Differential Equations and Department Stores

Posted on by John

Howard Aiken on the uses of computers, 1955:

If it should turn out that the basic logics of a machine designed for the numerical solution of differential equations coincide with the basic logics of a machine intended to make bills for a department store, I would regard this as the most amazing coincidence I have ever encountered.

Update: Some people have read the quote above and thought Aiken was ignorant of the work of Turing et al. I assumed he was speaking in terms of what was practical rather than what was possible, which is apparently correct.

Thanks to Anatoly Vorobey in the comments below, I found a paper that goes into more background. From that paper:

Aiken’s theme in the lecture … was that a machine designed primarily for scientific use was far from ideal for business computing. … For example, scientific computing (Aiken pointed out) involves relatively small amounts of data and complex processing, whereas business computing involves large amounts of data and relatively shallow processing.

4 thoughts on “Differential Equations and Department Stores”

  1. Andrew Dalke

    Could you provide more context? I only find secondary sources, eg, “Howard Aiken, quoted in Davis 1988a p152”, where Davis quotes Ceruzzi p43 – and I can’t easily get a copy of Ceruzzi.

    I ask because most places which reference the quote say something like “haha, it’s silly Aiken didn’t know about universal computing machines” (paraphrasing https://www.turing.org.uk/scrapbook/computer.html ). Or at https://ulud.pw/1622354943.pdf “Aiken was terribly mistaken …. cautious and unimaginative”.

    But I don’t see how that interpretation makes sense. In 1956, LEO – the first business computer – had already been in use for years, applied to tasks not much different than those needed for a department store. See https://en.wikipedia.org/wiki/LEO_(computer)

    I don’t think we can assume that Aiken didn’t know that business computers already existed. Instead, perhaps it’s more an engineering commentary about the traditional separation between a “business computer” (like the IBM 702) and a “scientific computer” (like the IBM 701), both released before 1956?

    For example, pre-decimalization British business computers could have special hardware for computing pounds, shillings and pence, which wouldn’t found in a machine made to find numerical solutions of differential equations. The latter could emulate the former, but the “basic logics” wouldn’t “coincide”.

  2. For years after the quote, certain computers were being offered and delivered with optional instruction-set extensions for “scientific” or “business” computing. Aiken was thinking at the level of abstraction at which, for example, COBOL and FORTRAN are miserable solutions for each other’s problems. One level down from that is instruction sets, which had not yet converged; and another level down is gates, which may be the first things that spring to our luxurious minds today when we read a quote form someone so long ago that the phrase “basic logic” had an evocative meaning.

  3. Anatoly Vorobey

    @Frank: this article has the complete story behind the quotation. Sci-hub has it.

    IEEE Annals of the History of Computing
    B. Jack Copeland
    “Unfair to Aiken”
    October-December 2004, pp. 35-37, vol. 26
    DOI: 10.1109/MAHC.2004.36

Comments are closed.