1Dear Reader,

2This is a special occasion. Instead of the traditional form of editorial, we are writing a letter today. It is addressed to you personally.

3Why this unusual way of introducing an issue? Because we would like to make you aware of one very important point: the editorial you are now starting to read was not produced by ChatGPT; it was written by one of DMS’s editors, then revised, improved and finally approved by the others. We would therefore ask you to bear in mind that the text you are reading is not the product of a generative intelligence; it is the result of the work of those who signed it.

4But you’ll probably want to check this out for yourself. Nothing could be more normal. To make sure it’s correct, it’s up to you to choose the right method from the four available.

5The first, and simplest, is to form your own opinion without any outside help. So, all you have to do is apply the same criteria to this text as you would to any other. Among other things, you will be looking at the relevance of the reasoning, the quality of the arguments, the coherence of the ideas, their sequence and the accuracy of the references. The problem is that if you do this, you’ll get some interesting indications, but they won’t enable you to clearly identify the author or authors of this editorial.

6So, for example, you will notice that the summary of articles presented here focuses on only one aspect. Others, many others, will be left out. That is why you could think that this summary is too partial and too subjective to have been produced with the help of software. On the other hand, the choice of terms and the extent of the lexicon will lead you to favour computer-generated writing. Another example: the quality of the syntax will persuade you to opt for a computer-generated document, while, on the other hand, the originality of the points of view will steer you towards a human-generated document. So no single argument will really convince you, no single clue will be discriminating enough to steer your judgement one way or the other.

7Your embarrassment will increase when you are faced with this question, which is not as obvious as it might seem: is what is man-made necessarily better than what is machine-made? In other words, will you attribute the qualities of this editorial to software power or human labour? In the absence of a precise answer to this question and the others that precede it, you will inevitably have to abandon the first option.

8The second one is the most traditional: put this text to the test of a plagiarism detection system. Excellent idea, but its implementation will immediately come up against a major obstacle: ChatGPT and other generative intelligence software don’t really plagiarise. To be more precise, they take extracts from here and there, but they take so many extracts from so many sources that the texts they compose give the illusion of originality. This means that borrowings are virtually undetectable. So the second option is out.

9The third one involves software recently developed, in particular by British and American scholars. The method they use (which, incidentally, also has other uses) is known as inference attack. It makes it possible to check whether a text has been used to produce another text and to measure what the latter owes to the former. Once again, however, difficulties arise.

10On the one hand, no matter how efficient these software programmes are, the results they produce contain a fairly large number of errors. On the other hand (and this is more serious), this method requires the prior availability of source documents or, at the very least, a corpus of documents of the type that could have been used at the outset. The contributions in this issue are far from being the only sources on which this editorial has been based. There are many others, in fact, from which ChatGPT or other artificial intelligence software may have drawn, if they were called upon. However, with a few exceptions, the designers of these softwares refuse to indicate the origin of the documents they use (or could have used). It is therefore impossible to start from the initial reference. This is why this third option is no more satisfactory than the previous two.

11After this triple failure, what option do you have left? Trust us.

12Admittedly, most of you (105,000 annual visits and 4,300 downloads, according to figures from Open Edition communicated by Caroline Rizza at the annual meeting of our Scientific Advisory Board) have never personally met the signatories of this editorial. Many of you may not even know exactly who we are, which laboratories and universities we work in, or what our areas of research are. The fact remains that DMS, the first French-language journal on distance learning, was founded just over twenty years ago and has been a household name ever since. So you could rely on the intellectual integrity of its editors. In other words, the magazine’s reputation and respectability could convince you that we would not be asserting the authenticity of this editorial if we had entrusted its writing to ChatGPT. This fourth option is the only solution available to you: all you have to do is put your trust in us…

13This preamble must obviously be read cum grano salis. It had no other purpose, through a slightly amusing diversions, than to draw your attention to the importance of the issue of trust in communication in general and in distance learning in particular. It so happens that this question is at the heart of this issue.

15Will you object that, by definition, a Varia is supposed to have no coherence or common thread? That is true. In retrospect, however, the problem of how trust arises and operates in distance learning systems and practices is one of the common denominators of the contributions in this issue. Three of the underlying questions bear witness to this.

16First question: is trust really so important in distance learning? The contributors unanimously answer in the affirmative: a climate of trust, they say in substance, is the sine qua non of all asynchronous (temporal distance) and/or mediatised (spatial distance) learning. In this respect, it makes no difference whether this learning takes place within the formal framework of instituted training or in other contexts.

17In fact, no distance learning course is viable if the learner does not have a great deal of trust in the system, the group and the institution to which he or she entrusts the care of his or her learning. But the teacher also needs to trust in the people he or she is teaching, and in the people in charge of the technical systems. And the latter must have no less trust in the designers of the systems and the institutions that authorise them. This confirms that trust is everyone’s business.

18This trust, however, is not spontaneously given or experienced. Nor can it be decreed. It is negotiated and built over time; it is gained and lost according to circumstances and contexts. What’s more, as Pierre Bourdieu says, it is based on “tacit indigenous presuppositions” (Bourdieu, 1982, p. 110). Now, for social or psychological reasons, some actors are more determined than others to accept these presuppositions. At the same time, some are more inclined, and have a greater interest than others, in instilling trust. Who are these actors? And who are the others?

19Among the answers below, you will find one that is somewhat unexpected, and which deserves a special mention. It comes from Hubert Boët’s article (“Artificial intelligence in Duolingo’s growth strategy…”). He says that the managers of Duolingo, the world’s number 1 distance language teaching company, are highlighting their investment in generative intelligence to attract users and improve the learning experience. But it is also, and perhaps above all, according to Duolingo, to attract investors. Because Duolingo, like all tech players, in education and elsewhere, absolutely needs stock market and speculative financing, and financialisation in general, to develop. Later, one day perhaps, it will be able to make their own money.

20This strategy could well succeed—we add. In fact, between January and December 2023, the US Consumer Technology Association’s Artificial Intelligence & Robotics index listed on the Nasdaq rose by nearly 35%, and a Bloomberg Intelligence report published in June 2023 forecasts that spending on artificial intelligence will grow by $1,300 billion in the United States over the next ten years. This shows the trust that references to artificial intelligence in general, and conversational robots in particular, inspire in business circles. And in this field, the education industries are at the forefront. Let’s not forget André Malraux’s famous phrase (1946, p. 16): “Besides, cinema is an industry.” To plagiarise it, we would say that distance learning is also a component of Edtech.

21Second question: where does trust come from? It requires a favourable state of mind. But how is this state of mind formed? For learners, it depends on their previous experiences and their motivations. But it also results from the content of the learning, the pedagogical and didactic methods used by the teachers and the design of the teaching. Trust depends no less on the actions and reactions of peers and on the quality of technical resources and systems, their reliability and affordances. In short, trust (like its opposite, distrust) is an individual and collective construct, rooted in a system of beliefs and values, in which human and social factors, ideal parameters and material aspects play a more or less equal role.

22Significant in this respect is the question posed by Yannick Stéphane Nleme Ze and Gaëlle Molinari (“Six main barriers to learner engagement in the University of Geneva’s MOOC discussion forums…”). The answer is no less significant. First of all, say the authors, there are “factors linked to the student’s personal life, those relating to the quality of supervision in the forums, as well as intra-individual and interpersonal factors.” To these are added “six other families of factors.”

23Not surprisingly, these “six other families” include trust in peers, trust in forums, and trust in the usefulness of the forums themselves (even though they are not compulsory). The authors also refer to “the degree to which the user believes that using a technology will enable them to meet their needs and improve their performance (perceived usefulness)”. And they insist on the influence exerted by “the quality of the potential results, the opinion of important people around them who recommend or do not recommend using this technology.” As we can see, trust is built on belief.

24Mutatis mutandis, it is this same question of trust that Mathieu Cisel (“Contrasting disciplines about the dissemination and the reuse of open educational resources in higher education”) raises when he wonders why some teachers are more willing than others to share access to their educational content. He specifically mentions the case of open educational resources. His answer draws on the theory of social exchange and its applications to e-learning. In particular, he cites the American sociologist George Caspar Homans who, while being one of the precursors of the theory of rational choice, nevertheless attributes to trust a central role in the way in which an actor anticipates the costs and benefits of the decision he has to make. So when a teacher makes his content available to colleagues he doesn’t know, he runs the risk of being assessed by judges outside his “circle of trust.” This is why, according to Mathieu Cisel, members of scientific communities whose interactions do not foster inter-knowledge and mutual trust tend to avoid this risk.

25Third question: why do the communicational, pedagogical, psychological, technological, social and institutional conditions of trust in distance learning differ from those in face-to-face training?

26To answer this question, Éric Bruillard, in a dialogue with Daniel Peraya, draws on his experience as a Mooc designer: “This is one of the difficulties of Moocs: you don’t control anything in a Mooc: you don’t know who will be there, how many will be there, who will react, what they will have understood, whether they are ready to do something, and in what direction it will go. Teaching mastery is impossible.”

27This statement about impossibility of mastery does not only apply to Moocs. In their review of their experience (“The unthinkables of moving image education in a distance learning context”), Valentine Favel-Kapoian and Fanny Lignon convincingly explain that, at a distance, “the teacher loses a set of cues on which he relies when he is face-to-face.” And they add: “Everything that has to do with non-verbal communication is lost, making it much more difficult to mediate and accompany the viewer’s gaze.”

28The question then becomes what tools need to be put in place to make specific forms of distance regulation. You will be interested to read the practical suggestions and proposals that the authors put forward to turn a disadvantage into an advantage. Above all, you will find in their contribution confirmation of what is one of the deepest convictions of the supporters and followers of DMS: whether formal or informal, distance learning is only useful and effective if it is not face-to-face training deported to a distance. It has its own rules, constraints and virtues. So let’s avoid indulging in lamentations about what is irretrievably lost when distance replaces face-to-face. From an educational, social and economic point of view, there are also many things to be gained from it, which should be highlighted without any qualms.

29The Debate and Discussion section includes a contribution from Jean-Luc Rinaudo. At Daniel Peraya’s request, Rinaudo revisits previous contributions to this year’s section: the re-conceptualisation of the opposition between presence and distance and, by extension, the theme of presence at a distance. His starting point is the observation that “presence takes many forms, and each researcher who takes up this notion endeavours to make it intelligible according to his or her own epistemological paradigm.” Jean-Luc Rinaudo is therefore endeavouring to identify the approaches and perspectives favoured by the contributors to 2023. As you will see, his own contribution is marked by an approach tinged with psychoanalysis. Above all, it is not intended to be either a conclusion or a synthesis. It is intended to be “an opening, an invitation to continue exploring what presence means in distance learning, or more generally, in the digital world.”

30As usual, this issue has its share of critical readings. The first, on Anne Cordier’s book, Grandir informés, was written by Florence Michaux-Colin; the second, on Vincent Chabault’s book, Le livre d’occasion, sociologie d’un commerce en transition, was written by Maël Rannou; and the third, on the book coordinated by Éric Bruillard, Apprendre avec les énigmes. La résolution collective d’énigmes comme levier pédagogique, was written by Daniel Peraya. It follows on from his discussions with the author.

32A final word—or rather a thought—for Guy Berger, whose death on 13 June leaves a void for those of us who were marked by the brilliant and incisive thinking of this outstanding researcher and professor, one of the great founding figures of Vincennes University. Although he was not a specialist in distance learning, Guy Berger left a number of noteworthy contributions on the subject. These include his paper at a conference organised by the Séminaire Industrialisation de la Formation (Sif) in 1994. In the context of international negotiations on the cultural exception at the time, he argued in favour of “the possibility of an educational exception, correlative to the almost inevitable process of the free movement of goods.” And to underpin his argument, he added: “we can only reach a clear definition of educational effectiveness and determine what an educational good would be if we […] reopen the whole question of the educational process and the services that condition it, and do not confine ourselves to a product logic” (Berger, 1994, p. 311 and p. 317). As you can see, thirty years on, the point has lost none of its relevance.

33That, dear reader, is the welcome we wanted to give you on the threshold of this final issue of 2023. We hope you will read it with all the attention it deserves. We trust in you.