Comments on the current avidity for modeling:
The computer age has brought with it a resurgence of the ancient art of soothsaying, that is, the prediction of the future. That magic phrase "a computer model" carries with it a certain aura of irrefutable fact, similar, no doubt to the pronouncements of an oracle in ancient Greece.
And yet, even the earliest forays into this new art brought forth the acronym "gigo", that is to say "garbage in, garbage out". Not withstanding this early recognition of problems, modeling still carries within itself some very serious problems. Circularity, that is, that a well crafted model can produce results that merely, and accurately, reflect the assumptions on which it was based, is something that can be so well concealed in the complexity of the model that even the authors of the model--or perhaps, especially the authors of the model-- may fail to recognize it.
A second common problem is that of concealed or false correlation. In other words, if "A" is a function of "C", and "B" is a function of "C" then in the absence of knowledge about "C"it may be reasonably considered, but falsely so, that "A" is a function of "B". This usually arises from failure to consider--or ignorance of--some very pertinent factors. To make up a wild-eyed hypothetical example: "People who wear leather shoes are prone to liver cancer, relative to those who wear wooden clogs. Therefore there is something about animal hide that contributes to the development of liver cancer." The concealed correlation lies in the fact that wearers of leather shoes polish them frequently (or at least used to) and the polishing is done while wearing the shoes, and that the polish is based on wax softened with nitrobenzene, a known liver toxicant.
A third hazard of modeling is that of advocacy. In many cases, a computer model is devised to prove a point, or to support a position. It is all too easy in this circumstance to down-play factors that do not support the position and to emphasize factors that do. All of this can be totally unconscious and completely without deliberate dishonesty, but the result is a skewed model.
A fourth pitfall, somewhat related to the previous one, is the deliberate development of a skewed model to predict some sort of dire circumstance. This can be a very useful ploy to support a grant application to fund a major research effort.
And finally, for this essay, there is the element of hindsight in both constructing the model and in testing its validity. This is closely related to the first problem of circularity. I dare say that all of the above problems were common also to the pronouncements of the oracles of ancient Greece.
All this is to say that the future is inherently unpredictable and the more distant the future , the more unpredictable it is. Many natural phenomena are chaotic, including weather and climate. In these circumstances, modeling is a vain hope for prediction. The experienced human brain is a much better short term predictor than any model for these and we all know the problems of planning for tomorrow based on weather forcasts.
"Shipping and Handling"
< I've been thinking a lot lately about the economic
potential of "shipping and handling". Inflation seems
to have hit this item pretty hard, with a 20 dollar
mail order item now carrying a S&H burden of up to
$6.95. Shipping I can understand, as this presumably
represents the amount paid to an outside delivery agent
to accomplish the transfer of an item from the supplier
to the consumer. And I suspect that the actual shipping
cost for, say, a couple of CD disks of hard rock might
be about six bits (unless, of course they happen to be
Heavy Metal.)
But handling now, that is a different matter. Exactly
what constitutes handling? And when did this get
separated out from just plain old fashioned business
overhead? I suspect this is mostly a figment of the
imagination of a merchandising genius.
Someone must have said "hey, the price we'd like to get
for this item, 27 dollars is going to sound a bit high.
But if we can knock off a few bucks and then charge
them back as a legitimate-sounding mail order business
expense, then the advertised price will sound a lot
better, and we'll come out even."
And so, a new industry was born "SHIPPING and HANDLING".
I suspect that this item comprises a major chunk of
the profit earned on a mail-order sale.
In fact, I'm thinking of establishing a new mail
order business based on this. Our product will be just
pure Shipping and Handling. Since our overhead will be
very low (just my salary and taxes thereon), and the
cost of the product will also be very low, the potential
for profit should be most attractive. If this comes to
fruition, I plan to take on a certain number of limited
partners, say 5000, at a subscription fee of only
twenty five dollars each. Of course, with your
subscription would come a unit of shipping and handling
at $5.95 and if you prefer we ll have it inscribed with your name as a charter partner and will install it on our front office wall. This IPO (initial private offering) is bound t go fast so if you would like to get aboard, do it now!
To be serious though, this addendum to the price of
mail-order items is getting to be a farce and it is
obviously is a racket intended to make the advertised
price look lower than it is. This gimmick is related
to the advertised prices for gasoline which always end
in 9/10 ths of a cent. Who are they kidding? That's
two cents on a twenty gallon fill up!
INFLATION:
Inflation is the fuel that government runs on. Since
government is non-productive, that is, nothing concrete
results from government's activities, the only way
government can be paid for is by borrowing high and
repaying low. This has been going on since the invention
of money.
It would seem that the only absolute value in the world
is the value of a day's honest labor, freely given at
the pay rate offered, and willingly paid for at that
rate. If all the world's nations based their
monetary unit on this value, then all of the problems
of rate of exchange, controlled or free, would
disappear.
No longer would manufacturers chase the
phantom of cheap labor and no longer would artificial
means of monetary manipulation such as cartels, labor
unions, manufacturers associations and the like be
necessary.
Of course, the monetary unit would be based on, let's
say, an hour of common, unskilled labor. This has always
had a de facto value which has changed as necessary to
keep up with inflation. A search of the economic
literature would reveal that an hour's wage for common
labor has risen consistently, as the value of the
worlds currencies has dropped. There are transient
disparities, to be sure, but these always even out in
the long run. Pay rates for skilled labor would be
higher, as is true now, based on the lowest rate, and
on the availability of the higher skills.
The gold standard, as an entirely artificial basis of
exchange, appeared to stabilize international dealings
--for a while. But eventually that too always had to
be adjusted, always upwards, until now it is no standard
at all.
We are presently on an energy standard,
controlled and manipulated by energy rich nations. With
time, this too will fail either as energy supplies are
depleted, or, conceivably, as energy becomes unlimited.
Inflation feeds on itself, especially in these days of
cost-of-living adjustments and other indexing measures.
After all, when everything is adjusted upwards to
compensate for inflation, this is inflation itself.
The cost of labor in the USA is a case in point.
Union strength has resulted in spectacular dollar gains in
wage rates for labor. Of course, this has resulted in
increases in production costs, distribution costs, and
merchandising costs and that in turn has raised the
cost of living. In terms of "real dollars", how much
has the union movement accomplished? A factory
worker (say at Ford Motor Company) earned a living wage
in, say, 1920. A factory worker earns a living wage at
Ford Motor Company today. These are self evident facts,
as at neither time would a worker have worked at less
than a living wage. But I doubt that, relative to the
general standard of living, today's worker is greatly
better off than his 1920's cohort. To be sure,
everybody's standard of living today is higher (is it?)
but I'm not so sure that the working man is any happier
today than in decades and centuries past.
WHUT IT IS IS CRICKUT
(With apologies to Andy Griffith *)
Crickut is a funny kind uv ball game played by some
high-falutin gents in England and a lot of other places
around th' world that England uster have ther mits on.
Hit's played on a close-cut meader which has a bunch of
white lines whitewarshed onto it. They's also a couple
of passels of stakes driv in th' ground a fair piece
apart (they call em stumps but they ain't got no roots
on em) and acrost th' tops of these stakes lays a couple
of loose sticks that they calls bails, but them sticks
don't look like any bucket handles I ever seed. They's
two teams, about a lean dozen good old boys on each,
an they's all got up in white rigs like they wuz goin
to a summer social or sumthin. Ice cream suits I'd call
em.
Anyhow, one team spreads themselfs all over th' meader
except for two fellers; one is called th' bowler and he's
goin to be th' thrower, and th' other, well I don't
rightly know exactly what he does, 'cept now and then
throws th' ball back to th' thrower. They say he's th'
wicketkeeper but I aint sure who he keeps it fer.
Th' thrower takes his place alongst one of them passels
of stakes, and t'other feller on his team stands a ways
behind th' other passel, mostly lookin bored. Th' other
team is astandin on th' side of th' meader 'cept fer two
who takes up places, one a little afront uv each passel
of stakes. They's the hitters, but they's called
batsmen. One of these, th' one facin th' thrower, is
called th' striker and he's th' only one th' thrower
throws th' ball at. Th' striker's job is to keep th'
thrower from hittin the striker's passel of stakes and
knockin them ther sticks offen th' top. If th' thrower
gits them sticks knocked off, that hitter is out, and
he leaves th' meader and is replaced by th' next old
boy of his team.
Well, now th' bowler backs up a piece, runs
hell-fer-leather toward th' hitter, makes like a windmill
with his arms and lets fly with th' itty bitty ball--tain't
nuthin like a bowlin ball--at th' striker's passel of
stakes tryin to knock off them ther sticks, only thing
is, he's gotta be kirful not to cross any uv them ther
whitewarsh lines.
Oh yes, I fergot to say that th' hitters has these mean
old paddles--They calls em "bats" but I reckon I know
what a proper bat oughter look like and I mean to say
these looks like the paddle ole Missus Thompson, th'
principal down at Piney Branch school, uster use on us
country boys;--and th' striker is guardin his passel of
stakes tryin to keep that flung ball from hittin em.
Here's one of th' funny bits, instid uv holdin that ther
paddle uv hisen like as to be all set to hit what ever
comes at im, he's got it sort of dug down in th' dirt
in front of them stakes.
Well, that ther thrower ain't got much control, cause
th' ball mostly alwys hits th' dirt a ways afront uv
th' hitter, but then maybe th' thrower ain't so dumb
after all cause that ther ball kin take a powerful wild
bounce and th' hitter kin have a sore old time tryin to
hit it with his paddle to keep it away frum them stakes.
Lets figger now th' hitter connects, but only jist barely
and th' ball jist sort of dribbles out afront uv im.
Nothin happens 'cept th' ball is throwed back to th'
thrower and he tries agin. If this time around th'
hitter wallops a good un way out yonder in th'
meader--oh yes, I fergot to tell yuh, th' other hitter
is not a striker so I reckon he must be a scab. Anyhow,
th' hitter who's jist walloped th' good un and th' scab
run tored each other like all git out, pass one and
t'other, and end up at each others stakes. If th' ball
is still a fur piece off they kin trade places agin,
runnin like their granpappy's ghost wuz after em and
keep up this tradin places until they figger th' ball's
been throwed in too close to risk another try. Each
time they trade places, that's a "run". Then they each
stop at th' nearest passel of stakes and th' whole
shebang begins agin, and agin, and agin, until th'
striker has bin put "out" by havin a hit ball caught or
by havin them little sticks called th' bails git knocked
off them stakes. This goes on for quite a spell until
someone yells "over"--an hits a right shame that don't
mean th' game's over--and th' players in th' meader
('cept th' hitters) run around like chickens with their
necks rung off and finally everybody settles down in th'
same layout only backwards. Every time a striker's put
out, he swaps places with th'next player uv his team at
th' side uv th' meader an this goes on until ten hitters
uv had their chanct to hit and have been put out. Then
th' two teams switch everythin and th' rukus continues.
After both teams have gone through th' whole rigamarole
someone yells "Innings" (even if they's only one of em)
, and th' game continues as it wuz begun.
One other thing, they's a whole bunch of rubberneckers
(maybe even a dozen or more) at th' side uv th' field,
and when a hitter really smacks th' ball and it goes
plum outen th' meader, th' crowd goes real wild, pattin
their hands together and mumblin "good show".
All this kin go on for days cause th' number of runs
can add up to hunderds fer each player. I don't rightly
know how the game is ended, exhaustion, I reckon.
Well, that's crickut fer yuh. *(Authors Note)--Andy Griffith, comedian,
character actor (The Andy Griffith Show,
otherwise known as Mayberry, U. S. A.) and later,
dramatic actor (Matlock), is said to have first become
nationally known through his humorous hillbilly
descriptions of the games football and baseball. I don't
think he ever did
cricket.
My view of this situation is that mankind is part
of nature, too. And everything that mankind does is a
part of this nature. Other natural beings share with
mankind the tendency to utilize extant facilities to
the utmost. And so we see that goats, introduced
(either by man or by nature) into a limited
environmental niche, such as a small island, will eat
every stitch of vegetation to the point of driving all
other fauna to near extinction and to the point of
starving themselves to a population size limited by
whatever residual, less appetizing, flora there may be.
In the smaller world, a microbial pathogen will
multiply to the point of killing its host. And even in
the smallest, living(?) world, a virus multiplies by
destroying the cell that nurtured it.
It is a fact that man as a species is spectacularly
successful. By his ability to modify his environment
to meet his every need, he has accomplished the
extinction or near extinction of many other species.
This is not all bad. The extinction of smallpox virus
is certainly a worth-while feat. And few would argue
that the loss of the dodo was a global catastrophe.
The disappearance of neanderthal homo sapiens has not
been a major loss to modern man, tho, of course the
typical neanderthal would undoubtedly have taken a
different view of this.
It is still controversial, but it is considered
possible that very early man brought about the
extinction of most of the large animals that once
roamed the Americas. Among these were the wooly mammoth,
the saber-tooth tiger, and the giant ground sloth which
among other giants were once inhabitants of the North
American continent. And so, it is not just modern man
that is wreaking havoc with the earthly flora and fauna.
The end result of man's success will be to attain
an equilibrium population size determined by his ability
to utilize the bounties of nature to the full. This
will undoubtedly be to the detriment of many other
species, but then, isn't mankind the most important
species that we know of?
Of course, there are other possible scenarios to
man's future. Monocultures are subject to all kinds of
threats. A new or resurgent disease might eliminate
mankind. Or man's competitive cupidity might result
in a true world holocaust. We might get buried in paper.
Or we may just atrophy to a computer keyboard button.
Whatever the eventual outcome, it probably will be a
direct result of man's success.
>P>
So what's new about combinatorial chemistry (CC)? Well,
strictly speaking, it must be the "chemistry" part.
Because the methodologies of CC have been used for
decades in microbiology. Let's coin a new term for a
well established technique. Let's call it
"combinatorial microbiology" (CM).
What are some of the characteristics of CC?
All of these operations are found in the
screening of microbial and other biological activities.
The best known and most extensive of these has been
the screening of soil microorganisms for antibiotic
activity. This has been in progress for about six
decades (since the observation of useful antibiotic
activity in Penicillium). Here, the initial search for
further new activities probably involved one
microbiologist and a minimal support staff. And the
birth of CM involved simply amplifying that search by
adding more microbiologists and more support (parallel
operations.)
These teams prepared soil samples from many sources,
plated these samples on nutritious (not just "nutrient")
agar plates, selected "promising" candidate microbial
colonies on the basis of one or more discernable
properties, propagated these candidates in larger
volume (shaker flasks) to produce larger quantities of
crude product, examined these crude products for
novelty, and finally cultivated still larger quantities
of the "interesting" candidates for attempted isolation
and characterization of the active principles, by
associated chemists.
As intimated above, the earliest approach to
CM was to examine the procedures of one microbiologist-team and
simply use more teams. But as time went on,
methodologies were improved to the point that one
microbiologist and a support staff, in one year, could
isolate 10,000 discrete microbial colonies from any given
number of soil samples and carry these to the
point of mass culture, where serious consideration
could be given to the most promising candidates.
The procedure was one of converging interest,
narrowing a large number of initial, easily derived,
isolates down to a much smaller number of most
interesting candidates for detailed examination. At
the 10,000 isolates per year level, promising candidate
organisms were relatively few and record keeping and
decision-making could be handled as a purely mental
process.
But it eventually became obvious that the
unfavorable statistics (only a few leads from many
isolates) was a major barrier to success, and that,
somehow, the initial isolations and primary screening
had to be expanded. And so, mechanization came to the
fore. By abandoning rational (read "expert") selection
from petri plates,and substituting purely statistical
selection (by diluting the soil suspension to the
point that a mechanical selection would have a reasonable
chance of containing only one viable unit, that is by,
Poisson distribution) and by substituting miniaturized
liquid fermentations in 96-well plates for shaker flasks,
and by instrumenting the agar diffusion assays, and
computer examination for potential novelty, the isolation and
screening rate was easily increased to 100,000 per team
per year.
The mechanics of this enhanced CM screening rate are
very similar to those which have since been developed
for CC. That is, large numbers of candidates have been
substituted for rational ("expert") synthesis, and
mechanization and instrumentation have been devised
to generate these larger numbers; and computer
assistance has enabled handling these large numbers--
numbers too large to be handled by pure cogitation.
More detailed comparing and contrasting of CC and
CM can be informative: Just as the methods of CC are
used for process improvement as well as for discovery,
so the CM methods of examining soil organisms have been
adapted to microbial preselection, to fermentation
process improvement, to mutant
selection, and microbial genetic manipulation. And, of
course, the application of massively parallel analysis, data
compilation, and decision making are exactly analogous
whether applied to microbiology or to chemistry.
Perhaps none of this is surprising. After all,
microbiology contains a large dollop of chemistry. The
moral to all this is that those doing combinatorial
chemistry should pay more attention to the old art of
combinatorial microbiology.
Buffers,
Associated with the probable involvement of p[CO2]
in optimal fermentation response is the requirement for a
high level of oxygen supply to the living organisms.
In large scale fermentations, many stratagems have
been utilized to attempt to increase the transport of
oxygen from the gas phase to the submerged organisms
that require it. Many combinations of agitators, baffles,
fermentor geometries, oxygen enrichment of the gas
phase, agitation power input, oxygen transfer agents,
oxygen sources other than O2, and even inversion of the
gas-liquid dispersion state have been investigated.
One of the simplest methods of oxygen transfer
control has been perhaps the most neglected, namely, to
increase (or decrease) the pressure under which the
fermentor is operated. Most industrial scale fermentors
are designed for sterilization with pressurized steam.
And yet, the usual operating condition has been to vent
the exhaust gas from the fermentor directly to the
atmosphere so that the gas phase in the top of the
fermentor itself is essentially at one atmosphere
pressure. In addition, the air supply for fermentors
is usually generated at well over one atmosphere,
the pressure being lowered to attain the desired input
flow rate by simple throttling at a control valve.
Combining these several facts, it is a straight forward
procedure to operate the fermentor at the sterilization
pressure by the use of a back pressure control valve on
the exhaust. In many cases, it is possible to double
the partial pressure of oxygen in the gas phase (total
pressure 14.7 psi above atmospheric pressure) and
thereby double the gradient for the transfer of oxygen
to the liquid phase, a goal difficult to achieve by the
other stratagems mentioned above, and at virtually no
increase in operating cost.
It is necessary, however, when pilot planting a
superpressure fermentor experiment, to keep in mind the
interactions of CO2, as discussed above. If the mass
flow of air through the fermentor is kept at the same
value as in the atmospheric fermentor, then not only
will the oxygen partial pressure be doubled, but that
of the biologically produced CO2 will at least be
doubled also. The effect of this change in pCO2 will
depend at any time on whether the fermentor is operating
in a CO2 deficient, CO2 adequate, or CO2 excess domain
(and the effect of these levels on pH and on crucial
reaction rates). The CO2 domain at any time can be
determined by examining the effect of a change in air
mass flow and/or by enriching the gas stream with
exogenous CO2. In summary, the effects of increased pO2
and of increased or decreased pCO2 can be studied by
operating a fermentor at significantly higher gas
pressure and at greater and less than reference
(standard temperature and pressure) air mass flow rates.
J. R. D. Mcormick Biographical Data and Experience:
Born St. Albans, WVa; Married, wife Catherine; son, Joshua.
Education: B. Sc.(Chemistry), Rensselaer Polytechnic
Institute, Ph. D. (Organic Chemistry),UCLA
Professional History: Pharmaceutical
Chemist, Winthrop Chemical Company, Rensselaer,NY.
Pharmaceutical Chemist, Lederle Laboratories of
American Cyanamid, Inc. Pearl River , NY
Internal Consultant (Research Fellow) Lederle
Laboratories.
Independent Consultant: General Chemistry and Biochemistry, Fermentation Technology,
Antibiotics, Tetracycline Antibiotics.
Member American Chemical Society, New York Academy of
Sciences; Fellow American Association for the Advancement
of Science, American Institute of Chemists.
Publications: Over forty articles and
presentations on various aspects of tetracyclines
chemistry and biochemistry. Forty U.S. Patents on
chemical processes and products in the fields of
vitamins, antibiotics, and unique chemical substances.
Special current interests: Cellulose chemistry,
alternative energy systems, biomass production and uses.
Availability: E-mail me at the address above
for information or comment. I am also available through
Teltech/Intota
Advisory: Bookmark this page for future technical and social essays.
Guest book
accesses: Last updated August 10, 2002 by J. R. D.
McCormick
