Aristotle’s saddle-maker

Or the importance of software in education technology

We have invested too much in hardware and not enough in skills. That, at least, is the message that Michael Gove has given in his two recent speeches on education technology[1].

He is probably right. Rows of gleaming white boxes have always made good ministerial visits. Ever since the ill-fated “modems in cupboards” initiative of the 1980s, we have tended to fill our schools with hardware, while the current debate around the “dull and boring” ICT curriculum has highlighted the inadequacy of many teachers’ technical skills.

We should beware, however, of believing that these two issues—hardware and teacher skills—are opposed to one another, like two sides of a coin. As with most antitheses[2], this would be to divide our world mistakenly into opposing hemispheres.

There are two reasons why such a dichotomy might be false.

The first is to assume wrongly that the two halves of the coin are mutually exclusive. I have always thought of this as the “Miranda fallacy”, after Prospero’s daughter in The Tempest, who asks her father why he was ejected from his position as Duke of Milan and marooned on a desert island:

Miranda What foul play had we, that we came from thence? Or blessed was’t we did?
Prospero Both, both, my girl.

The second fallacy is not to realise that the coin in question has more than two sides. Maybe Prospero was marooned as a result of a legitimate judicial enquiry into his misconduct during the time that he was Duke.

In the Nicomachean Ethics, Aristotle uses the example of a saddle-maker to illustrate how ends and means are frequently linked together into productive chains. The tools of the saddle maker are one of the means by which he produces the end of his craft: a saddle. From the perspective of the cavalryman, this saddle is one of the means by which he produces the end of his craft: a cavalry charge. From the perspective of the army general, the cavalry charge is one of the means by which he produces the end of his craft, military victory; which in its turn serves the political ends of the state.

Diagram showing series of linked processes illustrating Aristotle's saddle-maker

A similar chain of ends and means could be constructed for education technology:

  • hardware is a means of running software;
  • software is a means of supporting teaching;
  • teaching is a means of facilitating learning.

One of the fashionable insights of the ICT-in-education community is to oppose technology and learning, using a similarly false dichotomy. A previous Chairman of NAACE wrote a blog called “Never mind the technology, where is the learning?” From the education technology perspective, this is a bit like saying “Never mind the plumbing, where’s the water?” Of course, you could go and get it from the well with a bucket, but if you want it coming out of your tap, it will be because of, not despite the plumbing.

When the process of education is understood in terms of this interlinked chain of ends and means, it can be seen that the opposition of hardware and teaching skills is misleading not only because these are not mutually exclusive alternatives, but also—and more importantly in this case—because they miss out a vital connecting link in the chain: there is no mention of software.

Just as past initiatives that focus exclusively on hardware have often gone awry, so have ill-considered training programmes. The 2003 training associated with the “New Opportunities Fund”, worth £230 million, was roundly slated by Ofsted[3], the main problem being that no-one really knew what they were meant to be training teachers to do. Becta, always keen to look on the bright side of any technology-related initiative, reported with approval that one teacher had used the scheme to find out how to send emails to her relations in Australia. It is not clear what benefit her pupils derived from this achievement.

For the current purpose, it is sufficient to note that computer training generally involves learning how to use a specific piece of software. Even if you want to develop transferable skills, you will almost inevitably do this by using a variety of different pieces of software. And it is not helpful for someone who wants to learn to do their accounts to be trained in the use of a 3D drawing program. The fact that NOF trainees learnt little more than how to send emails suggests that that the training programme had not, in Ofsted’s words, identified “subject-specific ICT applications” (in both senses of the word “application”).

Some software is designed specifically for education, some for general use; some is more application-specific, some is useful in many different contexts. A word processing package can be used to create many different kinds of document and a browser is so multi-talented that non-technical people often have trouble recognising its existence. At the other extreme, business software systems, such as a payroll management system or the checkout software being run in the local supermarket, are very application-specific.

Freedom comes at a price. A word-processor is a great tool in the hands of J K Rowling but will not help the idiot write a successful novel. Conversely, the payroll management software aims to be idiot-proof. It restricts the choices available to the user, automating complex functions, internalising expertise, minimizing training costs, and resulting in more consistent and reliable outcomes. No wonder business likes structured, application-specific software.

Just as with hardware and skills, we must again beware of creating a false dichotomy: open-ended and application-specific software are often related, and in the same kind of linked chain as Aristotle’s saddle-maker:

  • operating system software provides a means of running desktop applications (e.g. a web-browser);
  • a web-browser provides a means of launching web-pages (e.g. Google Maps);
  • Google Maps is a means of viewing different maps (e.g. of your home town).

In this case, each stage of the chain has multiple sub-links. It is a branching chain. The upstream software (to the left of the diagram) tends to be generic and multi-purpose, while the downstream software (to the right of the diagram) tends to be structured and application-specific.  At each stage of bifurcation, the relationship between parent and child is the relationship between an infrastructure and… the only satisfactory word that I can find is, its “content”.

Diagram showing series of nested infrastructures

You may object that I make no distinction in this sequence between applications and data—but I would argue that this is not a fault. Infrastructural software has the same relationship to subordinate applications as it does to subordinate data; it is increasingly difficult to distinguish between the two, as in the case of an interactive Flash movie or when JavaScript is embedded within a web page; and from the point of view of the computer they both look exactly the same.

Understanding “content” as a generic term that only has meaning in relation to a particular kind of infrastructure (or “container”) shows how restricted is our thought in the education technology debate. We use “content” to refer exclusively to expositive text and pictures, contrasting instructionalist, “content-driven systems” unfavourably with dynamic tools like Facebook. From the point of view of the World Wide Web, Facebook is content; from the point of view of the operating system, a web-browser is content. The problem is not that people do not understand the “real” meaning of the word “content” (as I have argued already on this blog, what’s in a word?). The problem is that if they don’t have a word for it, they do not understand the fundamental importance of the infrastructure-to-content relationship in systems design.

A complete digital ecosystem might be pictured as a branching sea anemone. At the base are the generic infrastructures, and at the tips are the application-specific tools. Floating in the water  are the technology users, depending on their technical skills to interact with the top layer of the anemone (not, it is to be hoped, with fatal results).

Diagram of anemone comparing business and education software

While businesses spend large amounts of money customising and integrating generic software to their particular requirements and processes, in education, the technological anemone is a poor, stunted specimen. End users (only a small minority of whom are interested or skilled in technology) are required to bridge the gap between these generic technologies and the “subject-specific ICT applications” that will effect useful educational outcomes.

Much of the conversation on education technology forums is about Skype, Twitter, Facebook and iPads: none of them education-specific, still less application-specific technologies. The draw-backs of this approach to education technology was highlighted recently by a question which Diana Laurillard[4] asked John Laughton at a recent Association for Learning Technology conference at

I have been trying to apply what you have been talking about to our business of education—teaching and learning—and one of the difficulties we have in teaching and learning is that it is quite a complex transaction, there are quite complex human needs going on there, if indeed you can call them human needs. I mean there may be a human need to teach and, in so far as there is, it is fairly simply satisfied by things like PowerPoint or websites—just give it; tell the story by subject and that’s enough. There may be a human need to learn things we can not know of until we have learnt them, which is what formal education is all about, but that’s a pretty complex technology—it certainly isn’t Wikipedia—that doesn’t do it. So, what education has done has been to appropriate everybody else’s technologies for all the different facets that we need in the teaching and learning transaction. So, is that our fate?

What is interesting is not just Professor Laurillard’s question, which I believe illustrates exactly the argument I have been making in this post, but the fact that, judging by his answer, Professor Naughton—a prominent academic in the e-learning field—clearly does not understand what she is talking about.

Shortly after Labour came to power in 1997, I wrote a letter to the Times Education Supplement arguing that a greater priority should be given to software, and criticising a speech by the incoming Secretary of State for Education, David Blunkett, in which he eulogised he benefits of the internet. Once you were online, Mr Blunkett said, you could download everything for free. In his speech to the SSAT on 1 December 2011, Michael Gove spoke in a similar way of the video content on the Khan Academy, saying that “such content is being created daily, and the vast majority is free to anyone with an internet connection”. For all the current popularity of the Khan Academy, we already know and as Professor Laurillard explains, using computers to deliver expositive media “doesn’t do it”. In fifteen years, we still suffer from the same failure to appreciate the importance of software in the supporting education technology as a coherent process. If Tesco does not run its logistics using free software downloaded from the internet, then why should we expect to be able to manage education in such a haphazard way? Is the delivery of education so much simpler to manage than delivery of groceries?

In future posts, I shall explore what the government needs to do in order to encourage the production of the education-specific and application-specific software that will give education technology a realistic chance of improving the delivery of formal education:

  • reducing the need for training;
  • supporting consistent standards of delivery;
  • making life easier for teachers (rather than burdening them with more work);
  • making savings in school budgets (rather than burdening them with unproductive costs).

In the meantime, the government should refrain from making any statements about their strategy for education technology, so long as this does not give a prominent place to encouraging the development of made-for-purpose, education-specific software.

[1] In his 1 December speech to the SSAT, Michael Gove said “Part of the problem has been that investment has focused on hardware…rather than on training individuals to be technologically as literate and adept as they need to be” ( In his BETT speech, Mr Gove said that “rather than focusing on hardware or procurement, we are investing in training individuals” (

[2] For a robust argument against our habit seeing the world through the black-and-white perspective of antithesis, see Friedrich Nietzsche’s Beyond Good and Evil, Chapter 1, Prejudices of Philosophers: “HOW COULD anything originate out of its opposite?”

[3] “Teachers became resistant to training that was inappropriate to their needs. Much training made a limited contribution to their awareness of subject-specific ICT applications and did not encourage them to consider pedagogical issues of teaching and learning with ICT”. Ofsted, paragraph 54, page 22, ICT in schools 2004: the impact of government initiatives five years on at

[4] Diana Laurillard, Professor of Learning with Digital Technologies at Institute of Education, and previously Head, E-Learning Strategy Unit in the DfE, where she was responsible for supervising early versions of the Harnessing Technology strategy, that Becta was responsible for implementing.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s