Wikipedia talk:Notability (academics)

This discussion was begun at Wikipedia:Votes for deletion/Nicholas J. Hopper, where the early history of the discussion can be found.

---

See Wikipedia:Notability (academics)/Precedents for a collection of related AfD debates and related information from the early and pre- history of this guideline (2005-2006) and Wikipedia:WikiProject_Deletion_sorting/Academics_and_educators/archive, Wikipedia:WikiProject_Deletion_sorting/Academics_and_educators/archive 2 for lists of all sorted deletions regarding academics since 2007.

Proposed deletion of C5

The rationale of C5 is I think more than a little outdated. It was based upon the days when only a few full professors had named chairs. However, now in many major university at least 1/3 and often more Profs have them, named chairs have proved to be good fund raising tools. Even Oxford and Cambridge, who used to have only 2-3 "Professors" now have Profs of almost everything. KISS principle, anyone truly notable for C5 should have other means of verification so we don't need it. Ldm1954 (talk) 19:00, 6 April 2024 (UTC)

I'd agree. There are some historic and prestigious chairs (Rouse Ball, Lucasian, Regius...) but nowadays it seems like every large department at every large university has at least a couple. This criteria would be redundant for truly notable academics, since they would only have a chair because they already passed one of the other criteria. So seems like this criteria will only introduce 'false positives' when universities make up new chair positions for marketing, or reward academics who are simply good at fund-raising or politicking. WikiNukalito (talk) 20:26, 14 April 2024 (UTC)

The rationale may be outdated (as for C6 in the section above) in the Western world, but it may not be outside of it. It is hard enough for even the most notable academics outside the West to get published and cited in journals that are indexed online. and the other criteria of this guideline, without people here automatically saying that their universities are "minor" simply because of where they are located. Phil Bridger (talk) 20:38, 14 April 2024 (UTC)

I don't think this refers to positions like Vice Chancellor or heads of large departments. If anything, western Universities seem more likely to invent the Bob Roberts Professorship in Whateverology precisely in order to make their academics appear more notable. WikiNukalito (talk) 20:50, 14 April 2024 (UTC)

What if we add a date cutoff to account for this, like "was a named chair prior to 2015", or whatever year the recognition became discernibly devalued? BD2412 T 01:58, 15 April 2024 (UTC)

There are many kinds of named chairs. Named chairs given to junior faculty to give them a step up, named chairs given to administrators as a slush fund, named chairs given to people who hold some kind of service position, the named chairs that every single full professor in the German system gets, ... and then there are also named chairs reserved for faculty of outstanding scholarly achievement, beyond that of an ordinary full professor. It is only the last kind of chair that C5 is supposed to cover. "Distinguished professor" is not the same thing as a named chair but it also means roughly the same thing, a step beyond ordinary full professor. Similarly, Canada Research Chairs come in two tiers, tier 2 is for more junior faculty, and only tier 1 really corresponds to the intended meaning of C5. I would be opposed to getting rid of C5 altogether; I think that using it for distinguished professors and tier 1 Canada Research Chairs, for instance, is still valuable. But if there's some way of clarifying that only certain named chairs should count, I'd be in favor of that. —David Eppstein (talk) 20:47, 14 April 2024 (UTC)

Stepping back slightly, I think most respondents here are approximately in agreement about what academic notability is. However, if you have spent any time on NPP/AfD/AfC and probably a few other places, the definition of academic notability is often tweaked, or even manipulated. I would like to simplify as much as possible.

I am a scientist, and maybe in other areas it is less clear. All the higher level Canadian Research Chairs I know would amply qualify from their other achievements. Similarly for really major chairs, for instance those that come directly from University Trustees.

The same goes for #Clarify C8 below on being an editor. If there are special cases where being an editor is really highly prestigious and is done at the expense of other academic achievements, please correct me. My experience is that it is really expected as part of service, along with reviewing papers, proposals, appointments etc. Peer recognition via citations, awards, reviews and sometimes news and a few other is what matters -- peers are "secondary sources".

Less is more. Ldm1954 (talk) 22:23, 14 April 2024 (UTC)

There are several areas of academic biography that I sometimes work in where markers of success such as citation counts are totally worthless, either because everybody lists everybody else as authors and they all get thousands of citations (high energy physics) or because citations are handed out like very valuable keepsakes and consequentially nobody gets any (pure mathematics). For those areas, it is important to be able to point to indicators that within the field certain people are seen as standing out above the others. Buggy as they are, C5 and C8 both provide that sort of indicator. C3 is maybe more consistent, within a specific field, but it doesn't have the same consistency from one field to another. —David Eppstein (talk) 22:51, 14 April 2024 (UTC)

I think (hope) that everyone here knows that citation numbers are field dependent. If unsure, compare to others in the field as has been mentioned various times, with various recipes provided. To me high, discipline adjusted citations are a strong peer vote, and I will take that every day over university chairs which are often retention/hiring enticements. Similarly senior awards matter. Best is both peer citations and awards, although I know of several where the publications were weak but the awards were convincing, e.g. from industry, and another where the citations were suspicious; careful analysis is needed.

Being an editor, WP:MILL, not notable. I have never seen that claimed as a significant tenure achievement, it is expected service. My personal experience is that almost everyone will review/act as a editor for Nature or Science, but in strong, second level journals it is very hard to get academics to be good citizens. Ldm1954 (talk) 23:16, 14 April 2024 (UTC)

If they were merely "field dependent" we could adjust for field and be done. No, there are fields where they are totally useless. Do you want to kill off other notability criteria, because we have C1, and then lose most of our coverage of people in those fields, because C1 does not work for them?

Also, being an editor, meaning on the editorial board, is so run-of-the-mill that I prefer not to mention it at all. It is only being editor-in-chief of a major journal that counts for C8. And no, it is not true that "almost everyone will review/act as a editor for Nature or Science". Maybe almost everyone in experimental physical science will act as a referee for those two journals, but referee is even less exceptional than editor, and even some STEM fields like mathematics and computer science are very poorly covered by those two journals to the point where I think very few mathematicians and computer scientists have done any form of editorial work for them. —David Eppstein (talk) 23:53, 14 April 2024 (UTC)

Sorry, but I see no difference in editorial roles. I also have experience with people declining to be editors, so do not agree with you there. As Ghandi said,

"honest disagreement is often a good sign of progress."

Let's switch things slightly so we can perhaps come to an agreement of sorts. Do you agree with my argument that academic notability should be based upon peer review? If so we can start to discuss more definitively, including perhaps votes. Ldm1954 (talk) 00:13, 15 April 2024 (UTC)

You see no difference between editor-in-chief, member of editorial board, and referee? Ok.

I do think that using signs of recognition by other people in the same discipline is better than using hype for one's research (the usual way WP:GNG might be accomplished for academics) and better than trying to determine for ourselves based on its content alone whether some piece of research is significant. I would not call this "peer review" because as Wikipedia editors we do not actually seeing the things that usually count as peer review: referee reports on papers, reviews of grant proposals, letters of recommendation for tenure and hiring, etc.

One of those signs of recognition is being given an advanced title (C5). These things typically happen only after peer review, letters of recommendation, etc. We don't see the peer review but we see the result. Another sign is being chosen to head a major journal (C8). —David Eppstein (talk) 01:50, 15 April 2024 (UTC)

If there's no difference between referee and editor-in-chief, my CV just got a heck of a lot more impressive. XOR'easter (talk) 03:57, 15 April 2024 (UTC)

I at least meant editor-in-chief and editor. Of course not referee. Ldm1954 (talk) 04:22, 15 April 2024 (UTC)

@Ldm1954, I've tried to do quantitative analysis of HEP citations based on the Scopus metrics of a subject's coauthors. It is impossible to determine what is "well above the average" number of citations when hundreds of people of wildly varying writing/experimental involvement and seniority are cited in every new consortium update on form factor measurements or whatever, each of which gets thousands of citations. There are assistant professors with Scopus h-indexes close to 100 for whom I could not find a single paper with fewer than 25 coauthors, much less anything secondary highlighting specifically their contributions. C5 can at least serve as a proxy for how a subject's scholarship has been received within their subfield, as evaluated by their university; the nuance and subjectivity then passes to how prestigious the school is (overall or in that subfield) and how prestigious a named chair or distinguished professorship is at that school. If we're to adjust C5, I think it should be to clarify what shouldn't count and to discourage knee-jerk "passes C5" assessments when it's not clear what the status of the position and/or the university is. JoelleJay (talk) 02:25, 15 April 2024 (UTC)

I am happy with the idea of clarifying what C5 (and perhaps C8) are not, small steps (?). Related, maybe I should have used "peer recognition" as we generally do not see reviews as @David Eppstein says, and words matter. I understand @JoelleJay's point about HEP, which is why I '''always''' look for awards and similar '''plus''' the citations. Ldm1954 (talk) 03:22, 15 April 2024 (UTC)

I agree that the profusion of chairs would make a presumption of notability for professors of no note who only have mention of their named chair on their employer's website. We can't write a biographical article on something like that. The notable named chairs might provide notability for holders per ANYBIO but the usage of C5 is probably over. Chris Troutman (talk) 00:06, 15 April 2024 (UTC)

Oh boy, it's the GNG-fanatics again trying to rework everything into a clone of GNG even when it makes no sense. "We can't" = "I don't want to because GNG tells me not to", very different from "it is actually impossible". —David Eppstein (talk) 01:52, 15 April 2024 (UTC)

We could swap the order around and say something like, "The person has held a Distinguished Professor appointment, a named chair that indicates a comparable level of achievement, or an equivalent position in countries where...". XOR'easter (talk) 03:56, 15 April 2024 (UTC)

I like the way this wording suggests that only this kind of named chair should count. I think it more accurately reflects our actual practice in AfDs. For instance, named chairs like Cambridge's Dorwart Professorships aren't generally taken as satisfying C5. These are visiting assistant professorships, basically postdocs with a fancy title, not permanent or higher-rank positions. The criterion-specific notes, 5(b), make clear that this doesn't count, but your wording matches this better than the status quo. —David Eppstein (talk) 07:44, 15 April 2024 (UTC)

I agree with both of you on this. Qflib (talk) 16:28, 15 April 2024 (UTC)

I agree that WP:NPROF C5 is the most problematic of the NPROF criteria. General comment that I've seen a few cases going by recently where a named chair was tied to a dean-ship. It might be good to add language to indicate that such named chairs are not a pass of WP:NPROF C5. On another note, perhaps it would be good to explicitly say that WP:NPROF C5 is a proxy for WP:NPROF C1, and that there should be at least some plausible case for the kind of impact we're looking for in C1. (I think this is generally along the lines of what XOR'easter is suggesting just above.) Russ Woodroofe (talk) 15:06, 15 April 2024 (UTC)

Re "is a proxy of C1": plenty of distinguished professors are in non-journal fields such as the humanities where there is little hope of passing C1 (or C3). C1 works well only for journal fields. That's another reason why we need to keep C5; it helps normalize our standards across different fields of academia. —David Eppstein (talk) 18:18, 15 April 2024 (UTC)

@David Eppstein, surely a notable academic in these areas would have awards (C1d/e and C2) and/or C3 & C4 and C7. As has been said here (not just by you), citations are not everything. If there is no peer recognition, just some Dean's award of a named chair then I would not be convinced, no verifiable evidence. Ldm1954 (talk) 18:49, 15 April 2024 (UTC)

David Eppstein, when I refer to C1, I mean the broad sense of having had a substantial impact on an academic field, and not the narrow sense of having a high citation count. As Ldm1954 suggests, there is usually some other signal of this. Perhaps it would be good to think about C5 as a little more like C2, where we disregard a lot of early-career and otherwise more routine prizes. Russ Woodroofe (talk) 19:01, 15 April 2024 (UTC)

This discussion began with the comment that 1/3 of all profs at major universities have named chairs to which I would reply [citation needed]. Across most of the US a named chair requires an endowment in perpetuity for the position, which can be $1-8 million dollars, which is a huge stretch for all but the most significant positions. Yes, you can look at the Harvard or Stanford faculty lists and say, "See! 1/3 of their full-professor faculty have named chairs!" but I don't think that you'll find anything like this across the "major university" landscape (in the US R1, R2, and significant liberal arts colleges), where most departments are struggling to get their first endowed chair. And if this criterion makes 1/3 of the faculty at Harvard notable, are you really willing to make the argument that 1/3 of the Harvard faculty are not significantly more notable than the average professor?

Can we make a meta-rule for this guideline, that "changes to the guidelines without examples of 'professors who shouldn't be notable but pass a current criteria' will be rejected without discussion"? I'd really like to know one deletion argument for an existing article on a professor who passes this case. Then I could consider changing my views. -- Michael Scott Asato Cuthbert (talk) 09:56, 16 April 2024 (UTC)

I don't know about the 1/3 number, and I am sure that almost all Harvard professors are indeed notable. But it does not take much searching to find other places with a large number of endowed chairs e.g. this was the first result on Google for 'university of Arizona named chairs', and this is just one school/college. Now maybe all of these people are notable, but the proliferation of chairs does kind of devalue them and make them a weak indicator of notability. WikiNukalito (talk) 11:53, 16 April 2024 (UTC)

One data point is an anecdote, but: in my R1 department of some 50 regular-rank faculty members, we currently have zero of them with endowed chairs. (We have roughly two chairs empty from recent retirements that we can hire people into, but they are unavailable to already-hired professors; we are working to fill those empty chairs, but that is a slow and difficult process.) —David Eppstein (talk) 17:16, 16 April 2024 (UTC)

Different universities, different fund raising. Taking Northwestern University Computer Science here as a comparison, the university is R1 and aggressive about fundraising. I counted 18 full professors (exclude by courtesy and adjunct) or which 6 have endowed chairs. (A fortuitous match to my 1/3.) Ldm1954 (talk) 18:56, 16 April 2024 (UTC)

So, to clarify, instead of "1/3 of the faculty", as MSC interpreted your remark earlier, you really mean "1/3 of the full professors in certain STEM disciplines"? I imagine that at least 1/3 of the full professors in CS at Northwestern are notable by other standards as well. Having that many notable by C5 is non-problematic, and makes it easier to identify the notable ones both for article creators and in deletion discussions. —David Eppstein (talk) 20:35, 16 April 2024 (UTC)

Yes, I meant full profs, and not adjuncts/courtesy since the latter are padding. Northwestern has some junior chairs but less as it is harder to sell those to donors. Ldm1954 (talk) 22:54, 16 April 2024 (UTC)

Shifting back

I would like to return to XOR'easter's proposal and vote Support for it as a good compromise that seems (?) to satisfy most participants of this debate. Ldm1954 (talk) 18:59, 16 April 2024 (UTC)

I agree. XOR'easter's proposal seems reasonable. - Enos733 (talk) 19:29, 16 April 2024 (UTC)

Yes, if it makes things clearer. It seems to be the way C5 is interpreted at AfD anyway. Phil Bridger (talk) 21:01, 16 April 2024 (UTC)

I'm fine with the change. I don't think it affects how people who often participate in academic AfDs will interpret the guideline, but it will help people who are trying to read the guideline for the first time and understanding what it means. -- Michael Scott Asato Cuthbert (talk) 21:23, 16 April 2024 (UTC)

For clarity, I understand this to be the change to: "The person has held a Distinguished Professor appointment, a named chair that indicates a comparable level of achievement, or an equivalent position in countries where...". Assuming I understand that correctly, I support that, too. --Tryptofish (talk) 21:37, 16 April 2024 (UTC)

I strongly recommend and request that a formal RfC be launched and subsequently advertised in relevant places to discuss a the proposed substantive change to this notability guideline. ElKevbo (talk) 23:39, 16 April 2024 (UTC)

I would agree with that if we were discussing the deletion of C5. But (assuming I understand correctly) what we are discussing here is the wording change that I just quoted in my support comment. I have trouble seeing that as a substantive change. --Tryptofish (talk) 23:54, 16 April 2024 (UTC)

I'm fine with this change and fine with doing it without a formal RFC. I think it is a clarifying minor wording change without any intended change to the substance of the guideline. —David Eppstein (talk) 00:34, 17 April 2024 (UTC)

I went ahead and made the change on those grounds. XOR'easter (talk) 18:46, 17 April 2024 (UTC)

Looks good to me. Cheers, Qflib (talk) 19:58, 18 April 2024 (UTC)

Clarify C8

Point b mentions that fringe and pseudo-science journals should not be included, but there are by some measure over 45000 academic journals, most of which have editors, most of whom are not notable. I think there should be some more clarity on what counts as a major well-established journal. Perhaps some combination of age, impact factor or association with important works, authors or events (though the last is a little circular). WikiNukalito (talk) 20:12, 14 April 2024 (UTC)

Impact factor is only a used criteria for quality/notability/etc. in the sciences. What do you propose for the rest of academia? -- Michael Scott Asato Cuthbert (talk) 09:59, 16 April 2024 (UTC)

I am not necessarily crazy about Impact factor which is why I suggested some other ideas, though impact factor is a widely reported metric, even for many journals outside of the sciences. Another idea would be that notable editors edit notable journals? If a journal is notable enough to already have a page e.g. https://en.wikipedia.org/wiki/List_of_philosophy_journals, then no further evidence is required, otherwise there must be some evidence (including but not limited to Impact factor) to support the claim that the journal in question is a 'major well-established' one. WikiNukalito (talk) 11:34, 16 April 2024 (UTC)

I am strongly opposed to using impact factor for anything at all. But whether a journal is notable is also a matter or some controversy; see WT:NJOURNALS. —David Eppstein (talk) 17:24, 16 April 2024 (UTC)

I mean, I agree, impact factor is flawed but at present the word major is doing a lot of work (also true for the discussion about chairs above in C5). I am an academic, not a very notable one, but by replying to some of the half dozen emails I get a week I could soon be the editor of a journal Here's an example of a person who might not be a notable academic, but for editorial work at a MDPI journal, a questionable publisher. Perhaps simply a link between C8 and the notability guidelines for journals would be beneficial? WikiNukalito (talk) 18:12, 16 April 2024 (UTC)

In all, C8's not my favorite part of the guideline, and it's not used very much, but I'm trying to search my brain for a case where it has been invoked successfully to keep a biography that would otherwise have likely been deleted. I can't remember one. Do we have any such cases? -- Michael Scott Asato Cuthbert (talk) 21:14, 16 April 2024 (UTC)

I linked this one in the post above: https://en.wikipedia.org/wiki/Wikipedia:Articles_for_deletion/Kevin_Knuth WikiNukalito (talk) 22:00, 16 April 2024 (UTC)

That wasn't a keep, it was a no-consensus close. And as I wrote in the AfD, the case for the journal he edits counting as major is weak. —David Eppstein (talk) 00:35, 17 April 2024 (UTC)

Isn't that exactly the point? I happen to have reviewed for that journal and I don't think it's problematic but it isn't 'major and well established', it's mid-tier I suppose, though notable according to Notability_(academic_journals). So that's my question really, 'major and well established' is not the same as 'notable', so how should that be determined? You guys are much more experienced here, so I hesitate to offer any other suggestions, and maybe it doesn't arise much in practice, but surely it wouldn't hurt to have some guidance on what 'major and well established' means? WikiNukalito (talk) 07:27, 17 April 2024 (UTC)

Thanks WikiNukalito -- sorry I didn't see it. Whether delete or no consensus, it is an example of a case where discussing the criterion seems like a reasonable response. FWIW, my notion of "major" is substantially below "Science" or "Nature" but above the hundreds of journals that random predatory publishers create each year. Presumably there has to be prestige associated with editing the journal and numerous people who would be thrilled to be asked. In musicology (my field), out of the perhaps 150-300 serious, regularly publishing journals, I'd consider about 30 to be at the level where I'd make a C8 case (all of the editors are people I could make a C1 case for as well, but it just takes more time). -- Michael Scott Asato Cuthbert (talk) 00:16, 20 April 2024 (UTC)

In terms of a page that is marginal for notability and tests the limits of C8 (WP:TOOSOON) is Draft:Benjamin Schlein, which is being debated by myself, Jähmefyysikko and TensorProduct with also some input from David Eppstein. He was one of four editors-in-chief for 2 years, but is not now. There is no question in anyone's mind (I think) that he is a strong mathematician. Also, I don't think that there is much doubt that the draft that was created directly in main space needed significant cleaning, which is slowly taking place.

However, with no great publication record, only relatively junior awards, the question is whether 2 years as one of the editors in chief meets C8 and WP:SUSTAINED. I am posting here to push forward the discussion of C8 which I personally am not a fan of. Ldm1954 (talk) 08:38, 24 April 2024 (UTC)

In my opinion C8 should be limited to sole EiCs of prestigious journals, not every run-of-the-mill indexed journal. JoelleJay (talk) 00:47, 25 April 2024 (UTC)

C1,4,7 vs C2,3,5,6,8

I just went and re-read the whole C1-8 for the first time in a while and realized something that is quite confusing about the whole list. There are sort of two different types of criterion mixed together; three of which apply to 90% of academic notability cases, and five of which are specialized shortcuts. Compare the current list with some sort of reorganization like this:

General Criteria for Notability of Academics (passing any one is enough):

• The person's research has had a significant impact in their scholarly discipline, broadly construed, as demonstrated by independent reliable sources. (=1)
• The person's academic work has made a significant impact in the area of higher education, affecting a substantial number of academic institutions. (=4)
• The person has had a substantial impact outside academia in their academic capacity. (=7)

Specific Positions, Awards, and Roles

Academics who have held any of the following positions or received such awards and recognition in their own field have sufficiently shown that their work has had significant impact in either their discipline, higher education, or outside academia. Thus passing any of these categories also satisfies notability:

• The person has received a highly prestigious academic award or honor at a national or international level. (=2)
• The person has been an elected member of a highly selective and prestigious scholarly society or association...(etc.) (=3)
• The person has held a named chair appointment or distinguished professor appointment...(etc.) (=5)
• The person has held a highest-level elected or appointed administrative post...(etc.) (=6)
• The person has been the head or chief editor of a major, well-established academic journal in their subject area. (=8)

---

It seems that somehow distinguishing the more general from the more specific cases might be helpful especially for newcomers to understand that there is a general way that academics are judged and then there are specific ways that save time at AfD. Drawback of course would be renumbering the criteria many of us know so well (call them A1–3, B1–B5?). The category labels and transition words are just things I came up with really fast, they're not important to the proposal, just the notion that these are two different things. Thoughts? -- Michael Scott Asato Cuthbert (talk) 21:40, 16 April 2024 (UTC)

I think that although this might have made sense to do at the dawn of time, this change would create more confusion than it would cure if carried out at this point. Just my opinion. Qflib (talk) 21:32, 20 April 2024 (UTC)

Clarify/quantify C1 with regards to Citation Metrics

Despite there being an entire Wikipedia:Notability_(academics)#Citation_metrics section, the section itself doesn't actually clearly spell out a metric by which to qualify. This makes it hard for article authors to actually know, is this an academic that qualifies or not? Personally I considered a person with over 200 citations to be pretty highly cited, but recently learned, that is not so. So, I'm asking if someone could actually help quantify that section, so that article authors writing new articles actually have a measure to work against. Raladic (talk) 20:36, 2 September 2024 (UTC)

The answer is that it is not possible to provide a single number that works because the citation patterns in different fields are too different to directly compare from one field to another. For someone who works in taxonomy, philosophy, or pure mathematics, a publication with 100+ citations may be highly cited and count as evidence towards C1; for someone who works in high-energy physics or astronomy, a publication with 10000+ citations (but 1000 coauthors) may not be evidence of much. Even if one tries to factor out numbers of coauthors, machine learning, neuroscience, medical image processing, and popular psychology are all very heavily cited compared to the others I listed earlier. And when individual researchers cross discipline boundaries, their work in one area may be more heavily cited than their work in another for reasons unrelated to its significance. —David Eppstein (talk) 20:47, 2 September 2024 (UTC)

Could hints for a by-field area be added that at least give some kind of closer idea than the current abstract?

The examples you just provided here already appear useful to be included in the section. Raladic (talk) 20:52, 2 September 2024 (UTC)

Commenting on your original post: it would be very rare for a person with 200 total citations to be considered notable under C1. A decent number of highly cited papers is needed. As David Eppstein said, a single publication with 100 cites in (for instance) mathematics would qualify towards notability. However, it would need to be more than one, in math I suspect that 15 with more than 100 cites each would meet C1.

In most areas of physics, chemistry and materials science, something like 500-1000 cites of a single paper makes it notable. An h-factor of about 45 is about at the border, that is 45 papers each of which has been cited more than 45 times. (I would prefer a slightly higher number of > 50).

But this is very subjective, and with so much field variability I doubt that anyone will want to provide hard numbers.

N.B., in most cases when an academic is truly notable they have received national and/international awards. For me those matter as they indicate that the peers consider them notable. Ldm1954 (talk) 21:10, 2 September 2024 (UTC)

When creating new articles I try to look for passes of multiple WP:PROF criteria, to make such matters more clear. But when other articles are taken to AfD, I often do not have that luxury. —David Eppstein (talk) 21:13, 2 September 2024 (UTC)

Even within fields there is a lot of variation across subfields that makes it difficult to compare (and controversial). Author-level metrics lists a bunch of ways people have tried to compare citation counts and impact for an author, but I think the proliferation of metrics demonstrates that this is a difficult task. It would be nice if we could set a standard like the "field-weighted citation impact," but even this comes with its criticisms, and I am skeptical it would be that useful. Having a bright line rule for citation counts could lead to poor outcomes (as in the examples David Eppstein laid out). Rather, I think extraordinarily high citation counts carefully evaluated in the context of the academic being discussed is probably the best way to go. Malinaccier (talk) 23:42, 2 September 2024 (UTC)

Ok thanks, all this helps as further explanations. Though it then almost sounds that likely such an academic will most likely also have to have some of the other criteria as an inevitable side effect beyond just high citation count if it’s that hard to quantify where the line is, even on a per field basis.

Or satisfy GNG itself I guess.

Is that an accurate interpretation? Raladic (talk) 01:46, 3 September 2024 (UTC)

I'd say so in a very vague and general sense, but I think it's more accurate to say that there isn't really a line, so much as a general "the higher the citation count relative to other academics in the field, the more likely C1 is a winning argument at AfD". -- asilvering (talk) 01:52, 3 September 2024 (UTC)

It is more difficult to assess citations now because of the large increase in citation gaming with the advent of research paper mills and similar. The best preparation is knowledge of the fields concerned and experience at academic AfDs. Anything out of the ordinary should be treated with suspicion. See what you think about the over 100+ authors from over 50 different instutions in this paper [1]. Xxanthippe (talk) 01:35, 3 September 2024 (UTC).

I compare the subject's Scopus citation metrics to those of their professor-level coauthors. If they're below the median then it's unlikely they meet C1 via citations. This only works in fields where paper coauthorship is common and where it actually means something -- disciplines where journal publications are rare, where monographs are the norm, or where authorship is afforded to every single person who breathes on a particle detector or recruits a clinical subject aren't going to yield usable statistics. JoelleJay (talk) 00:29, 5 September 2024 (UTC)

My practice in assessing academic AfDs is to look for single-author or low-authors papers in reputable journals that are well cited. In this way one can tell of an author is capapble of working independently of is just piggybacking on the work of others. Some editors think that author order is important (although not for single author papers!) but I see no evidence for this, as practices vary so widely. Xxanthippe (talk) 01:32, 5 September 2024 (UTC).

Yeah, that's another route I take when looking at researchers in fields with few coauthors. However you have to be careful since in some disciplines the only time you see single- or two-author works is for reviews/surveys, which shouldn't count nearly as much as research articles. And I do think it's important to consider author order in cases where that's meaningful (this should be clearest when authorship isn't alphabetized or is only alphabetized after the first few names); first-author is much more meaningful than second etc., and when last author = senior author that's probably an even better indicator of importance. JoelleJay (talk) 18:22, 10 September 2024 (UTC)

Practices definitely vary widely. In science, collaboration is everything and single-paper authors are rare in my experience. And some single-author review papers (invited ones) are a recognition that someone has achieved preeminence, or demonstrated early but very important achievements in a particular field; others, maybe less so. But the social sciences are completely different - single-author books are more important than articles in many of the social science fields. Qflib (talk) 21:38, 10 September 2024 (UTC)

• I agree that there is too much variation (between fields, between countries, over time, etc.) to even come up with a ballpark figure. We also have to be careful to avoid slipping into thinking we're the tenure board: notability is an assessment of the ease of writing an encyclopaedia article, not the achievements of the subject. The author of one or more highly-cited works is someone we can probably write an article on, because within those citations there is probably enough significant coverage of the work. Whether the work is good or really attributable to the subject or deserves the attention is irrelevant. – Joe (talk) 05:06, 5 September 2024 (UTC)

  Whether the work is good or really attributable to the subject or deserves the attention is irrelevant.

  Well, the guideline does stipulate that for a contribution to count towards C1 it needs to be very impactful and the specific involvement of the subject needs to be widely recognized.

  In this case it is necessary to explicitly demonstrate, by a substantial number of references to academic publications of researchers other than the person in question, that this contribution is indeed widely considered to be significant and is widely attributed to the person in question. JoelleJay (talk) 18:11, 10 September 2024 (UTC)

Selected publication section in professor articles

Articles about professors often include a list of selected publications. In the guide I wrote on how to write such articles I recommend the section. One list that I included in an article is currently under discussion as inapproriate unverified material here. Since this has the potential to affect many articles, I would like to consider the issue here instead. StarryGrandma (talk) 00:32, 22 October 2024 (UTC)

A short list of a scholar's most important works is appropriate if the individual is notable for research and publishing their findings. Some editors would limit a publication list to books and six journal articles but I would go as high as ten or twelve journal articles if the publications are significant in the field or if the article represents significant scholarship or major discoveries. In other words, we can set a guideline but need to consider the subject and what is appropriate; I would not limit Albert Einstein to a publication list of six articles.

A Select Publications list does not require a citation for each item in the list, especially when the items are linked to a digital copy or have an ISBN. This practice is well-accepted and used in countless FA and FL class articles. MOS:LISTSOFWORKS notes that a publication list acts as [its own] citation, even though it is not technically a citation. Thus, it would seem out of line to remove a Select Publications list because it is unsourced.

Some editors believe a list of Select Publications is arbitrary and question how its contents are determined. To me, it is just like any other content that is included in an article. An editor reviews available information and decides what would be the most valuable to understanding the subject, while also fitting within the concept of being encyclopedic. One way to determine the importance of a published article is to see how many times it has been cited by other scholars. However, this can not be the only determining factor because it fails to consider important work in narrow fields or older works that lack digitized coverage. Rublamb (talk) 01:16, 22 October 2024 (UTC)

Along with citation counts, other criteria I sometimes use to determine inclusion of selected publications are whether the publication has made a splash in popular media, or whether it has been given some significant award. But I agree with being flexible on the number while keeping the number small. It also depends on how senior the scholar is; for someone near or past the end of their career one would expect more publications to be significant than for a younger researcher. One way to make clear why some publications are selected is to include footnotes to them in the article text, distinct from the reference footnote numbering; see for instance the lettered (not numbered) footnotes in Ronald Graham (for whom six selected articles would be too few). —David Eppstein (talk) 04:19, 22 October 2024 (UTC)

David Eppstein, I'm going to have a look at that article but let me say right now that anything that can verify that there is a valid reason for selecting something is a welcome addition. Drmies (talk) 16:23, 22 October 2024 (UTC)

I see a lot of AfC drafts where I would agree with the comment that the Selected works section is "little more than resume building", but for an emeritus professor or someone who has published monographs I think it would be really quite strange not to have this section. -- asilvering (talk) 04:31, 22 October 2024 (UTC)

Trimming lists of publication down to size is definitely something I've done more than once when a bio makes it to AfD. It's the opposite of "resume building": a resume is all-inclusive, whereas a short list of publications here is curated. XOR'easter (talk) 06:27, 22 October 2024 (UTC)

XOR'easter, in the example at hand there was no indication in the article or anywhere else that "curation" had taken place. Drmies (talk) 16:23, 22 October 2024 (UTC)

StarryGrandma, thanks for bringing this to wider discussion. I have always been of the opinion, not for terribly well-considered reasons, that including e.g. the three highest cited papers from GS is usually due. With reasons, that might be augmented to 4 or 5 papers; in rare cases more. If I'm putting in a selected papers section, then I usually include an html comment with the selection criteria I used. So the list in Laurence D. Marks seems overly long to me -- the selection criteria are unclear, and I'm not sure what a reader should take away from it. (But in a situation where the main notability comes from published papers, it does seem appropriate to include a short list of them.) Russ Woodroofe (talk) 07:37, 22 October 2024 (UTC)

Russ Woodroofe, thank you. I am NOT against including a select number of publications: it seems clear to me, even in my own career, that some articles are simply more important than others--by which I mean cited, acknowledged, etc. Even if I may not agree with someone's selection criteria, at least we should be able to see what criteria, if any, were used. Drmies (talk) 16:23, 22 October 2024 (UTC)

In the LDM article it seemed odd to me that the 1983 article introducing the Marks decahedron (not actually a decahedron) was omitted. Selected publications should not merely be "highest cited papers from GS" (although that is often a good start); we should take into account other factors like prizes, whether the paper started or finished a Wikipedia-notable topic, whether it's the author's main work on some subtopic or duplicative of other selections, spread of dates over the author's career, and (in fields where this matters) position in author list. —David Eppstein (talk) 17:14, 22 October 2024 (UTC)

@David Eppstein: I don't know who selected the original list for LDM. I just remember reviewing and trimming the list, leaving major publications and articles where he was the main contributor. You are right about including his early work. I would also suggest his 2024 article on static electricity that was mentioned in The Guardian. Now LDM has retired, it would be great for someone to review this list, looking for the articles that best summarize his career. That may require looking at his CV as all of his early work may not show up in Google Scholar.

Now that many months have passed, I would back your original idea of cutting the list, especially so that it can really showcase his career. However, list curation is very different from completely removing a Select Publications list, which is what happened here. This current list is not horrible; it just could be better. But as with anything in Wikipedia, it is a process that takes time. Rublamb (talk) 19:22, 22 October 2024 (UTC)

Curating such lists takes a certain level of subject-matter expertise, which I for one do not have for the research area of LDM. It would be good if an editor who knows this subject well could take charge of any such trim. I do not know offhand who might be likely. —David Eppstein (talk) 20:14, 22 October 2024 (UTC)

Wonder if flagging for an expert would help? Rublamb (talk) 20:24, 22 October 2024 (UTC)

@Drmies: Just because you don't "see" the selection criteria, doesn't mean one does not exist. The criteria could be articles in publications with a Wikipedia article; given the tough road to get a journal listed in Wikipedia, that can be a way to determine notability. It might be based on citation counts, a connection to the text, or a combination of all three. These are all reasonable approaches, but I would be challenged to "see" them. I rely on the principle of Good Faith, especially when looking at contributions by experienced editors.

In a perfect world, all academic articles would be B class or above, with extensive details of scholarly work that connects to the publication list. However, Wikipedia is a process, and everything is not added simultaneously. Removing a Selected Publications list is not particularly helpful when that content could contribute to expanding the article. The current list may be flawed, but with a quick scan, I can see that LDM was published in Nature and Science at a time when that meant a scholar's work was significant had become mainstream. Rublamb (talk) 19:53, 22 October 2024 (UTC)

Rublamb, I'm not looking for a perfect world--I'm just looking for a citation, a footnote, just like we do with all things that we put in articles. If there's no note, no explanation in an edit summary, then it looks like "selected" is OR/editorial attitude. And what you are explaining here is indeed a kind of editorial overreach: you are outlining possible criteria which may or may not have been used but were certainly not explicated. And the road to get a journal listed is really not that tough: we have hundreds if not thousands of articles on journals; I think I wrote a few myself. "Article published in journal with Wikipedia article" is not a valid criterion: there are tons and tons of articles on journals and those journals publish tons and tons of articles, some of them with tons and tons of co-authors. As for good faith, sure, but is this not an article with a COI past? Drmies (talk) 20:06, 22 October 2024 (UTC)

It's no different from any other editorial decision we might make. We include World Cup and Olympic appearances of footballers but don't detail their performance in minor games; do we need a footnote every time to a reference that specifically states that for this person these are their most important performances? —David Eppstein (talk) 20:16, 22 October 2024 (UTC)

@Drmies, You have said: Anyway, without proper citations that prove noteworthiness, it's an arbitrary selection., and you are removing such lists when you see them. However WP:NNC says that notability guidelines do not apply to content within articles. "Selected" just means a short list. I use the list to illustrate a scientist's development over time. Its an editor's choice of what to include, not original research. The list is self-referencing. A short bibliography of the subjects work is common in scientific encyclopedias such as the Dictionary of Scientific Biography. Wikipedia is many types of encyclopedia rolled into one, and over here in the academic and technical section we can follow the style of those encyclopedia. StarryGrandma (talk) 20:40, 22 October 2024 (UTC)

What you're saying is you can insert material whose importance is unverified just because you think it's worth including. I didn't say "notability", by the way. By the same token, you could insert a list of every single YouTube video some influencer has ever made, or a bibliography of every single article some journalist ever wrote--or every single opinion article some columnist ever wrote. If "editor's choice", without any kind of sourcing on which to base that, is going to be the new guideline, we might as well turn into Wikia. Drmies (talk) 23:12, 22 October 2024 (UTC)

Now I am confused. I think you are talking about "verifiability doesn't mean inclusion", rather than importance, and we don't have problems keeping that stuff out. We have articles about people who are notable. Does every event in the life of a such a person have to be important to the wider world, with a source that shows not just that it happened but that it was an important happening, before it can be in a biography? Wikipedia has never required that. If a person moved to a new university all we need is a source that shows that it happened, not a source that shows that it was important. If a person does a piece of research all we need is a source that shows that it happened, not another source that declares it important. I include a list of publications to show the development over time of a researcher, what they were working on and what that led to. It is a list of events in a person's life as they progress in their field, facts supportable by self citation. And helped by the fact that titles of technical papers tend toward the very descriptive. If there is a bit more information available it can be done as prose with the same sources. But biography. particularly in science, needs to present what people did and how they came to do it, not just a list of awards and positions.

Drmies, I understand your frustration. Wikipedia is continually inundated by promotional editors, in the academic area as well as everywhere else. Professors want their latest research highlighted, or insist that their entire publication list be included, or want an article about their latest pet theory which no one in their field takes seriously. University PR people want shiny articles about their institution and professors. But in trying to fend this off we need to allow experienced editors use such factual material as is available, even if others try to misuse it. StarryGrandma (talk) 04:42, 24 October 2024 (UTC)

"If a person does a piece of research all we need is a source that shows that it happened"--what you are saying is that every single thing that a notable person does should be included in their article, no? I disagree. Drmies (talk) 12:43, 24 October 2024 (UTC)

@Drmies: Exactly. The role of the editor is to review and select what information to include, whether it is factual details or a list of publications. Rublamb (talk) 13:12, 24 October 2024 (UTC)

@Drmies, have you considered the constraints I mentioned below in selecting which papers to list? I don't think anyone here is arguing we include every piece of research or even that we choose the list items based on what we think ought to be most important. What experienced editors are doing is adding the most impactful publications according to external sources -- the citations or reviews themselves surely demonstrate why a particular piece is DUE in the page -- and (hopefully) trimming them down to just those works that reflect serious research contributions from the subject. JoelleJay (talk) 17:20, 24 October 2024 (UTC)

I did, but given the idiosyncrasies of the various fields (noted by others in response), how are these things workable at all? "What we think ought to be the most important"--I know what my reasoning is for dropping a paper in someone's list of publications: a citation, or a comment, in another paper/book/etc., in a footnote... Drmies (talk) 20:11, 24 October 2024 (UTC)

Generally if the academic is actually notable, there will be some evidence of that in the form of discussion of their work by other academics. Most fields use citations as a metric to at least some degree, so identifying the papers that have the most citations is very workable. My recommendation was to filter that list further by generally ignoring reviews/surveys and to focus on papers where the subject is the first or senior author if their field uses ordinal authorship. I don't see this approach as any different from the way we (should) decide which of the subject's research topics to discuss in prose -- we can really only be sure something is BALASP if it's had attention elsewhere, and citations are the easiest metric to check. JoelleJay (talk) 20:46, 24 October 2024 (UTC)

For that reason I wouldn't recommend just copying a professor's own "selected publications" list from their home page (my own included). There is some likelihood that they chose some or all of these publications for reasons at cross-purposes to ours (to highlight publications that they think should stand out rather than the ones that already do; to focus on only their current interests rather than the old stuff they're notable for but no longer interested in; to present an appearance of being active in some now-hot topic; to show off high citation counts for papers that they had a minor role in; etc etc.) —David Eppstein (talk) 05:16, 24 October 2024 (UTC)

Going purely by journal prestige, I'd include papers published in Nature, Science, and Physical Review Letters. XOR'easter (talk) 22:24, 22 October 2024 (UTC)

That was what I originally went with but it is good to have a physicist confirm my selection criteria. Maybe you can help identify any missing or better articles? Rublamb (talk) 23:00, 22 October 2024 (UTC)

In addition to what others said, when possible I would limit these works to those publications where the subject is a major author. In many fields authorship is ordinal (or ordinal at the start and end and alphabetical in the middle), and often senior/corresponding authors are listed last. When this is the case I would select the highest-cited research works that were by the subject as either first or senior author. I specify "research" because frequently someone's highest-cited papers are just reviews, which don't really demonstrate the impact of the author's scholarship (unless they basically defined their subfield). JoelleJay (talk) 21:08, 22 October 2024 (UTC)

Yes. On Talk:Laurence D. Marks, Ldm1954 mentioned the selected publications on the article about me, David Eppstein#Selected publications. I didn't select them (because writing that article is not for me to do) and wouldn't have made exactly that selection, for one thing because the mesh generation paper listed is a survey rather than the research paper that I would have put in its place, "Provably good mesh generation". Different citation numbers mean different things for different kinds of papers. Another situation where this sort of thing comes up frequently is with statisticians, who often work as the statistical consultant to papers in other fields (getting big citation counts with a middle-position authorship) but whose more significant contributions to fundamental statistics are often less heavily cited. —David Eppstein (talk) 21:22, 22 October 2024 (UTC)

I agree, but other fields don't follow these conventions (or didn't historically), so it would be important to avoid applying this kind of rule without exception. – Joe (talk) 11:15, 23 October 2024 (UTC)

I'll confess that I often have a gut reaction akin to it simply being resumé-building, but I also recognize that it's really not as simple as that. As other editors already pointed out above, one consideration is the career stage of the page subject. I'll expand on that by pointing out a rather easy test for whether or not a publication list should be trimmed. Look at the current version of the page as a whole. (I should quickly add that this does not apply for a very newly created page, that may be significantly expanded in the near future.) If the person the page is about is someone with a particularly substantial life story, so that the page has a lot of paragraph text about them, a publications list may make up a fairly small part of the page as a whole. When that is the case, the list has a good chance of being justified. On the other hand, for an early-career academic, for whom there can only be a start-level page, it tends to look very WP:NOTRESUME when the publication list fills up about half of the entire page. When that happens, there's probably a good reason to remove at least some of the list.

Another issue that I've come across, that hasn't been mentioned yet, is that I sometimes see pages about persons in the fields where I'm particularly familiar with the literature, where there is good reason to list some selected publications. But what I actually see in the list are a bunch of the less important publications by that person, related by the fact that they are all coauthored by a (not notable) junior author. I get a strong feeling that there must have been some WP:COI editing of the page in the past. The solution, of course, is not to remove the list, but to rewrite it. --Tryptofish (talk) 23:39, 22 October 2024 (UTC)

• My thought is that the selected publication list should only contain blue-linked publications. This is generally similar to other lists in articles. Everything else should be in the prose. --Enos733 (talk) 00:35, 25 October 2024 (UTC)

  That's a super-extreme position that would prevent listing almost all research articles, even those winning major prizes, and many entire books. I don't think it's a good idea. —David Eppstein (talk) 01:01, 25 October 2024 (UTC)

  That is a great selection criterion, espeically for new academics who might pad their CV with non-notable online journals, but it cannot be the only deciding factor. Many defunct publications and state or federal series are not included in Wikipedia but were, nevertheless, important in their day. I just finished an article on the leading surgeon for pulmonary tuberculosis in the early 20th century—all of those TB journals and national organizations have been inactive for a long time and are not in Wikipedia. However, it was a major branch of medicine and a public health crisis; Time magazine wrote about this African American surgeon's innovations in 1940. I would not exclude his articles in Bulletin of the National Tuberculosis Assocation or the Pennsylvania Medical Journal just because those publication are not in Wikipedia and don't seem to be digitized yet. Rublamb (talk) 01:03, 25 October 2024 (UTC)

  Am I misreading Enos's comment? It seems to require that the individual research article have a blue link, not merely that it be published in a bluelinked journal. —David Eppstein (talk) 01:12, 25 October 2024 (UTC)

  I believe "blue-linked publications" refers to journals, not to individual articles or books. Rublamb (talk) 01:17, 25 October 2024 (UTC)

  That would make even less sense, since basically 100% of notable contemporary scientists will have many many papers in blue-linked journals, and a huge proportion will have only published in notable journals. That would be effectively the same as having no selection criteria at all. JoelleJay (talk) 02:09, 25 October 2024 (UTC)

  It depends on the situation. Of course, there is a difference between a journal meeting Wikipdia's notability (verified by having an article) and a journal being considered notable in a given academic field. And not all articles in notable publications represent significant advances in a given field or major research by a given scholar. Nevertheless, I still think this can be a useful way to start a review/reduction of an existing publication list, depending on the situation. I have edited articles for newer academics where none of the journals in their publication list were blue-linked. On closer examination, the cited publications lacked peer review or were predatory, meaning the lack of a blue-link was a pretty good indicator of worthiness for inclusion. Regardless, I don't believe this should be a "rule", just a useful tool. Rublamb (talk) 02:47, 25 October 2024 (UTC)

  I think this is a fine rule of thumb, but our inclusion criteria for journals is so random at the moment that I oppose its use as a hard indicator. Suriname0 (talk) 04:32, 25 October 2024 (UTC)

  I would agree with avoiding the inclusion of anything in a predatory journal. That is not going to help much as a filter for most notable scholars, though. I find User:Headbomb/unreliable to be helpful in catching links to predatory journals, so if the listed articles have dois or other links they should show up as unreliable using this script. —David Eppstein (talk) 05:37, 25 October 2024 (UTC)

  I was referring to the actual publication, not a blue-linked journal. I know that some academics may not have a separate publication list in my formulation, especially if they only publish journal articles, but it does provide a bright line (and a standard that is followed in many other areas including WP:NLIST, and generally followed in filmographies). I do also think that editors can (and should) talk about the academic's work in the body of the article (this is why we consider citations as a reason to keep - because other academics are talking about the subject and the subject's theories. - Enos733 (talk) 05:48, 25 October 2024 (UTC)

  All of the filmographies I sampled to verify your claim (e.g. Gary Cooper filmography, Lillian Gish filmography, several others) included blacklinked or redlinked entries. The same is true for all the discographies I sampled, and all the bibliographies of fiction authors I sampled. If this is a rule it is one very much not followed. For open-ended lists of people, yes, this is a typical rule, but that is not what we have here.

  In cases of academics in book fields, the (authored) books and their reviews are the main claim to notability of the subjects; omitting them (in cases where they do not have standalone articles already) could leave the article without a claim of notability, in an A7 deletion worthy state. —David Eppstein (talk) 06:52, 25 October 2024 (UTC)

  Moreover, WP:MERGE suggests that it could be better in many circumstances to e.g. cover notable books from a notable author in the article on the author. So in many circumstances, the books not only may not be blue-linked, but should not be blue-linked (even though they are notable). Russ Woodroofe (talk) 07:51, 25 October 2024 (UTC)

  @Enos733: Looks like I misinterpreted your suggestion. Only citing blue-linked books in a selected publication list is not a reasonable standard, for many reasons. Firstly, you assume that a book is more important than cutting-edge research in a journal and that a book has a wider reach. Secondly, scholarly monographs are less likely to have a Wikipedia article than best-selling "trashy" novel; being blue-linked does not mean the novel is more important academically. Thirdly, being blue-linked is not a Wikipedia standard for lists. WP:NLIST specifically says that all items in a list do not need to be notable (i.e. have an article in Wikipedia). The function of a publication list in a biographical article is to showcase the person's career and their important/influential work; the list's importance is relative to the subject of the article and their field of expertise, not being blue-linked. Rublamb (talk) 18:54, 25 October 2024 (UTC)

  I agree a black and white (blue and white) standard may not always lead to optimum results, but it does provide a certain degree of clarity for editors. At the end of the day, material on this project must first be verifiable, and second, should be written about. "Selected," to a large degree, implies subjectivity and an editor's judgement that the publication is important. So, we must make choices - first by deciding to limit the publication record to a handful of publications, and second which publications are "most" important and reflective of the subject's career. If we don't list all publications, we are making editorial choices, and if we want to give editors discretion, we are going to have lots of variety across subjects (as the publications I think are important may not either be an exemplar of the scholar's work or important to other editor [ and to avoid conflict, it is easier to add than subtract]).

  However, our community does have guidance across all subjects - a focus on the material that are discussed by other authors. Yes, it could be a blue-linked work (there are some journal articles that should meet our guidance for a stand-alone article), But even if we don't use blue links as a metric, the selected publication list should be based on how impactful those publications are on their respective field (or society at-large), and impact should be evaluated through the lens of whether those publications (and the theories contained within those publications) are discussed by others - and ideally, those impacts are discussed in the prose.

  Just my 2 cents. - Enos733 (talk) 00:21, 26 October 2024 (UTC)

  But there is a huge difference between a publication receiving sufficient coverage to be DUE and receiving sufficient coverage to be notable. The vast majority of even well-cited papers are not going to receive sustained GNG SIGCOV: citations will be at best a few sentences, and even in-depth reviews/editorial summaries will only occur in the short period around the paper's publication (and anyway are considered routine for some disciplines, like math). I agree that the selected works should reflect external appraisals of impact, but limiting them to blue-linked works would basically eliminate these lists for everyone in STEM, and even for book-based scholars would be totally inconsistent with our treatment of, e.g., novelists, for whom every book is listed. JoelleJay (talk) 16:35, 26 October 2024 (UTC)

  For most research journal papers, even when the contents of the paper is covered in-depth in a chapter of a textbook (say), it would be better to have a Wikipedia article about the topic of the paper than a Wikipedia article about the paper itself. There may be papers that are separately notable from their topic but they are so rare that, looking through our existing Wikipedia articles on journal papers, I didn't see any truly convincing examples. —David Eppstein (talk) 18:43, 26 October 2024 (UTC)

  My usual go-to example of a notable paper is "Molecular Structure of Nucleic Acids: A Structure for Deoxyribose Nucleic Acid". DMacks (talk) 18:50, 26 October 2024 (UTC)

  It's an important paper and one I wouldn't even consider the deletion of. But by "truly convincing" I meant that there is clear WP:GNG coverage of the paper itself, considered separately from its topic and its history of discovery. Which sources in that article do you think provide that? —David Eppstein (talk) 19:55, 26 October 2024 (UTC)

  Arguably the Andrew Wakefield Lancet paper would meet GNG, but as you said would still be better covered in an article on the larger scandal (as it is currently). I think coverage that actually critically discusses a paper's conclusions/rationale/narrative, rather than just the results of experiments, would probably hew to GNG better.

  The Market for Lemons might be an example of a paper that became synonymous with the concept it introduced and is best covered in its own article. JoelleJay (talk) 20:16, 26 October 2024 (UTC)

  The paper gets highlighted for its omission of Rosalind Franklin as an example of sexism. And it's also mentioned as an example of a major breakthrough (including that it led to a Nobel Prize) that is not one of the most highly-cited papers.[2] Both of these are coverage of the paper itself not of the scientific content. DMacks (talk) 20:22, 26 October 2024 (UTC)