Thoughts about the BBC Microbit and educational computing

Posted to the Computing at School (CAS) Forum on 22 Jun 2015
Requires CAS member login, so made available here (expanded)
for anyone who is not a CAS member.

Aaron Sloman
Honorary Professor of AI and Cognitive Science
School of Computer Science
University of Birmingham

24 Jun 2015: [2] Extended history with other 1980s home/educational computers.
23 Jun 2015: [1] Added brief history of BBC Micro 23 Jun 2015

Some thoughts and concerns about the Microbit

I watched part of the presentation on Saturday morning (20th June) at the CAS (Computing at School) 2015 conference and was not at all convinced that the Microbit had been designed by people with a deep understanding of the diversity of needs of learners or teachers, and I thought it quite absurd that something still being developed by possibly excellent programmers but obviously far from ready, was going to be offered to millions of school kids backed by the (in this case spurious) authority of the BBC without first having been tried, tested, and criticised by a cohort of experienced teachers with a wide range of backgrounds, expertise and confidence levels.

Apart from the risks of immaturity of the product (evident on Saturday) I think a nation that commits itself to monolithic educational strategies will produce yet another generation that has a blinkered view of what's possible. The future is so open and uncertain that we need a wide variety of approaches to teaching, especially in a field that changes as rapidly as computing. Thanks to CAS this variety seems to have been growing and flourishing in the last few years. There's nothing wrong with trying to add to the variety: but the motivation behind Microbit seems to be more like trying to replace everything else being used for beginners. That would not matter so much if it did not have the backing of the BBC!

Any experienced teacher will know that children vary widely in their interests and abilities. One child may like to learn how to make a machine flash lights and make noises. Another may enjoy exploring properties of numbers through programming. Another may wish to program a physical or graphical turtle, or group of turtles. Yet another may enjoy programming a chatbot using a simple pattern-based language (and later grow up to be a famous theoretical linguist, or psychologist, or philosopher).

Moreover, teachers also have different interests and aptitudes and creative ideas and a national educational policy should find ways to build on that, not squash all teachers down a narrow funnel.

Automatically giving such a system such wide distribution (millions of learners), without a proper period of trial and debugging of both design and implementation, implies that everything else is worthless, which is both insulting and potentially demoralising to the many people who have already been developing, testing or using teaching support packages of high quality, in many cases tried and used with great enthusiasm and success by teachers.

Yet another case of excessive hubris? That's ok for commercial products people can check out and decide to buy or not to buy. But not when the minds of our children are the target - and the BBC is doing the launching.

I hope it is clear that I am not criticising the Microbit, only the process by which it is being launched, which, among other things, risks being grossly unfair to the many others who have already poured time, resources (and their hearts) into developing tools, teaching ideas, and machinery in support of the wonderful process launched by the ComputingAtSchool group, but without having the advantage of the resources and prestige of the BBC. (I am not one of them, in case anyone is wondering whether I have a vested interest.)

I am also somewhat reminded of many IT projects (e.g. NHS projects among others) that have flopped (often at great public expense) because they were not based on adequate requirements analysis and testing in close collaboration with all the relevant types of user. Of course, sometimes a genius designer does not need that consultation (e.g. the designer of Velcro?). But that is rare.

A very important aspect of 'Computational Thinking', often ignored, or the difficulties underestimated, is understanding the need for deep requirements analysis, especially for systems designed to be used by a wide variety of types of people, or systems designed to model or explain some very complex natural phenomenon (e.g. cognitive development in toddlers [*]):
What is computational thinking?
What Forms of computational thinking will our children need when they grow up?

An analysis of a major disaster in government and the IT industry -- attempting to revise the NHS (yet another Tony Blair disaster?) -- can be found in this messy web site:
The iSoft affair:
Open Letter to my MP (Lynne Jones) about government IT procurements
(Originally sent August 2006)

(Not in the version on the CAS community web site.)
The BBC Micro was produced in a collaboration between the Acorn Computer company and the BBC in 1981, and was highly valued by many teachers and learners (though in many schools there was nobody who could teach computing, so the machines they acquired were wasted, unless bright pupils were let loose on them).

A new version of the already widely used BASIC programming language (see was developed for the machine, which supported procedures with proper local variables, and therefore full recursion, a great improvement on previous versions of BASIC. Unfortunately the opportunity to go even further and add support for list processing, which would have given far better resources for teaching AI programming (including modelling language understanding, and logical reasoning processes) was missed, even though people elsewhere were already using such languages successfully for teaching introductory AI courses, e.g. in the USA and at Edinburgh and Sussex Universities (among others) in the UK.

Despite the virtues of the BBC micro, its use did not last and grow as it should have done, mainly because ill informed critics (including, according to my memories, politicians, people in business and industry, and some parents) complained that school-kids were not being taught to use the machines that they would encounter in future jobs, which at that time meant IBM PCs running a microsoft operating system and business tools like word processors and spreadsheets, not programming tools. This produced 30 years of serious damage to our educational system, which CAS (Computing At School) has been undoing, with spectacular results in the last few years.

Other contributors to IT education and home computing in the 1980s:

-- Sinclair's contributions (ZX81, Spectrum)
Very popular with people who could not afford the BBC Micro, or who wanted to buy something earlier.
-- Atari computers
on which our children learnt to program in C and Pascal in the 1980s.
-- Commodore 64
There were several others, including machines sold mainly as word processors. Also during that period Unix based systems were taking off, especially Sun workstations and servers (using the highly prophetic slogan "The Network is the Computer"), and Bleasdale computers in the UK.

-- Bleasdale
(Report in New Scientist December 1982)
(Someone should create a Wikipedia entry for Bleasdale computers.)

This list is far from complete. A major piece of the story of that decade concerns IBM's failure (or their lawyers' failure!) to understand what they were allowing Microsoft to do to them.


Maintained by Aaron Sloman
School of Computer Science
The University of Birmingham