# Writing a paper as a single author in 1st person singular

I am writing papers in theoretical computer science sometimes being a nonanonymous single author. Earlier, I used the 1st person plural in such papers, e.g.:

We will show that the complexity classes X and Y coincide.

I am neither a native English speaker nor extremely good at English. Recently, I've gotten an advice from a native English speaker (who is into science but not a computer scientist) to use the 1st person singular instead:

I will show that the complexity classes X and Y coincide.

He claimed that this style is widespread in "Nature" and other toplevel journals. I'm not used to this style of writing at all; sentences in the style "I have shown that Γ is well-defined on ℂ \ ℤ⩽0" inside proofs seem very unnatural to me.

Which style is more appropriate for theoretical computer science papers?

It is very clear to me that this question is strongly biased by the area (otherwise, I'd have asked ell.se or referred to https://academia.stackexchange.com/q/2945/7734), so I'm asking at cstheory. My question is addressed to mature theoretical computer scientists who are also native English speakers with a perfect command of English.

• I ("we"?) think the plural can work well sometimes in a single-author paper, interpreted as including the author and reader(s) together. "We will (together) show..." But that might be my idiosyncrancy. – usul Mar 3 '17 at 18:41
• I am but wouldn't weight that fact too heavily... – usul Mar 3 '17 at 21:55
• For whatever it’s worth, here is an analogous question on Academia. – Wrzlprmft Mar 4 '17 at 13:19
• When writing a single author paper I have always thougt that "we" refer to me and the giant whose shoulders I'm standing on. – user35592 Mar 6 '17 at 6:58
• @usul I don't buy the argument that "we" refers to the author and the reader. It's certainly not what I think when reading or writing a paper. How do you understand the "our contributions" subsection? To me, using "we" for single-author papers makes sense because writing style should be consistent, regardless of how many authors there are. – Thomas Mar 6 '17 at 19:19

The pronoun used is not in reference to the author or authors but about the reader, about the audience. We is used to include the reader in the process of discovering and understanding the result. Write the paper for the audience, not for you. This seems like a good place to plug Knuth, Larrabee, and Robert's notes on writing style in Mathematics. Pretty much everything said there applies to writing in TCS.

• Great document. – Lamine Mar 4 '17 at 14:36
• @LeonMeir Maybe, but this isn't a question of English. This issue is about stylistic choices within the genre of TCS papers. Using "we" in the way I described is a way to engage the reader and humanize the work through appealing to the excitement of discovery. My answer, and the other answers here, should apply to non-English TCS writing. – Logan Mayfield Mar 6 '17 at 14:35

I am not a native English speaker, but I don't think this is important here. The appropriate thing is not what an English professor or a poet suggests, but what is standard in the particular field. And from my experience, "we" is the standard in theoretical computer science. As noted by usul in his comment, you can interpret the "we" as the author including the reader. I can think of several reasons why "we" is better than "I":

• It sounds more natural and less egocentric.
• A paper in theoretical computer science is not about your personal opinion, but you state and prove general facts. Replacing "we" with "I" in terms such as "we thus have" is therefore weird since what follows is generally true, not only for you.
• Submissions are often anonymous. You don't want to reveal that the paper has been written by a single author in the submission (and you don't want to change "we" to "I" in the final version).

Edit: An exception are parts of the paper that are subjective, e.g., the acknowledgement section: I would like to thank András Salamon and Boaz Barak for pointing this out.

• @LeonMeier Maybe the paper gets rejected and you have to submit it elsewhere, where it is anonymous. Also, you might want to be consistent with other papers you write and maybe submit anonymously. – Christian Matt Mar 4 '17 at 0:38
• I wonder why you think that other disciplines are more about opinions than facts? How much opinion is there in mathematics or physics? – Barmar Mar 4 '17 at 22:48
• @Barmar I was more thinking about the humanities. – Christian Matt Mar 4 '17 at 22:54
• I agree with the bulk of this answer, but feel that the "natural and less egocentric" part is too strong. To me, referring to "we" in a single-authored publication has a slight connotation of en.wikipedia.org/wiki/Royal_we in addition to the default I-and-the-reader, probably reinforced by the differing conventions in the many non-mathematical papers I read. This is usually not a problem and I don't hear the author's voice sounding majestic, unless the sole author then goes on to acknowledge others in the plural. Since the reader had nothing to do with the helpful discussions... – András Salamon Mar 5 '17 at 10:52
• I'm not sure what the relative frequency is, but for the acknowledgements there's also the option of using the third person: "The author would like to thank..." – Peter Taylor Mar 6 '17 at 8:21

I'm not a native English speaker but this is what I do in single-authored papers. (Or rather, what I would hypothetically do if I wrote single-authored papers :) )

1) Throughout most of a technical paper, I use "we" to refer to the joint effort of the author and reader. That is, my interpretation is that a sentence in the abstract such as "We show that C=D." talks about a technical result that is present or reported on by the paper.

2) In subjective parts, such as the acknowledgement section, or when talking about opinions or conjectures, then I would use the first person. So I think "I conjecture that this result can be extended to the X setting" is better than saying "we conjecture..."

I'm also not native english. However, I serve as a referee for about $20$ papers per year in theoretical computer science both for journals and conferences since a lot of years. This what I experienced up to now:

• I never came across a paper speaking in the first person, even when it was authored by a single person;

• even the first plural person ('we') seems inappropriate or too colloquial in some occasions (when it is used too often in the same paragraph for example).

Of course, in other disciplines it might be much different.

• I strongly support the second point: The papers that I usually read prefer sentences like "It is shown in this paper that the complexity classes X and Y coincide." and I try to avoid "we" and "I", too, as much as I can. – koalo Mar 4 '17 at 12:50
• @koalo This is actually very awkward and unnatural phrasing. Almost any paper I have read would say "In this paper we show that the complexity classes X and Y coincide." – Sasho Nikolov Mar 4 '17 at 16:28
• @SashoNikolov Obviously I read too many papers by non-native speakers :-) – koalo Mar 4 '17 at 16:51
• Indeed, most of papers is written by non-native speakers ;-) – Maczinga Mar 4 '17 at 16:53
• "It is shown that" seems common enough in 'pure math' papers, but uncommon in TCS. It might seem good because modest, but to me it reads as excessively awkward and even confusing. (It sounds too much like "It is known that", which points to prior work. And it strains so much to avoid "I/we" that the reader is left wondering, "who?".) – Andy Drucker Mar 8 '17 at 15:00

I am not native in English and whenever it is the case of single euthored work, first person of singular sounds clearer and more honest because the first person of plural might always give impression to the reader that the question is about multi authored paper or at least that exist other unquoted people who support author's views.