Appendix C -- Selected Correspondence
Copyright ©2006 by Paul Niquette, all rights reserved.

The Verb "Sofware"

Excerpts from e-mail messages in 2000 from Emil Borgers, an eminent figure in programming from the fifties onward...
Borgers: Programmers in 1957 were never called 'software engineers' and I corrected anyone who tried. I remember being insulted by its use --- an annoyance really. 

The common phrase 'to software around it' was how the term was used the most, much to my dismay -- basically, by engineers who had no more room for components (or money in the design budget) and wanted to slip the problem into someone else's domain (mine). Today, with processing speed as it is, 'software-arounding' things is normal but in those days it was not often a wise choice. 

The "Ralliac"

In 2006, another colleague from the fifties responded to my request for recollections about adventures with early computers.  Bill Paine was the programmer who wrote the bit-by-bit simulator for the RW-300 on the Univac 1103A and then became a regional sales manager for TRWP.
Paine: In the sifting through the shards of early 'computerdom' a modern archeologist finds...a lot of dust.  There must be a better use of time in one's final remaining years!  Churchill wrote his History of the English Speaking People in his later years. 
PN: How much do you recall about 1957 and the first international sports car rally in the U.S. -- and the "Ralliac"?
Paine: The idea for entering the rally was mine, and I invited you to be my partner.  The Ralliac was conceived by you when I explained the need for mastering on-board real-time calculations required in a sports car rally.  The main goal was to finish an initially unknown course, hitting unknown check points at the exact times specified by the rally designers.  Therefore it was critically important that the rally team (driver and navigator) have accurate real-time information on time and distance traveled. 

The Ralliac was designed to produce the critical mileage information using pulses generated by a magnet on a wheel passing the 'read head' mounted closed by.  Precise timing would be supplied by clock circuits within the Ralliac.

Sports car rallies were international events in those days.  They often lasted 24 hours and represented a total distance to be traveled of many hundreds of miles.  There would be dozens of contestants.  Counting the planning and checkpoint staffing for the rally, there must have been many other people spending a great deal of time in addition to the rally contestants. 

Our planning for the first application of the Ralliac included its installation in a new car donated by a local dealer...

PN: As the "hardware" guy, my assignment called for me to provide a demonstration at a Beverly Hills luncheon sponsored by a certain Renault dealer.  All I had at that time was one printed circuit board.  It was populated with advanced electronic components like diodes and transistors and was just small enough to fit in your jacket pocket. 

In his introduction to your presentation, the owner of the dealership exulted about cars ("the era of the small car is upon us") and about new electronic technologies (using the not uncommon mispronunciation "digitile computors"). At the appropriate moment in your presentation, you passed the board around the table.  That was all we needed.  The subject of your "software" never got mentioned.  The same afternoon, we were handed the keys to a brand new sedan.

Paine: We never entered the rally due to circumstances beyond our control, as I recall.
PN: Right you are, Bill.  The rally was suddenly cancelled by the sponsor.  European contestants were caught in mid-ocean with their sports cars covered with cosmoline. 

After several week-ends of frantic soldering and coding, you and I suffered the instant obsolescence of the Ralliac.  And I shall never forget the day we had to return the no longer new Renault to the dealer.

The Confusing Term: "Programming"

Exerpt from e-mail messages in 2006 from Stu Schy, colleague at TRW Computers, commenting on passage in Part 2: "The word 'programming' was serviceable enough, of course.  Renaming it 'software' did more to assure my reputation as an eccentric than to illuminate computer technology."
Schy: In the early 1960s, a group of us from TRW Computers attended a meeting at CBS headquarters in New York prior to the contract signing for an automatic switching system for the CBS network using the RW-300.  There were a number of CBS VPs there.  One of the CBS people asked "What do we do if we change from a CBS type fade to an NBC type fade?"  This referred to how running overtime was handled.  I answered, "We can handle that with a small change in the program (meaning the timing subroutine.)  At which, one of the men at one end of the table took the cigar out of his mouth, stood up, pounded his fist on the table and said, "No one is messing around with my programming!" 

He was the CBS VP for Programming.  This was one of a number of incidents that taught us the need to hire people from the industries we hoped to serve if we wanted to speak their language.  We did get the contract.  Some time later, a CBS technician said to me "Hey, you bastard, you cost me my job!  Now, how about giving me one?"  I said, "That's fair." and we hired him.  After a few months, he used the expertise he learned at TRWC to get a much better job at CBS as Manager of Computer Operations.

My Mentor: Dan Gerlough

While conducting my researches for this memoir, I received an e-mail message on March 24, 2001 from Walter F. Bauer, a renown figure in the early history of computers.  His recollections about Dan Gerlough stirred up my own memories.  I replied the same day.  Let this exchange serve as a tribute to Dan and to his enduring contributions to traffic engineering.
Bauer: In reading [the draft], I couldn't help recalling my contacts with Dan Gerlough.  In 1960 Dean Wooldridge [the W in TRW] got the idea to start a traffic control business (project).  I was chosen to start it.
PN: Coincidence Alert: During that time period, Walt, I was also with TRW, on a consulting assignment in air traffic control at the fledgling FAA at the National Aviation Facility Experimental Center (NAFEC) in Atlantic City, so I did not know about the [vehicular traffic control] initiative.
Bauer: The idea was to apply electronics and modern analysis to the emerging traffic problem, especially the freeways. I got a basic understanding of what was going on in traffic control technology, travelling around the country and talking to the cognoscenti.
PN: Ironically, just last month [April, 2001], I attended a "brown-bag" luncheon in San Francicso put on by one of my colleagues at Booz Allen Hamilton.  Our firm has a consulting contract with Caltrans and other highway jurisdictions for -- well, suffice it to say, I have no doubt that you would find the technical means for characterizing traffic congestion, and the remedies are all familiar enough.  The Law of Unintended Consequences prevails universally.  There is a tangle of conflicting interests that impede solutions, if indeed solutions there be (apart from bicycles, of course).
Bauer:  Not long after starting, I heard about Dan and his project at UCLA. I convinced him to head up the traffic control group in my department at TRW.  We got some prestigious traffic control contracts, and developed a road sensor.
PN: At UCLA's ITTE in 1953, a couple of us, with Dan Gerlough's approval and mentoring, invented an electric wheel sensor, using a rubberized tape with sinuous metal conductors inside, to replace the pneumatic tube contraptions of that time.  We placed them in pairs across traffic lanes, separated by a yard, and, with an electromechanical timing device, we were able to classify speeds, thus solving what I called the "density-lock deception" [flow-rates measured with single-point detectors will decrease with congestion tricking the analysis into thinking congestion is clearing out].  My guess is that your sensor was the "embedded loop"; if so, it is still in use today -- also deployed in pairs, but, alas, often not connected to any recording devices, due to budget constraints.
Bauer:  But our best claim to fame was applying the RW-300 for intersection traffic control. The computer handled some 50 intersections around Dodger Stadium.  It was the very first time a computer had ever been employed for automatic traffic control!  I wonder if any historian would be interested in that.
PN: One quasi-historian is interested; that's for sure.  More coincidence, Walt.  I was involved about a year later with an awkward, non-computerized system at Scantlin for controlling clusters of intersections around the Los Angeles Coliseum.
Bauer: Post Script:  I remember many discussions in 1961 with traffic engineers on the subject of on-ramp metering.  We were ready to install, but they rejected it out of hand.  It wasn't until about 10 years later that on-ramp metering became a reality.  And now it's commonplace.  The idea, as you well know, is that when density passes a critical stage, volume drops off precipitously.
    PN: Yes, I do know about this idea, which has its roots as follows:
  • In mid-1953 some of us at ITTE, with Dan Gerlough's approval and mentoring, did a study of afternoon congestion on the UCLA campus -- a time when departing vehicles should have been diverging and therefore more or less free-flowing.  Our hypothesis was that a goodly (badly?) number of cars parked in lots at the South end travelled through the campus to exit North on Sunset Blvd., and vice versa, North parkers exiting southbound onto Westwood Blvd.  We hired a bunch of "student workers" at $1.25 per hour to dab poster paint onto bumpers of parked cars, color coded by lot.  In the late afternoon, we stood around the campus, clicking our Veeder-Root counters and -- well, the rest is history.  The remedy was -- and is -- the one-way street system on the UCLA campus.
  • In late-1953, some of us at ITTE, with Dan Gerlough's approval and mentoring, did a study of the Hollywood Freeway, which suffered anomalous congestion in the morning.  You will recall that not all freeway entrances were on the right side in those days.  Our hypothesis was that a badly number of commuters came onto the freeway -- on the left side -- at Barham Blvd and upon seeing congestion ahead, proceeded to change lanes to the right in order to get off at the next exit ramp (Silverlake), inflicting thereby an increase in travel time upon every vehicle behind them -- about 1.75 seconds each, as I recall.  Poster paint was out of the question.  Instead, we used Tel-Audograph recorders (magnetic tape? -- what's that?) to record license numbers, along with time codes at Barham and at Silverlake. Then we punched up Hollerith cards for the SWAC and -- well, the rest is history.  The remedy for that time, by the way, was the extreme in vehicle metering: saw-horses across the on-ramp at Barham every morning.
  • In early-1954, I began my work on density lock with Dan Gerlough's approval and mentoring, which culminated in my undergraduate thesis paper in 1955, which was based on controlled experiments on Hildegard Blvd. and won the Deans Award (ahem), entitled "Experiments in Spacing of Stopped Vehicles for Improving Linear Flow."  Hey, I even got to do some crude modeling on the SWAC, using the line printer for a primitive "graphical display" showing platoons of vehicles, each depicted with a two-digit code representing their respective speeds.  Hot stuff, that.
  • In late 1954, I began promoting the idea of "vehicle metering" in speeches and campus forums, repeatedly refuting doubters ("It makes no sense to stop cars getting onto the freeway!") using a slogan, "If you don't stop them before they get on, density lock will make the freeway into a parking lot."  Dan Gerlough bought into my argument, but, like so many other innovations for which we were advocating (seatbelt interlock, radar speed meters, high intensity runway lights), metering signals would take years -- and your imprimatur along with Dan's, apparently -- before acceptance.
Bauer:  Dan Gerlough was one of my favorite people.  I remember well his easy, avuncular manner.
PN: So do I.  At that "brown bag" on traffic congestion last month, Walt, I saw a slide depicting a venerable scattergram -- that plotting of double-valued flow-rate versus speed (a horizontal "V" -- remember?) and got tears in my eyes.

Exchange with Ray Stanish

One student at Los Angeles Tech and later a close friend, Rig Currie, did act on my behalf by writing an e-mail message to another student, Ray Stanish.  Nice try, Rig.
October 27, 2000

Dear Paul;

All the time I worked at TRW, I don't recall hearing the term "software."

I left TRW to become a Professional Public Speaker.  I started out with the talk "Atomic Energy - Peasant Style" but soon thereafter developed one I called "Giant Nincompoops" -- about computers obviously.  I presented that talk off and on for 10 or 15 years, and I don't recall during all that time ever using the term
"software" in my talks, so I couldn't have been aware of it.

So, I have no idea when I became aware of it or the source. 

Sorry, I can't help you, Paul.  I would've been glad to if I could.  I always had high regard for your brains and your 'way-out-in-front knowledge of computers.

It's been a pleasure to have known you and to have learned from you.

Your friend, 

October 29, 2000

Thanks for your message, Ray.

Perhaps you do remember that in 1957, TRWP (pronounced "twerp") was solemnly engaged in launching a new business development that called for marketing unfamiliar technology in a risk-averse business environment --  the earliest applications of small, primitive digital computers ("digitile computors") that would be put into service exerting absolute real-time control over safety-critical systems, including chemical factories, oil refineries, and nuclear power plants.  "Closing the Loop," read the headlines.

    Nothing to kid around about. Soft things need not apply.

Ten years later, Ray, I had the privilege of attending a performance of "Giant Nincompoops" at a management club banquet at Scientific Data Systems.  I sat crimson-faced in the audience as the speaker singled me out for recognition as the pedagogical source (LA Tech, remember?) for the material being lampooned -- but not the comedic inspiration for the lampooning.
And no, I don't think the talk included so much as one utterance of the word 'software,' which for general audiences -- even in 1967 -- would have necessitated a pause for etymology (see "Cultured Laughter").

My own laugh-line in those days was not suitable for general audiences.

Imagine your wife complaining to her mother, "All he ever talks about is software."
My memories are vivid for that time, Ray.  Do you remember the "Million Dollar Order Party"?  And the song you wrote based on IBM's company anthem?
We are loyal and true, Ramo-Woo
For together we stand
To fight Remington Rand,
IBM and Bendix, too.
Other than in my class, there were really not many opportunities for you to have heard me say the word 'software'.  In the middle of my tenure, just before TRWP moved to Beverly Hills, I broke daily contact with you and others and went off to consult for the fledgling FAA in Atlantic City, NJ, applying the RW-300 in -- gasp! -- air traffic control. 

Well, while consulting for the FAA, I also wrote simulation software for the IBM 709 and commented with a straight face...

"After last night, I don't seem to have any more problems with my software."
Being smirked at by IBM programmers in their Florsheims and striped ties kept me from smirking.

Best regards,

Computer Languages (Plural).

Commenting on a draft of Softword, a historian of computer languages at Murdoch University, Diarmuid Pigott wrote:
Pigott: Just read your book on the word "software," and was very excited by it: I shall make sure that everyone learns of your coinage, and correct those who remain in ignorance. 

When ideas are named, as you no doubt realize, they act as seed crystals for inspiration, and only when something has a name can it become the subject of study, thought, and discipline. So if you hadn't come up with the name, all would have been different.  I also think that it was important that it be a neologism because some of the stupid debates about the meaning of "knowledge" and "information" would not have gotten off the ground if the term had not been clearly different (as in "real-information" vs. "computing information"). 

I have spent some considerable time over the last few years trying to work out the origin of the word "language" in the sense of instructions for computers. The word programming is older, obviously coming from optimization and linear programming, but the curious sense of "language" used -- which has gotten all sorts of people into all sorts of knots by making them think of human languages rather than thinking of machine codes and symbologies -- derives from a minor, restricted use that was formerly considered arcane. 

What we know of it is: 

  • Not used by Aitken, Booth, Church, Curry, Goedel, Laning, Lovelace, Post, Turing, Von Neumann, Wilkes. They use "system," "code," "instructions";
  • Prior use of the term to describe closed system of word – mathematical language, engineering language;
  • Use in this sense by Backus 1954 "Can a machine translate a sufficiently rich mathematical language into a sufficiently economical program?";
  • Prior use as system of signs (semiotic use);
  • Used in modern sense by 1958 (AlGOL, IAL, JOVIAL, CODASYL etc.);
  • Standard term for programming system by 1962.
The earliest usage I can find is Edmund Berkeley  1949 "Since it is not always easy to think, men have given much attention to devices for making thinking easier. They have worked out many systems for handling information, which we often call languages. Some languages are very complete and versatile and of great importance. Others cover only a narrow field—such as numbers alone—but in this field they may be remarkably efficient. Just what is a language? Every language is both a scheme for expressing meanings and physical equipment that can be handled."

I would love for you to draw on your rich storehouse of memories and help me unravel this problem! Thanks anyway for your most illuminating work.

The mental exercise to which I was invited by this message proved unavailing, except that I am now all the more sympathetic to the difficulty faced by my friends and family in trying to respond to my queries about the word 'software' (when did you first hear it?  learn its meaning? from whom?)
hus, a lifetime being subjected to the use of the word 'language' in diverse contexts (the language of aviation, culinary language, the language of birds, the language of logic, of love, of storms, medical language, cowboy lingo, diplomatic language) has obliterated my recollections of its earliest application in the realm of computers.

The word 'lan·guage' has made a long but, I suspect, rapid metaphorical journey from its literal origins...

Etymology: Middle English, from Old French langage, from langue, tongue, from Latin lingua. its appropriation for entities-without-tongues.
1a. Communication of thoughts and feelings through a system of arbitrary signals, such as voice sounds, gestures, or written symbols. b. Such a system including its rules for combining its components, such as words. c. Such a system as used by a nation, people, or other distinct community; often contrasted with dialect. 2a. A system of signs, symbols, gestures, or rules used in communicating: the language of algebra. b. Computer Science A system of symbols and rules used for communication with or between computers [emphasis added]. 3. Body language; kinesics. 4. The special vocabulary and usages of a scientific, professional, or other group: “his total mastery of screen language—camera placement, editing—and his handling of actors” (Jack Kroll). 5. A characteristic style of speech or writing: Shakespearean language. 6. A particular manner of expression: profane language; persuasive language. 7. The manner or means of communication between living creatures other than humans: the language of dolphins. 8. Verbal communication as a subject of study. 9. The wording of a legal document or statute as distinct from the spirit.
-- American Heritage Dictionary

Apparently just about any form of information exchange, with or without the use of a literal tongue, can be a language so long as sender and receiver agree on a protocol.  Now, according to my way of thinking, that incudes every form of expression except modern art.  But then, at a typical exhibition, I may be merely a defective receiver surrounded by exceptionally gifted senders using languages that I am unable to fathom.

0 | 1 | 2| 3 | 4 | 5 | 6 | 7 | 8 | 9 
Next Part >>
Home Page
Softword: Table of Contents
Top of Article