I’ve recently been reading several posts and articles about the current crises in scientific publishing, specifically about writing and publishing papers. Many current concerns stem from the use of AI, and others from the so-called draconian publish-or-perish system. A profound sense of discomfort exists in academia, especially among young scholars, a feeling that I’ve experienced and share. I agree with many of those criticisms, and I also believe we are in crisis. However, we need nuance and a bit of recentering in how we understand scientific publishing to keep the broader picture in mind. In these challenging times, it’s easy to lose direction.
Why do scientists write papers? What do we ultimately want to achieve? If the answer is practical- whether to increase H-index, to fulfil a research project outcome, or to achieve an academic qualification- as important as they might be, the risk is to lose the fundamental aim of scientific communication.
Scientific research, what do we do to answer our questions, is dialectically linked to scientific communication, as my mentor used to call it. This is to communicate the findings of our research to peers, allegedly one fundamental activity of every scientist. A responsibility that goes far beyond any practicality. Why? Because, as I like to repeat in my lectures, science is a collective endeavour. It differs dramatically from other beautiful and important activities, like art, that are essentially individual. Art doesn’t necessarily respond to anyone else, only to individual will. By contrast, science is collective by nature in two senses. First, to understand the knowledge that others have produced and, from that understanding, identify a knowledge gap to fill. This is how our research endeavours fit within a larger body of knowledge, not necessarily coherently (always in agreement), but articulated (in consideration of). This is the compromise scientific work owes to others, and what I normally call “standing on the shoulders of giants”: a deserved tribute to the knowledge that others have produced before us. In scientific communication, this happens by citing others’ work with substance[1] and using those ideas to build our own. Citing it is not, it never was, the last thing to do just before submission. It is the first thing to do, before even starting to think about research questions. That’s how classic breakthroughs have been achieved.
The knowledge others have produced is a fundamental input to our own research and must be paid back. Here is the second side: our contribution to the body of knowledge. That contribution must meet scientific standards: it must be produced using the scientific method, undergo adequate peer review, and be shared with the larger community as a published piece that can be used to push the frontiers of knowledge.
To meet the requirements of scientific communication, with the rigour expected of scientists, we need sufficient time. As my mentor, a best-in-class editor-in-chief, used to say, how can we expect to publish in a couple of weeks research that took several years and sometimes an important amount of money and people? We need to balance the time research deserves with the time to prepare a thoughtful, robust communication of findings to our peers.
Unfortunately, the academic system is becoming increasingly distorted. Tight funding schemes systematically shut down genuine scientific curiosity by predefining the problem and the answers we’re expected to produce. The demand for impactful research that solves our wicked problems leaves little room for creative exploration of new research avenues driven by the will to understand. Instead, it favours a short-sighted pragmatism that is killing the very spirit of science.
The AI’s hallucination of references is just the symptom. The problem lies elsewhere.
DISCLAIMER: I don’t use AI in my writing. Proud of my human mistakes; they show who I am and are my source of learning.
[1] Avoiding misattribution; for instance, see here.