Why do the terms digital and therefore digital computer seem to be so problematic? In all literature dealing with computer history, there is a certain indifference concerning the differently scaled function descriptions and terminologies of individual parts or the entire computer and its analogue, discreet or digital properties. “Digital” and “digital computer” was first used by George Robert Stibitz in 1942.” (Ceruzzi/Aspray 1990) So far, no author was able to state a direct source. In his closing report Report on electronic predictors for anti-aircraft fire control from 1942 about a computer-assisted anti-aircraft system Stibitz' pointed out the fundamental advantages of merely having to program and compute two discreet voltage levels.
There is a reference to this technical and historical gap, for example, in the first volume of the Encyclopedia of Computers and Computer History by Raul Rojas from 2001. It should however be possible to find at least a circuit directly succeeding the “trigger relay” from 1919 or a first application. Indeed, an early mention of the onomatopoetic name “flip-flop” can be found in A. T. Starr’s essay “A Trigger Peak Voltmeter Using »Hard« Valves“ from 1935. In April 1920, around six months after the trigger relay essay was published by Eccles and Jordan, Laurence Beddome Turner presented his so-called “Kallirotron, an Aperiodic Negative-Resistance Triode Combination”.
The term “digital computer” (Stibitz 1942) and the “flip-flop” (Turner 1920)
[T]he computer is not one thing but many different things, and the same holds true of computing (Michael Mahoney) and many indeed have struggled to give the computer and computation its historical due.
2nd international conference on the History and Philosophy of Computing (HaPoC), Paris 2013
In this article, the origins and genesis of the expressions “digital” and “digital computer” will be analysed. The technological term “digital” refers to a certain economy of electronic binary clocking. To explore the historical roots of this issue, for example Paul Ceruzzi compares the term “digital” with “analog” in the sense of an opposition in William Aspray’s “Computing Before Computers” from 1990. The author notes that the Atanasoff-Berry-Computer (ABC) from 1942 and the ENIAC from 1945 functioned digitally. But this observation is possible only because Ceruzzi replaces Atanasoff’s descriptive term “direct” with “digital”:
“If Atanasoff is the inventor of the electronic digital computer, as the courts judged in 1973, then it is in the restricted sense outlined here. At the same time […] Mauchly had only vague and ill-defined ideas about how to use vacuum tubes to build circuits that could perform digital calculation. Atanasoff, by contrast, was skilled at circuit design and had a thorough understanding of the difference between electronic circuits used for analog as opposed to digital applications.7 [Endnote] 7. Indeed, Atanasoff was the first to use the word »analogue« to describe that type of computer [built in 1936, not the ABC]; »digital« was first used by George Stibitz in 1942.”[1]
Little is known about Stibitz’ writings (figure 1) because his entire literary estate remains unpublished in the archives of Darthmouth College in Hanover in the US state of New Hampshire. During an awards ceremony held by the renowned Institute of Electrical and Electronic Engineers in 1977, one of the speakers discussed Stibitz’ merits and almost parenthetically defined the general characteristics of the digital computer using the terms binary logic, floating-point arithmetic, memory addressing and programme control:
“For pioneering contributions to the development of computers, utilizing binary and floating-point arithmetic, memory indexing, operation from a remote console, and program-controlled computations.”[2]
illustration not visible in this excerpt
Figure 1: George Robert Stibitz (1904–1995)
Indirect evidence of Ceruzzi’s endnote can be found in the dissertation of the physicist and sociologist of science Friedrich-Wilhelm Hagemeyer. In “Die Entstehung von Informationskonzepten in der Nachrichtentechnik” (The Genesis of Information Concepts in Communications Engineering) from 1979 he cites two unpublished documents in which Stibitz uses the expressions “binary places” and “digital computer”:
“Shortly after this conference [Conference on Electronic Fire Control Computers] George Stibitz summed up the advantages of digital computers again in a memorandum entitled »Digital Computation A(nti) A(ircraft) Directors« (23 April 1942). […] With this work Stibitz very probably coined the term »digital computer« as the counterpart to the analogue, consciously distancing himself from the term “pulse computer” that had occasionally been used up until then and still put more emphasis on communications processing and transmission. The fact that the phrase »binary places« […] is used here is just as remarkable.”[3]
For that “Conference on Electronic Fire Control Computers” on 16 April 1942 he wrote a 110-page “Report on Electronic Predictors for Anti-aircraft Fire Control” in which he distinguishes between two fundamental classes of computer systems – the “impulse system” and the “direct system” (figure 2). However, Stibitz also points out that in practice such a distinction is purely of an academic nature, because any machine that would actually be built would probably have both characteristics.
illustration not visible in this excerpt
Figure 2: Title page of “Report on Electronic Predictors for Anti-aircraft Fire Control” by Stibitz 1942.
Analogue computers add up continuously changing signals. The “direct system” (figure 3b) adds up binary signals produced from them for the logical values of zero and one. Stibitz explicitly addresses the economical aspect of the binary system, coining the term “binary places” and thus anticipating the term “bit”:
“The question of the most suitable number system was examined critically and it was found that for most purposes the binary system (radex 2) is the most economical. […] This number system lends itself very well to electrical computing devices where elements are used to represent binary places. Its advantage lies in the fact that each element only requires two states to represent the coefficients or digits.”[4]
Stibitz modifies the “impulse system” (figure 3a) by not adding up the binary impulses simultaneously via parallel data links as in the “direct system,” but after one another as a so-called “number train” (figure 3c).
“Fig. III-1 shows two computing devices, (a) being of the impulse type and (b) being of the direct type. A number is introduced into the direct device by applying voltage to the leads A. [...] The number remains in the device only so long as the voltage are [sic!] maintained. In the impulse device a number is introduced by applying an impulse to each wire representing a place having a digit 1 and no impulse on the others. [...] The number is only removed when the device is cleared. [...] The sum may be considered registered by the indicators i. With the direct calculator one number is introduced onto wires A and the other [number] onto the wires B, while their sum, as before, is registered on indicators i.
illustration not visible in this excerpt
illustration not visible in this excerpt
Fig. 3.I
[…] A modification of the impulse system employs number trains. Instead of transmitting the number as a simultaneous group of impulses on separate wires for each binary position, the number is a train of impulses whose ordering designates the binary places. […] Thus the number 10110101 [181] would be transmitted over a single wire by the impulse train shown in Fig. III-1c. Number train systems are inherently slower than ordinary impulse systems but frequently result in a considerable simplification of equipment.”[5]
In a “complete computer” that still had to be designed, Stibitz said that the advantages of both systems should be used, with the central computing element consisting of “multivibrator counters”. Such a computer would then be able to convert signals from outside in an analogue way and at the same time compute in a binary, discrete manner, or digitally, because inside computing would be done solely by means of flip-flops. The electronic computation of the zeros and ones would be carried out by “multivibrator counters” (figure 3.2):
[...]
[1] W. Aspray, Computing Before Computers, Iowa State University Press, Ames 1990, pp. 239, 247.
[2] E. Loveday, George Robert Stibitz and the Bell Labs Relay Computers, in: Datamnation, Vol. 23, 1977, p. 80.
[3] F.-W. Hagemeyer, Die Entstehung von Informationskonzepten in der Nachrichtentechnik. Eine Fallstudie zur Theoriebildung in der Technik in Industrie- und Kriegsforschung, Dissertation, Berlin 1979, p. 359.
[4] G. R. Stibitz, Report on Electronic Predictors for Anti-aircraft Fire Control (1942), in: Box 14, Inventory of the Papers of George Robert Stibitz concerning the Invention and Development of the Digital Computer, Dartmouth College Library, Hanover, New Hampshire 1973, pp. 3, 39.
[5] G. R. Stibitz, ibid., pp. 40f.
-
Upload your own papers! Earn money and win an iPhone X. -
Upload your own papers! Earn money and win an iPhone X. -
Upload your own papers! Earn money and win an iPhone X. -
Upload your own papers! Earn money and win an iPhone X. -
Upload your own papers! Earn money and win an iPhone X. -
Upload your own papers! Earn money and win an iPhone X.