By RACHEL MEISTER KO. FREITAG*

Integrity in research is built by training scientists who are aware of their role and responsible for their practices.

The scientist stereotype is that person alone who, out of nowhere and instantly, appears as the creator of something that changes humanity. Nothing could be further from reality, at least for some areas of knowledge. Science is not the result of a single, individual achievement. Perhaps the story told only reproduces that attempt that worked, suggesting a linear and peaceful path.

Perhaps there are geniuses, perhaps there are those who do science like this.

It's not my case. My starting point is that science is collective and collaborative work. There is no scientist, there is a group of scientists who come together to do science. The union can be in the laboratory, but it can also be in the cafe or at the bar. It is the encounters, views and diverse perspectives that contribute to the advancement of science.

I am one of those scientists who do collective science. More than a collective thing, I found out that I do the science of great teams. A very, very large group of people working towards a cause or research problem. In this way of doing science, the concept of authorship – understood as a unique and individual expression – loses its meaning. That's why we talk about collaboration, not authorship. There are proposals for a collaborative taxonomy (one of them is the Credit) that recognize the contribution of each person who worked on a discovery. The person who designed the experiment has the same value as the person who collected the data; who did the statistical analysis has the same value as who wrote the final version; who validated the experiment has the same value as who raised funding, and so on. The authorship belongs to all the people who participated, without exception or hierarchy. Furthermore, complex problems are not solved at once: if someone ever announces a cure for cancer, for example, it is because there was a whole engendered set of previous scientific products fitted together like a jigsaw puzzle, whose final result is the such a discovery.

Since the final product is collective, and there are different products, in this way of doing science it is inconceivable to think that people who worked on the same project, developing the same task, report their findings differently, especially methodologies and problem propositions. What science strands that are based on the concept of authorship, the science strand of large teams that is based on collaboration calls textual recycling. Journals have a clear policy regarding the difference between textual recycling and plagiarism (in Brazil, see the publications approved by the Associação Brasileira de Linguística: Linguistics notebooks e Abralin magazine). It makes no sense for a search problem to be paraphrased in order to bypass similarity-detection applications. Even worse: the reproducibility crisis in science, which makes results obtained in experiments not possible to be reproduced, is highly affected by the patrolling due to the application of concepts such as authorship and originality. In large team projects, textual recycling to explain the research problem or detail the methodology is a constant. Unnecessary effort is expended to “say differently” what has already been said by a collaborating collective voice; not only is it unnecessary, but it can be harmful to “say differently” a methodological routine that has been developed, with the result that reproduction is unfeasible.

Software that detects similarities between texts is used to support decisions regarding research integrity. In products derived from the same project of a large team, it would be unexpected to identify rates of dissimilarity. The motto of Google Scholar is “On the shoulders of giants”, a phrase used by Sir Isaac Newton (but that's not his) when he “discovered” the laws of general mechanics. We can only move forward because there is a solid foundation built, gradually. If each new product needs to reinvent the wheel, or rather say differently that the wheel has already been invented, just to satisfy percentages of similarity, science does not advance, science lives the circularity of reaffirming what has already been said, with words many different.

It is not difficult to paraphrase what has already been said. Incidentally, nowadays, with large language models such as those that underlie GPT3, paraphrases can be achieved with a click, and more: similarity software does not capture that it was the artificial intelligence that produced it. The big debate in science today is how to use artificial intelligence to advance science; our posture is that it should be used (because we work for it to exist!), but always with transparency, explaining its use (Chat GPT has been listed as a research collaborator, including access to Scopus). Technological innovations demand reflections on practices: if artificial intelligence can produce scientific writing, how to equate the role of authorship? If similarity software is still not capable of identifying whether there was interference by artificial intelligence in the preparation of the text, how can we know whether or not it was used? ChatGPT-type applications in science automate work (they are capable of summarizing, paraphrasing, systematizing information, organizing references, preparing programming codes, reviewing programming codes, among other uses) and free up demand for time and cognition for scientists to do that which is its primary role: to think of solutions to problems! Use them in the flow of scientific research, from the initial stage (https://elicit.org, for example, writes a systematic review), to correcting the final version of the text (https://edit.paperpal.com, for example), is an irreversible path that has the potential for great revolutions. The discussion about authorship, in this sense, becomes even less relevant. But the discussion about ethical and integrity research practices, explaining the use of these applications in each of the stages of the development of the work, is a path to be followed.

Integrity in research is built by training scientists who are aware of their role and responsible for their practices. As a member of the Open Science movement, I advocate transparency in practices and the democratization of science.

*Raquel Meister Ko. Freitag is a professor at the Department of Vernacular Letters at UFS.

The A Terra é Redonda website exists thanks to our readers and supporters.
Help us keep this idea going.
Click here and find how