Refereeing a lazy, but correct paper [duplicate]

Question

When refereeing a paper, I often have objections to the work that do not rise to the level of using the coercive powers of a referee to delay or reject the work. For instance, I find myself wanting to write: "If I were you, I would be embarrassed to publish a paper with [such unclear writing/such a silly comparison to experiment/so few parameters studied/incredibly ugly figures]."*

If the paper is not wrong or incoherent, just lazy or badly written, I don't feel justified in rejecting the paper or requiring a revision. I also don't want to invest hours of time acting as a language editor or graphic designer for figures.

Is there a politer way to express this sentiment and encourage the authors to improve their own work? Or should I simply take on good faith that the authors have done the work to the standards they are willing to publish, and that these may differ from my own?

*note: this is distinct from the case where the paper is so unclear that you can't tell whether their work is correct or not - where I think rejection is appropriate. I'm also assuming that this is not a terribly selective journal - in fact, it could be a journal like PLOS ONE, which explicitly says that importance to the field is not a relevant criterion.

Note: I've edited this to address a couple of points made by answers, and to focus away from the writing part, which has been answered in the past. Bad writing is one aspect of this question, but general author laziness is the larger one.

Answer 1

I think there are two important factors here:

1. Reviewers are typically the closest subject experts a journal can locate.
2. Most journals have (or want) a target audience that is considerably broader than the expert reviewers they enlist. Think of the industrial practitioners, novice grad students, and other interested non-experts.

If it takes you a lot of effort to parse and understand a paper as the related expert, it seems entirely plausible that a non-expert would not follow the paper at all. Particularly if the poor exposition is so bad that you can only follow the train of thought because you as an expert can fix errors in your head or make logical leaps, then it's definitely beyond the scope of the general reader. Both are valid complaints that could lead to a rejection of a paper, depending on the particulars of the work and the venue.

There are other avenues to address these problems than simply rejecting the paper. If the work otherwise is adequate, you might say so but tell the editor that the work does not meet the language standard for this journal and let them handle it. Some journals have a notion of "conditional acceptance" or "shepherding" where an agreement is worked out between the authors and editors that certain items must be addressed or revised prior to publication (however, they need someone to oversee such papers, so you might be volunteering for the role if you suggest it!).

Nobody wants a rejection, but most authors do want their work to be read and appreciated by others. Addressing these issues is a short term frustration for authors but a long-term win in terms of the visibility and impact of their work.

Answer 2

You seem to be assuming the premise that a paper ought to be rejected solely for being wrong or for being so poorly written that its correctness cannot be verified. I reject this premise - it is decidedly wrong in my discipline (pure math) and I'm pretty sure it's wrong in most other disciplines, certainly in the case of prestigious journals (Nature, Science, PNAS, etc.) that have to be selective and reject the large majority of papers they receive, many of which are not only correct but also quite well-written and significant and still end up being rejected.

The bottom line is that you as the referee are being asked to evaluate the paper for its overall contribution to your area of research. That includes not only correctness but also significance, originality, and the ability of the paper to effectively communicate its findings to the research community. You are being asked to apply your standards (possibly calibrated based on any guidelines you receive from the journal), not the authors' or anyone else's, and you can and should reject the paper for any flaw(s) that significantly diminish its impact and contribution. In my opinion that can also include poor writing, except possibly if the results are so groundbreaking (e.g., a poorly written but still verifiable proof of the Riemann Hypothesis) as to overshadow all other considerations.

Edit: following your clarification and citing PLOS ONE as an example of a journal whose standards you have in mind, I checked out their publication criteria. They include the following criterion:

The article is presented in an intelligible fashion and is written in standard English.

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. We may reject papers that do not meet these standards.

If the language of a paper is difficult to understand or includes many errors, we may recommend that authors seek independent editorial help before submitting a revision. These services can be found on the web using search terms like “scientific editing service” or “manuscript editing service.”

So, as you can see, my answer above clearly applies to PLOS ONE, and, I'm pretty confident, to many other journals of lesser stature than Nature and Science.

Answer 3

The role of peer review is not just to check that the manuscript is original, correct, and significant, but also whether its clear, concise, and well presented. In fact, from my own experience as author, reviewer, and editor, as well as from anecdotal evidence from other editors (also in other fields), owing to reviewing process, almost every revised manuscript is greatly improved over the original submission.

The authors of a manuscript that suffers in the way you have described will greatly benefit from improving their style, because more people will read, understand, and cite it. Hence they should appreciate your comments.