Each English word encodes its concept -- concepts, plural -- using 26 discrete symbols, much like a decimal number encodes a quantity using ten discrete symbols. Indeed, it is possible to represent the number of subatomic particles in the Milky Way Galaxy with a string of fewer than 38 cyphers, each confined to only ten different shapes. Likewise, with variable-length strings of letters, each confined to only 26 different shapes, English can represent all 38,000 words, which is quite a remarkable feat, considering that other exceptionally popular languages (need I say Chinese?) require the use of almost that many written -- drawn -- symbols.

In early 2001, Fred Shapiro, Associate Librarian for Public Services and Lecturer in Legal Research at Yale Law School expressed concern that my estimate of 38,000 is too low, suggesting that the Oxford English Dictionary lists many times that amount. I shall take his word for that.

The estimate of 38,000 English words stuck in my memory from some long time ago (38,006, perhaps, since I have added a half-dozen myself). Upon receiving Shapiro's query, though, I took down from my shelf The American Heritage Dictionary of the English Language, which is not the OED but one I happen to like a whole lot. Now, there are 1,491 pages with definitions on them. I did a quick count of the words defined on a couple of randomly selected pages and got an estimate of 25.5 words. You can do the indicated arithmetic, of course, but I'll save you the trouble: 1,491 times 25.5 equals 38,020 words.  In the present context, I consider that number to be quite generous, since it includes entries like Aalborg and Zola, which ought to be excluded since they do not qualify as invented words capable of offsetting the paucity being lamented here.

Mr. Shapiro expressed the belief to me that William Shakespeare used 33,000 separately definable words. I shall take his word for that, too. Assuming the Bard used all the English words available in his time, then people have succeeded in adding only about one word-per-month to the language throughout the past four centuries. One of them, as I have noted elsewhere, is "byte."

Another, of course, is 'software.'

{Return to citation in Part 0}

Once, I needed to select a clickable word to appear on a computer screen, and "run" seemed to fit the requirement. I looked up "run" in my favorite dictionary, The American Heritage Dictionary of the English Language, and discovered that its definitions took up three-quarters of page. I tentatively concluded that "run" must be the English word with the most definitions.
 
Some authorities have since nominated the word "set" for that distinction, but I remain unconvinced. The controversy is further analyzed elsewhere.


Of course, both words are cases in my point here -- that English would surely be enriched if more lexical creativity were at play, by which heavily laden little words would be gratefully unburdened. By the way, I used "run" on the screen since the word "execute" would not fit in the limited space.

{Return to citation in Part 0}

Although various abbreviations were accomplished through symbols, even in the works of Aristotle himself, the use of symbols in a formal system, the precursor of modern symbolic logic, began with George Boole (1847) and Ernst Schröder (1890-1905). The system was developed further by Gottlob Frege (1879) and finally culminated in the Principia Mathematica of Bertrand Russell and Alfred North Whitehead (1910-13).

in logic, a valid deductive argument having two premises and a conclusion. The traditional type is the categorical syllogism, in which both premises and the conclusion are simple declarative statements that are constructed using only three terms between them, each term appearing twice (as a subject and as a predicate): "All men are mortal; no gods are mortal; therefore no men are gods." The argument in such syllogisms is valid by virtue of the fact that it would not be possible to assert the premises and to deny the conclusion without contradicting oneself.
-- Encyclopedia Britannca
{Return to citation in Part 1}

noun [Latin, from Greek synekdoche, from. syn- + ekdoche sense, interpretation, from ekdechesthai to receive, understand; akin to Greek dokein to seem good]: a figure of speech by which a part is put for the whole (as fifty sail for fifty ships), the whole for a part (as society for high society), the species for the genus (as cutthroat for assassin), the genus for the species (as a creature for a man), or the name of the material for the thing made (as boards for stage).
-- Merriam-Webster's Collegiate Dictionary


In the seventies "Detroit iron" was the synecdoche used to pejorate vehicles manufactured by domestic automobile companies. "Silicon Valley" is a synecdoche, so is "dot-com."  Likewise "hardware" for computer equipment -- the concrete things you can heft and haul.

{Return to citation in Part 1}

The 1999 edition of the Oxford English Dictionary includes the following definition:
 
hardware  the physical components of a system or device as opposed to the procedures required for its operation; opp. software.

The earliest citations include...
 
1947 D. R. Hartree Calculating Machines "The ENIAC. I shall give a brief account of it, since it will make the later discussion more realistic if you have an idea of some ‘hardware’ and how it is used, and this is the equipment with which I am best acquainted."

1953 A. D. & K. H. V. Booth Automatic Digital Calculators "The engineering difficulties encountered in this type of machine are great, and a considerable increase in the size and complexity of the ‘hardware’ seems inevitable."


Accordingly, the word "hardware" was decidedly not a retronym, having made its attested appearance as applied to computing machines by 1947 -- a half-dozen years before the coinage of the word 'software.'   That's six years by my reckoning -- not 13 years as depicted in the OED citations.

Readers are invited to observe the exclamation point at the end of the next sentence.  In a rapidly developing realm, it seems doubtful indeed that such an obvious word relationship would have taken more than a dozen years to be recognized!

{Return to citation in Part 1}

A biographical sketch of one of the first computer programmers, Kay McNulty Mauchly Antonelli, appeared in the October 27, 2000 edition of Atlanta Journal-Constitution. She is the widow of John Mauchly who with Presper Eckert built the ENIAC in 1945. The passage most relevant to the present subject: "There was no such thing as software. ENIAC was programmed by wiring."
{Return to citation in Part 1}

In about 1949, according to the mythology of the 1950s, a hapless moth flew into the Harvard Mark II. The technician taped the insect's corpse alongside the entry in his log "first actual case of a bug being found."

There was no verbal distinction in those days between a hardware failure and a programming error, both merely being deplored as "bugs." For some people the distinction is still pending (see Software Does Not Fail).

{Return to citation in Part 2}

In "Digital Computers" (American Mathematical Monthly, Vol. 62, No. 6. Jun. - Jul., 1955, pp. 414-423), Mina Rees quoted the late Claude Shannon (1916-2001)  as saying "a digital computer must be instructed in words of one microsyllable."

My friend Rich Alexander sent me Claude Shannon's obituary accompanied by this note:
 
Regarding the first use of 'software,' you might be interested to know, I checked one of my all-time favorite books, Symbols, Signals and Noise (Harper & Row, 1961), written by J. R. Pierce, the top guy at Bell Labs and a buddy of Claude Shannon.  The book focuses on information/communication theory, not computers, but there are passages about computers in which he uses these software-related terms: commands, compiler, instructions, program (noun), program (verb), programmer, and sequence of instructions.  Pierce, the head of Bell Labs, does not use the word 'software.'
{Return to citation in Part 2}

Here is a partial list of the projects on which I worked during my two years (1953-1955) at the Institute for Transportation and Traffic Engineering.

• Automobile Crash Injury Research -- The world's earliest crash tests using real automobiles and anthropometric dummies to ascertain the efficacy of various restraining devices (lap belt, shoulder harness, chest-level strap) in preventing injuries to passengers resulting from frontal collisions; the research was cited prominently in Ralph Nader's landmark book Unsafe At Any Speed.  A certain undergraduate authored the first published article on this research in the April 1954 edition of California Engineer, and the paper entitled Engineered Automobile Crashes re-appears on this website 50 years later by permission of the publisher. 
• Radar Speed Meter -- The first such instrument being larger than a breadbox mounted on a tripod, with primitive klystron oscillator; controlled experiments confirmed the predicted accuracy and "cosine error in favor of the driver"; conducted successful demonstrations for law enforcement officials and legal experts from the eleven western states.
• Padded Dashboard -- Laboratory testing (accelerometry using an impacting pendulum) of various materials to determine the most effective, asymmetrical resiliencies for minimizing injuries from the "secondary" (inside-the-vehicle) collision; drafted recommendations to lawmakers resulting in Federal mandates for inclusion as standard equipment in new cars beginning in the sixties.
• Vehicular Metering -- Computer analyses ("Monte Carlo simulation") and experimental confirmation of various causes of traffic viscosity, including marginal friction and critical absorption volumes that cause "density lock" (now known by the misnomer "gridlock"), which adversely affects both the capacity and the mean transiting speed of a given highway system; the work resulted in locally deployed metering lights, which are now widely applied to relieve congestion.
• Cable Barrier -- Analysis and testing of a UCLA-initiated system to assure opposing lane separation on expressways, designed to prevent both head-on intrusion hazards and rebounding damage resulting from post-crash control-loss; a system characterized by longitudinal cable supported by frangible aluminum posts, that were ultimately applied nationwide throughout the 1980s and into the 1990s.
• Signage and Signal Standards -- Experimental work in support of the earliest "human factors" research into traffic safety impacts attributable to destination signs and direction indicators, lane striping and signal lights; also conducted controlled experiments in a unique driving simulator with a wide-angle, cinerama-like screen to gauge effects on drivers subjected to fatigue and to various physiologically active substances.


There can be no doubt that the innovative work carried out by the research team at UCLA's ITTE during the 1950s was profoundly influential and will be viewed with respect by generations to come. It was an awsome privilege for an otherwise undistinguished verbal prankster to be included on that team.

My reputation was first established during the automobile crash injury research at ITTE. Derwyn Severy, the project leader, subjected himself to a low-speed rear-end collision experiment in a donated 1939 Plymouth, sort of in the tradition of Col. John Stapp and his contemporary rocket-sled tests.  Remember those "news reels"? I was in charge of the high-speed motion picture photography.

Derwyn wore earplugs to prevent anticipation of the impact by the test car approaching from behind, and to assure that he would keep his eyes straight ahead, I affixed a purloined copy of the famous -- notorious -- Marilyn Monroe calendar upright on the hood. When the film was developed, we all gaped at the screen, astonished to see that, prior to the intended whiplash, the camera had faithfully recorded a hundred feet of Derwyn Severy grinning at a picture, which, though foreshortened, was all too immodestly discernible in each frame.

It was my turn to get a sore neck when the experiment was repeated for presentation in public forums. Without the calendar, of course.

At the conclusion of our first seatbelt study, I drafted the instrumentation chapter of the final report about the effectiveness of restraining devices on vehicular safety and prepared another section with a diagram describing one of my own early inventions. It was a micro-switch mounted under the cushion of a car-seat which was activated by the weight of a passenger and, in conjunction with the corresponding seatbelt switch, sounded a warning when the belt is left unfastened. In the diagram, I labeled it "API."

Nobody on the research team bothered to ask me what the letters stood for, doubtless assuming it was an arcane term-of-art. Finally, after publication of the first report, Dan Gerlough got suspicious.

"Ass-Presence Indicator," I explained cheerfully. Suffering profondément consterné, Derwyn Severy promptly issued a memorandum mandating that replies to all outside queries must define API as "Auxiliary Passenger Interlock." To the best of my recollection, no reviewer or journalist ever asked.

{Return to citation in Part 2}

...might be offered herein as a self-referent addition to Sniglets by Rich Hall and Friends, Illustrated by Arnie Ten (The MacMillan Company, 1984), probably the most popular of several humorous collections of frivolous neologisms ("Any word that doesn't appear in the dictionary, but should").

In a separate work entitled 101 Words I Don't Use, I have confessed to several coinages, including 'edutainment', 'circloid', 'holomorph', 'pluplural', 'patientoid', 'polycut', 'reprographics', 'totorial', and 'tridecabillion'.  Not all of them were frivolous.

{Return to citation in Part 7}

The hardware in the SWAC was always breaking down. But not the software. Once I got a program checked out (the expression "debugged," a term credited to the Navy's Grace Hopper, was not in common use in 1953), then it would never break.
 
Software never wears out, never decays, never gets weak. In terms of durability, then, software is not soft.


Words and phrases that start with ‘soft’ often imply attractive – even essential – attributes in some contexts.  ‘Soft’ means offering little resistance; easily molded, cut, or worked; malleable: plastic; not hard; yielding readily to pressure or weight; not firm; smooth or fine to the touch; not harsh or coarse; bland, not irritating; low-toned, not loud or strident; subdued not glaring or overly brilliant;  mild, gentle, and caressing; mild; balmy; yielding; easily touched; compassionate…

soft+boil, soft+copy, soft+drink, soft+focus, soft+footed, 
soft+goods, soft+hands, soft+key, soft+lighting, soft+liner, 
soft+money, soft+pedal, soft+rock, soft+shell, soft+shoe, 
soft+skin, soft+soap, soft+spoken, soft+spot, soft+style.

Some of these, along with undesirable properties of softness that are implied from other realms, act together to make ‘software’ a misnomer.  ‘Soft’ means out of condition; flabby; not sharply drawn or delineated; lenient, not stern; weak; unmanly or, for that matter, unwomanly; informal; simple; feeble; easy; diminished in value or importance…

soft+back, soft+ball, soft+cover, soft+headed, soft+hearted, 
soft+job, soft+news, soft+nosed,  soft+rock, soft+science, 
soft+target, soft+tissue, soft+touch, soft+witted, soft+y.

Excuse the immodesty, but shortly after I graduated from UCLA in 1955, I foresaw the incomparable "hardness" of software. That rather controversial concept is celebrated in my iconoclastic screed entitled Software Does Not Fail. First drafted in the seventies, the piece was initially offered in various versions to any number of publications throughout the eighties. Finally I published it myself in Sophisticated: The Magazine.  Here is an exerpt...
 

Hardware is 'hard,' which is to say concrete, not abstract. Hardware can be made, in some sense harder, more reliable. Hardware can be insulated, hermetically sealed, ruggedized, bullet-proofed.  Harder the better, presumably. But never hard enough. Sooner or later, hardware fails.  Software, being abstract, is -- well, 'soft'.   No reason to make software hard, though. Truth be known: It is because of its softness that software does not fail. Nothing Else Is Software. There is nothing softer than software. Other things may be as abstract as software but hardly softer.

It has given me considerable pleasure that such a little polemic has over the years been cited in technical articles, popular texts, and system specifications. The title, which was once reproached as unsound, is now widely recited as a simple declarative (always with the "does not," never with a "doesn't").

That 'software' is a misnomer may be the reason for an apparent resistance to it in other languages.

 

Catalan: programari Dutch: programmatuur  Finnish: ohjelmisto French: logiciel

Irish: bogearraí Indonesian: piranti Norwegian: programvare Persian: narmafzar

Russian: programmioye Spanish: programa Swedish: mjukvara Vietnamese: phan mem

Little wonder, considering the tepid connotations in 'softwords' -- softball, softcover, softener, softheaded, softhearted, softish, softling, softwood, softy.

Meanwhile, you have the strength and perminence of 'hardwords' -- hardback, hardball, hardbeam, hardcore, hardcover, harden, hardfisted, hardhead, hardhearted, hardihood, hardline, hardness, hardscrabble, hardstand, hardtop, hardwired, hardwood, hardworking, hardy.

Some 30 venerable examples attest to the popularity of 'warewords' -- agateware, barware, brassware, chinaware, clayware, cogware, cookware, copperware, crackleware, dinnerware, earthenware, enamelware, flatware, glassware, graniteware, greenware, henware, hollowware, honeyware, ironware, kitchenware, lacquerware, lusterware, metalware, redware, seaware, silverware, slipware, stemware, stoneware, tableware, tinware, tupperware, willowware, woodenware.

The most colorful 'warewords' are all derived from 'software': abandonware, adware, beerware, bloatware, brochureware, careware, crippleware, firmware, freeware, groupware, malware, middleware, nagware, netware, postcardware, shareware, shovelware, spyware, vaporware, wetware.  The following sentence has the exclamation point that belongs at the end of the previous sentence.  For a misnomer, that's a fabulous foundation for creative coinage!

{Return to citation in Part 7}

The computer I used for teaching maintenance was the prototype RW-300. The machine had a weakness that became well-known to me -- a thermal run-away in one of the germanium transistors on a small printed circuit board called the "C-register read amplifier." Do I have a memory or what! I did not bother to fix the thing. Instead, every few days the computer would halt, and I took advantage of the opportunity to guide my students through a trouble-shooting procedure that invariably wound up with pulling out the board, feeling the hot transistor, waving it around to cool it off and plugging it back in again. Problem solved.

That computer was also used by programmers for debugging process control systems. They were instructed to call me whenever the computer stopped. Soon a necktied figure would be seen running into the room, pulling out some printed circuit board, touching one of its components, waving it around, plugging it back in, and running back out again. My stratospheric reputation, though undeserved, was nevertheless thoroughly enjoyed, not to say exploited.

Our first magnetic tape machine came with a "degausser," literally a black box with nothing more than a switch on it. To restore a tape to its pristine, unmagnetized state, one simply place the reel on top of the box, turned it on and back off again. An oscillating magnetic field, which radiated from inside the box, did the work.

One day, when I was sure a couple of programmers were watching, I carried out a more elaborate procedure. I put a reel on the box and turned it on. I turned the reel first clockwise then counter-clockwise. Next, I slowly lifted the reel with both hands upward, higher and higher, then over my head, turning gracefully in a circle while stepping away in a solemn pirouette. Finally, I placed the tape on a nearby table and reached back and turned off the degausser. "Gradually shrinks the hysteresis loop," I muttered. "Eliminates glitches."

You surely know the rest of that story. The degaussing dance was faithfully performed for years thereafter by succeeding generations of programmers, and then, and then... Witnesses are plentiful who will testify that the following incident actually occurred:

Returning to the California offices of TRW from a year of research on air traffic control in New Jersey, I strolled into the computer laboratory. During my absence, the C-register read amplifier on the prototype RW-300 had behaved itself perfectly. A few minutes after my arrival, though, the computer stopped. Tah-dah. A new batch of programmers were waiting to be astonished. I could not let them down. Out of such coincidences, gods are created.

{Return to citation in Part 7}

American Institute of Electrical and Electronic Engineers (AIEE) and Institute of Radio Engineers (IRE) merged to become the Institnte of Electrical and Electronic Engineers (IEEE, pronounced I-triple-E).
{Return to citation in Part 7}

There must have been more than seven, come to think of it, since any list at about that time would surely have included Alwac, Bendix, Burroughs, Control Data, Data General, Digital Equipment Corporation, General Electric, Honeywell, National Cash Register, Radio Corporation of America, Remington Rand Univac, Scientific Data Systems, all aspiring to challenge IBM in computers -- each, as we used to say, "like a lecherous mouse climbing the hindleg of an elephant with a smirk on his face."
{Return to citation in Part 7}

In American Mathematical Monthly (Vol. 72, No. 2, Part 2: Computers and Computing. Feb., 1965, pp. 8-14), R. D. Richtmyer wrote about the emerging utilities that were necessarily being provided by hardware purveyors, invoking thereby a narrow, not to say parochial definition of 'software.'
 
The compiler is usually embedded in an elaborate system for using the computer, which sequences the problems through the computer, one after another, and supervises them in various ways. This system, together with the library of subroutines, routines for control of input and output formats, and so on, has come to be called software in the industry (as opposed to hardware); a manufacturer is expected to supply a considerable amount of software with the machines he sells.

A curious phenomenon that has accompanied the development of software is a tendency for the hardware to become dependent on it. The amount and complexity of the software have increased enormously in the last few years, and its preparation has become a rather large industry.
 

Thus, from the perspective of a mathematician in the middle '60s, computer hardware depended on a class of programming -- language processing and operating systems that provided support services for the programmer and called that class of non-application programming 'software,' whereas in a broader context, computer hardware had from the earliest days been called upon to execute software-intensive applications -- science, business, military -- without the benefit of that class of non-application programming.
{Return to citation in Part 7}

Throughout the seventies, Xerox was perceived as one of the two best run enterprises on the planet, the other being IBM. One could not help wondering if the mere mortals running the company read those rhapsodic pieces each day about themselves in The Wallstreet Journal and Forbes, Barrons and Business Week and asked, "Who me?"  At corporate headquarters of The Grand Old Duplicator Company, I can tell you, self-doubt was nowhere in evidence, and in its place there was plenty of unmerited self-approval.

Perhaps I might have been more discrete, but to my close friends I described Xerox top managers as grim-faced officers standing in the bridge of a venerable vessel, steadily plowing the commercial ocean toward some unseen Valhalla in the distance, each in his sharply pressed uniform ornamented with gleeming brass and rows of ribbons, their caps and collars festooned with scrambled eggs, all eyes gazing at the horizon, hands confidently operating levers and wheels, each unaware that the controls were not connected to anything, that instead the equipment in the bowels of the ship had been welded firmly in place, and that a unique technological marvel -- a xerographic engine, protected by "the thicket of patents" -- would assure more than enough financial propulsion to conceal their own buffoonery.

What good are metaphors if you don't mix them?

{Return to citation in Part 7}

---


 

 
The first giant brain equipped with a Teletype was the MANIAC.  In a secret demonstration for a group of high-level Pentagon officials, a four-star general was invited to ask the machine a question.

With two fingers, the general pecked the keys solemnly, "WILL THERE BE A THIRD WORLD WAR?"

For a full minute, the MANIAC made whirring sounds, its panel lights flickering, then typed one word, "YES." 

The general and his entourage gasped.  He leaned over the keyboard, frowning.  "YES, WHAT?" he pounded.

Taking longer this time, the MANIAC made more whirring sounds, its panel lights flickering wildly, then typed, "YES, SIR!" 

{Return to citation in Part 7}

---

 
Lost from my collection of satirical cartoons is one from 1954 in which the scene was a destroyed city.  Smoldering ruins are strewn as far as the eye can see.  Rubble and broken bricks are spread out in the foreground alongside smashed cars and toppled utility poles.  Twisted steel girders are all that remains of a tall building, with a bathtub and a toilet held aloft by their plumbing.  A public address speaker can be seen dangling by its wires.  The caption reads, "All clear."
{Return to citation in Part 5}
Another is one from 1974 featuring a row of vending machines.  A sign in the background reads "Baggage Claim," establishing the setting as a transportation terminal.  A terse placard appears on one of the machines: "CHANGE."  On the floor in the foreground atop an open paint can droops a brush dripping red paint.  Defacing the machine in crude, red lettering is a forlorn battle-cry, "CONTINUITY!"