EWD 997

Computing Science in the United Kingdom

This is just a sad reminder that in the Old World not all is well, not even in the United Kingdom that has so often been leading in Computing Science.

In the Computing Newsletter, October 1986 —a publication of The British Computer Society— I encountered in an article titled “Second Manchester Conference on Careers with Computers. Are we Getting it Right” by Geoff Hayes the following:

“From returned questionnaires attempting to gauge interest and relevance, the conference was a singular success, especially as it has also highlighted some significant differences of needs as perceived by educators and employers. Many delegates found this profoundly disturbing [...].”

But the discrepancy between what the world asks for and what the world needs has been computing’s traditional curse for more than a quarter of a century! That “many delegates” can still be surprised when faced with a manifestation of this fact, well, that’s what I think rather disturbing....

In The Computer Journal, Vol.29 No.5, October 1986 —the most prestigious publication of The British Computer Society— I found in an article titled “A Knowledge-Engineering Approach to Instructional Design” by Nick Rushby the following revealing diagram:

EWD997 figure 2

For those who think this insufficiently baroque, the text adds: “This shows some of the possible topics and linkages: readers are encouraged to add their own as they read through this paper and further ideas occur to them.”.

Another sentence from that paper to be conserved for posteriority is: “Clearly, there is a need for a better understanding of the learning process, and a need for a language with which to describe it.”.

That same issue of The Computer Journal contains a nice specimen of junk science: IBMology. It is the paper titled “IBM makes usability as important as functionality” by B.Shackel. In its first column it contains what we have to consider an educated guess “In my opinion the decision for usability and human factors was probably taken at the same time.”; in its second column the author gets really scientific: “A corporate decision must have been made in 1980 or early 1981, with substantial funding assigned to various programmes. The result, below, provide the evidence.”. I always thought the “Phenomena breed science” of Newell, Perlis, and Simon (1967) a bit superficial; the Shackel paper is what we get when the dictum is taken to heart.

It is a sad world.

Austin, 8 January 1987

prof.dr.Edsger W.Dijkstra
Department of Computer Sciences
The University of Texas at Austin
Austin, TX 78712 - 1188
United States of America