on 11/28/04 10:26 AM, Michael Pettersen at [log in to unmask] wrote:


> Your list of engineering applications to medicine only includes efforts to
> recreate the human body with replacement parts; it omits the tools of
> analysis, such as X-rays, MRI, the microscope, EKG, and so on, which are much
> more widely used, it seems to me.

Ah, but these are all diagnostic or observational tools, which, while
invaluable to us as physicians, do not affect the function of the body. The
analogy I was constructing has to do with the irreducible complexity and
unpredictability that is the nature of  biologic systems. The engineering
approach to such systems, which by their diverse genetic nature do not
behave in a reproducible manner, tends to fall foul of the law of unintended
consequences.

To get back to harpsichords, recall yet again the experiments of the last
century, where engineers addressed the issue of tuning stability by
increasing the mass of the instrument's frame and case. This measure not
only failed to obtain the desired benefit, but also destroyed the volume and
tonal quality of the harpsichord.

Remember also the attempt to maintain regulation by attaching a long metal
spring to the side of the jack, designed to keep the plectrum in precise
position in an oversize slot. That didn't exactly work either.

Those of you too young to remember such creations would do well to find
somehow a copy of Wally Zuckermann's book, _The Modern Harpsichord_. It is
important that you know from what you have been delivered. It was a
protracted and nasty battle.

(I don't intend to demean those who enjoy these instruments, but they have
no relationship to historical harpsichords. They don't look like 'em and
they don't sound like 'em. Someone should think up another name for the
beasts.)

There is no way, even with every measuring modality known to modern physics,
that a harpsichord maker can know in advance what effect a given piece of
wood will have on an instrument. Only by building the thing, by thinning the
soundboard (a trial-and-error process), by voicing the plectra (another
trial-and-error process), and by playing the instrument in over time will he
know what tonal results have been achieved. No two are alike, no matter how
carefully one controls the process. Too many variables.


--
James R. (Jay) McCarty, MD
Fort Worth, TX

"Sine arte, scientia nihil est"