Electronic brain
« previous post | next post »
On Facebook, this conversation thread followed from a post by Bill Benzon, commenting on his recent blog post, "Once more around the merry-go-round: Is the brain a computer?"
Here's the graph from Google Books Ngrams Viewer that Bill shared:
As long as I've been learning Mandarin (since 1967), computers have always been referred to as diànnǎo 電腦 / 电脑, though in the sense of "calculator" they are also called jìsuànjī 計算機 / 计算机. I recall that there was slippage between these two terms back in those days, with some people preferring one and some the other. Nowadays, diànnǎo 電腦 / 电脑 is overwhelmingly predominant over jìsuànjī 計算機 / 计算机.
According to the etymology in Wiktionary, the term diànnǎo 電腦 ("electronic brain") was coined by Taiwanese computer expert Fan Kuanling (范光陵) in 1965 in his book 《Diànnǎo hé nǐ 電腦和你》 (Computers and You).
Selected readings
- "Sinographic inputting: 'it's nothing' — not" (2/22/21) — with a lengthy bibliography
- Victor H. Mair and Yongquan Liu, eds., Characters and Computers (Amsterdam, Oxford, Washington, Tokyo: IOS, 1991)
chris said,
July 28, 2022 @ 2:29 pm
The assertion that the computer was built as a model of the human brain seems pretty suspect to me. Even accounting for the limited knowledge of the workings of the brain at the time, neither Turing's work nor Von Neumann's seems to have much to do with brains.
Mathematics doesn't come naturally to humans and has to be learned, often at great effort.
I got the impression that when Asimov referred to the "positronic brain" he was *contrasting* it with conventional computers, which were mechanistic and not that brain-like. But maybe he was just spicing up an already-prevalent metaphor of the day which vastly underestimated the structural differences between the two?
Frans said,
July 28, 2022 @ 2:50 pm
@chris
I'm fairly sure that the word positronic in Asimov is comparable to the word quantum today.
Gregory Kusnick said,
July 28, 2022 @ 3:46 pm
It's not clear what Lawler means by "the first instance", but it seems pretty clear that Babbage's machines were meant to model systems of equations, not brains.
Perhaps Lawler is thinking of cybernetics and Perceptron-like neural network devices from the 1950s and 60s, which were explicitly inspired by biological analogies, but that was a pretty distinct research program from the design and construction of programmable calculating machines for artillery ranging, payroll processing, etc.
norgie said,
July 28, 2022 @ 4:04 pm
@chris, agreed, I don't see much overlap between computers and brains, nor am I aware of any claims that the brain was an inspiration for Babbage, Turing, or the ENIAC designers.
Interestingly, some of the mentions of "electronic brain" from the 50s also point out that computers are not "electronic brains":
– "An electronic computer is not an electronic “ brain . ” "
– "The EDPS is often called an electronic brain , but actually it is relatively stupid and unimaginative"
– "Computers Information from the cards can thus be fed into a computer or data-processing machine — frequently miscalled an " electronic brain.""
Seth said,
July 28, 2022 @ 4:16 pm
Yes, the brain is a (very complicated) computer, but as chris and Gregory Kusnick have already said, the development of computers had nothing to do with modeling brains. The "electronic brain" phrase should be analyzed as a popular science term, akin to "flying saucer" or "robot". It went out of style as computers became more mundane. Note I don't mean personal computers, just that ordinary people had business dealings and learned the phrase "it was a computer error". It'd say that made computers seem not like brains at all.
Bill Benzon said,
July 28, 2022 @ 4:24 pm
I don't know what John Lawler had in mind, but in 1943 Warren McCulloch and Walter Pitts published a very famous essay, A Logical Calculus of the Ideas Immanent in Nervous Activity, in which they argued the neurons were binary devices and showed how to construct logical circuits out of neurons. That has influenced the design of circuitry in computers.
Charles in Toronto said,
July 28, 2022 @ 5:42 pm
Let us not forget that humans who performed calculations used to be called "computers". If you've seen the film Hidden Figures then you have heard this usage.
Chester Draws said,
July 28, 2022 @ 5:54 pm
in which they argued the neurons were binary devices and showed how to construct logical circuits out of neurons. That has influenced the design of circuitry in computers.
You can build logical circuits out of neurons, but that is not the same as building "brains" out of circuits. In fact it is the opposite direction.
Computers have set circuits, brains grow new neurons and paths. Computers will provide the same answer every time, brains will not. Brains come with hard programming, which is near impossible to change, computers have software. They are utterly different at a physical level.
If a computer is going to model human brains, then it will be at the software level, never at the hardware level. So neural nets, for example, are software emulations, not actual neural nets.
Jerry Packard said,
July 28, 2022 @ 10:05 pm
I agree with the general tenor of your remarks, but not sure I’d agree that brains come with immutable hard programming. The most obvious counterexample would be learning in general, in which case we modify our ‘software’ programming by modifying the hard circuitry.
Stephen Hart said,
July 28, 2022 @ 10:52 pm
Jerry Pakarrd: "not sure I’d agree that brains come with immutable hard programming"
Absolutely. "Programming" is inappropriate with respect to brains, let alone "hard programming," whatever that's supposed to mean. "Hard wired" is even worse. Also "x changes your brain." (Reading this comment changed your brain, for example.)
GH said,
July 29, 2022 @ 4:04 am
Babbage may not have designed his machines in explicit mimicry of the human brain (however he might have imagined that the brain worked), but Lady Byron reportedly referred to a model of his Difference Engine as a "thinking machine" in her diary in 1833, before he even started designing the Analytical Engine.
And in the modern computer age there were some early luminaries who did consider them to be models of the human brain. Edmund C. Berkeley (co-founder of ACM, who worked on one of the first UNIVAC machines), wrote an influential book called Giant Brains, or machines that think (1949) promoting this idea.
Gregory Kusnick said,
July 29, 2022 @ 9:22 am
But as Chester says, the strategy of (most) AI pioneers was to model cognitive processes in software, on computers that had already been built for other purposes. AI research was not the raison d'etre of computer design "in the first instance", as Lawler claims.
Victor Mair said,
July 29, 2022 @ 11:21 am
See now:
"Infinitely malleable electronic brain — software and hardware" (7/29/22)
https://languagelog.ldc.upenn.edu/nll/?p=55476
BillR said,
July 29, 2022 @ 12:28 pm
My recollection is that computer = electronic brain was a science fictional trope that leaked into the popular zeitgeist in the 40s and 50s, maybe earlier. Most of the time the gist of the stories was that this was not a good thing. “The Terminator” is probably the epitome of the sub-genre.
What you all are going on about is how the terms were used by the folks who were inventing and building the things.
bks said,
July 31, 2022 @ 7:02 am
George Boole's book "An Investigation of the Laws of Thought on Which are Founded the Mathematical Theories of Logic and Probabilities" (1854) and von Neumann's book "The Computer and the Brain" (1958, published posthumously) are worthy of mention in this discussion.