Sholes filed for a patent in 1867 and spent the next six years in trial-and-error efforts to improve his machine. QWERTY emerged from this period of tinkering and compromise. As another added wrinkle (and fine illustration of history’s odd quirks), R joined the top row as a last-minute entry, and for a somewhat capricious motive according to one common tale (perhaps apocryphal)—for salesmen could then impress potential buyers by smooth and rapid production of the brand name TYPE WRITER, all on one row. (Although I wonder how many sales were lost when TYPE EEEEEE appeared after a jam!)
The Survival of QWERTY: We can all accept this story of QWERTY’s origin, but why did it persist after the introduction of the modern platen roller and front-stroke key? (The first typewriter with a fully visible printing point was introduced in 1890.) In fact, the situation is even more puzzling. I thought that alternatives to keystroke typing only became available, with the IBM electric ball, but none other than Thomas Edison filed a patent for an electric print-wheel machine as early as 1872, and L. S. Crandall marketed a writing machine without typebars in 1879. (Crandall arranged his type on a cylindrical sleeve and made the sleeve revolve to the required letter before striking the printing point.)
The 1880s were boom years for the fledgling typewriter industry, a period when a hundred flowers bloomed and a hundred schools of thought contended. Alternatives to QWERTY were touted by several companies, and both the variety of printing designs (several without typebars) and the improvement of keystroke typewriters completely removed the original rationale for QWERTY. Yet during the 1890s, more and more companies made the switch to QWERTY, which became an industry standard by the early years of our century. And QWERTY has held on stubbornly, through the introduction of the IBM Selectric and the Hollerith punch card machine to that ultimate example of its nonnecessity, the microcomputer terminal.
To understand the survival (and domination to this day) of drastically suboptimal QWERTY, we must recognize two other commonplaces of history, as applicable to life in geological time as to technology over decades—contingency and incumbency. We call a historical event—the rise of mammals or the dominance of QWERTY—contingent when it occurs as the chancy result of a long string of unpredictable antecedents, rather than as a necessary outcome of nature’s laws. Such contingent events often depend crucially upon choices from a distant past that seemed tiny and trivial at the time. Minor perturbations early in the game can nudge a process into a new pathway, with cascading consequences that produce an outcome vastly different from any alternative.
Incumbency also reinforces the stability of a pathway once the little quirks of early flexibility push a sequence into a firm channel. Suboptimal politicians often prevail nearly forever once they gain office and grab the reins of privilege, patronage, and visibility. Mammals waited 100 million years to become the dominant animals on land and only got a chance because dinosaurs succumbed during a mass extinction. If every typist in the world stopped using QWERTY tomorrow and began to learn Dvorak, we would all be winners, but who will bell the cat or start the ball rolling? (Choose your cliché, for they all record this evident truth.) Stasis is the norm for complex systems; change, when provoked at all, is usually rapid and episodic.
QWERTY’s fortunate and improbable ascent to incumbency occurred by a concatenation of circumstances, each indecisive in itself, but all probably necessary for the eventual outcome. Remington had marketed the Sholes machine with its QWERTY keyboard, but this early tie with a major firm did not secure QWERTY’s victory. Competition was tough, and no lead meant much with such small numbers in an expanding market. David estimates that only 5,000 or so QWERTY machines existed at the beginning of the 1880s.
The push to incumbency was complex and multifaceted, dependent more upon the software of teachers and promoters than upon the hardware of improving machines. Most early typists used idiosyncratic hunt-and-peck, few-fingered methods. In 1882, Ms. Longley, founder of the Shorthand and Typewriter Institute in Cincinnati, developed and began to teach the eight-finger typing that professionals use today. She happened to teach with a QWERTY keyboard, although many competing arrangements would have served her purposes as well. She also published a popular do-it-yourself pamphlet. At the same time, Remington began to set up schools for typewriting using (of course) its QWERTY standard. The QWERTY ball was rolling but this head start did not guarantee a place at the summit. Many other schools taught rival methods on different machines and might have gained an edge.
Then a crucial event in 1888 probably added the decisive increment to QWERTY’s small advantage. Longley was challenged to prove the superiority of her eight-finger method by Louis Taub, another Cincinnati typing teacher, who worked with four fingers on a rival non-QWERTY keyboard with six rows, no shift action, and (therefore) separate keys for upper-and lowercase letters. As her champion, Longley engaged Frank E. McGurrin, an experienced QWERTY typist who had given himself a decisive advantage that, apparently, no one had utilized before. He had memorized the QWERTY keyboard and could therefore operate his machine as all competent typists do today—by what we now call touch-typing. McGurrin trounced Taub in a well-advertised and well-reported public competition.
In public perception, and (more important) in the eyes of those who ran typing schools and published typing manuals, QWERTY had proved its superiority. But no such victory had really occurred. The tie of McGurrin to QWERTY was fortuitous and a good break for Longley and for Remington. We shall never know why McGurrin won, but reasons quite independent of QWERTY cry out for recognition: touch-typing over hunt-and-peck, eight fingers over four fingers, the three-row letter board with a shift key versus the six-row board with two separate keys for each letter. An array of competitions that would have tested QWERTY were never held—QWERTY versus other arrangements of letters with both contestants using eight-finger touch-typing on a three-row keyboard, or McGurrin’s method of eight-finger touch-typing on a non-QWERTY three-row keyboard versus Taub’s procedure to see whether the QWERTY arrangement (as I doubt) or McGurrin’s method (as I suspect) had secured his success.
In any case, the QWERTY steamroller now gained crucial momentum and prevailed early in our century. As touch-typing by QWERTY became the norm in America’s typing schools, rival manufacturers (especially in a rapidly expanding market) could adapt their machines more easily than people could change their habits—and the industry settled upon the wrong standard.
If Sholes had not gained his tie to Remington, if the first typist who decided to memorize a keyboard had used a non-QWERTY design, if McGurrin had a bellyache or drank too much the night before, if Longley had not been so zealous, if a hundred other perfectly possible things had happened, then I might be typing this essay with more speed and much greater economy of finger motion.
But why fret over lost optimality. History always works this way. If Montcalm had won a battle on the Plains of Abraham, perhaps I would be typing en français. If a portion of the African jungles had not dried to savannas, I might still be an ape up a tree. If some comets had not struck the earth (if they did) some 60 million years ago, dinosaurs might still rule the land, and all mammals would be rat-sized creatures scurrying about in the dark corners of their world. If Pikaia, the only chordate of the Burgess Shale, had not survived the great sorting out of body plans after the Cambrian explosion, mammals might not exist at all. If multicellular creatures had never evolved after five-sixths of life’s history had yielded nothing more complicated than an algal mat, the sun might explode a few billion years hence with no multicellular witness to the earth’s destruction.
Compared with these weighty possibilities, my indenture to QWERTY seems a small price indeed for the rewards of history. For if history were not so maddeningly quirky, we would not be here to enjoy it. Streamlined optimality contains no seeds for change. We need our odd little world, where QWERTY rules and the quick brown fox jumps over the lazy dog.*
Postscript
Since typing falls into the category of things
that many, if not most of us, can do (like walking and chewing gum simultaneously) this essay elicited more commentary than most of my more obscure ramblings.
Some queried the central premises and logic. An interesting letter from Folsom Prison made a valid point in the tough humor of such institutions. (I receive many letters from prisoners and am always delighted by such reminders that, at least for many people, the quest for knowledge never abates, even in most uncongenial temporary domiciles):
Some of us were left with a nagging question: If the hunt ’n peck method prevailed until around 1882, how could Sholes or his cohorts have “relegated common letters to weak fingers” when there were no weak fingers, just hunt ’n peck type fingers? At least none of the hunt ’n peck typing clerks or cops around here use the weak fingers. If you could find the time to answer this it would really be appreciated and could serve to reduce the likelihood of increased violence at Folsom between opposing QWERTY origin factions.
My correspondent is quite right, and I misspoke (I also trust that recent tension at Folsom had sources other than the great typewriter wars—yes, I did answer the letter promptly). Fortunately, my hypothesis is secure against my own carelessness—for Sholes needed simply to separate frequently struck keys to avoid jamming. The finger used to strike mattered little (I also rather suspect that many people were experimenting with many-fingered typing before the full four-fingered methods became canonical).
But the vast bulk of correspondence, more than 80 percent, took issue with my throwaway and tangential last line—thanks to our long-standing and happy fascination with words and word games. I gave the conventional typist’s sentence as being the shortest phrase using all letters:
The quick brown fox jumps over the lazy dog.
I have since learned that sentences containing all letters of the alphabet are called “pangrams,” and that the quest for the shortest represents at least a minor industry, with much effort spent, and opposing factions with strong passions. Many readers suggested, as a well-known alternative with three fewer letters (32 versus 35),
Pack my box with five dozen liquor jugs.
Zoological enthusiasts and prohibitionists then retort that the fox-dog classic can still tie by dropping the first article and becoming only slightly less grammatical:
Quick brown fox jumps over the lazy dog.
But Ted Leather wins this limited derby for shortest sensible pangram with the 31-stroke
Jackdaws love my big sphinx of quartz.
We now enter the world of arcana. Can shorter pangrams be made? Can the ultimate 26-letter sentence be constructed? This quest has so far stymied all wordsmiths. Using common words only, we can get down to 28 (but only by the slightly dishonorable route of using proper names):
Waltz, nymph, for quick jigs vex Bud.
And to 27, with some archaic orthography:
Frowzy things plumb vex’d Jack Q.
But for the ultimate of 26, we either use initials in abundance (which doesn’t seem quite fair),
J. Q_. Schwartz flung V. D. Pike my box,
or we avoid names and initials, but employ such unfamiliar and marginally admissable words that an equal feeling of dissatisfaction arises,
Zing! Vext cwm fly jabs Kurd qoph.
A cwm is a mountain hollow in Wales, while qoph, the nineteenth letter of the Hebrew alphabet, has been drawn (and has attracted the ire of an immigrant fly) by a member of an Iranian minority. Sounds awfully improbable.
My favorite proposal for a 26-letter pangram requires an entire story for comprehension (thanks to Dan Lufkin of Hood College):
During World War I, Lawrence’s Arab Legion was operating on the southern flank of the Ottoman Empire. Hampered by artillery fire from across a river, Lawrence asked for a volunteer to cross the river at night and locate the enemy guns. An Egyptian soldier stepped forward. The man was assigned to Lawrence’s headquarters [G.H.Q. for “general headquarters”—this becomes important later] and had a reputation for bringing bad luck. But Lawrence decided to send him. The mission was successful and the soldier appeared, at dawn the next morning, at a remote sentry post near the river, dripping wet, shivering, and clad in nothing but his underwear and native regimental headgear. The sentry wired to Lawrence for instructions, and he replied:
Warm plucky G.H.Q. jinx, fez to B.V.D.’s.
A free copy of this and all my subsequent books to anyone who can construct a 26-letter pangram with common words only and no proper names.
2 | Dinomania
5 | Bully for Brontosaurus
QUESTION: What do Catherine the Great, Attila the Hun, and Bozo the Clown have in common? Answer: They all have the same middle name.
Question: What do San Marino, Tannu Tuva, and Monaco have in common? Answer: They all realized that they could print pretty pieces of perforated paper, call them stamps, and sell them at remarkable prices to philatelists throughout the world. (Did these items ever bear any relationship to postage or utility? Does anyone own a canceled stamp from Tannu Tuva?) Some differences, however, must be admitted. Although San Marino (a tiny principality within Italy) and Tannu Tuva (a former state adjacent to Mongolia but now annexed to the Soviet Union) may rely on stamps for a significant fraction of their GNP, Monaco, as we all know, has another considerable source of outside income—the casino of Monte Carlo (nurtured by all the hype and elegance of the Grimaldis—Prince Rainier, Grace Kelly, and all that).
So completely do we identify Monaco with Monte Carlo that we can scarcely imagine any other activity, particularly something productive, taking place in this little land of fantasy and fractured finances.
Nonetheless, people are born, work, and die in Monaco. And this tiny nation boasts, among other amenities, a fine station for oceanographic research. This combination of science and hostelry makes Monaco an excellent place for large professional meetings. In 1913, Monaco hosted the International Zoological Congress, the largest of all meetings within my clan. This 1913 gathering adopted the important Article 79, or “plenary powers decision,” stating that “when stability of nomenclature is threatened in an individual case, the strict application of the Code may under specified conditions be suspended by the International Commission on Zoological Nomenclature.”
Now I will not blame any reader for puzzlement over the last paragraph. The topic—rules for giving scientific names to organisms—is easy enough to infer. But why should we be concerned with such legalistic arcana? Bear with me. We shall detour around the coils of Boa constrictor, meet the International Code of Zoological Nomenclature head-on, and finally arrive at a hot issue now generating much passion and acrimony at the heart of our greatest contemporary fad. You may deny all concern for rules of taxonomy, our last domain of active Latin (now that Catholicism has embraced the vernacular), but millions of Americans are now het up about the proper name of Brontosaurus, the canonical dinosaur. And you can’t grasp the name of the beast without engaging the beastly rules of naming.
Nonprofessionals often bridle at the complex Latin titles used by naturalists as official designations for organisms. Latin is a historical legacy from the foundation of modern taxonomy in the mid-eighteenth century—a precomputer age when Romespeak was the only language shared by scientists throughout the world. The names may seem cumbersome, now that most of us pass our youthful years before a television set, rather than declaiming hic-haec-hoc and amo-amas-amat. But the principle remains sound. Effective communication demands that organisms have official names, uniformly recognized in all countries, while a world of changing concepts and increasing knowledge requires that rules of naming foster maximal stability and minimal disruption.
New species are discovered every day; old names must often change as we correct past errors and add new information. If every change of concept demanded a redesignation of all names and a reordering of all categories, natural history would devolve into chaos. Our communications would fail as species, the basic units of all our discourse, would have no recognized labe
ls. All past literature would be a tangle of changing designations, and we could not read without a concordance longer than the twenty volumes of the Oxford English Dictionary.
The rules for naming animals are codified in the International Code of Zoological Nomenclature, as adopted and continually revised by the International Union of Biological Sciences (plant people have a different code based on similar principles). The latest edition (1985), bound in bright red, runs to 338 pages. I will not attempt to summarize the contents, but only state the primary goal: to promote maximal stability as new knowledge demands revision.
Consider the most prevalent problem demanding a solution in the service of stability: When a single species has been given two or more names, how do we decide which to validate and which to reject? This common situation can arise for several reasons: Two scientists, each unaware of the other’s work, may name the same animal; or a single scientist, mistaking a variable species for two or more separate entities, may give more than one name to members of the same species. A simple and commonsensical approach might attempt to resolve all such disputes with a principle of priority—let the oldest name prevail. In practice, such “obvious” solutions rarely work. The history of taxonomy since Linnaeus has featured three sequential approaches to this classic problem.
1. Appropriateness. Modern nomenclature dates from the publication, in 1758, of the tenth edition of Linnaeus’s Systema Naturae. In principle, Linnaeus endorsed the rule of priority. In practice, he and most of his immediate successors commonly changed names for reasons, often idiosyncratic, of supposed “appropriateness.” If the literal Latin of an original name ceased to be an accurate descriptor, new names were often devised. (For example, a species originally named floridensis to denote a restricted geographic domain might be renamed americanus if it later spread throughout the country.)
Bully for Brontosaurus Page 7