For the past year, I have conducted a weekly “data physicalization” project with my colleagues at the Urban Institute. Each week, I would devise a new research question and give my colleagues the opportunity to input their own data using toothpicks in foam, stickers on the wall, wooden disks on a map, string tied together, dried beans in glass tubes, and more. Together, we would create a data visualization, be it an isotype chart, bar chart, scatterplot, histogram, or other illustration.

Because the project led me to new insights on how people approach visualizing data, I started to think about writing on my process and insights to empower others to experiment with similar ideas. When an academic journal invited me to write a paper on a topic of my choosing, I thought it was the perfect opportunity to do so.

Having made it to the other side of that process and ultimately deciding to publish the paper through the Urban Institute, I am now much less inclined to try to publish in academic journals again. Not because the process is difficult (it is), or expensive (also true—this journal wanted me to pay $1,800 for open access publishing rights), or tedious (it takes time to write, edit, and get the paper reviewed), or even because academic publishing is essentially broken (oh, it clearly is), but because the journal’s editors wanted me to make the paper less accessible and more difficult for people to read.

In case you’re not familiar, academic publishing generally follows a formulaic routine: After the paper is submitted, an editor or co-editor reviews for basic quality (known as a “desk review”), then sends it out for peer review—typically, other experts in the topic or field who will give the paper a more detailed, critical review. Peer reviewers can outright reject a paper or ask for major or minor reviews before publication, depending on the subject area and journal guidelines.

In my case, the paper was reviewed by two peer reviewers and an associate editor. While I found one of the reviewer’s comments useful, I felt the comments from the second reviewer and associated editor were off-the-mark and emblematic of misaligned priorities in the academic publishing world.

In both reviewers sets of comments, they focused on changing the language and style of the paper ways that seemingly intended to make the writing more difficult and convoluted. In addition to minor quibbles with style (e.g., change “Research by Bae and colleagues shows …” to “Bae et al. [12] shows …”), the associate editor wrote the following:

“In the paper, the author used the first-person pronoun “I” throughout the paper. It is understandable that only one author is in the paper. However, using passive voice or third-person perspectives to emphasize the results and the science rather than the researcher is better.”

The reviewer took it a step further by writing, “So, while the content of the paper is of interest, it needs a more scientific approach of writing (avoid also “I”, write more technical).”

This, in a single sentence, encapsulates one of the major problems with academic publishing: A desire to make language more technical, more difficult to read, and less accessible.

In my entire career, I have never heard anyone argue for passive voice. Writing in the active voice makes the work clearer and consistently more engaging than passive voice. Why the editors of this special call and the journal more broadly would promote less clear and more technical language is beyond me.

I emailed the associate editor to push back on the style requests, to which the lead editor responded: “For the “I” pronoun, it is true that there is no universal rule against the use of the first person in scientific writing. However, most researchers (including Dr. Shultz in Eloquent Science) mentioned that limited use of first-person pronouns in scientific writing is more acceptable.”

The editor’s citation of Shulz’s book, however, is misguided. Instead of arguing that limited use of first-person pronouns is “more acceptable,” Shulz writes (emphasis added),

Some teacher or professor in your past might have taught you to avoid the use of the first person (I or we), leading to a forced marriage with the passive voice. To appear disconnected from the research, common practice among authors of scientific and technical documents is to favor the passive voice, with the person who performed the simulation unstated and irrelevant. Such obtuse writing style has not always been the preferred style. Prior to the 1920s in the United States, active voice and first-person pronouns were quite common in scientific writing. Because science is done by individuals who make conscious decisions in designing, implementing, and communicating their research, such an air of impersonality, frankly, is disingenuous. We are intimately tied to our research and bias creeps in. The least we can do is acknowledge it…. I believe the first person can be quite effective when used sparingly and with purpose.

Regardless of Schulz’s actual recommendations, I wasn’t going to win this argument with the journal or its editors. They were clear that all of the reviewers’ and editors’ comments needed to be addressed, including those on writing style. I may not be the best writer in the world, but I couldn’t in good conscience make the paper more difficult to read just to sound more scientific. I will never believe that less accessible writing is a good way to present analysis or research.

There is one other quote from Shulz’s book that I think aptly summarizes my perspective on why being a better writer (and presenter and data communicator) makes for better science:

“When we write or speak, we fail to convey our enthusiasm and to personalize our science within a proper context. Purging our personalities from our work sterilizes it. We scientists individually need to find our voices, our creativity, and our originality.”