And besides, why do we prefer creation myths to evolutionary stories? I find all the usual reasons hollow. Yes, heroes and shrines are all very well, but is there not grandeur in the sweep of continuity? Shall we revel in a story for all humanity that may include the sacred ball courts of the Aztecs, and perhaps, for all we know, a group of Homo erectus hitting rocks or skulls with a stick or a femur? Or shall we halt beside the mythical Abner Doubleday, standing behind the tailor’s shop in Cooperstown, and say “behold the man”—thereby violating truth and, perhaps even worse, extinguishing both thought and wonder?
4 | The Panda’s Thumb of Technology
THE BRIEF STORY of Jephthah and his daughter (Judg. 11:30–40) is, to my mind and heart, the saddest of all biblical tragedies. Jephthah makes an intemperate vow, yet all must abide by its consequences. He promises that if God grant him victory in a forthcoming battle, he will sacrifice by fire the first living thing that passes through his gate to greet him upon his return. Expecting (I suppose) a dog or a goat, he returns victorious to find his daughter, and only child, waiting to meet him “with timbrels and with dances.”
Handel’s last oratorio, Jephtha, treats this tale with great power (although his librettist couldn’t bear the weight of the original and gave the story a happy ending, with angelic intervention to spare Jephthah’s daughter at the price of her lifelong chastity). At the end of Part 2, while all still think that the terrible vow must be fulfilled, the chorus sings one of Handel’s wonderful “philosophical” choruses. It begins with a frank account of the tragic circumstance:
How dark, O Lord, are thy decrees!…
No certain bliss, no solid peace,
We mortals know on earth below.
Yet the last two lines, in a curious about-face, proclaim (with magnificent musical solidity as well):
Yet on this maxim still obey:
WHATEVER IS, IS RIGHT
This odd reversal, from frank acknowledgment to unreasonable acceptance, reflects one of the greatest biases (“hopes” I like to call them) that human thought imposes upon a world indifferent to our suffering. Humans are pattern-seeking animals. We must find cause and meaning in all events (quite apart from the probable reality that the universe both doesn’t care much about us and often operates in a random manner). I call this bias “adaptationism”—the notion that everything must fit, must have a purpose, and in the strongest version, must be for the best.
The final line of Handel’s chorus is, of course, a quote from Alexander Pope, the last statement of the first epistle of his Essay on Man, published twenty years before Handel’s oratorio. Pope’s text contains (in heroic couplets to boot) the most striking paean I know to the bias of adaptationism. In my favorite lines, Pope chastises those people who may be unsatisfied with the senses that nature bestowed upon us. We may wish for more acute vision, hearing, or smell, but consider the consequences.
If nature thunder’d in his op’ning ears
And stunn’d him with the music of the spheres
How would he wish that Heav’n had left him still
The whisp’ring zephyr, and the purling rill!
And my favorite couplet, on olfaction:
Or, quick effluvia darting thro’ the brain,
Die of a rose in aromatic pain.
What we have is best for us—whatever is, is right.
By 1859, most educated people were prepared to accept evolution as the reason behind similarities and differences among organisms—thus accounting for Darwin’s rapid conquest of the intellectual world. But they were decidedly not ready to acknowledge the radical implications of Darwin’s proposed mechanism of change, natural selection, thus explaining the brouhaha that the Origin of Species provoked—and still elicits (at least before our courts and school boards).
Darwin’s world is full of “terrible truths,” two in particular. First, when things do fit and make sense (good design of organisms, harmony of ecosystems), they did not arise because the laws of nature entail such order as a primary effect. They are, rather, only epiphenomena, side consequences of the basic causal process at work in natural populations—the purely “selfish” struggle among organisms for personal reproductive success. Second, the complex and curious pathways of history guarantee that most organisms and ecosystems cannot be designed optimally. Indeed, to make an even stronger statement, imperfections are the primary proofs that evolution has occurred, since optimal designs erase all signposts of history.
This principle of imperfection has been a major theme of my essays for several years. I call it the panda principle to honor my favorite example, the panda’s false thumb. Pandas are the herbivorous descendants of carnivorous bears. Their true anatomical thumbs were, long ago during ancestral days of meat eating, irrevocably committed to the limited motion appropriate for this mode of life and universally evolved by mammalian Carnivora. When adaptation to a diet of bamboo required more flexibility in manipulation, pandas could not redesign their thumbs but had to make do with a makeshift substitute—an enlarged radial sesamoid bone of the wrist, the panda’s false thumb. The sesamoid thumb is a clumsy, suboptimal structure, but it works. Pathways of history (commitment of the true thumb to other roles during an irreversible past) impose such jury-rigged solutions upon all creatures. History inheres in the imperfections of living organisms—and thus we know that modern creatures had a different past, converted by evolution to their current state.
We can accept this argument for organisms (we know, after all, about our own appendixes and aching backs). But is the panda principle more pervasive? Is it a general statement about all historical systems? Will it apply, for example, to the products of technology? We might deem this principle irrelevant to the manufactured objects of human ingenuity—and for good reason. After all, constraints of genealogy do not apply to steel, glass, and plastic. The panda cannot shuck its digits (and can only build its future upon an inherited ground plan), but we can abandon gas lamps for electricity and horse carriages for motor cars. Consider, for example, the difference between organic architecture and human buildings. Complex organic structures cannot be reevolved following their loss; no snake will redevelop front legs. But the apostles of post-modern architecture, in reaction to the sterility of so many glass-box buildings of the international style, have juggled together all the classical forms of history in a cascading effort to rediscover the virtues of ornamentation. Thus, Philip Johnson could place a broken pediment atop a New York skyscraper and raise a medieval castle of plate glass in downtown Pittsburgh. Organisms cannot recruit the virtues of their lost pasts.
Yet I am not so sure that technology is exempt from the panda principle of history, for I am now sitting face to face with the best example of its application. Indeed, I am in most intimate (and striking) contact with this object—the typewriter keyboard.
I could type before I could write. My father was a court stenographer, and my mother is a typist. I learned proper eight-finger touch-typing when I was about nine years old and still endowed with small hands and weak, tiny pinky fingers. I was thus, from the first, in a particularly good position to appreciate the irrationality of placement for letters on the standard keyboard—called QWERTY by all aficionados in honor of the first six letters on the top letter row.
Clearly, QWERTY makes no sense (beyond the whiz and joy of typing QWERTY itself). More than 70 percent of English words can be typed with the letters DHIATENSOR, and these should be on the most accessible second, or home, row—as they were in a failed competitor to QWERTY introduced as early as 1893. But in QWERTY, the most common English letter, E, requires a reach to the top row, as do the vowels U, I, and O (with O struck by the weak fourth finger), while A remains in the home row but must be typed with the weakest finger of all (at least for the dexterous majority of right-handers)—the left pinky. (How I struggled with this as a boy. I just couldn’t depress that key. I once tried to type the Declaration of Independence and ended up with: th t ll men re cre ted equ l.)
As a dramatic illustration of this irrationality, consider the accompanying photograph, the keyboard of an ancient Smith-Corona upright, identical with the one (my dad’s original) that I use to type these essays (a magnificent machine—no breakdown in twenty years and a fluidity of motion unmatched by any manual typewriter since). After more than half a century of use, some of the most commonly struck keys have been worn right through the surface into the soft pad below (they weren’t solid plastic in those days). Note that E, A, and S are worn in this way—but note also that all three are either not in the home row or are struck with the weak fourth and pinky fingers in QWERTY.
This claim is not just a conjecture based on idiosyncratic personal experience. Evidence clearly shows that QWERTY is drastically suboptimal. Competitors have abounded since the early days of typewriting, but none has supplanted or even dented the universal dominance of QWERTY for English typewriters. The best-known alternative, DSK, for Dvorak Simplified Keyboard, was introduced in 1932. Since then, virtually all records for speed typing have been held by DSK, not QWERTY, typists. During the 1940s, the U.S. Navy, ever mindful of efficiency, found that the increased speed of DSK would amortize the cost of retraining typists within ten days of full employment. (Mr. Dvorak was not Anton of the New World Symphony, but August, a professor of education at the University of Washington, who died disappointed in 1975. Dvorak was a disciple of Frank B. Gilbreth, pioneer of time and motion studies in industrial management.)
Since I have a special interest in typewriters (my affection for them dates to childhood days of splendor in the grass and glory in the flower), I have wanted to write such an essay for years. But I never had the data I needed until Paul A. David, Coe Professor of American Economic History at Stanford University, kindly sent me his fascinating article, “Understanding the Economics of QWERTY: The Necessity of History” (in Economic History and the Modern Economist, edited by W. N. Parker, New York, Basil Blackwell Inc., 1986, pp. 30–49). Virtually all the nonidiosyncratic data in this essay come from David’s work, and I thank him for this opportunity to satiate an old desire.
The puzzle of QWERTY’s dominance resides in two separate questions: Why did QWERTY ever arise in the first place? And why has QWERTY survived in the face of superior competitors?
A classic upright typewriter of World War I vintage. Brother to the machine that I use to write these essays.
Notice the patterns of wear for most frequently used keys, as illustrated by breakage through the surface after so many years of striking. In QWERTY, all the most common keys are either not in the home row, or are hit by weak fingers in the home row—thus illustrating the suboptimality of this standard arrangement.
A keyboard for a typewriter made in the 1880’s, illustrating one of the many competing non-QWERTY arrangements so common at the time.
My answers to these questions will invoke analogies to principles of evolutionary theory. Let me, then, state some ground rules for such a questionable enterprise. I am convinced that comparisons between biological evolution and human cultural or technological change have done vastly more harm than good—and examples abound of this most common of all intellectual traps. Biological evolution is a bad analogue for cultural change because the two systems are so different for three major reasons that could hardly be more fundamental.
First, cultural evolution can be faster by orders of magnitude than biological change at its maximal Darwinian rate—and questions of timing are of the essence in evolutionary arguments. Second, cultural evolution is direct and Lamarckian in form: The achievements of one generation are passed by education and publication directly to descendants, thus producing the great potential speed of cultural change. Biological evolution is indirect and Darwinian, as favorable traits do not descend to the next generation unless, by good fortune, they arise as products of genetic change. Third, the basic topologies of biological and cultural change are completely different. Biological evolution is a system of constant divergence without subsequent joining of branches. Lineages, once distinct, are separate forever. In human history, transmission across lineages is, perhaps, the major source of cultural change. Europeans learned about corn and potatoes from Native Americans and gave them smallpox in return.
So, when I compare the panda’s thumb with a typewriter keyboard, I am not attempting to derive or explain technological change by biological principles. Rather, I ask if both systems might not record common, deeper principles of organization. Biological evolution is powered by natural selection, cultural evolution by a different set of principles that I understand but dimly. But both are systems of historical change. More general principles of structure must underlie all systems that proceed through history (perhaps I now only show my own bias for intelligibility in our complex world)—and I rather suspect that the panda principle of imperfection might reside among them.
My main point, in other words, is not that typewriters are like biological evolution (for such an argument would fall right into the nonsense of false analogy), but that both keyboards and the panda’s thumb, as products of history, must be subject to some regularities governing the nature of temporal connections. As scientists, we must believe that general principles underlie structurally related systems that proceed by different overt rules. The proper unity lies not in false applications of these overt rules (like natural selection) to alien domains (like technological change), but in seeking the more general rules of structure and change themselves.
The Origin of QWERTY: True randomness has limited power to intrude itself into the forms of organisms. Small and unimportant changes, unrelated to the working integrity of a complex creature, may drift in and out of populations by a process akin to throwing dice. But intricate structures, involving the coordination of many separate parts, must arise for an active reason—since the bounds of mathematical probability for fortuitous association are soon exceeded as the number of working parts grows.
But if complex structures must arise for a reason, history may soon overtake the original purpose—and what was once a sensible solution becomes an oddity or imperfection in the altered context of a new future. Thus, the panda’s true thumb permanently lost its ability to manipulate objects when carnivorous ancestors found a better use for this digit in the limited motions appropriate for creatures that run and claw. This altered thumb then becomes a constraint imposed by past history upon the panda’s ability to adapt in an optimal way to its new context of herbivory. The panda’s thumb, in short, becomes an emblem of its different past, a sign of history.
Similarly, QWERTY had an eminently sensible rationale in the early technology of typewriting but soon became a constraint upon faster typing as advances in construction erased the reason for QWERTY’s origin. The key (pardon the pun) to QWERTY’s origin lies in another historical vestige easily visible on the second row of letters. Note the sequence: DFGHJKL—a good stretch of the alphabet in order, with the vowels E and I removed. The original concept must have simply arrayed the letters in alphabetical order. Why were the two most common letters of this sequence removed from the most accessible home row? And why were other letters dispersed to odd positions?
Those who remember the foibles of manual typewriters (or, if as hidebound as yours truly, still use them) know that excessive speed or unevenness of stroke may cause two or more keys to jam near the striking point. You also know that if you don’t reach in and pull the keys apart, any subsequent stroke will type a repetition of the key leading the jam—as any key subsequently struck will hit the back of the jammed keys and drive them closer to the striking point.
These problems were magnified in the crude technology of early machines—and too much speed became a hazard rather than a blessing, as key jams canceled the benefits of celerity. Thus, in the great human traditions of tinkering and pragmatism, keys were moved around to find a proper balance between speed and jamming. In other words—and here comes the epitome of the tale in a phrase—QWERTY arose in order to slow down the maximal speed
of typing and prevent jamming of keys. Common letters were either allotted to weak fingers or dispersed to positions requiring a long stretch from the home row.
This basic story has gotten around, thanks to short takes in Time and other popular magazines, but the details are enlightening, and few people have the story straight. I have asked nine typists who knew this outline of QWERTY’s origin and all (plus me for an even ten) had the same misconception. The old machines that imposed QWERTY were, we thought, of modern design—with keys in front typing a visible line on paper rolled around a platen. This leads to a minor puzzle: Key jams may be a pain in the butt, but you see them right away and can easily reach in and pull them apart. So why QWERTY?
As David points out, the prototype of QWERTY, a machine invented by C. L. Sholes in the 1860s, was quite different in form from modern typewriters. It had a flat paper carriage and did not roll paper right around the platen. Keys struck the paper invisibly from beneath, not patently from the front as in all modern typewriters. You could not view what you were typing unless you stopped to raise the carriage and inspect your product. Keys jammed frequently, but you could not see (and often did not feel) the aggregation. Thus, you might type a whole page of deathless prose and emerge only with a long string of E’s.