Son of Devil's Advocate, January 2004

Click here to make UNIX performance tuning easier!

Previous SODA Index of SODA Columns Next SODA

Son of Devil's Advocate

Stan Kelly-Bootle

Picture of Stan Kelly-Bootle

Resolutions Considered Harmless

Looking back over past sins, I detect a certain sameness (or in PostModSpeak, make that certain samenesses -- extra marks for spurious, singular plurals? [ref 1]) in my columnar treatment of the EOY (End of Year) thingy. And there are now, as archived proof, about twenty DA or SODA files with that pass-the-bottle, drop-the-ball for auld-lang-syne extension .jan.

When Lotus Magazine paid me 25 cents a word, I would avoid epistemology, miners' lung diseases, and long-winded Welsh Train Stations, throwing in the odd $1.25's worth of "be that as it may" at the drop of a paragraph. I'm now paid, if at all, per laugh, a complex formula involving volume, duration, and sincerity, divided by a hidden scatological factor. Hits (on the sarcheck website) are also understandably important, including, of course, the very visitation you are now making dear reader, almost by definition. Strictly speaking, as us mathematicians nearly always do when using measure-theoretic terms such as AE (Almost Everywhere), you may have hit me in error &/or given up before now, finding my sets too dense by far. Higher website hit-rates, I gather, can be guaranteed by a suitable cascade of meta-tags, or with appropriate bribes, or the purchase of sufficient stock in Googol Inc., when its awaited IPO is launched. Kevin G. Barkes's site enjoys high hit rates more simply and legally by exploiting his precious initials: KGB. Many search-mad Sovietologists reach and are assuaged and tempted to linger and link to Kevin's incomparable collection of over 11,000 quotes in [ref 2]

The Lenny Bruce Paradox is that we cannot publish the list of forbidden strings, and you are forced to deduce these by their unexpected absence. Nor is the weasel use of cartoon characters encouraged unless they happen to make sense as regular expressions. Thus if find myself unwilling to write out in full an unspeakable soccer team, I'm allowed to curse F*ng M?nche?ter Unit*.

Be all that as it may or may not be ($2.50 in hair-down Lotus days), I resolve to avoid the obvious year-end topics such as resolutions, this sentence notwithstanding.

Yet, it must be noted that at my Valley Orchards active- retirement abode (also known as Chateau Despair), the inmates have resolved sous peine de mort (not the deterrent that it used to be) never to utter phrases including the words "doctor," "medication," "pill-count," "operation," and "Medicare." Then there's "Where are my hearing-aid batteries" which precludes detection of the verboten topics.

These calendric transitions are, of course, replete with numbing arbitrarinesses, from nation to nation, timezone to timezone, and within these, dare I say arbitrary ethnic and religious conventions? The oldest latch-lifter [ref 3] still traps the unwary: "When did Christmas Day and New Years Day fall in the same year?" More testing is "When did Ash Wednesday fall on a Saturday?" Usual prizes! No trick, but profoundly puzzling are the years when Hannukah starts on Christmas Day (that is December 25, to avoid confusion with rival Christmas dates elsewhere). I have in my pocket Robert Fowler's small prompt card showing compact algorithms for determining the Day of Week given the date. They cover Julian and Gregorian calendars and, unlike many computer date programs, allow for the infamous discontinuities. Once mastered you can dazzle your friends, no end. Beyond the scope of Fowler's card are party tricks involving the Jewish, Muslim, and Mayan calendars! Donald Knuth famously asserts that fixing that damned "moveable" Easter was a powerful impetus in developing the concept of the algorasm, bringing Arabic mathematics to the West.

The Jewish solar-lunar calendar copes (some say miraculously) with establishing a date for the Feast of the Passover, held one day before the Feast of the Unleavened Bread. In Exodus 12 and 23, Jahweh strictly commands the how and why, and more vexing, insists that Passover be commemorated in Springtime. Without a precise knowledge of the moon's months and the sun's seasons, this springtime restriction would be impossible. The lunar-only calendar of Islam, for example, lets Ramadan fall where it will each year.


The response to my last ROSODA (November, 2003 on a page near you) has been most encouraging. Keep those emails rolling in. How's 2004 so far? As we say when half-way through the EPL Soccer Season: "Early days, mate. Early days."

ref 1: See NB, Times Literary Supplement, October 17, 2003 where the unknown (to me) columnist JC lists a conference under the heading "Interrogating the Gerund." The November conference at the University of Maryland covered "Attending to Early Modern Women Structures and Subjectivities," the latter being one of the vogue plurals under discussion. Other catching topics include "Gendering Encounters with the Ethnic Other," where "Other" by common consent must remain single.

ref 2: Kevin G. Barkes's printed selection risibly titled "Eff the Ineffable, Scrute the Inscrutable" is available via his website or by email: I'm reluctant (not really) to boast of four quotation by my goodself, two more than Walter Cronkite, three more than Wolfgang Pauli (stung by his own Exclusion Principle?), and, quite unfairly, three more than Doc Cypher. Doc's solo quote, though, has the quality: "The day Microsoft makes something that doesn't suck is probably the day that they start making vacuum clearners." But I digress.

ref 3: A custom whose origins are lost in the mists of iniquity is still honored in many Liverpool pubs. The latch-lifter is the minimun amount needed to buy your first drink, after which the crafty Scouser can booze on ("filling yer legs, like" in the local funicular) by posing diverse challenges for money or drinks. "I'll bet you a pint..." or "I'll bet you a quid..." are the standard preambles. My father-the-plumber had the largest fund of teasers, and often staggered home with a welcome surplus of cash. Plumbing in those days was not the well-paid trade we see now in the USA.

This column 16 years ago:

Devil's Advocate - UNIX Review - January 1988 - Stan Kelly-Bootle


I stopped smoking, dindon froid, two months ago when I realized it was hurting my nearest and dearest (the wife wasn't too keen on it either). I did not wait until some arbitary quirk of the calendar, Julian or otherwise, popped a cork in Times Square. I have discovered through painful experiment that making life- changing resolutions to herald a so-called New Year just does not work for me. If you need the added incentive of (year++), by all means resolve away. However, as a new subscriber to the Skeptical Inquirer and its policy, I now avoid such pagan rituals and shun all magical formulae except those incanted by the Amazing Randi, the Prince of Pure Reason. This wonderful enemy of psychic charlatans and crooked spoon-benders named James Randi is worthy of your attention if you have not yet encountered him. He is a leading Fellow of CSICOP, the Committee for the Scientific Investigation of Claims of the Paranormal, of which the Skeptical Inquirer is the official journal (published quarterly at $20 per year; details from Box 229, Buffalo, NY 14215-0229). The majority of CSICOP Fellows are prominent academics such as philosophers Paul Kurtz, Antony Flew, and W.V. Quine; psychologist B.F. Skinner; Nobel Prize winners F.H.C. Crick (of DNA fame) and Murray Gell-Mann (physics); biologist Stephen Jay Gould; and astronomer Carl Sagan, among other household names. Randi, however, is by trade a professional magician whose keen wit and eye have uncovered many a bogus flim-flam (the title of one of his many books, in fact, is Flim-Flam![fs1]), ranging from card-sharping ESP and clairvoyance to claims of water divining, faith healing, and bare-hand, psychic surgery. The McArthur "genius" prize he recently won will certainly give him the opportunity to investigate more outlandish claims, and publish the results. Randi always carries with him a certified check for $10,000, which he will give to anyone who can demonstrate paranormal powers. The published rules are quite simple, but the prize remains unawarded in spite of several botched attempts.

His most widely publicized success came in exposing the methods of Uri Geller, who also started as a stage illusionist. Geller and other tricksters now refuse to "perform" if Randi is present. Their specious argument, of course, is that doubting spectators somehow disturb subtle psychokinetic force fields.

Randi and other well-known practicing magicians who are members of CSICOP, including Henry Gordon and Robert Steiner, are particularly incensed that Geller invokes a mish-mash of psychic babble to cover what are in fact simple stock-in-trade tricks. The interesting point is that Geller has succeeded in fooling not only millions of gullible TV viewers, but also many highly trained scientists sitting just a few feet away. Scientific training, it seems, does not always provide its beneficiaries with the magician's critical instinct for spotting sleight of hand. In fact, scientists have a sort of trusting attitude toward data, since dishonesty normally plays no role in their hypotheses.

I used to fondly imagine that computer programmers, who daily hone their deductive skills to spot a dangling ELSE six blocks away, would somehow constitute a growing pool of skeptical rationalists, a bulwark against the onslaught of pseudoscience. In the absence of any statistically sound surveys on this prospect, I can only report my own personal small-scale study.

Living in a notoriously New Age California county clearly distorts my sample, yet I find that an alarming number of my programmer friends and acquaintances are deeply involved in Astrology, ESP, channeling, and allied astrally resonant nonsense--and not merely on a harmless, playful level, but with the conviction that Shirley MacLaine, Uri Geller, and the Stanford Research Institute have brought all this flim-flammery into the bosom of respectable science.

CSIOCP members are expected to keep an open mind concerning "fringe" science--after all, segmented memory models might conceivably be better than linear memory models, there may be a levitation mechanism yet to be revealed, and--damn it--there could be a monster in Loch Ness. Our philosophy, though, is that theories demand a level of proof somehow commensurate with their wildness. I would probably accept Mark Riordan's anecdotal evidence that he has a brother (why should he lie about that, Lansing resident though he be?), but if I'm told that Jesus took a sabbatical in Tibet, I feel the need for more solid verification.

Carl Sagan offers the following paradox: an over-open mind lets in incredible facts without due verification, while an over-closed mind will reject the intuitive leaps that keep science on the move.

Wheel In the Expert System, or If You're So Damned Clever, What Are You Doing on This 39-Cent Floppy?

Not entirely unrelated (or not unentirely related) to the aforementioned topic is my recent reading on the subject of Knowledge Engineering and Expert Systems--to wit, a paper called "Interactive Elicitation of Knowledge from Experts" in the Future Computing Systems Journal (Vol. 1, No. 2; 1986). The subject must be one of intense study by a small number of experts since the paper cites four references by Gaines, B.R.; 7 by Shaw, M.L.G.; and 15 by both Gaines, B.R., and Shaw, M.L.G. The authors of the paper are. . .drumroll. . .Shaw, M.L.G, and Gaines, B.R. Gaines, B.R., is also one of the Journal's editors. Without looking it up, I suspect that the etymology of "expert" is: Latin ex, "out", and perdere, "to lose", whence "to lose out." The new high-rolling Knowledge bandwagon certainly seems to be heading for boggy ground. Experts, apparently, are appallingly dumb outside their own tiny corrals. The so-called Knowledge Engineers are having trouble extracting both what the Experts know and how the Experts know what they know. One snag is that Experts consider some Knowledge so painfully obvious that it defies elicitation, whether interactive or otherwise. Or maybe the Experts quite justifiably say to themselves, "I'll be buggered if I tell this nerd too much." A further problem is that experts disagree among themselves. One might almost include this in the definition of "expert"; otherwise you would need only one per specialty, so there. A good example of this emerged last week in the San Francisco Chronicle when the editors solicited the views of four leading economists on the causes of the October 1987 market crash and how the budget deficit should be handled. There was violent disagreement--not just over whether this or that parameter should be tweaked up or down to solve the crisis, but major schisms on basic issues. So what's new, you may ask. Well, the four experts were all Nobel laureates in Economics. Perhaps the Knowledge Engineers are not yet ready to tackle such disputatious situations. Yet some say that the October 19 market crash was itself the result of 50,000 PCs all running Intellivest's Expertfolio program, and all bleeping "Sell! Sell! Sell!" at the same instant. Which leads me to this numbing quote from William R. Swartout's article titled "Knowledge Needed for Expert System Evaluation" in the same issue of the Future Computing Systems Journal:

"Answering questions about what the system cannot do appears also to be difficult because it requires the system to have knowledge of its own limitations. Research into providing systems with such knowledge is still at a primitive stage."

Stan's bio:

Liverpool-born Stan Kelly-Bootle has been exposed to computing, on and off and vice-versa, since 1953 when, after graduating in Pure Mathematics at Cambridge University, he switched to impure post-grad work on the wondrous EDSAC I. After some trenching with IBM and Univac in the 1960s and 70s, Stan opted for self-employment as a consultant, writer, folk-song revivalist, after-dinner entertainer, and cunning linguist.

His monthly DA ("Devil's Advocate") column ran and ran in UNIX Review (aka Performance Computing) from 1984 until January 2000 (a date that will live in infamy) but lives on as SODA ("Son of DA") via the homepage devoted to UNIX performance.

The latest of his umpteen books are "The Computer Contradictionary" (MIT Press) and "UNIX Complete" (Sybex). More on his biblio- and disco-graphy can be found on soon due for its millennial update.

Stan welcomes reader reaction:

To the SarCheck home page

The URL of this page may change in the future. Please bookmark the home page instead of this one.

Portions © copyright Stan Kelly-Bootle 2004.
Portions © copyright Aptitune Corporation 1997-2004, all rights reserved.
This column is sponsored by Aptitune Corporation, makers of SarCheck. The opinions of Stan Kelly-Bootle are his own and do not necessarily reflect the opinions of Aptitune Corporation. Other products and companies referred to herein are trademarks or registered trademarks of their respective companies or mark holders.