Monday, 23 May 2011

Are We Trapped by our Training?

I have decided to return to the research I abandoned 20 years ago, in order to reassess it in the light of recent developments. Currently I am going through boxes of research papers stored in my garage. Today I came across one which is very relevant to the theme of this blog - and is also relevant to the decision I made to abandon the research. In many ways the paper is still very relevant - and if writing it today I might well also point to much Artificial Intelligence where advances in the last 20 years are due more to the availability of increased computer power than to a better understanding of human intelligence.

The paper is Are We Trapped by our Training?  which I presented at the conference Information Technology for Training and Education held at the University of Queensland in 1991. The paper reviews the way that learning one programming language can act as a blinkering effect in learning a new programming language - and in particular the ways in which a knowledge of conventional programming, and conventional programming systems, make it hard for people to relax and use the flexibility that CODIL offered.
For example one experienced COBOL programmer decided to learn CODIL (MicroCODIL's main frame predecessor) and set out to implement a simple application in both CODIL and COBOL. After a short time he reported that it was so easy in CODIL that he wouldn't bother to do it in COBOL. On examining his working package I observed with surprise that he had tried to use CODIL as if it were a COBOL interpreter which did not require a predefined data division. When I explained that virtually all his "code" was redundant because CODIL would would carry out many of the operations automatically he threw up his hands in utter disbelief. 
To put it simply I had identified a real problem. The more someone has been brain-washed by learning to program a computer the harder it is for them to approach a problem without first trying to map it into algorithmic form. The fact that computers were turning up everywhere - including in the schools, suggested that the real task was not the technology of producing a fully operational CODIL system - that was comparatively easy. But to be successful the system would need to be compatible with existing computer systems and data bases (which represent a very significant investment).  - and there would have to be a major unlearning exercise. We all know that the QWERTY keyboard continued because it was too expensive in human skills investment to change. I felt I faced an even bigger barrier in trying to take CODIL forward, and challenge what is by now trillions of dollars investment in relevant existing computer technology. I looked at the number of years I had devoted to the research, my mental stage after the death of my daughter, and what still needed to be done, and decided that enough was enough.

The full text of the paper (with a few typographical errors corrected) is directly available in the right hand column of this blog. Currently other papers are being prepared for online access and once this has been done links will be added to this paper.

No comments:

Post a Comment