In a recent issue of Nature, an editorial featured pronouns, not vis a vis gender but rather concerning that hoary problem of whether to use the first person (Scientific language is becoming more informal, 2016, vol. 539, p. 140). The editors were responding to a paper that charted rising numbers of “we” and “I” in journal articles, particularly in biology (Hyland & Jiang Engl. Specif. Purp.). I responded with a letter, which of course they didn’t publish, so here is an expanded version of what I wrote to them.

Dear Editor,

The first word of the double helix paper by Watson and Crick is “We” (a free copy of the paper is here). Given the iconic stature of this paper, the claim that scientists do not, or should not, use first person pronouns needs to be dismissed as myth. But Watson and Crick use the personal pronoun for mental acts, where objectivity demands accepting the inescapable subjectivity of thought. This contrasts with physical acts (i.e., experimental methods), where the scientific method stipulates that agency is irrelevant. This view is so pervasive that the scientific literature brims with sentences that start like this: “To ensure accuracy, samples were measured…” Here, main and subordinate clauses have different subjects, a violation that all other kinds of (English) literature forbid. What would be interesting to learn, and not addressed by Hyland and Jiang, is whether personal pronouns are gaining ground on agent-free action in the descriptions of experiments.

The Nature editorial writer considers using “we” to be an example of informal usage, apparently because they imagine using personal pronouns to be frowned on by the authority. But as shown above, authority is not frowning. The notion that “we” is somehow louche (OK, informal) is likewise shown to be imaginary by the “To ensure accuracy, samples were measured…” type of sentence structure: how can such pronoun avoiding usage be called “formal” when it violates actual rules of grammar? The structure is sloppy, one that every teacher would mark wrong.

Using pronouns is not a matter of formality or its absence but about agency. To write “We think…” is to claim agency for an act (thinking) that requires an agent. Not since Moses have thoughts emerged from a bush. Methods are more complex. Not only is there an expectation that who ever uses the measuring tape would get the same length, true attribution of agency in the methods is likely to be tricky. It would be odd to write: “To ensure accuracy, the third author measured…” but therefore, strictly speaking, false to write “To ensure accuracy, we measured…” Nevertheless, just as writing “We think…” allows agency for a thought that probably occurred in only the third author’s mind to be shared by all the authors, representing the polite fiction that all the names under the title are—actually—authors, it seems reasonable for “We measured…” to allow agency for an action to be similarly shared. I’d pick claimed agency over bad grammar any day.