First-person pronoun for things other than mathematical steps - singular or plural?

Question

The question Choice of personal pronoun in single-author papers has a good discussion of the motivation for using first-person plural pronouns ("we") instead of first-person singular (“I”), especially in math-type papers.

But here are three edge cases where the first-person plural seems a little funny for a single-author paper (but not as flagrantly wrong as thanking "our wife" in the acknowledgments):

1. To (our/my/the author's) knowledge, this result is new.
2. (We/I/the author) performed a careful review of the previous literature, but did not find any examples of X.
3. (We/I/the author) believe(s) that this method is easier to employ and more illuminating than the method outlined in the previous section.

Any thoughts on which is the best option for these three constructions?

Answer 1

I usually remove the speaker entirely in these cases. Certainly I would not use 'we' in any of those situations. 'We' is used in proofs because the author and the reader are (hopefully) both proving the result together, it isn't just the author. Indeed, if the author wrote "and then I see that (1) implies that the lemma is true" it sounds very much as if the reader won't see it!

"The author(s) believe(s) this result is new" is absolutely fine, but I normally solve it by using "This result appears to be new". So your three examples in your question become:

1. This result appears to be new.

2. A careful search of the literature did not find any examples of X.

3. This method seems to be easier to employ and more illuminating than the method outlined in the previous section.

Answer 2

To be honest, unlike the other answers, I wouldn't bat an eye if "we" were used in any of these three examples in a math paper (they seem more jarring out of context, but in context it's just a convention of the discipline). That being said, if one wanted to avoid both "we" and passive constructions, I think "the author" works just fine.

Answer 3

For pure math: to counter most of the other answers here (except Greg Martin's), I still find the use of first-person singular in research papers strange (excluding perhaps some special acknowledgements), simply because it's not the norm. To check this, I just browsed the new arXiv posts in my area (Number Theory), and all of the single-authored papers there use "we" including in situations like yours where it doesn't make traditional grammatical sense (e.g., abstract starts with "We describe...").

My thoughts are:

• first-person plural is perfectly fine in your examples; to me it's also a little more direct and less stiff than "the author", so preferable

• in general don't mix first-person singular and plural forms to refer to yourself

• try not to mix first-person pronoun usage and "the author" in nearby areas of text

• it's of course a matter of personal style, and as far as I know, there are no pure math research journals which would prohibit any of the options you give based on their style guidelines

Answer 4

Fashions change, and there is no hard and fast rule ... but I agree that to use "we" in any of the situations you have described would be odd.

The use of "we" seems reasonably well suited to cases where the pronoun is intended to include the author and the reader but the three situations that you describe are not like that. It is you who have done a literature search, you who are or are not aware of previous results, and you who found a particular method easier.

The use of the passive (as an alternative to any first-person pronoun) is still common but increasingly unfashionable ... and all round, "I" seems very well suited to the situations you've described.

Answer 5

Try using "the present author"

For academic writing, I also like using "we" instead of first-person singular, when discussing findings, proofs, etc. I have always thought of the "we" as referring to the reader and I, embarking on a journey of discovery together. We find that there is evidence for A, we find that the present method works well on problems of class B, we can see that the proof of Theorem D can be extended to deal with generalisation D, etc. r However, as you note, there are instances where one is making a statement that only applies to the singular author and which should not be attributed more broadly. For these cases a useful construction is to refer to yourself as "the present author" (see related answer here). This leads to the following framing for the cases you note, all of which sound quite pleasing and maintain the formality of academic writing.

To the knowledge of the present author, this result is new.

The present author performed a careful review of the previous literature, but did not find any examples of X.

The present author is of the view that this method is easier to employ and more illuminating than the method outlined in the previous section.

Answer 6

I do not like "we" at all in single author papers. In my view, "we" implicitly says that the author expects all readers to agree and basically synchronise themselves with what the author is doing, and this in my view is patronising.

Personally I don't have much problems with "I" as I think that in this way the author takes personal responsibility for what they're writing, and I think this is appropriate. I do however accept that "I" can come over as arrogant and self-centered, so I'm very careful with it. I have no issues with "I" in the three cases provided in the question, however I'd probably say "this method seems easier" in case 3. In case 2 one could say "A careful review did not..." (your sentences there seem longer than needed and the "I/we" could be removed as a side effect of shortening).

I do think avoiding first personal personal pronouns as far as possible without becoming too awkward is a good thing. George Spencer-Brown wrote somewhere that a reader, in order to properly appreciate and understand a mathematical proof, needs to do the steps themselves and observe the consequences, and he stated that the imperative is the best form for this, which I found very convincing. So in proofs I'll write "Assume... set... get... observe..."

Answer 7

In a single author paper, it does not matter much if one uses the royal “we” or the singular “I”. I tend to go with what is more common in the journal.

What is important IMO is proper attribution.

It makes a difference in a thesis for instance, or in other situations where the author has to be clear on who did what. A thesis written entirely using “we” could be problematic as it becomes impossible to identify the contribution of the candidate, especially if the candidate has published with collaborators. An interview talk where the presenter only uses “we” could also be problematic if the person was part of a team.

In your specific examples I would probably use “I” as I can’t see using the royal “we”. You did the literature review, and it’s difficult to see how you can attribute this to someone else or draw the reader into thinking they are involved in this review.