Criticism of von Neumann historiography
Dear SIGCIS researchers, Dear SIGCIS researchers, In my lecture at the ICOHTEC annual meeting in St. Etienne, I point to the dominance of the subject of the Neumann Computer in the historiography and show on the basis of numerous sources that in the Californian cluster of the aircraft industry, in addition to the IAS in Princeton, a second birthplace of the computing machines (mainly with analog computers) was created after 1944, which was previously considered by the research little. The aircraft cluster found no use for the digital von Neumann machine. Since von Neumann's machine was merely academic and was not used in industry, John von Neumann had to invent fields of application in the future as the legitimacy of his project. He opted for meteorology - although it is not clear that analog computers could not offer solutions here as well. Until 1960, the race between analog and digital computers in aircraft and missile industry was not decided in favor of the latter; in 1961, NASA commissioned a large general-purpose analog computer for the Saturn V project. My thesis of the aviation industry as the source of computing machines after 1944 can also be applied to other countries with strong aircraft industries, such as England and the Soviet Union. It was also no coincidence that Zuse‘s Z3, the first electric digital computer in Germany, was created in 1941 in the context of the Berlin aircraft industry. With kind regards from Berlin Richard Vahrenkamp -- ******************************************** Prof. Dr. Richard Vahrenkamp Logistik Consulting Berlin Phone 0177- 628 3325 E-Mail: Vahrenkamp2016@gmx.de Web: www.vahrenkamp.org Trendelenburgstr. 16 14057 Berlin *********************************************
Hello Richard, Your message seems to me to be asserting three potential distinguishable things. * You argue that digital computers were worse than analog ones for industrial problems * You argue that the contribution of von Neumann and the IAS computer was important only in academic areas * You argue that the Californian aerospace industry was a “second birthplace of computing machines,” challenging the East Coast focus of conventional historiography. Those are all big and bold claims to make, so for a short presentation I would suggest focusing on just one of them. Particularly as they all seem to me rather questionable. Analog computers have not been entirely overlooked, particularly in the book The Analogue Alternative by James S. Small but also, for example, in work by David Hemmendinger, Charles Care, etc. So the SIGCIS community is well aware that analog computers were sold commercially well into the 1960s and had significant advantages over digital machines for some applications. The application that von Neumann had in mind when coming up with the First Draft design for EDVAC was, more than anything else, hydrodynamic calculations for atomic weapon design. Not an entirely academic pastime, and Los Alamos was one of the major users of digital compute time in the second half of the 1940s. As for industrial use, digital computers could take on jobs already being performed with punched cards, as a result of which by the mid-1950s the balance of computer user was already shifting from scientific and technical use to business administration. Analog computers could not do clerical or accounting work. Your statement that “The aircraft cluster found no use for the digital von Neumann machine” is just not true. Aerospace firms used analog computers, but they were also the single most important early adopters of digital computers. Several were clamoring for ENIAC time, they commissioned BINAC (one of the very first commercially supplied digital computers), and the prominent Southern Californian cluster of aerospace users of IBM’s 701 (originally called the Defense Calculator) formed PACT, which later developed into SHARE. Atsushi Akera has written about this. The story of the CPC is also well known. BRL, which commissioned ENIAC, was not an aircraft firm but it was concerned with things that fly through the air and with the flow of air around objects (including wings) at supersonic speeds. I think the aerospace industry is already reasonably prominent in the literature on computing in the 1940s and 1950s. Thus my advice is to focus your argument on one of the areas, and to make clearer how your contribution builds on and extends existing work in the area rather than implying that these topics have previously been ignored. Best wishes, Tom From: Members [mailto:members-bounces@lists.sigcis.org] On Behalf Of Richard Vahrenkamp Sent: Friday, July 06, 2018 1:55 AM To: members@sigcis.org Subject: [SIGCIS-Members] Criticism of von Neumann historiography Dear SIGCIS researchers, Dear SIGCIS researchers, In my lecture at the ICOHTEC annual meeting in St. Etienne, I point to the dominance of the subject of the Neumann Computer in the historiography and show on the basis of numerous sources that in the Californian cluster of the aircraft industry, in addition to the IAS in Princeton, a second birthplace of the computing machines (mainly with analog computers) was created after 1944, which was previously considered by the research little. The aircraft cluster found no use for the digital von Neumann machine. Since von Neumann's machine was merely academic and was not used in industry, John von Neumann had to invent fields of application in the future as the legitimacy of his project. He opted for meteorology - although it is not clear that analog computers could not offer solutions here as well. Until 1960, the race between analog and digital computers in aircraft and missile industry was not decided in favor of the latter; in 1961, NASA commissioned a large general-purpose analog computer for the Saturn V project. My thesis of the aviation industry as the source of computing machines after 1944 can also be applied to other countries with strong aircraft industries, such as England and the Soviet Union. It was also no coincidence that Zuse‘s Z3, the first electric digital computer in Germany, was created in 1941 in the context of the Berlin aircraft industry. With kind regards from Berlin Richard Vahrenkamp -- ******************************************** Prof. Dr. Richard Vahrenkamp Logistik Consulting Berlin Phone 0177- 628 3325 E-Mail: Vahrenkamp2016@gmx.de <mailto:Vahrenkamp2016@gmx.de> Web: www.vahrenkamp.org <http://www.vahrenkamp.org> Trendelenburgstr. 16 14057 Berlin *********************************************
Dear Richard, The (comparatively) explicit emphasis on the analog side of computing in California after World War II is indeed well documented (see message by Tom Haigh). As I understand it, this is supporting rather than weakening your point. This Californian emphasis actually has a much deeper history (which prepares for and explains it). A history that goes back to the interwar period and even earlier. I had the opportunity to argue so in Aristotle Tympas, ‘A Deep Tradition of Computing Technology: Calculating Electrification in the American West’, in Where Minds and Matters Meet: Technology in California and the West, Volker Janssen (editor), University of California Press, Oakland, CA, 2012, 71-101 (if you are interested, I would gladly provide you with an electronic copy). The comparison between interwar (and even earlier) state-of-the-art computing in California and state-of-the-art computing in the East Coast (available through the GE-MIT partnership, and exemplified by Vannevar Bush), is, in my opinion, very suggestive. Because it shows that the Von Neumann approach to computing in the post-World War II context, which relied on decontextualized comparisons that introduced a version of computing as supposedly independent of computing analogy, is similar to the way electrical engineer Bush described computing in comparison to interwar Californian electrical engineers who challenged the way he (Bush) had presented his artificial lines (the state-of-the-art computing artifact at the time). The challenge by Californian engineers to the rhetoric of universalism advanced by well known masters (Neumann and before him Bush) suggests that locality is crucial to the history (and therefore, also, historiography) of computing. Local histories of computing in use are needed to balance a historiography based on a rhetoric based on abstract laboratory computing epistemologies. These local histories seem to offer a direct criticism to what you are referring to as the "Von Neumann historiography". Aristotle Dear SIGCIS researchers, Dear SIGCIS researchers, In my lecture at the ICOHTEC annual meeting in St. Etienne, I point to the dominance of the subject of the Neumann Computer in the historiography and show on the basis of numerous sources that in the Californian cluster of the aircraft industry, in addition to the IAS in Princeton, a second birthplace of the computing machines (mainly with analog computers) was created after 1944, which was previously considered by the research little. The aircraft cluster found no use for the digital von Neumann machine. Since von Neumann's machine was merely academic and was not used in industry, John von Neumann had to invent fields of application in the future as the legitimacy of his project. He opted for meteorology - although it is not clear that analog computers could not offer solutions here as well. Until 1960, the race between analog and digital computers in aircraft and missile industry was not decided in favor of the latter; in 1961, NASA commissioned a large general-purpose analog computer for the Saturn V project. My thesis of the aviation industry as the source of computing machines after 1944 can also be applied to other countries with strong aircraft industries, such as England and the Soviet Union. It was also no coincidence that Zuse‘s Z3, the first electric digital computer in Germany, was created in 1941 in the context of the Berlin aircraft industry. With kind regards from Berlin Richard Vahrenkamp -- ******************************************** Prof. Dr. Richard Vahrenkamp Logistik Consulting Berlin Phone 0177- 628 3325 E-Mail: Vahrenkamp2016@gmx.de Web: www.vahrenkamp.org Trendelenburgstr. 16 14057 Berlin ********************************************* -- Aristotle Tympas Associate Professor Department of Philosophy and History of Science School of Science National and Kapodistrian University of Athens tympas@phs.uoa.gr http://scholar.uoa.gr/tympas
participants (3)
-
Aristotle Tympas -
Richard Vahrenkamp -
Thomas Haigh