The F-word

Over at Whewell’s Ghost, Becky Higgitt has been pondering how far academic historians of science simply come across as a big bunch of spoilsports to general audiences who enjoy popular writing of the Dava Sobel school. We don’t believe in lone geniuses, we scorn heroic stories of upward progress, and we positively abominate endowing past people with present-day motives and beliefs.

Much the same applies to historians of technology. In the hist of tech, though, there’s one common popular-history device which offends on all three of the above points in a very neat and seductive fashion. If you want to cause the maximum of annoyance to a scholarly historian of tech with the minimum of effort, show her a First.

All claims to most of the classic Firsts – the First Computer, the First Television, the First Steam Engine – are a pain in the neck. They divorce inventors from their worlds; they mistake chronology for causation; they make it the historian’s job to praise famous men, rather than to explain why the past was as it was, or how the present came to be as it is. They demand that everything proceeds linearly from one place, one famous man, and one incident; and they crash with monotonous regularity against other, equally threadbare stories which put the impossible starting-point somewhere else.

Of course, we’re constantly told (I have had this message from people in museums, broadcasting, and various points on the journalist-flack spectrum) that our ivory-tower obsessions won’t play to the public. The public, apparently, crave Firsts.

Which is why I punched the air when I saw comment #9 on this piece on the 1951 Ferranti NIMROD machine, sometimes called the “first computer games console”. I’m pretty sure this was not written by a serving academic:

Anytime you have an article about an invention “first” like this, it just ends up being a p*ssing contest between various claimants and wannabe’s anyway, especially if an American tries to claim a first (Europeans seem to have a real inferiority complex about that). It always comes down to definitions and someone is always there to point out Lord Hugh Givezashiz, who invented a mechanical videogame 10 years earlier (but never exhibited it). Then someone counters with the writings and drawings of Sir John Bloohard which proceeded Lord Givezashiz’s work by several years (though Sir Bloohard never actually built it). And so on, ad naseum.

I’m paid a lot of money not to write like that, but he’s saying what I’m thinking. First-talk, far too often, reduces to an annoying game which gets out of the historical record pretty much what it decides to put in. It’s a distraction. Real technical change is gradual, and rich in independent overlapping discoveries. That’s not a fussy academic quibble: it’s a point small children can grasp.

Mind, while many cases certainly involve nationalist chest-beating, I’m not convinced of the commenter’s claim that European status anxiety is the chief offender. The “pissing contest” can kick off anywhere, often within national borders. The most active theatre in the apparently interminable “invention of the computer” conflict, for instance, pitches the Eckert-Mauchly ENIAC design (U Penn) against the Atanasoff-Berry machine (Iowa State). The history of the First Computer question – which involves a patent ruling, some ferociously partisan internal history, a sponsored TV documentary and a Pulitzer-winning novelist – is a great deal more interesting than the question itself. Who invented the computer? Loads of people. It’s complicated. It’s a computer.

There are, of course, First Computer contenders from Europe, including Manchester’s 1948 SSEM prototype. The greatest demonstration I have ever seen of Firstism as a rhetorical battering ram appears in an off-the-cuff 1998 interview with the legendary Tony Wilson, who was a moving figure in the machine’s fiftieth-anniversary commemorations. Characteristically, his take on the matter is refreshing, incredibly charismatic, and top-to-bottom wrong.

The great shame is that interesting stories could be told about most of this technology without dragging First-talk into it at all. The SSEM is a great example; the NIMROD machine, mentioned above, is another. Built for the 1951 Festival of Britain, it was a vast electronic toy designed purely for visual impact. The original plan had been to display a fully functioning Ferranti computer, but this proved impossible in the time available: NIMROD was the most impressive single-purpose machine that the engineers knew they could rig up in a hurry.

The salesmen’s real aim, though, was to get visitors to understand the nature of the general-purpose computers Ferranti was trying to market. The awkward compromise they came up with was to sell a booklet at the exhibition which explained the difference between NIMROD and a ‘real’ computer. This indirect approach turned out not to have much effect: Vivian Bowden of Ferranti had more luck when he worked the material up into a book-length treatment, Faster Than Thought, which became the first widely read introduction to computers in Britain. Nobody, in this period, was trying to “invent the computer game”. They were trying to get some say in defining what a computer was.

The F-word is not always obscene. Sputnik 1was the First Artificial Satellite in a sense that matters. Even here, it would be naïve to conclude anything about the relative state of US and Soviet space technology from the fact of that First. But the fact of the First sent a vast number of individuals and a great deal of money down paths they otherwise wouldn’t have gone down. “The first satellite was Soviet” is a statement which explains things; “the first games console was built in 1951” isn’t.

First-mindedness is difficult to fight. It seems to have worked its way into a lot of scripts for institutional behaviour. The effect is sometimes to frustrate communication at the most basic level. Far from making things simple, it can make them far too complicated.

A museum-based colleague contacted me recently about a planned display on some early computer designs. The proposed text focused on one particular model as having introduced something called the “general register set”, an approach to machine architecture used in most modern computers. This is the kind of concept whose origins may be interesting to informed computer scientists; for the intended audience (families in general), though, it’s impossibly abstruse. So why mention it? Because it’s the best available technological First associated with that machine. The people responsible for marketing, apparently, need the description to include some reference to “ground-breaking” novelty, as this increases the chances that journalists will pick it up. I suspect they are really going to have their work cut out getting the Birth Of The General Register Set into the Metro.

But it doesn’t necessarily take a First to change the world. Indeed, sometimes it takes anything but. The point of releasing a machine without significant technical novelty is usually that it’s a safe, affordable proposition – the kind of thing, in other words, that people might buy and use a lot. In computing, the extreme example of this is the IBM PC 5150, cobbled together very quickly from not-very-powerful bits and pieces other firms had been using for years. The people who designed it knew exactly what they were doing. The “original PC” sold by the bucketload, was cloned by everyone in sight, and stands as a rare example of a design which really did have an identifiable, single-handed directing influence on an entire technological category.

As David Edgerton has pointed out, much hist of tech privileges whatever was considered “new”, at any given time, over what was actually having an effect: too many Concordes, not enough trishaws. This enthusiastic literature, he laments, is written “for boys of all ages”. When we consider whether our academic sensibilities address the needs of “the public”, it’s as well to remember that there are sections of that public who don’t feel the schoolboy’s fascination, who see Firstist history as a rather pointless succession of interchangeable Givezashizzes and Bloohards.

This entry was posted in Uncategorized. Bookmark the permalink.

12 Responses to The F-word

  1. Thony C. says:

    But James everybody knows that the computer was invented by Turing oops Zuse oops Babbage…

    Your tittle should be The Two F-words, “Father of” is just as obnoxious as “First”.

    • James says:

      Ah, yes, the “Father of the Computer”. I should probably have been clearer in separating Firstism, which at least allows the possibility of a group effort (sometimes), from the “Who invented…” approach, which doesn’t.

      It is possible to get some degree of nuance into an answer to the “Who invented…” question: Andrew Hodges does this in nominating Turing. But I don’t think it’s a wise question. (Interestingly, the orthodox view in the University of Manchester’s Department — latterly School — of Computer Science, until very recently, was that Turing had nothing at all to do with inventing the computer, because the institutional culture encouraged viewing “invention” as almost purely a question of engineering.)

    • Thony C. says:

      James I haven’t had time to read all of the Andrew Hodges’ page that you linked to but I’ve bookmarked it for later. However I did notice that he dismisses Babbage’s claim by stating that the Analytical Engine was not a stored programme computer.

      I found this very interesting as my friend Günther Görz who is professor for informatic here in Erlangen and a historian of science currently serving as guest professor at the Max Planck Institute for the History of Science in Berlin says that the Analytical Engine was a stored programme computer!

      Any thoughts on the subject?

      • James says:

        I see both claims as a bit odd, because “stored-program” doesn’t have any established widely accepted meaning for the pre-electronic age.

        The term is usually used as a shorthand for “electronic digital stored-program.” Some of the 1940s computers stored the data they were working on electronically, but had the programs fixed as wired circuits. The point of “stored-program” machines like the SSEM and EDVAC was that they stored the programs electronically too, so didn’t need to be rewired for new tasks.

        In the most literal sense, all programmed devices store their programs: if they didn’t, well… the programs wouldn’t be there… The literal definition doesn’t do anything useful.

        If, on the other hand, you extrapolate the real-world usage backwards as best you can, you may decide that any machine is “stored-program” if it stores its code and data in the same way. This certainly excludes the Analytical Engine (data on cogs, code on cards). But the definition, again, doesn’t capture anything particularly useful. Nobody wanted or tried to build an all-card or all-cog machine.

  2. Rebekah Higgitt says:

    Nice post James! This is an interesting one to think about from a museums point of view, of course. One of the problems is that prestigious objects in collections have often been kept, conserved, researched and displayed because of good and exciting provenance, and because they are firsts or earliest known examples etc. It would often be very difficult to get these off display and to replace them with what was actually being used. Perhaps placing both together would be a reasonable and interesting option?

    There is also the question, especially for smaller or specialist museums, of just what one *can* say. If talking about the first use of ‘general register set’ is too complex to mean anything, what did you suggest for that particular label?

    • James says:

      I’m not a museum insider, of course, but I seem more often to deal with kinds of artefact (usually machinery or electronics) where the reverse is the case. It’s the production models which survive, usually having been donated following clear-out of a defunct site where they were in regular use. Prototypes, on the other hand, often don’t survive, because their heritage significance only develops in retrospect.

      The SSEM (aka the “Manchester Baby”) is a good example. The machine barely had any lifetime in the form in which it famously first functioned: by the time anybody thought to photograph or publicise it, it was about twice its original size and had been redesignated as the Mark 1. It was scrapped in 1959 (see Jon Agar’s article on this, paywalled).

      The machine with the first general register set was the Ferranti Pegasus. Around 40 were built, of which one — held by the Science Museum — is in working order.

      In fairness, the particular museum project I’m talking about is a temporary display based around documentation, rather than actual machines — and most of the documentation is by nature technical; there will also be a discussion session with a friendly computer scientist, which might give better scope to discuss why the architectural details matter. As a first approach to promoting general interest in Pegasus, though, I’d definitely focus on how it was used. It was intended, by the standards of its time, as a reliable, medium-sized machine to attract commercial sales — cutting-edge fast computing for military and scientific use was specifically the fiefdom of other divisions of Ferranti — and found uses in banking, aviation, the steel industry, and so forth. Plenty of possible stories there about industrial change and its effects on everyday life.

  3. Beto Pimentel says:

    This is closely related to another common disease that affects some texts on history of science that guys here call “precursoritis”: an incredibly uncontrollable impulse to find “precursors” to virtually anything.

  4. Pingback: Geek mythology and Nikola Tesla « News in Briefs

  5. Pingback: Nobody invented the scientific method. | The Renaissance Mathematicus

  6. Pingback: Nobody invented the scientific method | Whewell's Ghost

  7. Pingback: Geek mythology and Nikola Tesla | teleskopos

  8. Pingback: The “first” Atlas | The Renaissance Mathematicus

Leave a Reply

Your email address will not be published. Required fields are marked *