4.0421 Responses: Memory; Digital Computers; Big Science (3/61)
Elaine Brennan & Allen Renear (EDITORS@BROWNVM.BITNET)
Fri, 24 Aug 90 22:26:44 EDT
Humanist Discussion Group, Vol. 4, No. 0421. Friday, 24 Aug 1990.
(1) Date: Fri, 24 Aug 90 13:02:37 GMT+0100 (21 lines)
From: Stavros Macrakis <macrakis@ri.osf.fr>
Subject: Technology/Memory
(2) Date: Friday, 24 Aug 1990 01:08:08 EDT (19 lines)
From: "Patrick W. Conner" <U47C2@WVNVM>
Subject: 4.0418 Digital Computers and Binary Representation
(3) Date: Fri, 24 Aug 90 17:20:55 GMT+0100 (21 lines)
From: Stavros Macrakis <macrakis@ri.osf.fr>
Subject: 4.0414 Big Science
(1) --------------------------------------------------------------------
Date: Fri, 24 Aug 90 13:02:37 GMT+0100
From: Stavros Macrakis <macrakis@ri.osf.fr>
Subject: Technology/Memory
The Brits prefer (or perhaps preferred) `store' to `memory', precisely
because it is less anthropomorphic. There are informaticians (notably
Edsgar Dijkstra, the "humble programmer") who insist on avoiding
anthropomorphic names in informatics.
Too much information is not only a problem for librarians and
historians, it's also a problem for users today. As Herbert Simon
pointed out many years ago, the scarce resource in information
retrieval is not data, but the user's attention. And then there's the
temptation to want to use as much data as possible rather than to get
the best understanding possible.
Stavros Macrakis
PS All digital computing is alike, to first approximation. Ternary
(balanced or otherwise) won't display Greek any better, either.
(2) --------------------------------------------------------------26----
Date: Friday, 24 Aug 1990 01:08:08 EDT
From: "Patrick W. Conner" <U47C2@WVNVM>
Subject: 4.0418 Digital Computers and Binary Representation (
I don't know enough even to ask this question, I suspect. But John
Lavagnino tells of of Claude Shannon's computer <that used Roman numerals
internally.> The numbers or numerals must be representations of
electrical states, mustn't they? The shape of the symbol isn't
important, but rather the number of symbols. How many Roman numerals
were used and how did Shannon achieve more than two states? As someone
has already said, there is no pure ON nor pure OFF state, but for
practical purposes we have to assume that. I don't care which numerical
symbols are used by those who schematize the electrical states in the
machine, but I am interested in knowing how more than two states can be
defined in any one circuit. I suspect that Knuth's ternary system is a
refinement of the system of representing on/off symbolically and does
not represent a change in the number of electrical states which the
machine recognizes, but I admit to having no background in such things,
just a curiosity.
(3) --------------------------------------------------------------46----
Date: Fri, 24 Aug 90 17:20:55 GMT+0100
From: Stavros Macrakis <macrakis@ri.osf.fr>
Subject: 4.0414 Big Science
Those interested in Kessler's remarks about the funding and publishing
"racket" might want to look at C.J. Sinderman's _Winning the Games
Scientists Play_ (how to); Derek de Solla Price's _Big Science, Little
Science_ (sociology of science); Sir Peter Medawar's _Advice to a
Young Scientist_ (paternal). Eugene Garfield, the publisher of the
various citation indexes useful both for research and for counting
citation brownie points, also writes essays on the subject--mostly
based on statistics, occasionally containing interesting
insights--duly collected, published, and indexed by his firm.
A propos, how many Humanists use Citation Indexes, and what do you
think of them? On paper, it looks enticing to be able to pull up the
articles referring to, say, _Billy Budd_. (Important in-text
references are counted as though they were footnotes.)
Stavros Macrakis