The power of language was already well understood by ancient civilizations. The Greek philosopher Aristotle for instance, or the Roman statesman Cicero both investigated the structure of language in their scholarly works and delved into rhetoric — the art of persuasion or influence through speech or writing. Among the different instruments adopted by rhetoric, hyperbole is a central figure of speech, which is based on the deliberate use of exaggerations as an intensifier to impress the audience. Here, a sort of implicit agreement between speaker and listener — or between writer and reader — is that those exaggerated figures are not meant to be taken literally.

For many researchers, writing scientific articles represents a challenge. In fact, training on writing skills is seldom part of conventional curricula. Over the years, the principles of scientific writing have been the subject of lively disquisitions1, and certain canons have eventually become widely accepted2. Large numbers of publications, tutorials and journal guidelines have highlighted the need for a simple and accessible style, which avoids pompous prose as well as jargon, lengthy acronyms and subjective statements. All in all, a scientific article should hardly come across as a masterpiece of rhetoric. And yet too often scientific literature features styles and linguistic elements that are not quite effective, besides being inappropriate. Hyperbolic language is one of these.

“Science needs to maintain a high level of rigor and the data should speak for themselves.”

There may be a number of reasons why some researchers adopt hyperbolic language in scientific articles, and we recently tried to get a sense of the community’s opinions in this regard with a short and informal survey on our X (formerly Twitter) account. The participants were given the possibility of choosing between four different options. Considering the granting schemes currently operating in academia and the pressure to publish, the first two options (“to impress the editors” and “to impress the reviewers”) refer directly to the possibility of facilitating the publication process. The other two, namely “to emphasize impact” and “to impress the community” are more nuanced. The former may well imply a genuine belief in the extraordinary value of the study; the latter is again a little more cynical and reflects on the effect that language can have on the legacy of a paper and its echo within the community once this has been published.

Admittedly, the different answers are somewhat interconnected: an emphasis on impact may intentionally be sought with the peer-review process in mind; moreover, editors and reviewers are often members of a relevant community themselves. Therefore, all those aspects cannot be truly considered in isolation and may well complement each other. Interestingly, about 40% of the respondents interpreted the use of hyperbolic language as a means to affect the publication process, either at editor assessment (“to impress editors”: 21.4%) or during technical evaluation (“to impress reviewers”: 17%). The majority of the voters instead opted for “emphasize impact” (39.1%) possibly considering the intrinsic value of a piece of research, while the remaining 22.6% opted for “to impress the community” focusing on ways to amplify its resonance among peers.

The statistical significance of this poll is obviously limited (based on sample size — 407 votes — and sample selection), and the narrow list of available choices just allowed us to scratch the surface of the phenomenon, preventing any firm conclusion. Nevertheless, we would like to thank all people who took part, and those who shared comments and considerations with our editorial office, which are useful to reflect on this issue more deeply. From the feedback we received, we realized once more how journals can play an important role in popularizing good practice in scientific writing. Therefore, in the following, we would like to mention a few aspects of our policies that relate to the use of hyperbolic language.

The use of expressions such as extraordinary, unprecedented, paradigm-shift and related terms to qualify the findings of a given study is relatively frequent in today’s papers, and some studies even suggest an increasing trend over the past years3. However, one of the first things to notice is that such hyperbolic statements almost always contradict one of the most important principles of sciences, which is objectivity.

Our editorial office does not perform a stringent screening to spot the presence of hyped language or catchy phrases during initial article evaluation. We do, however, assess the strength of evidence for the major claims of the article at length. In doing so, an editor carefully considers the advances of a study against the background of the published literature, performing an independent screening of related articles. Accordingly, claims of novelty or unprecedented levels of performance are not taken as granted, but are evaluated against relevant benchmarks. Sometimes authors may inadvertently choose reference studies somewhat arbitrarily: a work may appear as an important advance over the references covered in the paper, but may turn out to be less disruptive upon a more thorough literature search. Therefore, it is always preferable to avoid a narrative that is excessively based on superlatives, and focus rather on justifying the advances and explaining the implications of a work for the community. When articles receive positive evaluations during peer review and are accepted in principle for publication at Nature Catalysis we will edit all phrases that are considered excessive or subjective and discuss with authors more appropriate forms to describe their own findings: eventually, advances and novel aspects of a study should be clear from the context, rather than being advertised though overzealous marketing.

In a culture that induces many individuals to self-promotion and competition, the habit to resort to hyperbolic language may be taken with a certain lightness. It is therefore important to remind ourselves that science needs to maintain a high level of rigor and the data should speak for themselves.