Oxford University Press's
Academic Insights for the Thinking World

First person pronouns and the passive voice in scientific writing

Imagine you are explaining your research to a friend. You might say “I tested this factor” or “We examined that effect”. But when you later prepare a written version for a scientific journal, you would probably eliminate the “I” and “we” in favour of the passive voice, which, unfortunately, can sometimes present a challenge. Here is an example from a chemistry journal, but the discipline is immaterial:

The influence of residual chloride ions on the catalytic activity, the kinetic aspects of the oxidation of methane over these catalysts, the nature of the active sites, the influence of metal particle size and reaction products on the activity, the observed changes in catalytic activity with reaction time and the effect of sulphur containing compounds are examined. (Appl. Catal., B, 2002; 39: 1)

The 55 words before the verb “are examined” at the end require the reader to maintain an exceptional commitment to the content, and many would fail. Although this example is extreme, unwieldy passive constructions are common in scientific research articles and not peculiar either to native or to non-native English speakers. Yet top-heavy sentences—those with a very long subject and a short predicate—are unnatural in English outside scientific writing. Normal practice, according to the principle of end-weight, is to put the complex material, the detail, towards the end of the sentence, not at the beginning.

Why, then, do authors avoid “I” and “we” and routinely embrace the passive voice? After all, there is firm advice to the contrary. Authoritative style guides such as Day and Sakaduski’s Scientific English and Montgomery’s The Chicago Guide to Communicating Science argue that introducing the first person removes uncertainty about the agent of the action; it reminds the reader of a human presence, the person with whom the knowledge should be associated; and, practically, it avoids having to manage complicated passive sentences. The journal Nature is explicit, insisting that short reports should contain a one-sentence statement starting “Here we show” or an equivalent phrase. And, not to be discounted, using the first person in a string of statements “I compared”, “I tested”, “we found” does make writing easier, even automatic.

Shorter Oxford English Dictionary, 2007. Oxford University Press.

One compelling reason for eschewing “I” and “we” is to preserve objectivity, or at least its appearance, as is customary in other areas of writing. Thus, in the influential guide On Writing Well, the journalist and teacher William Zinsser points out that newspapers do not want “I” in their news stories, and magazines do not want it in their articles. Readers expect to read the news and be informed objectively—an aspiration that surely extends to readers of research journals. Intrusion of the first person is a distraction from the content and can distort the message. Consider the following example, which, initially, does not contain the first person:

A random effects meta-regression showed that [the] proportion of women in the sample was not significantly related to gender difference effect size. (Psychol. Bull. 2014; 140: 165)

Apart from the technical language, the content is clear enough. Now introduce the first person by adding “We”, thus:

We showed in a random effects meta-regression that the proportion of women in the sample was not significantly related to gender difference effect size.

Although the content is preserved, the focus of the sentence shifts from what is important to the reader, the random effects meta-regression, to what is unimportant, the role of the authors. Simultaneously the sentence becomes wordier.

In his essay How to Write Mathematics, the great expositor and mathematician Paul Halmos described the use of “I” as sometimes having “a repellent effect, as arrogance or ex-cathedra preaching”. When it appears in the present rather than past tense, the preachiness of the first person becomes more pointed still, as in this example:

I derive and compare two new estimators that help correct this small-sample bias. (Ecology 2015; 96: 2056).

The emphasis is on the author’s action: “I derive and compare”. Yet it is unnecessary. In the following rephrasing, the emphasis is on what is important, the new estimators:

Two new estimators are derived and compared that help correct this small-sample bias.

As a device, declarations of the form “I show”, “I derive”, “I compare” do make easy writing, though their repeated use can transform an exposition into a testimonial, not softened by using the plural “we show”, “we derive”, “we compare” instead. When “we” refers to the sole author of an article, the testimonial becomes a regal pronouncement.

Of course, there are circumstances where the first person is entirely appropriate, for example, in articles comprising reminiscences (“I first met”), in position statements (“We believe”), and in reviews reflecting a personal view (“I interpret”). In all of these uses, the author is central to the account. The use of “we” is also apposite in referring to the research community (“How can we explain?”) and to humankind (“How do we perceive?”). It is also deployed to effect in mathematical and related expositions, where it does not mean the authors alone but the authors and reader in joint activity to develop the argument (“If we substitute x for y, we see”). Other special uses of “we” are enumerated by Quirk and his colleagues in A Comprehensive Grammar of the English Language.

What of the rationale that the first person avoids top-heavy passive constructions? The example at the beginning of this article can indeed be rescued by introducing “We” and moving the verb from the end to the beginning, thus:

We examine the influence of residual chloride ions on the catalytic activity, the kinetic aspects of the oxidation of methane over these catalysts, the nature of the active sites, the influence of metal particle size and reaction products on the activity, the observed changes in catalytic activity with reaction time and the effect of sulphur containing compounds.

But, as elsewhere, the cost is the shift in focus. Since the material in the sentence is essentially a list, that fact can be exploited in a rephrasing that avoids “We” and better prepares the reader for what comes next:

Several effects were examined: the influence of residual chloride ions on the catalytic activity, the kinetic aspects of the oxidation of methane over these catalysts, the nature of the active sites, the influence of metal particle size and reaction products on the activity, the observed changes in catalytic activity with reaction time, and the effect of sulphur containing compounds.

The moral of all this is that explaining your research to a friend is not the same as reporting it in a scientific journal. Your friend is interested in you, whereas the reader is interested in what you have found, in other words, the “news”. The two audiences for your account are different, and so is the need for “I” and “we”. Eliminating the first person from a written report does not, though, require top-heavy passive constructions, just rephrasing that goes beyond the merely automatic.

Featured image credit: Hand by nattanan23. Public domain via Pixabay

Recent Comments

  1. Fumie Costen

    This explains one of the very fundamentals in writing scientific reports in a very logical manner extremely well !!

  2. Aimin Song

    Very enlightening indeed!

  3. josh

    right. next paper i’m using the second person. YOU’RE not going to believe this but…

  4. Paul Martin

    I like it!

  5. Frans Cornelissen

    Nice read. One argument I’ve heard and that is not explicitly mentioned here for using the “we” form is that it establishes a sense of ownership (both in writer and through attribution, also in the reader) of the presented work. Which would enhance the writing and reading experience (and it does, in my personal experience).

  6. Andrew Stockman

    The sole use of passive voice in Methods sections can get pretty clumsy and tedious? At least there, I prefer to mix up active and passive. (And often elsewhere.)

  7. Rafael Huertas

    Nice reflection and very true. In Spanish all the arguments works similarly.

  8. Scott Stevenson

    Very helpful examples, here.
    In scientific writing one often sees use of “this study” as a substitute for “we”, putting the agency into a project or activity rather than a person or persons. For example, “This study addressed the role of …” This compromise allows more fluid sentence construction while avoiding the first person.
    We didn’t do it. The study did. :-)

  9. adam reeves

    the final comma in the last re-write is critical for understanding. Thanks to David Foster for leaving it there, and resisting the copy-editor who (no doubt) tried to remove it.

  10. Hema Radhakrishnan

    Excellent article! I (first person) also feel that writing in first person favours established researchers. When fairly junior scientists, someone new to the field writes in first person, a certain amount of unconscious bias creeps in from the readers end.

  11. Rhea Eskew

    Nice article. Good writing in English with avoidance of strange constructions and passive voice is enthusiastically and personally endorsed.

  12. Ipek Oruc

    Great examples and a wonderful read!

  13. Cong Yu

    The post is very helpful. I passed it to my students and some colleagues. Passive sentences are much less often used in Chinese. The sentence “Two new estimators are derived and compared that help correct this small-sample bias” would sound very odd if translated literally in Chinese. Most likely it would be translated as “We derived and compared two new estimators that help correct this small-sample bias”. Plus there is no past tense in Chinese. I am not sure whether Chinese scientists would use less passive sentences in their writings though.

  14. Bevil Conway

    One possible risk of the passive voice is that it relieves the author of some responsibility for the work. But maybe the passive voice raises the moral bar, since we are aiming to report “news”? I wonder if articles using the passive voice are any more likely to be those in which we find scientific misconduct. Or rather, is it conceivable that scientific misconduct is found in articles that use the passive voice? In the first formulation, you can discount the query because it’s just me asking.

  15. Ramina Shrestha

    Very useful article specially for new researchers.

  16. Ravu Verenagi

    Great article. Having spent many years working in the bureaucracy, I am now in a role with a research institute, not as a researcher, but a manager and find research writing interesting. In a few weeks, we will screen abstracts for papers that are going to be presented at a consortium conference we are organizing. A good read and tip in preparation for that task ahead.

Comments are closed.