Free (e.g. arxiv) vs. paid journals?

[bjnartowt]

Suppose I wanted to publish a physics paper. It costs nothing to publish to arxiv. Why (setting aside the obvious advantage offered by peer review and the mistakes it can catch) would I publish in a pay-journal if I have something that is free?

Sorry if this is an ignorant question. I am trying to learn the ropes in this field.

 

[Orodruin]

Putting your paper on the arXiv is not publishing it. There is no peer-review and the peer-review is typically seen as a quality stamp. That being said, depending on the field it is free to publish in many journals that make their money on subscription fees.

 

[S.G. Janssens]

Why (setting aside the obvious advantage offered by peer review and the mistakes it can catch) would I publish in a pay-journal if I have something that is free?

Please be aware of so-called predatory journals (some may look quite bonafide) that charge you money but actually hardly do any peer-review. Do not publish in such journals. It is much better to upload to the arxiv first (the arxiv is moderated, but not peer-reviewed) with the aim of publishing in a peer-reviewed journal not long afterwards.

A number of bonafide journals offer open access in exchange for a publication fee. If in doubt, carefully investigate the journal, its publisher and editorial board.

 

[Dr. Courtney]

Putting your paper on the arXiv is not publishing it.

I disagree. The meaning of publish is "to make public." Material posted to arXiv is made public, it establishes priority for the author(s), it may be cited by other works, and it is subject to review and discussion by the broader community after it is posted. It may not be subject to peer-review prior to posting.

Sure, the average quality of material at arXiv may not be the same as material published by Physical Review, but I've known a number of journals and conference proceedings whose standards are not much higher than arXiv.

Some sub-categories of arXiv seem to have even higher standards than the peer-reviewed literature in the field. For example, after this paper was accepted and published in the educational journal, European Journal of Physics, it was rejected from the Physics Education section of arXiv and instead posted in the Popular Physics section. Most folks familiar with arXiv know that the Popular Physics section is where questionable material gets re-categorized. Somehow, the peer-review process of the journal was not sufficient for posting in the Physics Education section. The arXiv posting was favorably reviewed by the MIT Technology Review: https://www.technologyreview.com/s/514636/us-air-force-measures-potato-cannon-muzzle-velocities/ .

A number of our group's papers have been published to arXiv without submitting them to peer-reviewed journals. Some of these are comments pointing out mistakes in papers of other parties which the original journals chose not to publish, perhaps due to editorial policy that prevents publishing of comments, regardless of merit. In other cases, we simply have too many irons in the fire to shepherd every paper through the peer-review process. Readers who know our reputation will read and cite our papers even if they only appear at arXiv.

Most of the pay journals don't offer any benefits over arXiv. Having had a close look at peer-review from both sides over the past 25 years (I review more papers than I publish), in many cases it has become more of a mechanism for controlling contrary views than assuring quality. Most of our submission choices are based more on the intended audience and the career needs of our younger co-authors than a deep and abiding respect for peer-review. When I need respected third party opinions on our work products, I email the paper to respected colleagues and solicit their honest feedback.

 

[Orodruin]

I disagree. The meaning of publish is "to make public." Material posted to arXiv is made public, it establishes priority for the author(s), it may be cited by other works, and it is subject to review and discussion by the broader community after it is posted. It may not be subject to peer-review prior to posting.

Don't start with this. You are very well aware that selection committees, grant agencies, etc do not agree with this and that the general jargong of "publishing" in physics is different from "to make public".

Sure, the average quality of material at arXiv may not be the same as material published by Physical Review, but I've known a number of journals and conference proceedings whose standards are not much higher than arXiv.

Yes, which is why the journal you publish in matters!

Anecdotal level evidence is also the worst kind of evidence.

 

[Dr. Courtney]

Don't start with this. You are very well aware that selection committees, grant agencies, etc do not agree with this and that the general jargong of "publishing" in physics is different from "to make public".

Good work stands and is recognized as such, even if it only appears at arXiv. Consider our paper:
Sheep Collisions: The Good, the Bad, and the TBI which only appears at arXiv.

My co-author was recognized by the Physics Department at West Point for this paper, and it played an important role in my hiring and promotion at the Air Force Academy. Further, this paper has been cited by peer-reviewed papers a number of times. It has also received favorable attention here at PF and at ZapperZ's Physics blog. Jearl Walker also cites it at his Flying Circus of Physics web site.

I am not saying arXiv is the equivalent of the peer-reviewed literature. I tend to consider it as a valuable subset of the Grey Literature, which also includes the vast array of technical reports and theses that have not been peer reviewed.

"Peer-reviewed" venues are viewed as a higher tier, but a number of our co-authors have received favorable treatments in their applications based on material appearing only at arXiv. One co-author reported the lab director who hired them as an undergraduate lab assistant dancing with glee based on their published papers available only at arXiv. This co-author also received a coveted REU position at a prominent government lab and had hiring managers at that government lab already lining up the offers for permanent positions before the first week of the REU was completed.

 

[Nik_2213]

Slightly tangential, the proportion of eg Phys.Org reports whose 'source' link points to an EXPENSIVELY pay-walled document makes me cringe...

Retired, I still have a wide spread of interests, but no academic or industry affiliation to obtain a discount on such. So, to 'dig a little deeper', I must rely on serendipitous publication in 'open source' journals, or find an Arxiv version.

Latter, of course, comes with proviso that it is NOT fully peer-reviewed. Yes, the author(s) probably used 'Due Care', bounced notion and/or findings off trusted colleagues. And, yes, Arxiv versions do get amended. Still...

More tangential, I remember my city's formerly-famed Sci/Tech library cruelly culling its shelved stock as, 'everything is now on computer'. Sadly, many publishers then used their copyright ownership to pay-wall those removed reference books...

 

[atyy]

Putting your paper on the arXiv is not publishing it.

This is not correct, as @Dr. Courtney explains above.

 

[Choppy]

Typically in physics when we talk about "publication," we're talking about publication in peer-reviewed journals, and ones that are recognized in the field. And when I say "are recognized" I mean journals that my peers and I (i.e. the groups working in this field) most often read from, cite, and submit our own work to. Most of the recognized journals in my field are free to submit to with the option of paying to make the article publicly accessible.

Posting something on a preprint server like arxiv is not considered a peer-reviewed publication. And I think it's important that someone starting out understand this, because even with all their faults, peer-reviewed publications are the currency of academia. I've seen many CVs where people try to pass off something that's posted on arxiv as a something more than it is. That's a red flag when it comes to hiring.

 

[jtbell]

Do peer-reviewed journals allow citations (references) to arxiv posts? I suspect not. My experiences with publishing pre-date arxiv, but I can't imagine anyone in those days citing (in a journal) a preprint distributed by the old photocopy and snail-mail method.

 

[Dr. Courtney]

Do peer-reviewed journals allow citations (references) to arxiv posts? I suspect not. My experiences with publishing pre-date arxiv, but I can't imagine anyone in those days citing (in a journal) a preprint distributed by the old photocopy and snail-mail method.

Most of our publications in the grey literature (arXiv and DTIC) have been cited in peer-reviewed journals. We even have one paper that was only published on ResearchGate and our company website that has been cited in peer-reviewed journals. As a reviewer, I've reviewed for lots of journals and very few have a policy preventing citation of material that is not in a peer-reviewed journal. Most often, the criteria is that cited material needs to be in the open literature and available to reviewers, editors, and readers. For example, secret documents and material that is not in the open literature may not be cited. Material that is not in the open literature may usually be referenced as a "personal communication" or similar subject to the journal's policy on that.

Let me also mention that the potential career impacts depend more strongly on the whole CV rather than a single publication. One undergraduate I am mentoring has seven peer-reviewed papers (so far) and one arXiv-only paper. Does anyone think that one arXiv-only paper is going to hurt them when applying to graduate school? Later in their career?

I suspect some folks may be trying to suggest the standards for reviewing applications for tenure-track positions at R1 universities apply more broadly than they really do for other jobs (teaching-centered faculty jobs at lower-tier schools, industry jobs, etc.) I've been on too many search committees and been involved in too many promotion processes to be under that illusion. Sure, some folks on committees may still strongly favor peer-reviewed papers on CVs and promotion and tenure applications. But at institutions where the main focus of the job is teaching rather than research, an even mix of peer-reviewed and grey literature publications is unlikely to hurt and applicant for hiring, promotion, or tenure. For jobs where there is only a preference for PhDs (no firm requirement), even a lower proportion of peer-reviewed papers will not be significant.

But discussing career impacts would make more sense had this been posted under Career Guidance rather than Academic Guidance. Under Academic Guidance, my advice is whether a single paper is published at arXiv or in a paid journal is not a big deal. Under Career Guidance, my advice is by the time you complete your PhD, your career would be well-served if at least half your publications are in peer-reviewed journals (rather than conference proceedings, arXiv, or other grey literature venues.)

 

[atyy]

Do peer-reviewed journals allow citations (references) to arxiv posts? I suspect not. My experiences with publishing pre-date arxiv, but I can't imagine anyone in those days citing (in a journal) a preprint distributed by the old photocopy and snail-mail method.

Yes, PRL accepts arXiv papers as references. Old academic tradition considered PhD theses to be publications, in the sense that they are publicly available, the criterion mentioned by @Dr. Courtney. It is interesting that PRL instructs references that are theses not to be described as "unpublished".

https://journals.aps.org/authors/references-physical-review-physical-review-letters

 

[ZapperZ]

Suppose I wanted to publish a physics paper. It costs nothing to publish to arxiv. Why (setting aside the obvious advantage offered by peer review and the mistakes it can catch) would I publish in a pay-journal if I have something that is free?

Sorry if this is an ignorant question. I am trying to learn the ropes in this field.

When you seek a job that requires such expertise, or if you are looking to get tenure, or if you seek funding, they will ask for publications in peer-reviewed journals, and the most prestigious the journal, the stronger your application will be.

arXiv is NOT meant to replace peer-reviewed publication. That was never why it was created in the first place. It is a place for rapid communication between physicists so that early reports can be disseminated. It is not a scientific journal, and has never pretended to be. Ask Paul Ginsparg!

Zz.

 

[atyy]

Do peer-reviewed journals allow citations (references) to arxiv posts? I suspect not. My experiences with publishing pre-date arxiv, but I can't imagine anyone in those days citing (in a journal) a preprint distributed by the old photocopy and snail-mail method.

More generally, regardless of the publication rubrics, scientific ethics requires that one cites one's sources. Thus if a blog post happens to be such a source, then it must be acknowledge as such in the paper.

 

[f95toli]

Moreover, there is a difference between a well-established group/author with a good track record putting papers on the arXiv and a less-known author/group doing the same.
People working in a given field generally have a very good idea of which groups publish relevant papers and wil mostly l read them as soon as the go up on the arXiv, whether they have been peer-reviewed or not.
Hence, peer-review and publishing in well-known journal is more important in terms of attracting attention if you are NOT well-known; simply because the review process/journal serves as a mark of quality.
A paper published in PRL is probably worth reading regardless of the author, but the same is certainly not true for something uploaded to the arXiv.