Informatics Education

Why talk about education

The "self-conception" of a discipline is closely tied to how it is being taught, as "it" is a sum of its scholars, students and public reception. This relationship of course is bidirectionally – how a discipline is taught influences how it is perceived as well as the other way round.
Thus, tracing the history of how Informatics education has changed since its origin also paints a picture of how the subject understood itself throughout time. Discussing today's methods might shed some light on its current understanding and, vice versa, debating how Informatics can (and maybe should) be understood might help inform how it can and should be taught.

Informatics, Computer Science, Computer Engineering ...

Throughout this, I will refer to the discipline discussed here as Informatics. This might be due to my German language-influenced perspective, it is also, however, due to the fact that, throughout this text, any other possible name for the subject will be used to express a certain connotation.

The term was first used by Karl Steinbuch¹ in his publishings on a data processing system for the mail-order company Quelle.

The term was well received and thus, ten years later, the French began using the term ”informatique” which was then adopted by many other European languages such as Dutch (informatika), Italian (informatica), Polish (informatyka) as well as Czech and Russian. In Germany, the Federal Minister of Research, Gerhard Stoltenberg, officially introduced the term for the scientific discipline in 1968, describing it as the scientific background of a federal research programme in the field of data processing [127].

Traditions

Before jumping into the hard-facts of Informatics' genesis and decades and their respective views, a more general remark on traditions is in order. Matti Tedre, tracing the history of the discipline, discusses the traditions of mathematics, engineering and science [128] brought forth by Wegner in 1970 [146].

Tedre [2015] points to the different understandings of these traditions. Science, e.g., as argued by many notable scientists, should merely aim to describe and predict, whereas engineering, as argued by philosophers of engineering and technology, finds its essence in its aims [128].

Tedre [2015] quotes Mitcham like so:

"Engineering as a profession is identified with the systematic knowledge of how to design useful artifacts or processes, a discipline that (as the standard engineering educational curriculum illustrates) includes some pure science and mathematics, the “applied” or “engineering sciences” (e.g., strength of materials, thermodynamics, electronics), and is directed toward some social need or desire. But while engineering involves a relationship to these other elements, artifact design is what constitutes the essence of engineering, because it is design that establishes and orders the unique engineering framework that integrates other elements." [129, pp.146–147]

Tedre [2015] further points towards Koen's description of the engineering method:

"the use of heuristics to cause the best change in a poorly understood situation within the available resources." [130, p.94]

While one might agree or disagree with this view of engineering, I wonder if, out of those three traditions (mathematics, engineering and science), I would pick engineering when describing Informatics – or rather, if I would pick any one of them at all.

A short history

Before going into this short history, it’s important to mention that it is quite USA centric. This is due to multiple factors. One of them being that it was primarily private US companies providing computers to universities, another that many of the “pioneers of computing” (again pointing to killer stories) were from the US and dominated the fields for decades.

The story of Informatics' genesis is convoluted to say the least. Like many other disciplines, Informatics was born out of necessity. However, sticking with the analogy of birth, unlike many others, in the grand timeline of academic disciplines, Informatics’ parent has barely left the hospital yet. In other words, Informatics is a latecomer. The aforementioned necessity was that of computing proving more and more useful in many a different discipline around the 50s and early 60s.

However, computing did not yet have any sort of unified curriculum to teach under, a fact that heavily influenced its perception as a field, or as Tedre [2015] puts it:

"What kind of a field is one that cannot agree about its own curriculum?" [128, p. 41]

So, while Informatics was trying to establish itself as a discipline, its understanding of itself was still emerging, arguably hindering it in doing so. Thus was it, e.g., that pioneers of programming reminisce about the 50s' as a shoot from the hip time [131]. Tedre [2015] mentions that at the same time, and maybe not entirely unrelated to this, I believe, Hammer and Hamming use the terms "computer bums", "educational asses" and "computniks" for colleagues in computing [128, p. 43].

Likewise, Fein recounts, in an interview conducted by Pamela McCorduck, that computers were thought of as "exotic calculators, that had hardly anything to do with science", further nudging Informatics into the corner of a mere engineering matter [147].

It was also Fein who, in the 50s in a report trying to define what computing in an academic context is, proposing a curriculum, urged universities to buy computers rather than build them [132]. The field of computing had thus far focused on the possibility of building computers, chasing after potentially better (for whatever that may mean) machines, rather than taking the next step of figuring out what could be done with them. This, to me, sounds like an early version of the killer story, we describe in this critique and thus points towards a pattern of issues (more on this later).

Aside from how scholars and students were perceived, however, a good indicator of the general understanding of the field is found elsewhere.

It starts in the 1960s.

Curricula and the right questions

Tedre [2015] picks the ACM and IEEE curricula as milestones in the development of computing as a discipline, adding what authorities on the field saw as the core questions of Informatics at the respective times.

The 60s

The CC'68 (ACM's 1968 Computing Curriculum) [133], summed up by Tedre [2015], describes Informatics as "the study of information structures and processes and how those structures and processes can be implemented on a digital computer" [128, p. 45]. Thus, Fein's [132] advice seemingly was heard. Computing moved away from building computers, towards theory. Heavily focussing on the discipline it was ever so closely tied to – Mathematics.

In the dynamic between those two disciplines, however, Informatics, once again, was the latecomer, the little sibling of the well-established Mathematics. Adding injury to insult, the academic world mocked Informatics, accusing the field of a "deep nostalgia for mathematics by people who would like to label themselves mathematicians but certainly were not" [128, p. 45], thus creating the foundation for one of the, in my opinion, core insecurities the field is still struggling with today (however, again, more on this later).

As for the main question of Informatics, Forsythe [1968] put it this way:

"The question 'What can be automated?' is one of the most inspiring philosophical and practical questions of contemporary civilisation." [134]

The 70s

Towards the 70s, critics of the CC'68 noted a lack of laboratory work in it. One example of such a critique is Richard W. Hamming [1968], proposing, during his Turing Award speech, he'd require every student, major, undergraduate or graduate, to, in the course of their studies, "design, build, debug and document a reasonably sized program" [135]. Tedre [2015] points out that the CC'78 followed that notion, in turn moving away from Maths.

Where the CC'68 read:

"The committee feels that an academic program in computer science must be well based in mathematics since computer science draws so heavily upon mathematical ideas and methods." [133]

The CC'78 read:

". . . no mathematical background beyond the ability to perform simple algebraic manipulation is a prerequisite to an understanding of the topics \[. . . ] mathematics is not required as a prerequisite for any of that [core] material." [136]

As a further result of this move away from Mathematics, and likely enough also due to the zealous hostility which Informatics scholars encountered from Mathematicians, an image of academic, mathematical ivory towers spread through the Informatics community. This was supported by industry representatives and academics alike – as Wishner [1968] put it, discussing the sort of academics the CC'68 would produce:

"A good education should not be solely directed toward academicians whose only economic justification is to teach in order to turn out recursively new generations of academicians." [137]

Thus, industry pushed for more practical education, i.e. programming, as of course, it was not mathematical, theoretical proof that was driving the economic revolutions that came with the computer (at least not on a surface level) but coders.

Towards the 80s

Furthermore, and maybe relevant to our cause here, the CC'78 [136], for the first time, recommended HCI as "a minor part of social implications" [128, p. 168]. Related to this, during the 70s and 80s, the psychology of programming was an important topic. This, however, was not exactly what we refer to as HCI today but rather dealt with program comprehension, debugging, software development phases and maintenance – topics both important for education (when thought of as a practical matter) as well as industry. As for the apparent self-confidence issues of the field, opinion pieces of the 80s started to change the narrative:

"There is nothing laughable about calling computer science a science [anymore]." [138]

As for the aforementioned question "What can be automated?", this was adjusted, fitting the shift towards practical application seen in the 70s, by the Task Force on the Core of Computer Science, to read:

"What can be (efficiently) automated?" [139]

The 90s

By the time the CC'91 [140] came around, seemingly, the effort to create a coherent curriculum had been given up. The CC'91 [140] did not provide a course list, rather, it gave recommendations on courses that would be necessary and left it up for universities to fill up the rest of the program how they see fit. The spirit of finding what exactly Informatics is, was still there but maybe the idea that there would be one definitive answer was slowly fading. The introduction to the CC'91, by the chair of the ACM Education Board, read:

"There is still a need for diversity and well-intentioned experimentation in computing curricula." [140]

The general notion of computing reverted back to the concepts Wegner [1970] had brought forth in 1970. The CC'91 [140] explicitly addressed them. It acknowledged computing to be a combination of those paradigms (of theory, abstraction and design), stating that, in Tedre's [2015] words:

"A theoretically oriented educational program would, naturally, foreground the theoretical paradigm, a computational science program might emphasize abstraction more, and a software engineering program could focus its attention on the design aspects of computing. No more bickering about a one-size-fits-all model of computing education." [128, p. 53]

However, the CC'91 [140] also brought another, major change. It first considered social, ethical and legal issues – even philosophical and aesthetic questions of computing. However, as Tedre [2015] points out, they seemed to be more of an "tacked-on afterthought" [151, cited in 128, p.54].

The 2000s

Around the 2000s, the IEEE Computer Society started dividing computing into six different fields, as a one-size-fits-all report apparently was not possible. Those fields were:

1. Electrical Engineering – Electronics, Circuit Design, Telecommunications

2. Computer Engineering – Computer architecture, Embedded systems, Signal processing

3. Computer Science – Algorithms, Programming languages, Intelligent systems

4. Software Engineering – Software design, Software management, Professional practice

5. Information Systems – Information management, Enterprise architecture, Data mining

6. Information Technologies – web technologies, Information security, system administration

[128, 140]

It was also around this time that a new question, besides "What can be (efficiently) automated?", did not exactly replace the old one, but was at least floated around. That was what Tedre [2015] calls "The Human-Centred Question" of "What should be automated?". It was during the mid 2000s that some Informatics scholars started questioning their field's impact on people. A shift Tedre [2015] describes as such:

"Someone whose work is concerned with efficiency or reliability can always set aside ethical and social concerns by arguing that science and technology are value-free (a questionable argument, though). But someone whose work is concerned with changing people’s lives cannot ignore ethical and social concerns." [128, p. 168]

Which, as said before, is not to say of course, that Informatics completely turned away from its "What can be (efficiently) automated?" core. In 2006 and 2008, e.g., Wing [2006, 2008] presented her "Deep Questions", reminiscent of Forsythe's [1968] view of the discipline. Based on the idea of computational or algorithmic thinking, computing was yet another tool such as language or mathematics at our disposal.
Her questions, hence, were:

1. "P = NP?"

2. "What is computable?"

3. "What is intelligence?"

4. "What is information?"

5. "(How) can we build complex systems simply?"

[76, 142]

Wing also acknowledges the multidisciplinary aspects of computing, advocating for computational thinking for “all fields of endeavour.”

About Today

And while Wing was shortly before the 2010s, I think it is safe to say we're closing in on the present. What is it then, that remains today? Tedre [2015] wrote his history of Computing in 2014. Now, yet another 11 years later, a lot has changed and even more seems to have stayed the same. Before, I promised two digression that I want to pick up on now, as they seem to answer both question – what changed and what remained the same.

Is this real science?

"Deep nostalgia for mathematics by people who would like to label themselves mathematicians but certainly were not" [128, p. 45]

About this quote, I said that it was the foundation for one of the discipline's biggest insecurities. That insecurity is that of not being considered a proper science. It is this fundamental issue that, in my opinion, drives many of the issues we see in Informatics education today, and beyond education, many of the issues we aim to discuss with this critique.

Such a fundamental insecurity, the one of potentially being a fraud, rather than a real scientific discipline, is what divides Informatics today. The separate fields, which slowly emerged during the 2000s (though not precisely in the same configurations), are still a rough guideline for the different sorts of Informatics specialisations. Oftentimes they even roughly constitute the different Informatics studies one can choose from when starting out in the field. However, the spirit of seeing Informatics as more than a one-size-fits-all concept is somewhat gone. Rather, today, different fields within Informatics repeat what was being done to the whole of Informatics back in the 60s and 70s. Different fields within the discipline are now questioning each other's value as a field, accusing each other of merely "creating academicians creating recursively new generations of academicians". Repeating also the industry mantra, that so effectively changed the landscape of Informatics education in the 70s, that only that which creates economically feasible applications is of value.

When I first read "deep nostalgia for mathematics by people who would like to label themselves as mathematicians but certainly were not" [128. p. 45], I laughed. For I was told, a mere half year prior, by a Data Science professor in a discussion on HCI, that he views HCI as "those who think of themselves as better Sociologists". He warned me that in academia, one should respect boundaries – competences. Our competence, as engineers (as which he sees us), was that of engineering. If only I had read that quote earlier, it would have made for such a nice debate.

Interestingly enough, this image of Informatics is quite well established – not just throughout the field but even beyond it. Such is it that, Mass and Wiesner [2006], in their 2006 paper "Programmieren, Mathe und ein bisschen Hardware ... Wen lockt dies Bild der Informatik?" (Programming, Maths and a bit of hardware ... who's drawn to that image of Informatics?) find the same assumptions established among high school students. People who study Informatics, the students assume, end up engineers, writing programs. Furthermore, in their perception, the kind of person that succeeds in their Informatics study is male, well-versed in different programming languages even before starting their studies, and more likely than not has a bad haircut and smells [143]. Maybe something that, a few decades prior, would have been described as "computer bums", "educational asses" and "computniks". [128, p.43]

Those who do, against all odds they think to be against them, decide to study Informatics, are then frequently surprised by how abstract the field really is.

And, again, anecdotally, as somebody who's been teaching in Informatics courses for the last five years, this is true. We do frequently see students surprised by the diversity of the field. That unfortunately goes in both directions though. While those, not perfectly fitting the well-established image of the Informatics student. Those who are not versed in different programming languages before starting to study, who are not necessarily male or who, beyond the "how do we" questions, are also intrigued by the "why do we" or "should we do at all" aspects of the field usually rejoice at the complex and holistic approach of computational thinking. They enjoy the pondering upon societal implications, philosophical debates and the like. The others, however, unfortunately don't. Others in that case being the ones who do fit the image. Those who are already, for the better or worse, "engineers", those who already know all the "hows" and, more likely than not, don't want to consider there even are "whys". Like its topics, the study is diverse though, and the students perfectly fitting this very dated cliché are outnumbered by diversity.

Interestingly enough though, it is still those students, who seemingly rule the field. Even when being outnumbered by diversity, even when realising their image of Informatics is limited at best, it is still those students that judge others on how well they fit into the studies. Those that sometimes drive others out of the studies for their lack of programming skills, their mindset, or just their diversity. This phenomenon is also shown in Håpnes' and Rasmussen's [1991] work. Those, who fit the image of the classic Informatics student, also dictate it [144].

Of course, this is a vicious circle and not the students' fault alone. This self-perpetuating, hurtful image of Informatics needs to be addressed with both groups, students as well as educators. However, it is also this misconception, in my opinion, that leads to another, even bigger issue in Informatics education.

Implementierungswut

That is the one of what we refer to as "Implementierungswut" – which roughly translates to "desperate need for implementing". This concludes my promises to further digress, as it is, in my opinion, related to what was mentioned before about the 50s: The field of computing had thus far focused on the possibility of building computers, chasing after potentially better (for whatever that may mean) machines, rather than taking the next step of figuring out what could be done with them.

as well as:

"Pioneers of programming reminisce about the 50s' as a shoot from the hip time." [131]

Again, reading about this made me chuckle. Especially considering the subsequent move away from building the next big thing, the next fastest, cheapest or most complex computer. Moving away from these goals that may an authority of the time made out to be nonsensical, Informatics moved away from building the next big computer to implementing the next big application. For a very recent example, see the developments in "AI". Industry, once again focusing on the "how", drives to create bigger, better, faster, cheaper machines, never once stopping to ask "why" or "should we". This trend, of course, is portrayed in media, is repeated by people. It is stuck in the heads of high school students, thus carried into universities and there repeated by some of the educational staff. It is that dire, desperate need for implementation that is blind to a bigger context of technology. While I agree with the notion of computational thinking and see it as a key competence that Informatics scholars should possess, it is very easy to stop there. It's simple to force this idea of implementing a solution onto a complex reality that technical implementations are simply not fit to work in. I believe, it is this, that Burstall [1992] means, when he says:

"Working with computers and working with other people who work with computers, we develop our own particular attitude to our emotions and our own particular view of reality, or own particular style as the computing community. This style affects both our working lives and our personal lives. It runs a good deal less deep than our inheritance of Western European (Greek/Jewish/Christian) culture or Japanese culture, or whatever we happen to be landed with; but because the computer metaphor is a powerful one it probably has a significant influence on the way we see our world." [145]

It is perfectly relatable to fall for this. For it isn't just easy to do so but also feels good:

"But what is the style of thinking inculcated by the usual practice of computing? Computing systems seem to be totally understandable; you start with a general notion of what is supposed to happen and work your way down through module documentation to the high level language code and eventually even to the object code produced by the compiler. It is all there; you can spell it out. So the computational viewpoint suggests the possibility of completely grasping and mastering our world in a formal framework of concepts and definitions." [145]

To completely grasp a small world, is partially the fallacy of computing that we refer to as Killer Story. It's a combination of technological determinism and a hero story of computing. It's the law of the instrument, the Birmingham Algorithm, if you want, mixed with a skewed public perception of Informatics. It is also, I believe, the reason for this angst of ever touching another field's jurisdiction. This separation of concerns (funnily enough a core design principle taught in many programming classes) is also applied to science as a whole, thus fuelling ancient debates on what is and isn't science, Maths and Engineering and gatekeeping methods within their respective fields.

The carrier bag theory of Informatics

Our approach to Informatics Education is less strict. We argue for Informatics to be seen as "Gestaltungswissenschaft", a view that demands what goes missing in the aforementioned approach – the awareness, responsibility and respect implied by the word Gestaltungswissenschaft.

Access, participation and inclusion

We see the need for access, participation and inclusion (see e.g. [141]). As mentioned before, in Western countries, access disparities when it comes to gender have long been reported as problematic. Despite many initiatives by governments and funding institutions the gender gap in computer science education is rather stable. Furthermore, intersectional perspectives are important to acknowledge – like the underrepresentation of BIPoC (Black, Indigenous, People of Color), people from low-income backgrounds and others. Systemic inequities, such as lack of access to computers, tech resources and computer science programs in schools, exacerbate gaps in access and participation. Stereotypes and myths on who belongs in computer science serve as gatekeeping.

Didactic Approaches

Traditional lecture-based approaches struggle to engage students, particularly with abstract computer science concepts. Alternative didactic concepts that address the interplay between abstraction and concentrisation could help. Also, the integration of Universal Learning Design in higher education programs could support more inclusive learning environments. To realise new approaches teaching staff need adequate support and resources. Besides training, this could also include the exchange and support of educators, teachers, and professors from other disciplines, fostering an exchange not only between researchers but also encouraging the inclusion of “outside perspectives” for students.

Curricula development and interdisciplinary integration

The pervasiveness of digital technology in all life domains requires expertise that goes beyond traditional understandings of “core” computer science skills. For socially acceptable, democratic value-oriented, digital systems design, knowledge from the social sciences and humanities needs to be considered. Understanding computer science as Gestaltungswissenschaft means working with design studies and artistic approaches. Understanding the computer as a very special, a semiotic machine, requires theory building through media studies. This might also mean, depending on the national context, a broadening or easier inclusion of media studies perspectives, including new media studies, into the conversations (and faculties) of computing. This “interdisciplinary integration” would require balancing what is considered core topics and interdisciplinary topics. This forms a field of tension, given how “crowded” the typical computer science programs already are and how the study duration has shrunk since the Bologna reform. An integration of new knowledges then also needs a re-evaluation of learning goals within computing curricula.

Relating to our work, for the carrier bag, we would like to bring together constructionism (or other alternative learning concepts) and new materialism to computer science education to focus on material agency, feminist ethics of care and the co-constructive role of technology in learning.

While the hunt for a one-fits-all curriculum of Informatics can't be the right approach, a shared post-disciplinary one might help. Maybe, we need to ask ourselves, what the goal of studying computing or Informatics in a university context is and what skills we want to equip graduates with to empower them as informed citizens. This should be an education focused on care, cooperation, and critical engagement. An Informatics student should be trained in computational thinking, they should, however, also be trained to estimate the impact of computation. They shouldn't fall for the fallacy of trying to reduce complex issues to technically solvable simplifications, instead they should act, sticking with Informatics terms, as interpreter, between computation and a real, complex world.

“In short, Bildung is still a key idea in the digital university, not as a subject trained in a vocational fashion to perform instrumental labour, nor as a subject skilled in a national literary culture, but rather as subject that can unify the information that society is now producing at increasing rates, and which understands new methods and practices of critical reading (code, data visualisation, patterns, narrative) and is subject to new methods of pedagogy to facilitate it.” [61, p. 26]

¹ Steinbuch was involved in war crimes during the Second World War and, from his diary, it becomes clear that he was both an anti-Semite as well as racist [126].