Coding or informatics: that is the question

by Enrico Nardelli

(versione italiana qua)

At the start of this year, the European Commission launched an ambitious Digital Education Action Plan. Starting from the entirely sound premise that "education and training are the best investments in Europe's future," it proposes, in order to bring education and training in line with the digital age, to act along three lines of action:

1. making better use of digital technology for teaching and learning;
2. developing relevant digital competences and skills for the digital transformation;
3. improving education through better data analysis and foresight.

There is nothing to object to in the first line: it is well established that a well-calibrated use of digital technologies can improve both teaching and learning. The challenges all lie in getting that calibration right, but I trust that the new political leadership at MIUR and the experience of our teaching profession — among the finest in the world — will be able to identify and find the right solutions.

The third line is also highly valid: we live in the age of data, which we collect at every level. When properly analysed, data can tell us a great deal about the current state of affairs, at a level of granularity never before achieved. Here too, the devil is in the details: on one side, the need to guarantee people's privacy; on the other, the awareness that there is always a gap between what we measure and reality — a gap that must never be forgotten. Generals who have confused the map with the territory have never distinguished themselves in military campaigns.

I would like to say a few words about the second line, because for its implementation — beyond the accompanying measures relating to awareness-raising, communication, and closing the gender gap (all of them critically important!) — the concrete action proposed is to "introduce programming classes in all schools." Here, "programming" means "computer programming," that is, what the English call coding, a term that by now enjoys considerable currency in Italy too. I have already written (here and here) about why I do not think it wise to use the English term, but I will use it throughout the rest of this article for the sake of brevity.

My central argument is that this approach by the European Commission — centred on teaching coding rather than teaching informatics — is reductive. I draw a comparison between teaching coding and teaching informatics because teaching coding is, to use an engineering analogy, like teaching someone to build a bridge, whereas teaching informatics means teaching someone to design and build the bridge.

Designing means working within a space of needs and constraints and "creating" the solution — inventing a synthesis between competing requirements. This calls for a body of knowledge that goes well beyond what is needed simply to build. As someone trained as an engineer, I know this from experience. Two years of my university education were spent studying mathematics, physics, chemistry, and other foundational subjects that had no direct application to the construction of any complex artefact. Any doctor or lawyer will readily confirm that what sets the designer apart from the operative is precisely the breadth and depth of their foundational training.

Building means following a more or less defined plan that someone else has created. There is no requirement to know much beyond that, because someone else is handling — or has handled, or will handle — everything else. Talking only about "teaching coding" is therefore a reductive approach, because it focuses solely on the final, operational stage of a much more complex design process. It amounts to training a technically sophisticated workforce that will nonetheless remain a working class, subject to the directions of those who decide what to build, for whom, and when.

I note in passing — because this takes us into a political and social dimension that certainly deserves attention, but not here — that history (now more than ever magistra vitae) has shown that the basic levels of the labour force are the first to be replaced by machines. If we want our young people to acquire skills that will serve them throughout their working lives, it is better to train them as designers, not merely as executors.

I do not deny that programming can be a highly creative activity. But if we are talking about mass education, and if we teach only programming without educating young people about all the other aspects that make informatics the rich and multifaceted discipline it is, then we genuinely risk producing nothing more than an army of workers — one that will be easily sidelined by technological advances or turned into an "industrial reserve army."

For young people to have a real chance of finding a fulfilling path in the digital society — within a constantly evolving world of work — they need to be introduced from the outset to the scientific ideas underpinning the digital world, so that they can grasp its full significance.

One might object that there are many examples of successful start-ups born from the ideas of young people who had certainly never studied informatics formally. But if you look at the overall numbers, these are merely the exceptions that prove the rule. For a start, while the founder of Facebook was certainly not a computer science graduate, Google was born from the ideas of two Stanford informatics PhD students — one of the great cathedrals of computing research. Furthermore, for every start-up that makes it, there are at least a thousand that go nowhere, and so at a macroeconomic level — that is, in terms of growing a country — betting everything on start-ups is not a sensible strategy (another development policy debate worth having).

Similarly, there is a tendency to treat exceptions as if they were the norm. When it comes to programming, people cite geniuses like Linus Torvalds (creator of Linux) or Guido van Rossum (creator of Python), who single-handedly produced true masterworks. First of all, genius is an extremely rare quality, and what works for a genius cannot be used as a teaching method for mass education. Second, even those who are genuinely gifted still design: they do it in their heads, and very rapidly, as they develop — but the design phase is there all the same. Third, it is obvious that every project, at the point where it is expressed, requires a language through which to express it — which represents the "encoding" of the project. But this does not mean that the most appropriate term for someone who designs is "coder," or that the word "coding" encompasses the entire area of software design and development.

Let me be clear: incorporating informatics education into schools matters for every citizen's formation, regardless of whether they go on to become doctors or lawyers, musicians or writers. The digital society is here. Failing to teach in schools the science underlying its mechanisms would be like teaching children that babies are found under cabbage leaves, that the plague is spread by poisoners, and that stones move by horror vacui.

What makes a country great in the long run is not its exceptional talents, but the quality of its mass education. Europe's digital development requires the teaching of Informatics in schools, starting from primary level. This is the path that the United Kingdom has begun to follow; these are the trends in many advanced countries, both East and West; and this is what we have proposed as European associations of computer scientists.

--
The original version (in italian) has been published by "Il Fatto Quotidiano" on 17 September 2018.