John Searle considered the question of what a "computer" is to be vacuous and that anything can be given a computational interpretation (Searle articulated this in the context of computational neuroscience, believing that proposition that "the brain is a computer" was not even wrong).[1] There is, in fact, very little to define what a computer is and little agreement among philosophers.[2]
Brunnock (talk · contribs) has claimed that [m]emory aids such as abaci, tally sticks, chalkboards, and paper are not capable of computation and therefore not computing hardware
and that [a]lgorithms and mathematical concepts or techniques are not hardware
, but I am not so sure that either of these are tenable. I'm not sure what Bubba73 (talk · contribs) originally was looking at when they stated that certain items were not hardware, but I assume both Brunnock and Bubb73 were talking about similar items.
Michael Williams includes knotted cords, tally sticks, finger reckoning, abaci, and other such similar examples in his history of computing technology. Likewise, Arno Borst discusses abaci and other similar technologies in his history, which also gives a lovely description of the ancient and medieval computus. Note that our page on the history of computing hardware discusses everything from the Lebombo bone to Greek combination locks, though of course WP:OTHERCONTENT applies. Also, I'm not sure that there is such a stark difference between an abacus and Napier's bones. Even slide rules don't seem that much different and nomograms, which are entirely diagrammatic, work on the exact same principles.
While it is harder to defend algorithms and mathematical techniques as "hardware", there is a significant grey area here, and from the standpoint of diagrammatics, I think there is an argument to be made that certain tables and graphs can be called "hardware" if only because their materiality is central to their operation.[8] For example, volvelles (not mentioned on this page) are accepted as early computers,[9] but the earliest example in Europe actually only uses rotation to expedite the calculations,a so that they can be done without moving the manuscript or the reader's body to read upside-down text.[10] Does this make it computing hardware? What about slightly later volvelles, like that made by the same scholar (Matthew Paris)b in MS. Digby 46, which is practically just a roulette wheel?[11] Further, is it correct to excise mnemonics and heuristics if they have had significant influence on computing hardware? For example, Ramon Llull's figuraec are something between traditional computer hardware and mnemonics, but heavily influenced "true" analog computing technology through Leibniz and Kircher.[12][13] I don't believe a timeline which ignores these precursors is complete, and it might even be misleading.
I'm overly caffinated and this is way too long for such a minor page, so to summarize:
- It is difficult, maybe impossible, to define what a "computer" is in anything but vague or vacuous terms;
- Scholarly histories consider memory aids and other technologies as forms of computing hardware;
- Diagrams and other visual devices blur the line between "hardware" and "software" (or whatever you want to call abstract machines); and
- It is improper to excise certain memory aids or algorithms from a timeline of computing hardware because of their disproportionate influence.
My argument here is less that we should broaden our definitions of "hardware" and moreso that a hardware/software distinction is not useful for historical or chronological purposes. Why is this page a Timeline of computing hardware before 1950
anyway, instead of Timeline of computer science before 1950? Michael Hardy (talk · contribs) moved the page all the way back in 2010, based on the state of the page back then, trying to reconcile the page's contents rather than trying to cut down on irrelevant entries. In fact he writes, "A century ago computers were people who used pencils and paper. How that was done is part of the history of computing.
"
It seems to me that a return to the original page, assuming that each event is properly sourced within scholarly histories of computer science to prevent it from becoming too broad, would be better. There is already a Timeline of computational mathematics, which strangely begins in the 1940s; and a Timeline of scientific computing, neither of which discuss e.g. Church–Turing thesis. Note also that Timeline of computing links here as its earliest page, and that none of the other pages make reference to "hardware". In fact some of the most important events on this page, like George Boole's algebra, Kurt Gödel's fundamental result of theoretical computer science and the Turing machine (!), have little if nothing to do with hardware except their influence.
- Notes
- ^aThis is technically WP:SYNTH and I'm only using it here for the sake of argument. Don't incorporate this into the article until I (or another scholar) make this claim in secondary literature.
- ^bTechnically this is a faithful medieval copy of Bodelian Library, MS. Ashmole 304, done in Paris' hand, which is unfortunately mutilated.
- ^cThe state of Llullian scholarship in English is abysmal, largely due to Martin Gardner's massively incorrect and vitriolic treatment of Llull in his Logic Machines and Diagrams. See Bonner 2007, pp. 303–305 for a description of the "Martin Gardner problem": "for every non-medievalist who tackles Llull from the angle of the history of logic, mathematics, or combinatorics [this chapter] would seem to constitute their first introduction to the subject, one which colors most of their subsequent thinking thereon". Unfortunately many writers now make very inaccurate claims about what Llull's figures actually did. Indeed this page says that Llull invented something called the
Lullian Circle
, which is a notional machine
. Pace Borges, Llull never described any of his figures as "machines" nor did he invent any single device called a "Lullian Circle". In the Ars brevis, his most influential work and probably the only genuine Llullian work that Leibniz ever read, precisely one figure rotates, and the same results are also rendered as tables.[14] While not exactly mnemonics, Llull made it explicit that a user was expected to memorize their operation and thus were not entirely mechanical calculators either.[15]
- Bibliography
References
E.g., Llull 1994, p. 304: "So much for the four figures, which the artist must learn by heart, for without them he cannot use or practice this Art."
wound theology◈ 09:51, 15 May 2026 (UTC)
- Also, I am working on a new draft here that goes beyond hardware, which someone should have done in 2010. wound theology◈ 05:29, 16 May 2026 (UTC)
