthink" in his 6th edition of _Social
    Psychology_ (1999. McGraw Hill College. Boston, p. 316-325). He refers to it
    as a sort of collective dissonance reduction (shades of Festinger?) and has
    a little spotlight sidebar where the originator of the concept (Irving
    Janis) says pondering the Bay of Pigs fiasco led him towards the idea and he
    even says he was wondering what sort of "psychological contagion" might have
    been at play :-/

    I guess in a nutshell people of like mind harden in a consensus view and
    ignore dissenting views even at their peril.

    MSN Photos is the easiest way to share and print your photos:

    This was distributed via the memetics list associated with the
    Journal of Memetics - Evolutionary Models of Information Transmission
    For information about the journal and the list (e.g. unsubscribing)

    This archive was generated by hypermail 2b29 : Sat May 11 2002 - 00:15:22 BST