An extended lunch break: a response to Wolfgang Klein [Journal of Pragmatics]

In his brilliant article published 35 years ago, Wolfgang Klein (1989) drew attention to the increasing disparity between writing and reading. Humankind’s unceasing inventive spirit has helped to facilitate the process of writing to get from the laborious and time-consuming hammer and chisel to today’s ease of text processing which does a great deal of writing almost automatically. But so far, he claims, nothing has been done on the side of reception, which leads to an ever-growing imbalance between the amount of text production and the amount of reading that fits into a single scholar’s lifetime. As a solution, he first proposes, tongue-in-cheek, to return to more arduous production methods involving goose quills and hand-rubbed ink or indeed hammer and chisel, but then proceeds to suggest a more realistic reduction of the output of scholarly material to just 30 pages per scholar per year. This is coupled with the hope that scholars will invest more time and care into the production of these 30 pages and that readers will more easily be able to cope.

Quite clearly, this suggestion has failed completely. Scholars, in their attempt to meet the quotas required for tenure and promotion and other academic honors and rewards, continue to publish ever increasing volumes of academic publications, and thus the conundrum of far too many publications and far too little time to read them all has grown to proportions that must have been difficult to predict at the time when Klein wrote his piece. And this does not even include what is awaiting us in the near future. It turns out that Klein’s comment in 1989 that “[i]n a few years, we’ll most likely be able to feed our old essays – along with one or two articles from another author – into a machine that will then automatically generate a new composition or even an entire book” (Klein 1989) was remarkably prescient. The recent emergence of Large Language Models that can imitate the writing styles of anyone whose data they have been trained on will lead to a further exponential growth of the output of academic texts.

Moreover, even limiting each scholar to Klein’s suggested 30 pages a year would not have solved the problem. The amount of research produced in a single year across all disciplines would still be huge and the data base of publications would grow every year, as presumably we don’t only want to read this year’s crop. Even if we confine ourselves to humanities scholars in the USA, the 30-page limit would generate about three and a half million pages a year and almost 18 million over a five-year period. If we confine ourselves to philosophers in the US, we are still talking about 180,000 pages a year and about a million pages over a five-year period.

A more radical solution is clearly needed, and we are happy to note that such a solution appears to be close at hand. Many scholars have already started to implement it of their own accord. What is needed at this stage is just a little prodding to make it more widely accepted and more generally implemented. The solution is as obvious as it is simple. Instead of Wolfgang Klein’s well-meant but ultimately futile suggestion to reduce the amount of writing, we propose a drastic reduction of the amount of reading. Instead of being expected to read more or less everything that has been published on the topic of a new article, scholars are expected to read no more and possibly much less than 30 pages of academic material on that topic per year.

The advantages appear to be self-evident but let us spell out the most important ones. For the authors themselves, it is a time-saving measure on an unprecedented scale. Why should they bother with the tedious and time-consuming activity of reading boring papers written by other scholars, when they can use the time much more efficiently to thrash out any number of new papers by referring mostly to their own previous publications? They can produce a lot more papers in a fraction of the time. This will not present any problems to the system because – under the new scheme – nobody else (or at least no other human) will read their papers anyway. At the time of Klein’s paper, some environmentally aware nitpicker might have objected that this increase in publications might be bad for all the trees that are bound to fall victim to the surge in academic publications created under this scheme, and librarians across the world would have had to apply for extra funding to multiply the storage capacities of their libraries. But in today’s world of electronic publications, this is a trifle and can be disregarded.

The increased number of papers produced by any scholar, especially early career scholars, also has the effect that they can present much more impressive lists of publications to search committees or funding agencies. After all, nobody is interested in how much they have read. What counts is how much they have published. For search committees and funding agencies this will also make life much easier. They can give up the pretense that only quality counts, and evaluators can abandon the tedious and time-consuming work of reading the applicants’ work in a pointless attempt to assess its quality. It is much easier and faster to just do the counts. As a consequence, the new system will also be much fairer. The traditional system generally provided an advantage to clever and gifted students. This is clearly biased and not in line with equal opportunities for everybody. But if articles are only counted rather than read, everybody has an equal chance for success, and we are another step closer to a perfect world.

Perhaps one of the most important advantages for the individual scholar is that it will be much easier to make “new” discoveries. If you do not have to read the existing literature on a topic, it is a piece of cake to present a snippet of insight as a groundbreaking new discovery. It does not really matter if earlier generations already had the same insight. The earlier generation may even have shown that the snippet of insight is altogether faulty and ill conceived. But in a system in which nobody reads other people’s publications anyway, this is of little consequence, and, as already Klein has pointed out, “not without a slight touch of satisfaction, (…) our work, in contrast to that of the natural scientists, does not harm anyone” (Klein 1989).

For journal editors and peer reviewers this would also be a great help. No need to point out to hopeful authors that they should have read the relevant literature (the old classics as well as the most recent research in the area) because nobody expects them to do that. This would also relieve journal editors of the burdensome task to write all those disheartening rejection letters beginning with “We have read your paper with great interest, but unfortunately…”. In fact, journal editors and peer reviewers would soon become obsolete altogether. The authors would just upload their most recent papers to a dedicated webserver and include the publication in their CV. Done and dusted, and a lot of effort saved.

In contrast to Wolfgang Klein’s proposal, there are clear indications that we are already making a great deal of progress on this road to a much better world. The number of scholars joining forces to make this come true appears to be growing every day. As yet, the old system still mixes with the new one, for instance in the form of editors who still try to insist on thorough background reading, but these are the losing battles of the old guard. The choice between writing or reading will soon cease to be a major predicament, and we can all happily engage in writing texts without any tedious interruptions necessitated by having to consult some source or check some reference.

Klein worried that his solution would hobble progress in science, and he suggested certain tweaks to his system that would ensure that scientific progress continues, for example, by allowing the quota of annual pages to be raised for Nobel Laureates. Some might think that our scheme to limit reading will have even more devastating consequences on scientific progress. But here too recent technological advances and trends in scholarship could help, without the need to lift the proposed limits on required reading. In Google’s recent launch of their AI model Gemini, they claimed that Gemini 1.0’s “remarkable ability to extract insights from hundreds of thousands of documents through reading, filtering and understanding information will help deliver new breakthroughs at digital speeds in many fields from science to finance” (https://blog.google/technology/ai/google-gemini-ai/). This model can allegedly search a database of over 200,000 scientific articles, find 250 of the most relevant to some topic, summarize and annotate those 250 articles, and then produce figures or charts that illustrate trends in the data reported in the summarized articles. And it can do all this while the scientists who queried the model are on their lunch break. One assumes that if things went more slowly because of lack of bandwidth, scientists could always extend their lunch breaks.

Anne Bezuidenhout and Andreas H. Jucker are the Co-Editors-in-Chief of the Journal of Pragmatics.