Good description of our complex system.  Ahh for the good old days of vacuum
tubes and things that can be tightened, loosened, and touched and
understood.

Any "long jumps" that citizens can initiate?  Long jumps that citizens can
begin rather than respond to.  

arthur


-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Mike Spencer
Sent: Sunday, February 28, 2010 1:36 AM
To: [email protected]
Subject: [Futurework] Re: consistency



Keith wrote:

> Our present politicians are still mainly composed of lawyers --
> products of a medieval, agricultural, religious age. Asking them to
> point the way forward is like asking a car mechanic to repair a
> computer.

As I am (or rather, I once was) a hot-shot car mechanic and (more
recently) I have "repaired" numerous computers, that got my attention.
Moreover, I've been reading Steven Pinker's ramblings on metaphor and
his critique of Lakoff & Johnson so I'm (more than usually) alert to
what we might mean by locutions such as, "A is like B".

I think this is a yarn or a rant or something of the sort so if you're
keen on "terse" and "cogent", you may want to hit DELETE just here.

I once worked at B.J.'s Foreign Auto in a university town where
eccentric academics and hip students drove a bizarre assortment of
auto marques you've never heard of or at least specimens of which
you've never seen.  Finding parts or specs for a 10 year old
Borgward, an even older Panhard or a model of DKW that was never sold
this side of the pond was, predictably, a constant problem.

Bud Jarvis -- the boss -- was just barely tall enough to have passed
the US Navy physical and had spent WWII on carriers in the Pacific,
repairing airplanes.  After the war, he spent some time in England as
a mechanic, maintaining the pitiful fleet of domestic vehicles with
which the Brits were left after prolonged wartime austerity.

Now, not infrequently, we'd call Bud over to look at a broken, frayed,
chafed, cracked or otherwise nonfunctional bit of some weird car and
complain that it was, well, a problem.  The company that made the car
had been out of biz for a decade and hadn't imported any parts for
twice that long.  And bud would say, "A man made it.  A man can fix
it." So I've made parts for a Lancia out of wood, hand-forged parts
(when I was still a novice blacksmith), adapted, hand-filed, glued and
fabricated parts.

But no more.  Cars, you see, are now made by robots, not by men.
Bud's dictum is evicerated and you *can't* always fix them.  And
that's the reason for the scare quotes, above, when I said I'd
"repaired" computers.  I lied.  You can't "repair" computers.  You
can't even properly diagnose them without an oscilloscope, a circuit
analyzer and other electronic devices of which I don't even know the
names.  You probably can't determine the true source of a fault
without a metallurgy lab and an electron microscope.  The faulty
capacitors of recent years that can be identified visually are a rare
exception and even then, it's equally rare that you can "fix" them.

So: You can't repair computers because the interiors of half dozen
major assemblies of which they're composed simply aren't accessible to
the brightest -- or even most experienced -- humans.  Even for the few
thousand highly educated, fully equipped and experienced engineers in
the world, an accurate determination of etiology -- what's causing a
malfunction -- could take weeks.  That's vaguely similar to a mechanic
taking a fortnight to diagnose a maladjusted carburetor float valve.

To follow this metaphor, finally, to it's end, even if the MIT solid
state physicist and his colleague, the electronic design engineer can,
after a month's work, say something like, "Yes, a heisenbug in the
clock chip causes the timing pulse stochastically to desynch in a
certain way from from the CPU and the latter is designed to treat this
as a fatal error", there's still no way to actually *fix* it other
than to do what your 14 year old nephew could have prescribed: replace
the whole motherboard.

Okay, back on topic: I submit that the whole global economy has passed
through a sort of phase boundary of the same sort that separates
repairing a 1955 Borgward from repairing a Dell laptop.  The mechanisms
of the socio-economic structure that dominate and control your life
just aren't accessible, even if you're quite bright and well educated.

If you are, in fact, quite bright and well educated (or a voracious
learner) and you make a serious avocation of studying, dissecting,
analyzing, comparing, assimilating, you may arrive at a meaningful
truth or a verifiably valid principle.  But the magnitude of the rest
of the system -- a "rest" that remains a black box -- is so great that
there's nothing you can do with your hardly won truth or principle
that is predictably better than anything else you might do. [See "long
jump", infra.]

If you make a *real* career of this -- become a policy analyst,
economist, politician, corporate, government or political strategist,
investment banker or the like -- you become, after a decade or three,
so deeply immured in so narrow a domain that you forget that the
"rest" of the system even exists.

Reiterating Keith's remark:

> Our present politicians are still mainly composed of lawyers --
> products of a medieval, agricultural, religious age. Asking them to
> point the way forward is like asking a car mechanic to repair a
> computer.

The number of people, especially influential or powerful people, who
operate with an essentially pre-enlightenment mind set is shocking.
The consequences are equally shocking. But that may not be the best
characterization of the problem.  There are numerous other
shortcomings that are (I surmise, without any hard evidence)
over-represented among "present politicians": clinically diagnosable
psychopathy, fulminating greed, pathological yearning for approval
etc. You know the list.  It may not be those, but rather a feature of
the system itself, of a system that reaches a certain (if ill defined)
level of complexity, that engenders the apparently refractory problem.


Stuart Kauffman:

  ...the onset of a novel complexity catastrophe which limits
  selection.  It is the consequence of attempting to optimize in
  systems with increasingly many conflicting constraints  among the
  components: Accessible optima become ever poorer, and fitness peaks
  dwindle in height. 

  [....]

  I believe this to be a genuinely fundamental restraint facing
  adaptive evolution.  As systems with many parts increase both the
  number of those parts and the richness of interactions among the
  parts, it is typical that the number of conflicting design
  constraints among the parts increases rapidly.  Those conflicting
  constraints imply that optimization can attain only ever poorer
  compromises.   [1]

Standard bearers for American exceptionalism, and to a lesser extent,
those for all the functioning democracies, proclaim that with "free
markets and free elections" [2] we have achieved the indisputable
global optimum, a peak from which others, still lost on the rugged
landscape of evolving systems, try to pull us down when they should be
climbing up to join us.

But it's not like that.  The rugged landscape in Kauffman's model of a
complex system *changes*.  What may well have been something like a
global optimum in the age of sail and 2 billion population, may, in
the age of the Airbus, the internet and 7 billion population, be a
small and declining local optimum that, by its situation and nature,
obstructs access to other potential peaks in a landscape of dwindling
optima.

What's the conclusion of all this abstraction?  One of the strategies
to deal with being stuck on a local optimum is the long jump, the
"punt" in jock terms.  Jumping to a random place on the systemic
landscape offers a chance of an improved situation or at least, a
chance to start climbing toward a better and hitherto unreachable
optimum.  Of course, almost no politician or nation would just do
random things but, in the context of a complex system, a carefully
planned and executed strategy for dramatic change will be executed in
that context with the result that it might as well be random for the
planners cannot reliably predict outcomes.  Long jump is not a
demonstrably or even theoretically promising move but it may be the
only one to preserve even vestiges of what we think of as Western
civilization from spiraling into chaos or freezing into deadlock.

So we're back to my computer metaphor again: Even the insiders and the
experts can't access enough of the global black box to diagnose and
prescribe with any certainty.  Yank the motherboard and stick in a new
one?  "Revolution" and the 60s revisited?  Would the Royal Society and
the Enlightenment ever have happened in Britain without Cromwell?
Would we have the internet without the treasonous Declaration of
Independence?  Some long jumps we now regard as unthinkable: Revoking
the US Constitution; a land war in China; nuclear destruction of
Israel; total collapse of the US or global financial system; global
pandemic. Do we just have to mark time til one of them happens?

Is there a better way to cope with a system that is so complex that it
has, so to speak, a mind of its own? One that is intrinsically
inaccessible to us?


Well.  I had a second point but this is already too long.  More later.

- Mike


[1] The Origins of Order, pp. 52-3.  Recommended reading, especially if
    the metaphor of "landscape", above, is obscure and you're
    comfortable with pre-calculus-level math.

[2] In that order, in the words of Bush 41.

-- 
Michael Spencer                  Nova Scotia, Canada       .~. 
                                                           /V\ 
[email protected]                                     /( )\
http://home.tallships.ca/mspencer/                        ^^-^^


_______________________________________________
Futurework mailing list
[email protected]
https://lists.uwaterloo.ca/mailman/listinfo/futurework

_______________________________________________
Futurework mailing list
[email protected]
https://lists.uwaterloo.ca/mailman/listinfo/futurework

Reply via email to