Wikipedia talk:Make technical articles understandable

Archives

Archive 1 Archive 2 Archive 3

This page has archives. Sections older than 730 days may be automatically archived by Lowercase sigmabot III when more than 5 sections are present.

While some articles can be, in places, difficult to understand without a backing in the topic that they discuss, this often has more to do with the subject matter of the article, than it has to do with the word choice of its authors. In order to be factually robust, many articles must talk about things that could come off as complicated. In the past, Wikipedians have been known to insert tone tags into articles which contained language they found overly technical. If you are unsatisfied with the article's readability, a better solution would be to tag it as needing help from an expert-- somebody with a backing in the topic, who can re-write portions of it so that they're easier to read, without compromising their educational value. Generally, the tone tag is reserved for articles which fail to meet a basic standard-- articles that are poorly written, needlessly verbose or flowery, or which are clearly biased. As technical articles are typically created by professionals and tend to be of higher quality, the tone tag is not usually appropriate. I encourage Wikipedians who are interested in helping with this project to remove tone tags from articles which do not need them, so as to avoid confusing copyeditors. Many copyeditors will see the tone tag and look for something completely different. Atomic putty? Rien! (talk) (talk) 20:41, 21 June 2022 (UTC)[reply]

Exactly. The idea that people need knowlege pulped and chewed up is a misconception. If one understands everything in a Wikipedia article on a technical subject, that means they know the subject well. Then why bother reading this particular article? It's a waste of time. One of the first books I ever read was an encyclopedia of technology and science. Of course I did not understand, pretty much anything, but I got an overview of how much there is TO learn! Did it bother me? No, of course not. I think there are people who might be bothered by it, and the question is why. Did they get traumatized by someone who did not like answering questions? There are probably plenty such people, but hopefully, they will in time heal. It won't help them heal if we pretend that information is just not there to bother them. I doubt they are visiting technical articles just to feel bad. Morycm (talk) 23:02, 4 February 2023 (UTC)[reply]

@Atomic putty? Rien!: I don't think anyone's suggesting dumbing down articles or removing technical content. And an "expert needed" tag won't help, because it is the experts who write these incomprehensible articles. There's a difference between being an expert, and being an expert and a good encyclopedia editor. --Chetvorno^TALK 19:40, 6 February 2023 (UTC)[reply]

@Chetvorno Then I have an idea: use team approach. Get an expert, a writer, and a Wikipedia stylistic advocate. They will make sure it is technically correct, easy to understand, and reaches the widest possible audience for that subject. But if you expect to maximize audience, just do what top Youtube channels do: stick to sports, girls, beer, dogs, cats, and comedy. Because for those who look for information, a wikipedia that appeals to an average person from the global crowd would simply be a useless babble. Morycm (talk) 02:17, 13 February 2023 (UTC)[reply]

I agree with @Chetvorno 100%. Looking back, my opinion's definitely changed. When I wrote the earlier note, about a year ago, I felt strongly that accuracy shouldn't be compromised for readability-- but honestly, just like they said, a good encyclopedia editor can tone down the heavy detail, improve readability, and conserve accuracy.

Congrats to all parties for participating in the discussion. When in doubt, let's turn to our already-established best practices. I'm not sure why I felt the need to defend the conservation of detail so strongly before. Looking over the article, it does a really good job of keeping to the consensus with regard to detail and readability. ^-^ Atomic putty? Rien! 19:51, 13 February 2023 (UTC)[reply]

I also agree with Chetvorno. My particular concern is the lede paragraph of technical, or merely semi-technical articles. All too often Wikipedia writers leap instantly into textbook mode in the opening sentences, as if readers could understand arcane jargon, calculus or otherwise advanced mathematics. Such practice violates multiple policies and guidelines that article ledes should be written for the general reader and should not presume the reader is familiar with a technical subject. The "Bessel function" lede that Chetvorno showed is a very good example. When I do a Google search for something, I want a brief comprehensible summary of an unfamiliar word or concept. But frequently in Google results, I'm confronted with a blurb like Bessel function, and I will do exactly what Chetvorno did--I'll look for the Britannica version, because the Wikipedia introduction, though probably accurate, is nevertheless impenetrable, and therefore useless for my purpose. I also oppose the practice, in the lede, of linking jargon to separate articles. I added text to this Guideline to say that "Terminology in the lead section should be understandable on sight to general readers". Loading up a lede with linked terminology forces readers unfamiliar to stop and start multiple times as they hover or click on the terms, when all they wanted was an accessible summary of the topic which they could read uninterrupted. I don't object when editors include advanced technical material in the body of an article, but doing so in the lede violates policy and is a disservice to readers that should be avoided. DonFB (talk) 11:46, 14 February 2023 (UTC)[reply]

@DonFB: Thanks for the addition to the article. I have no objection to a lot of links in the lede, as long as it is not overly dependent on links: it should have brief explanations in ordinary language of the jargon used.

I absolutely agree with you about the confusing ledes in our technical articles. They are often ridiculously abstract and esoteric, when they are the part of the article that should be most widely understandable. I have rewritten dozens of ledes to be clearer. I think what happens is technically-minded people progressively expand the definition in the lede to be more abstract and cover more special cases, borderline cases and fields, until the definition becomes completely incomprehensible to ordinary readers. The ironic thing is that it is usually possible to incorporate the advanced stuff they want in the lede and still have it be adequately comprehensible to general readers, using ordinary good writing techniques. Most technically-educated Wikipedia editors are used to writing only for others in their specialty. Writing an encyclopedia requires a slightly different style of writing, but one which any educated person can do. Our editors just don't want to. --Chetvorno^TALK 21:01, 15 May 2023 (UTC)[reply]

I prefer to see no linked jargon in a lead, although I recognize that it is possible to show jargon with a useful parenthetical definition. However, I want to see such practice held to a bare minimum. It's also reasonable to link from an ordinary word which has special meaning in the context. An example is the link I made in the Tidal Locking lead from the word "variablity". Originally, "libration" appeared in the lead as a link with no parenthetical explanation, an example of exactly what not to do (link only from jargon, with no other help). Recently, I took a somewhat novel approach to making the case for rewriting a jargon-filled lead in the Femur article. The result was a successful revision with no objections and a small followup tweak by another editor. See the Talk topics beginning with Turgid Intro and the next two headings; note the (tl;dr) collapsed Virtual Discussion. DonFB (talk) 04:36, 19 May 2023 (UTC)[reply]

In a sweeping set of changes, CactiStaccingCrane rewrote, in a substantive not copy-editing way, nearly everything in this guideline, and that is of course not how guideline change happens. Each substantive proposed change to the guidance and its meaning should be individually discussed and consensus sought for it, as even seemingly trivial changes to guidelines and policies can affect thousands, even potentially millions, of pages and their content.  — SMcCandlish ☏ ¢ 😼  19:43, 17 October 2023 (UTC)[reply]

PS: For the record, I dispute that many of these changes were improvements. Many of them are clumsily, even misleadingly worded, and much of the material is rather emotive and laced with superlatives and over-broad or exaggeratory statements. Perhaps CactiStaccingCrane should consider writing a user essay on this subject to express their heated opinions.  — SMcCandlish ☏ ¢ 😼  19:48, 17 October 2023 (UTC)[reply]

I think a good rule of thumb when editing guidelines might be: "If you're thinking about/trying to remedy particular editors' particularized classes of mistakes, you shouldn't be writing a guideline right now". Remsense聊 20:02, 17 October 2023 (UTC)[reply]

@Remsense Then what should I do instead then? Isn't a guideline all about trying to do just that? CactiStaccingCrane (talk) 17:59, 21 October 2023 (UTC)[reply]

i would disagree with this actually, a guideline written in language as general as possible is preferable, if you're targeting specific editors, that's a tone much more suited to an essay imo Remsense聊 19:01, 21 October 2023 (UTC)[reply]

Thanks @SMcCandlish for the revert. I do think it's worth making changes to this guideline though (with discussion), especially to its "lead" section, which doesn't do a great job introducing or summarizing the rest of the page, in my opinion. I'm not sure quite the right tone we should adopt in this kind of guideline, but I somewhat liked Cacti's idea of boldly leading with an emphasis on readability. I really dislike the current sentence, "When adding content and creating new articles, an encyclopedic style with a formal tone is important.", which seems vague, passive, and generally weak. The current mention of "essay-like, argumentative, or opinionated writing" seems somewhat off topic; the problem this guideline is trying to address is overly technical writing, not overly informal or opinionated writing. Etc. –jacobolus (t) 20:05, 17 October 2023 (UTC)[reply]

Aye, 'encyclopedic' and 'formal' seem a bit like tautologies. There's some way to phrase that, even in the lead, that's actively helpful. Remsense聊 20:10, 17 October 2023 (UTC)[reply]

I agree that CactiStaccingCrane's edit/rewrite was excessive. Also agree with Jacobolus about the off-topic tone in some wording in the opening section. DonFB (talk) 20:21, 17 October 2023 (UTC)[reply]

Sure. I actually thought about restoring a copyedited version of some of CactiStaccingCranes lead changes, but I don't want to be acting unilaterally while complaining of someone acting unilaterally.  — SMcCandlish ☏ ¢ 😼  20:29, 17 October 2023 (UTC)[reply]

How about this: Wikipedia articles should be written for the widest possible general audience that the subject matter of those articles will allow. Wikipedia serves readers with a diverse range of backgrounds and covers subjects from the everyday to the arcane. Some articles will naturally be targeted to specialists, while others need to be comprehensible by a wider population, and it can happen that a single article should provide value both for novices and for experts. This affects how we organize articles. The introduction of an article should ideally be comprehensible to a broad readership. When organizing the sections that follow the introduction, it is typically a good idea to put the least technical part of an article first. It can be helpful to consider the stage in a student's education when they are introduced to a subject, and to write with an earlier stage in mind. XOR'easter (talk) 02:14, 18 October 2023 (UTC)[reply]

Well-expressed. I would differ, though, on the matter of "targeted". Certain subjects are highly technical, and the article text will reflect as much. I don't think targeting is needed, per se. The text of any article should be written to be as accessible as possible. My overriding concern is article lead sections, excerpts of which appear in Google search results. As a reader and seeker of information, I am all-too-often disappointed and annoyed when I see jargon-laden or needlessly dense text in those results. DonFB (talk) 02:36, 18 October 2023 (UTC)[reply]

How about, "Some articles will naturally be mostly of interest to specialists"? XOR'easter (talk) 03:03, 18 October 2023 (UTC)[reply]

Here's my take (anyone rewriting this page should feel free to chop any bits you like):

Wikipedia articles should be written for the widest possible audience.

As a free encyclopedia with the goal of democratizing knowledge, Wikipedia serves readers with a wide range in background, preparation, interests, and goals. Even for articles about the most technically demanding subjects, these readers include not only subject experts but also students and curious laypeople. While upholding the goals of accuracy, neutrality, and full coverage of the most important aspects of a topic, every effort should be made to also render articles accessible and pleasant to read for less-prepared readers.

It is especially important to make the lead section understandable using plain language, but it can be helpful more generally to begin with more common and accessible subtopics, and proceed later to those requiring advanced knowledge or addressing niche specialties.

Articles should be written in encyclopedic style, but this differs from the spare and technically precise style found in scholarly monographs and peer-reviewed papers aimed at specialists. Articles should stay on topic without twisting the truth or telling "lies-to-children", but they should also be self contained when possible, and should not take prerequisite knowledge for granted or gratuitously use unexplained jargon or advanced technical notation: shortcuts which save time and effort for experts can be barriers to the uninitiated.

–jacobolus (t) 03:07, 18 October 2023 (UTC)[reply]

I am generally happy with that. XOR'easter (talk) 03:15, 18 October 2023 (UTC)[reply]

Generally works for me, but it could use some trimming for concision. E.g. drop "with the goal of democratizing knowledge", "not only subject experts but also", "but it can be helpful more generally to", etc. There's some redundancy here and there; "proceed later to" → "proceed to". However, the fact that WP has "the widest possible general audience", in XOR's original, is valuable, and a point we don't make frequently enough.  — SMcCandlish ☏ ¢ 😼  08:05, 18 October 2023 (UTC)[reply]

I don't have any objections to Jacobolus' suggested text. I'm not sure, though, where in the Guideline it would be used, as it seems to reiterate points already present in the current version. My preoccupation with the lead section of an article, to be clear, is not merely due to my own preferences, but is because that's the part of an article that all readers will see, and may be the only part they see. We owe it to them to make the lead readily comprehensible, even for highly technical subjects, to say nothing of subjects that may be merely semi-technical. I'm frankly not overly concerned about highly technical descriptions deep inside some articles. But I want the Introduction (lead) of every article to be clear, free of jargon, and not dependent on links to other articles for meaning. DonFB (talk) 18:45, 18 October 2023 (UTC)[reply]

Does the current § Lead section section cover it, or is there more (or something different) we should say? One of CactiStaccingCrane's ideas was to include some examples. Is that something we should try to do here (if so does anyone have recommendations of good examples to look at), or are they a distraction? –jacobolus (t) 19:13, 18 October 2023 (UTC)[reply]

Disclosure: the wording that terminology should be recognizable "on sight" is my contribution. I'm not fond of the last sentence in the section, which undercuts the emphasis I favor, but it was retained in a previous discussion. I'm not averse to including an example; elsewhere, I have pointed to revision of the lead of Tidal locking, which I initiated. (Before revisions: https://en.m.wikipedia.org/w/index.php?title=Tidal_locking&oldid=904381140)

DonFB (talk) 19:55, 18 October 2023 (UTC)[reply]

I'm happy with the section as is. The last sentence accurately describes how highly technical topics are addressed on Wikipedia, so there is nothing wrong with it. The "on sight" claim may actually be a bit too much; we routinely link potentially unfamiliar keywords in the lead section, and as said elsewhere we do not use a "lies-to-children" approach to dumb down the material. There's a fine line between doing that an avoiding unnecessary jargon. Some jargon is actually necessary, and just has to be made explicable in-context one way or another.  — SMcCandlish ☏ ¢ 😼  20:35, 18 October 2023 (UTC)[reply]

"On sight" is, indeed, rather rigorous, but it can be considered aspirational--a reminder that links should be used to supplement info in the lead, not to explain a term that may unfamiliar or incomprehensible, or both. When I read an introduction, I want to stay on the page, not chase after multiple linked terms to figure out what the text means--I believe that's probably the preference of most readers. Such terms appear too often in lead sections, because editors did not endeavor to explain or define the term in plain English, but simply defaulted to language found in textbooks or technical journals with which they're familiar--language that's often not appropriate, in my view, for the introductory section of an encyclopedia article. You said, "we routinely link potentially unfamiliar keywords in the lead section," which is true and is both symptom and cause of problems evident in too many leads. I think the practice should be much less frequent, rather than routine. Yes, we shouldn't dumb down the text, but the Guideline does suggest writing down one level when appropriate. I think a common reason the problem happens is because people want to put in information they know without much regard whether the text helps general readers or befuddles them. One approach, when possible, is to link a plain English term to an article about the technical term or concept,* or put the linked technical term in parentheses (though I am not a fan of parenthetical stage whispers in lead sections).

• An example, from the aforementioned Tidal Locking, was use in the Intro of "variability", linked to the article Libration.

DonFB (talk) 23:48, 18 October 2023 (UTC)[reply]

Guidelines and policies generally don't work where they contain aspirational language that can't really be concretized; it simply provides levers for disruptive wikiawyering.  — SMcCandlish ☏ ¢ 😼  00:33, 19 October 2023 (UTC)[reply]

That is why we should not encapsulate a complete definition in a sentence. In the case of tidal-locking:

Tidal locking between a pair of co-orbiting astronomical bodies occurs when one of the objects reaches a state where there is no longer any net change in its rotation rate over the course of a complete orbit. In the case where a tidally locked body possesses synchronous rotation, the object takes just as long to rotate around its own axis as it does to revolve around its partner.

Why don't abstractize the idea of "spin–orbit resonance" in the first sentence and explain what is it in the second sentence?

Tidal locking occurs when one astronomical body has a fixed resonance ratio between its rotational and orbital period. The most common resonance ratio is 1:1, meaning that the tidally locked body completes one rotation around its own axis for every revolution around the parent body. This specific phenomenon is also called synchronous rotation.

The first sentence can serve as a formal definition for the topic. It might have a lot of jargon ("resonance ratio", "rotational period", "orbital period") but it is ok when a newcomer to the topic just need to read more to understand what these terms meant. CactiStaccingCrane (talk) 13:37, 21 October 2023 (UTC)[reply]

Many highly technical topics on Wikipedia would benefit from putting (significantly) more effort into accessibility for a wider audience. It's not always possible to eliminate jargon, formulas, etc., but especially for basic topics of wide interest, we should try try try.

As a concrete example, I just stopped by Computer programming, and wow what a mess the lead is there. –jacobolus (t) 15:38, 19 October 2023 (UTC)[reply]

I tried doing something with the first paragraph, at least. XOR'easter (talk) 23:12, 19 October 2023 (UTC)[reply]

Please feel free to write something more concise! (With or without copying bits from my proposed text above.) –jacobolus (t) 19:12, 18 October 2023 (UTC)[reply]

Oh, and I did write a user essay of my own that I intended to be a one-stop shop for physics and math people wanting to get started on Wikipedia (and for Wikipedia generalists who have to deal with mathematicians and physicists showing up). XOR'easter (talk) 02:30, 18 October 2023 (UTC)[reply]

Your essay is the main motivation I want to rewrite this page. When I asked what makes people hates Wikipedia, almost all said that Wikipedia's article is so so unreadable. It triggers me a lot that we are spending all of our time verifying the facts, copyediting, debating about inane things when we are spending so little time actually making the text understandable to the lay reader. If the lay reader could not even understand what are we talking about and expert readers are just skimming Wiki articles for citations, then we are writing encyclopedic articles for no one really. Too many articles start with false balance filler:

Criticism of Israel is a subject of journalistic and scholarly commentary and research within the scope of international relations theory, expressed in terms of political science)

or punch the reader eyes with jargons for the sake of technical accuracy:

Pharmacology is a branch of medicine, biology, and pharmaceutical sciences concerned with drug or medication action, where a drug may be defined as any artificial, natural, or endogenous (from within the body) molecule which exerts a biochemical or physiological effect on the cell, tissue, organ, or organism (sometimes the word pharmacon is used as a term to encompass these endogenous and exogenous bioactive species).

Yes, I do agree that my rewrite is a bit opinionated. But I want to show in the how-to very clear that readability should be our number 1 priority and why it trumps all other metrics of quality. In my opinion, how-tos needs clear instructions and concrete examples on what to do, and they can be a bit opinionated because it is not an encyclopedic article. CactiStaccingCrane (talk) 13:19, 21 October 2023 (UTC)[reply]

Perhaps it would be better to add an additional essay that is more explicitly "how to" / more explicitly opinionated; I agree with @SMcCandlish that a page called a "guideline" is prone to being mined for "levers for disruptive wikilawyering". –jacobolus (t) 15:29, 21 October 2023 (UTC)[reply]

[ below comment rethreaded after extracting tangential idea –jacobolus ]

However, I still do think that this guide should be made more understandable and be more nuanced. The guideline/policy status is pretty arbitrary, we should not follow instructions in pages with template {{guidelines}} and {{policy}} to the letter anyways, because that is against the spirit of the rule. If a person is Wikilawyering then we should get ban of them and not trying to bend our guideline/policy make it needlessly complicated. CactiStaccingCrane (talk) 16:48, 21 October 2023 (UTC)[reply]

Sounds nice in theory, but basically never happens. People can wikilawyer the living F out of everyone for literally years in this sphere before ANI, ArbCom, or AE will do anything about it. It is very close to impossible to get anyone topic-banned from doing disruptive things that have anything to do with style or article titles. Anything in a guideline that has to do with writing style has to be very, very carefully crafted to avoid providing wikilawyering leverage. PS: I really don't know why this is categorized as an "editing guideline" instead of part of MoS, since it's entirely about the style of writing technical articles.  — SMcCandlish ☏ ¢ 😼  21:33, 21 October 2023 (UTC)[reply]

Alternately, instead of guideline pages, perhaps it would be better to try to start some kind of WikiProject type social group targeted toward lay accessibility of technical content, which could try to encourage / organize collaboration specifically around roaming around locating and improving gratuitously inaccessible pages (especially lead sections), explicitly reviewing technical articles for accessibility, perhaps starting with Vital articles, and/or working with existing WikiProjects related to technical subjects to help them focus on the same. –jacobolus (t) 15:35, 21 October 2023 (UTC)[reply]

That's a great idea. CactiStaccingCrane (talk) 16:41, 21 October 2023 (UTC)[reply]

The ambitious (but perhaps necessary) thing would be to recruit new editors who are enthusiastic about this kind of improvement. XOR'easter (talk) 17:38, 21 October 2023 (UTC)[reply]

I also think this idea has merit. I have mulled creating a Signpost article on the topic, but have not summoned the necessary energy. A Project would certainly be more systematic and seems as though it would gain more attention. A project also has the potential of marshaling forces needed to counter the likely resistance among technically-educated editors who know their stuff, but don't always know how or care to translate their knowledge into comprehensible prose in the Introduction of an article. DonFB (talk) 06:09, 24 October 2023 (UTC)[reply]

@Jacobolus, with my Wikipedia:WikiProject Council hat on, I implore you not to start another group (yet). It's far better to join an existing group like Wikipedia:WikiProject Guild of Copy Editors, or, if you can't find a relevant one, to gather up all of your best friends and WP:REVIVE one. The important point is that you really do have to have at least half a dozen editors involved, or the group won't be able to sustain itself. If it's just you and maybe one or two other editors, then the infrastructure (e.g., around tagging articles and describing the group's interest areas) will swamp your actual goal. WhatamIdoing (talk) 01:03, 31 October 2023 (UTC)[reply]

I certainly don't want to spend a lot of effort on organizing. Some kind of complicated new bureaucratic system "tagging articles" etc. sounds like a pain. Mostly I just don't know any current page(s) or social structure for people discussing or working on this deliberately. So if such a page doesn't exist, it might be worth making one. But if there are existing relevant pages, leaning on those certainly might be better. My general impression is that the typical WikiProject is over-focused on leaving its tags and ratings around, and not focused enough on discussion and concrete collaboration. YMMV. –jacobolus (t) 01:35, 31 October 2023 (UTC)[reply]

New WikiProjects, in particular, tend to fall into that trap.

Dhtwiki could probably tell us whether GOCE would be willing to host such efforts. There are a few thousand pages in Category:Wikipedia articles that are too technical. These 15 articles are tagged for general copyediting (=GOCE's main concern) as well as jargon, so that might be an easy place to start. WhatamIdoing (talk) 01:59, 31 October 2023 (UTC)[reply]

I myself would be against suddenly adding potentially thousands of articles to our load, as the GOCE has a year-long backlog already, unless doing so resulted in a flood of new copy editors to our ranks that would more than make up for the increased burden. But I don't speak for the guild, and a talk-page discussion there might yield other opinions. The articles marked as "too technical" are often in need of general copy editing even if not marked as such. So, there's definitely an overlap between our aims.

I can't be sure what constitutes a "gratuitously inaccessible" article, but this quote, at 36 Signal Regiment (Canada), is a likely candidate: The role of 36 Signal Regiment is to force generate combat capable signallers and Communication Information Service (CIS) capabilities to enable command and control in support of Canadian Forces domestic and expeditionary operations. Some articles have the explanatory material but it needs to be made clear. You would expect an article named Ahaetulla prasina to contain recondite terms. However, the attempt to explain "preocular" there ...preocular scale in front of the eye... adds confusion for not having the explanatory text set off. Then there's the attempt to explain by linking to other articles, some of which seem more technical, without being marked as such (compare Transactional memory, linked from Advanced Synchronization Facility), which suggests that either the former article is appropriately technical (I belive in the possibility of that being the case) or that people just haven't gotten around to tagging it yet. Dhtwiki (talk) 07:46, 31 October 2023 (UTC) (edited 05:41, 11 November 2023 (UTC))[reply]

More examples (they are legion!) of knowledgeable editors simply making no attempt to convert jargon/bureaucratese/tech manual lingo into ordinary English words and syntax that would be recognizable to non-specialist readers. Their lack of effort is not malign; they simply don't have the concept of writing for a general non-technical readership—especially in the opening sentence or two—and thus don't do it. Jargon terms are linked to other articles, in which the ledes are laden with more jargon, linked to still more jargony articles...a reader can easily click/tap three, four or more levels deep with no guarantee of clarity, but only an excursion into a wilderness of techno-babble. DonFB (talk) 08:17, 31 October 2023 (UTC)[reply]

I agree with everything DonFB says here. However I don't think the answer is to prohibit Wikilinks in the lead, as he advocated. Discouraging links in the lead is not going to force jargon-obsessed technical editors to write plain-language explanations for general readers, believe me.

I also don't think another WikiProject is the answer. We should rewrite the Introduction guideline here to strengthen the need for plain language explanation of jargon. I would be willing to bet that 95% of the visitors to most technical articles don't read beyond the introduction. --Chetvorno^TALK 21:13, 8 November 2023 (UTC)[reply]

I would be open to strengthening the text in Wikipedia:Make technical articles understandable#The lead section somehow, though to me it seems rather clear already. Do you have thoughts for new phrasing that could augment or replace what currently stands there? XOR'easter (talk) 21:36, 8 November 2023 (UTC)[reply]

95% of the visitors to most technical articles don't read beyond the introduction. – My impression is that this is true of all articles, not just technical articles. Which is why ideally the lead section should be a reasonable summary as well as broadly accessible, especially for articles about fundamental concepts. –jacobolus (t) 21:47, 8 November 2023 (UTC)[reply]

Quite right, and that's why my focus has consistently been on lede sections; I don't fret about highly technical text/jargon/formulae/etc within the body of articles. It's to be expected that material there can be challenging. But most people don't get that far, and we need to serve them much better in article intros. DonFB (talk) 22:52, 8 November 2023 (UTC)[reply]

It may be true that most jargon obsessed editors won't desist, but I think other editors should be encouraged and supported by Policy/Guideline to fix problematic ledes. To me, the biggest problem is unfamiliar or unrecognizable jargon, so discouraging its use would seem to be the approach to take. I don't object to links from things like names of places, events, or people; the basic meaning of the text does not depend on knowing their details; those links are a courtesy to readers who want to know more. DonFB (talk) 23:22, 8 November 2023 (UTC)[reply]

The Introduction section already discourages jargon. But I think most editors are going to say it is impossible to write introductions to technical articles without using the appropriate terminology (i.e. jargon). That's my opinion too. So the Introduction section should also strongly suggest that jargon terms that must be used be accompanied by brief plain-language definitions. --Chetvorno^TALK 01:32, 9 November 2023 (UTC)[reply]

Maybe what that section of the guideline needs is an example or two. XOR'easter (talk) 05:56, 9 November 2023 (UTC)[reply]

I was surprised by all the pushback from expert editors against making articles easier to understand in the thread Yes, they are too hard to understand. No, really above. Although most of this stuff is already in the article, maybe it would help to put a list of their objections, and answers, in the guidelines. Here are some --Chetvorno^TALK 06:50, 9 November 2023 (UTC)[reply]

• "Some articles are about topics too advanced to make them comprehensible to general readers; it is a waste of time to try"
  Look at the Encyclopedia Britannica entry for the topic: a good encyclopedia writer can explain any topic to general readers. A concept can always be made more comprehensible. And even if the concept can't be fully explained in the introduction, it should include context: What categories does this concept fit in? Does it have any analogies to simpler concepts? What properties does it have? What are some simple examples? Who discovered it? When? What is it used for? Why is it important? This is a large part of what general readers want to know. Context is a major part of understanding. --Chetvorno^TALK 06:50, 9 November 2023 (UTC)[reply]

  It is simply not true that there are some topics that can be made comprehensible to general readers, especially without serious violations of our content policies and guidelines. (E.g., Galois cohomology.) Articles should be written in a way that complies with our content policies and guidelines. If no source exists that explains a topic for general readers, then it is typically the case that a Wikipedia article cannot either, without at the very least engaging in original research, and at worst being completely wrong. For articles about very abstract mathematical concepts, often the best "introductory" sources available are the AMS Notices "What is..." columns, written by subject experts for a "general mathematical" audience (i.e., graduate students already working in the area or mathematics PhDs who are not specialists.) That's often the best we can hope for. Tito Omburo (talk) 12:31, 10 November 2023 (UTC)[reply]

  A subject of interest primarily to e.g. math PhDs working on unsolved problems, lawyers working in the corners of inter-state contract law, or industrial chemists trying to eke an extra percent of yield in some obscure reaction can still usually be made accessible to advanced undergraduates or early grad students. I think we can often do better at accessibility than the "What is..." columns for mathematical topics, but those are indeed in many cases more accessible than what we currently have at Wikipedia, so are a good source to take inspiration from and cite. –jacobolus (t) 15:35, 10 November 2023 (UTC)[reply]

  There are often no sources as good as "What is..." columns, particularly for abstract subjects that are only of interest to specialists. Our own coverage should of course aim to be as accessible to as wide a likely audience as possible, but the expectation needs to be tempered. Firstly, Wikipedia follows sources. That is policy, and there is a policy against original research. To some extent, this is relaxed for explanations aimed at making things accessible, but still it is a policy demanding some deference. You can find examples (like homology theory) where there are loads of accessible sources out there. I think the typical editor fails to appreciate that this tends to be the exception rather than the rule, and expectations should be set appropriately. Tito Omburo (talk) 16:17, 10 November 2023 (UTC)[reply]

  "Follow the sources" doesn't mean you can't ever put a brief definition of a jargon term inline, can't include examples or pictures, can't describe historical/mathematical context, etc. –jacobolus (t) 17:37, 10 November 2023 (UTC)[reply]

  No, but not every subject has a picture that explains it (most don't). Some jargon terms require still more jargon terms to define them. The article Galois cohomology you recently complained about has historical context out the wazoo. An explanation of cohomology as "cocycles modulo coboundaries" is unlikely to be availing, a Galois group as "group of automorphisms of a field fixing a subfield", etc. One has to be reasonable, and to communicate with other wikipedians that in many cases an article will not be understandable at all to someone without a good background in mathematics (and certainly not without doing OR). I'd like you to at least acknowledge this point. It seems like it should be pretty uncontroversial. Tito Omburo (talk) 18:15, 10 November 2023 (UTC)[reply]

  Agree with jacobolus above. Tito Omburo: If you read what I wrote above carefully, you might find there is not as much difference between your position and ours as you think. I said that a good encyclopedia writer can explain any topic to general readers; not make it comprehensible. This is the key point that science writers understand but many specialist WP editors miss: there are many levels of "explanation" (Mathematician explains infinity in 5 levels of difficulty, Wired mag, Harvard professor explains algorithms in 5 levels of difficulty; I'm not saying these vids are good, just that they illustrate the idea of multiple levels of explanation) --Chetvorno^TALK 00:29, 15 November 2023 (UTC)[reply]

  • Sorry, I'm lost on that one. What does it mean that the good writer can explain the topic, but not necessarily make it comprehensible? And what good would it do? --Trovatore (talk) 00:38, 15 November 2023 (UTC)[reply]

In the Galois cohomology article, jacobolus and I would like to see more effort to explain the actual concept simply; a diagram or example, say. But for nonmathematical readers an adequate "explanation" could be just its history, and why it is important. As you said, the article has historical context out the wazoo. Why is none in simple language in the introduction? The history and significance of a concept can usually be summarized for general readers without an understanding of the concept; that's what White House briefing books are for. --Chetvorno^TALK 00:29, 15 November 2023 (UTC)[reply]

What non-mathematical reader is going to look up Galois cohomology? Or Local Euler characteristic formula? Or Sylvester matroid, Lusternik–Schnirelmann category, or Symplectic spinor bundle? These aren't pi or prime number. Even the history and motivation of some of these topics is along the lines of, "Take a way of thinking about classical mechanics that nobody learns until at least three years into college, and generalize it to a higher level of abstraction so that even the physicists are lost". Sometimes, the best one can do is to take a topic that nobody even hears about until well into their PhD and make the opening lines approachable to a first-year grad student. XOR'easter (talk) 03:39, 15 November 2023 (UTC)[reply]

People who ask college students "So, what are you studying at university right now?"

I think you're misunderstanding the level of detail we (or at least I) am hoping for. Something like "Incomprehensible is a type of abstract algebra related to classical mechanics" would suit me. I'm looking for "Punica granatum is a fruit-bearing plant". You seem to be "But you didn't tell them about the sarcotestas yet! They don't know anything if they don't know about the sarcotestas!" The sarcotestas may well be important (and delicious), but we've first got to establish the basics, like it's not a kind of Granite and it doesn't have anything to do with the Punic Wars. WhatamIdoing (talk) 03:50, 15 November 2023 (UTC)[reply]

As a veteran of many such conversations over the past twenty-odd years, involving first my studies and then my research and teaching, and as a witness of many colleagues having such conversations too, this is how they go:

"So, what are you studying at university right now?"

"Math."

Even if the answer is more specific than that, it will be something like topology or number theory. The articles I mentioned above (picked at random out of many similar possibilities) are all much more specialized. And most articles of that sort do tend to establish their context, either with the standard "In mathematics..." opening gambit or by saying what part of mathematics they belong to (combinatorics, differential geometry, etc.). So, they do the job of indicating that they're not talking about Roman coins or Cambodian verb forms. XOR'easter (talk) 04:03, 15 November 2023 (UTC)[reply]

• "Explaining things in simple terms is 'dumbing down' the article"
  Adding simple explanations does not mean removing any of the advanced content. Many expert editors regard any simplified explanation as ‘lies to children’. This is probably too extreme (and haughty) a view. Simplified explanations given in introductory-level textbooks should be acceptable, as long as the ways it is inadequate are mentioned. --Chetvorno^TALK 06:50, 9 November 2023 (UTC)[reply]

  Many many topics are not covered in "introductory-level textbooks" (and certainly not for textbooks you would consider "introductory": for example, the book "Modular forms and Galois cohomology" is an introductory level textbook on Galois cohomology, but I doubt a non-expert would recognize it as such). If topics are covered in (what you're thinking of as) introductory textbooks, it is certainly possible to give introductory descriptions of them. But otherwise, probably not without violation policies at best, and being completely wrong at worst. Tito Omburo (talk) 12:31, 10 November 2023 (UTC)[reply]

  The article Galois cohomology makes very little effort to be accessible beyond a trivially tiny audience of specialists. Its accessibility could be dramatically improved without sacrificing any usefulness to a specialist audience. –jacobolus (t) 15:38, 10 November 2023 (UTC)[reply]

  That's just not true. The lede and most of the article is perfectly understandable to a non-specialist. If there are reliable sources that do a better job of introduction, they can be used to improve the presentation. Tito Omburo (talk) 15:44, 10 November 2023 (UTC)[reply]

  The lead section consists entirely of jargon. Someone who hasn't yet taken a course in homological algebra is going to be able to make no sense of it whatsoever. Edit from after the following comment: Worse, if someone clicks through to group cohomology, homological algebra, Galois module, Galois group, Field extension, Galois representation (redirects to Galois module), or Exact functor, they get an equally inaccessible description. This entire cluster of articles is a mess from an accessibility perspective. The article Homology (mathematics) stands out for being much better than any of those, but e.g. a curious undergrad physics student arriving at Galois cohomology is never going to manage to make it there. –jacobolus (t) 15:52, 10 November 2023 (UTC)[reply]

  Look, the context-setting in the article is at a more accessible level than Serre's introductory textbook on the subject, and Hida's introductory book, and even the encyclopedia entry at EoM. In fact, arguably our article should include more technical detail to be useful to as wide an audience as possible (e.g., transcribed from [2]). The problem with a topic like this is that what is one to explain in a short lede section? All of homological algebra? All of cohomology theory? All of Galois theory? Actions of groups on modules? The lede even leaves out any discussion of class field theory (which is actually quite essential), so in this way is pitched "one level down" from an intended readership. At some level it's going to be "turtles all the way down". Understanding the lede (or even the rest of the article) does not require a detailed understanding of homological algebra, contrary to your contention. Tito Omburo (talk) 15:57, 10 November 2023 (UTC)[reply]

  arguably our article should include moretechnical detail – this would also be fine. These two motivations are not in conflict here. This article has plenty of space to grow to serve all of these audiences. –jacobolus (t) 16:49, 10 November 2023 (UTC)[reply]

  What I'd like to see out of all of these articles listed above is for someone who just finished a typical introductory undergraduate abstract algebra course to be able to either directly understand the majority of the lead section, or else be linked within 1 hop to articles about the prerequisite jargon such that those articles' introductory sections are largely understandable.

  My experience is that mathematicians in particular have this problem more than almost anyone else I have run into, that they forget what it's like to not already know the meanings of the technical terms they spent years of toil making sense of, so take it for granted that other people will be able to understand a long string of conceptually sophisticated jargon without any assistance. –jacobolus (t) 17:06, 10 November 2023 (UTC)[reply]

  One hop from a specialist article is a non-specialist, not someone who just finished an undergraduate algebra course. Tito Omburo (talk) 17:16, 10 November 2023 (UTC)[reply]

  When I say "trivially tiny audience of specialists" above, what I mean is "people with math PhDs", and when I say "general audience", I mean e.g. physicists, economists, computer programmers, and undergraduate math students. Wikipedia's advanced math articles, on the whole, do a very poor job at including a wide audience. They are written largely for people who already know the subjects under discussion in detail, and could in most cases be made at least somewhat legible to a much wider audience (like 100x as many people) without much compromise. –jacobolus (t) 17:19, 10 November 2023 (UTC)[reply]

  Perhaps the mathematicians, who have spent their lives explaining mathematics to others, might actually be right? I'd like you to at least entertain this possibility rather than simply dismissing it by saying they are doing a "very poor job".

  Regarding the example of the Galois cohomology article, what would you propose we add to the lede to make it more accessible to your hypothetical undergraduate? Would the following be adequate? "Historically the first, and one of the most basic applications of Galois cohomology is to describe the structure of the group of elements of norm one in a cyclic field extension of degree ${\displaystyle n}$, that is, an extension field where the Galois group ${\displaystyle G}$ is generated by a single element ${\displaystyle \sigma }$ of order ${\displaystyle n}$. The cohomology group ${\displaystyle H^{1}(G,L^{*})}$ with values in the group of non-zero elements of ${\displaystyle L}$ can be described as the quotient of the cochains by the coboundaries of dimension one. A cochain is a group homomorphism from ${\displaystyle G}$ to ${\displaystyle L^{*}}$, which is completely determined by its value on ${\displaystyle \sigma }$, say ${\displaystyle b=f(\sigma )\in L^{*}}$, such that the product ${\displaystyle b\sigma (b)\cdots \sigma ^{n-1}(b)=1}$. This product is just the relative norm of the element ${\displaystyle b}$. A coboundary is an element of the form ${\displaystyle b/\sigma (b)}$. Hilbert's theorem 90 states that every such cochain is a coboundary, i.e., ${\displaystyle H^{1}(G,L^{*})=1}$ is the trivial group. Therefore, every norm one element of the field ${\displaystyle L}$ can be expressed as ${\displaystyle b/\sigma (b)}$." Tito Omburo (talk) 17:44, 10 November 2023 (UTC)[reply]

  My impression is that mathematics teachers (as a broad social group under their current societal constraints; I'm not calling out any specific person here) don't generally do a good enough job of explaining mathematics: the field is notorious for chasing away promising and interested students by making them feel stupid and not giving them enough support, at every level from primary school up through grad school. It's very much a sink-or-swim go-figure-it-out-for-yourself kind of subject.

  My impression is that the world's best mathematical expositors and popularizers are mostly not writing Wikipedia articles (because they are busy doing something else), and most of the people writing math articles on Wikipedia at every level are not in general making legibility to the broadest reasonable audience one of their top priorities. But I think it should be.

  Galois cohomology was your example, not mine, and is probably not the best place to start since this is a topic that has a lot of prerequisite concepts, even in the best case is only going to be vaguely comprehensible to non-mathematicians, and only gets about 20 page views per day. Energy directed at making lower-level topics legible will have better payoff.

  But if a hypothetical physics undergrad did want to figure out even roughly what Galois cohomology was talking about, and they tried clicking through any of the jargon words in the first two sections, the prerequisite articles they'd arrive at are all equally inscrutable, as are the prerequisites to those; our hypothetical student would have to go probably 5+ hops deep in any given direction to get to a page they could make any sense of at all. That's unfortunately pretty exclusionary. Most of our more basic articles like field extension, normal subgroup, exact sequence, automorphism group, etc. etc. etc. are still written in a very spare and inaccessible style.

  If you want to figure out if a particular explanation is clear enough, the best people to try it on would be some students who are a few years away from taking the course in question. Ask them if they can figure out what all the jargon roughly means, see if they can restate the main idea in their own words, etc.

  might actually be right? – what does this mean? I claim that our articles within 1–2 hops of your original example are almost entirely inaccessible to anyone who isn't at least a math grad student. Is your claim that any other result is impossible? You think this is the best we can do? –jacobolus (t) 19:14, 10 November 2023 (UTC)[reply]

  Again, you're dismissing not just mathematics researchers but even mathematics pedagogy as inscrutable. You start to sound like a classic crank. I'm just trying to inject some realism into this discussion. Can we do better? Certainly. Can we make everything accessible to everyone? Absolutely not.

  Regarding the examples of hops, the articles homological algebra, Galois group, field extension, and even group cohomology do include an accessible description in all cases. Definitely not enough to understand what a Galois group or field extension "really is" (because that requires knowing what a field is, what a group is, what an automorphism is). But they set their topics in context, and include links for further browsing. I don't claim that any of these are perfect, but blanket claims that everything is terrible strike me as very overwrought.

  Also, I am still curious whether you think my proposed addition to the Galois cohomology article enhances its accessibility, which you seem not to have addressed. Tito Omburo (talk) 19:25, 10 November 2023 (UTC)[reply]

  What you call "accessible description" I call "completely illegible to anyone who isn't already a math grad student". There's a very wide gulf between this state and making "everything accessible to everyone", a straw man which nobody is proposing. I urge you to go hand a printout of the first couple sections of any of these articles to randomly selected undergraduates in technical subjects or technical professionals from other fields (or even advanced math undergrads), and directly ask them whether they can make sense of them. I would expect almost none of them to find these articles legible. This is not just my personal opinion; people complaining about how illegible math articles are on Wikipedia is so common it has become cliché. As for your proposed addition, I'm not the person to ask about this. I don't know or care too much about algebraic number theory, and it's been 15 years since I last spent any time learning about cohomology or Galois theory. –jacobolus (t) 19:55, 10 November 2023 (UTC)[reply]

  I don't have much contact with undergraduates these days. But if I were to compare our article on group cohomology to the entry in the eom, or Serre's book, or, pretty much any other source that discusses the topic, ours is easily more accessibl, informing the reader that it studies groups using methods of topology by regarding groups as spaces. That's accessible! Similarly, our article on homological algebra informs the reader that the subject uses techniques from topology (homology theory, an article you recently praised) to study algebraic structures. That's accessible! (The eom article unhelpfully tells the reader that it is the study of derived functors.) The article Galois group appropriately points a reader to the more elementary account in Galois theory, but the first sentence eludes some technical details (without telling lies to children). You say I am attacking a staw man, but you seem unable to assess an appropriate level of accessibility for a technical article. In my experience, most Wikipedians suffer the same deficiency. People regularly complain that articles on technical subjects are technical. If they want to make them less technical, and have the expertise to do so, they are welcome to. But instead you have collectively dismissed the expertise of those that do have the expertise to do so as doing a "very bad job", or even those responsible for mathematics pedagogy (!), dismissing their expertise in such matters, even when more often than not such people are telling you they're doing the best they can. (Omg cut it out with the stupid templates, is there any evidence that a block of text at the top of a technical article improves its accessibility by obscuring the article contents?) That attitude is very counterproductive, and often leads to misguided threads like this one. Tito Omburo (talk) 21:21, 10 November 2023 (UTC)[reply]

  I will certainly agree with you that we're less opaque to the uninitiated than Serre's book (wide accessibility was clearly not among Serre's goals), and I will also agree with you that these Encyclopedia of Math articles by V.E. Govorov / A.V. Mikhalev and L.V. Kuz'min also don't do a good job at this. But "less inaccessible than some completely inaccessible sources" is not really where anyone should be aiming.

  Calling people who are frustrated by articles they can't read "deficient" isn't very welcoming. But note that these "deficient" people even include professional mathematicians sometimes: Math is so big that nobody understands every corner of it from top to bottom, it's of general benefit to all of us to make every article as broadly accessible as practical, even if it takes adding a couple extra paragraphs of introductory explanation that seem a bit redundant, or going out of our way to write overview articles that do more hand-holding than any advanced textbook would bother with.

  I don't doubt page authors are generally "doing their best", but writing for broad legibility takes significant dedicated effort and if it isn't a priority it just won't happen. I don't mean offense, and am not trying to call out anyone specific or "dismiss" anyone's efforts. There's just a significant gap between where we currently are and an encyclopedia that strongly prioritizes helping non-experts make sense of unfamiliar topics.

  I don't understand what you mean by "block of text" or what templates you are referring to. If you are talking about navigation templates, I typically find those to be space and attention wasters, and never add them to articles. Some people seem to like them though. –jacobolus (t) 22:55, 10 November 2023 (UTC)[reply]

  It appears to be a widely held belief that templates such as {{technical}} improve the accessibility of technical articles, but I am confused how that is supposed to work. A block of irrelevant text before the lede of an article does not seem to be an improvement for accessibility (or otherwise). Tito Omburo (talk) 23:41, 10 November 2023 (UTC)[reply]

  I don't think adding banners has ever helped any article. Cf. User:Jorge Stolfi/Templates that I sorely miss. –jacobolus (t) 23:46, 10 November 2023 (UTC)[reply]

  Looking at the group cohomology article, I'd guess that a third- or fourth-year physics undergraduate would get as far as tools used to study groups before giving up. There's a parenthetical before the bold text which is at least as obscure as the article name itself. The first sentence ends by mentioning algebraic topology, which just looks like the combination of two math words to make an even mathier word. The second sentence brings in group representations, which are an advanced undergraduate or early graduate topic, then drops in the term G-module. It could be worse, but it's not easy going.

  I share the concern voiced above that focusing on articles that get infrequently read and are targeted at the highest levels of education may be misplaced effort. I'd like to see Wikipedia have really top-notch coverage of math topics taught in high school and the first couple years of college. To me, that seems the way to aim for benefiting a wide population (and the people learning things at even younger ages probably aren't looking them up in an encyclopedia). This means Pythagorean theorem, quadratic formula, sine and cosine, linear equation, ..., on up through calculus, Matrix (mathematics), Eigenvalues and eigenvectors, and the like. XOR'easter (talk) 02:06, 11 November 2023 (UTC)[reply]

  About the audience:

  Our general rule of thumb is that we want to write one level down. That is, a subject that is typical for a PhD in math should be written so that it is understandable to someone who is starting a master's in math; a subject typical for a master's in math should be written so that it is understandable to someone who is starting a bachelor's in math; the subjects in the math classes taken by first-year math majors should be written so that high school students can make sense of them, etc.

  I suggest, though, that for a more or less sequential study like math, a different way to think about the audience is: Students in class n, who are looking ahead by a couple of semesters, should be able to understand the articles about classes n+2, for the very simple and obvious reasons that (a) they are likely to read these articles and (b) if they can't understand what the subject is, they may not be interested in taking that class. Galois cohomology should be understandable to someone who is currently studying Algebraic topology. Abstract algebra should be understandable to someone who has finished the section in calculus on vectors. Don't think of it as "dumbing down", because it's not. Think of it as telling math majors why all the cool people pick your favorite subject. WhatamIdoing (talk) 20:23, 14 November 2023 (UTC)[reply]

  XOR'easter makes a good point above that probably applies to me:

  "focusing on articles that get infrequently read and are targeted at the highest levels of education may be misplaced effort".

  Nevertheless, when I see an introductory section like that for Group cohomology, my blood tends to boil. My view of writing one level down is to apply it, if appropriate, to the body of the article. The introduction, I insist, should be aimed at a broad public audience, without regard for education levels. DonFB (talk) 21:03, 14 November 2023 (UTC)[reply]

  I'd prefer that it started with something like "It's a way to compare two geometrical objects" (if you can't explain what it is, then explain what it does), but yes, focusing on something that gets 1,000 page views per day might have more value than something that (as this article does) gets 100 page views per day. OTOH, there are a lot of websites that talk about basic concepts, and we might be one of the only options for really advanced subjects. WhatamIdoing (talk) 21:59, 14 November 2023 (UTC)[reply]

  "It's a way to compare two geometrical objects" I don't think that's a true description at all, a group isn't a geometrical object at all (Group (mathematics) actually does imo a pretty good job of explaining what it is) but there's no nontechnical replacement for the term "group" simply because it is such an abstract concept (and the basis of abstract algebra).

  I actually would agree with what you said above about writing for the person taking n-2 class. I searched and group co-homology is taught in the second grad class on abstract algebra at my undergrad institution (Berkeley) and at Stanford; the prereqs just for the first grad class is two semesters of undergrad abstract algebra. I have taken one semester of abstract algebra, and it could probably be more accessible to people like me (I don't really understand what it's about either), but I don't know why people are so pressed that articles like Group cohomology are non-understandable to people without the relevant background. There's a reason why math classes are hard and people don't skip prereqs - advanced math concepts are highly abstract and require knowing the background. Even the examples are going to be complicated math objects (even if Group cohomology#Basic examples isn't exactly comprehensible, it does look more understandable to me).

  I think maybe non-math people don't appreciate how abstract and complicated certain math concepts are - sometimes a concept is only relevant to a math phd student, and that's fine. There's no way to teach a semester of abstract algebra in the just the lead of an article. Galobtter (talk) 00:58, 15 November 2023 (UTC)[reply]

  Sure, there's no royal road to mathematics, but there are two considerations:

  1. That particular article really ought to be written in a way that you and your classmates can understand it, at least enough to make you think either "Now there's a class I'd really like to take some day" or "Eh, I'd rather do more with combinatorics and set theory".
  2. The most important thing is that people can get the information they need. You can get decent information on some subjects from almost anywhere. If (and I don't know whether this is true, in this particular instance, but if) Wikipedia is one of the few places that provides freely accessible information on ______, then we should make that information as good as we can because it's not popular.

  On that last point, the way that I think of articles on rare diseases is this: If you get a text message from someone you care about, and it says "We're at the hospital with the baby. The doctors say it's Scaryitis", then you should, with a quick glance at Wikipedia, be able to figure out whether your reply should sound closer to "What a relief" or "I'm so sorry". The rarer the disease, the more important it is for us to have something. You can get information about fevers and ear infections and flu anywhere. Your options for learning about unusual subjects might be much more limited. All the deep understanding can come later, but anyone should be able to get something out of the article. The "something" for Group cohomology might have to be as small as "So, I guess you're a math major?", but we should manage something. WhatamIdoing (talk) 02:47, 15 November 2023 (UTC)[reply]

  I think standard "In mathematics" opening is enough for the "So, I guess you're a math major" message. Could the article do more for, e.g., first-year grad students? Probably. Would that benefit come from adding generalities to the opening paragraph? Maybe, but not necessarily. XOR'easter (talk) 03:53, 15 November 2023 (UTC)[reply]

  Probably, because if the less specialized information isn't in the lead, it won't be read. WhatamIdoing (talk) 04:01, 15 November 2023 (UTC)[reply]

  I mean, when the topic is already hyperspecialized, I'm not sure that "less specialized information" is what its readers want. People who aren't there for the substance can nope out in the opening line; those who stick around might need something other than extra words of generalities. For example, lots of our higher-level math articles are under-referenced. They typically don't need a footnote after every line, but they ought to be brought up to a standard of one per section at least. Knowing that a particular chapter in a particular monograph is a thorough survey of some aspect of a topic is going to help the actual audience more than repeating the summary of the prereq's prereq. XOR'easter (talk) 04:27, 15 November 2023 (UTC)[reply]

  It's a relatively common complaint by even people with math degrees (including PhDs sometimes) that a significant proportion of Wikipedia math articles are inscrutable until after the reader already knows quite a lot about the subject. For instance see the ongoing discussion (started after this one) WPM talk § Making mathematical articles more broadly accessible. –jacobolus (t) 04:33, 15 November 2023 (UTC)[reply]

  Yes, I've experienced that myself. But the improvements that often seem necessary in that regard seem orthogonal to the kinds of changes being discussed here. XOR'easter (talk) 04:37, 15 November 2023 (UTC)[reply]

  I don't think so. WhatamIdoing (talk) 04:42, 15 November 2023 (UTC)[reply]

  I think they're related, but you're right there's also a lot of work to do beyond making the lead section as accessible as practical. The next step after that is to make sure that someone who has more or less satisfied the expected prerequisites (which might be quite advanced) but is new to the specific topic can make sense of most of the article including all of the most important high-level overview material. –jacobolus (t) 04:45, 15 November 2023 (UTC)[reply]

  My sense of the overall situation is that the top priority for some articles, like short ones on highly specialized upper-level topics, should be to expand, organize, and reference them. A lot of improvement in those corners of the encyclopedia would involve writing at a similar level to what the existing text presumes, but just being less half-ass about it. In other places, the top priority ought to be providing a solid opening. XOR'easter (talk) 07:38, 15 November 2023 (UTC)[reply]

  I definitely agree though that clearly citing clearly written sources in every section of every topic would be extremely helpful. (E.g. the most influential papers, survey papers, relevant textbook chapters, student theses with nice introductory context, historical discussions, ...). For most technical Wikipedia articles I have ever looked at carefully, 20 or 30 minutes of searching turns up better sources (sometimes much, much better) than the ones previously cited. –jacobolus (t) 04:45, 15 November 2023 (UTC)[reply]

  DonFB, for a lot of topics, to aim the introduction at a broad public audience is essentially to waste the introduction. You'd need to write this pro-forma pseudo-introduction for the broad public, and then write a separate one for the people who are actually going to read the article. Because they still do need an introduction, to summarize the logical plan of the content being presented. I don't see the point of this.

  I don't want to minimize the problem of articles being written in a way that they are harder to understand than they need to be. That really does happen. But a lot of the content in technical articles is just not even describable to a "broad public audience" without either making it so vague as to be useless, or repeating so much information that the audience that can actually use the article has trouble finding what they're looking for. --Trovatore (talk) 22:58, 14 November 2023 (UTC)[reply]

  My view, which I believe is supported by policy, is that this an encyclopedia for general readership. It seems to me that students in science/math/med/tech subjects should be getting their knowledge in school/college, and that Wikipedia is not the place for them to do serious learning in those subjects. Yet, many articles in those areas seem to be written as though this is a place that offers educational material, above and beyond information for the public. DonFB (talk) 22:35, 14 November 2023 (UTC)[reply]

  There are plenty of subjects where by that standard you'll just entirely eliminate the article from consideration based on too high a level of prerequisite knowledge. That doesn't seem useful to me. Better in my opinion is to do our best to make the articles as broadly accessible as possible, but also try to make them as useful as practical to a specialist audience. –jacobolus (t) 22:55, 14 November 2023 (UTC)[reply]

  Exactly. The corpus of scientific articles on extremely specialist topics is a very valuable part of Wikipedia's body of knowledge. Why would you want to get rid of that? It strikes me as a dog in the manger attitude. There are tons of Wikipedia articles that are unreadable to me at my level of background in the subject; I don't assume that means they shouldn't be there. --Trovatore (talk) 23:06, 14 November 2023 (UTC)[reply]

  I'm not saying articles should be eliminated, only that they be made comprehensible to more than a narrowly defined segment of the population. That's the challenge in writing about the abstruse in a general readership encyclopedia. On the other hand, there are many articles on more mainstream topics in which the intros are needlessly dense and send readers chasing after numerous links instead of using everyday language, as I've previously belabored. DonFB (talk) 23:21, 14 November 2023 (UTC)[reply]

  It's just not possible. Some of these topics, only a narrow segment of the population has any hope of understanding, even in fairly vague terms, what they're about. Now, we can do better than we're doing at making it a less narrow segment. --Trovatore (talk) 23:31, 14 November 2023 (UTC)[reply]

  Even if we accept that this is true ("Some of these topics, only a narrow segment of the population has any hope of understanding, even in fairly vague terms"), I'm not convinced that our articles have done what we can to make even a single sentence understandable in some of our advanced mathematics articles. Our articles are all "Two bodies are considered equivalent if equal torques will produce equal acceleration", and we need at least a little bit of "discussion about how hard it is to push a door open when you put heavy weights on the outside, compared to when you put them near the hinge". WhatamIdoing (talk) 23:50, 14 November 2023 (UTC)[reply]

  Yes, and I was getting at the same general idea here. If we have (or can make) simpler material on the same topic, nothing seems to prevent us combining it with the harder-to-follow but more technical material.  — SMcCandlish ☏ ¢ 😼  00:01, 15 November 2023 (UTC)[reply]

  You said You'd need to write this pro-forma pseudo-introduction for the broad public, and then write a separate one for the people who are actually going to read the article. That's actually not such a bad idea. It sort of combines the Intro To gambit into the main article. DonFB (talk) 23:32, 14 November 2023 (UTC)[reply]

• "Mathematical articles require equations in the introduction":
  Even if true, that doesn't mean it should not also include a simple word explanation. One technique is after giving the equation, to rephrase it in words for nonmathematicians. --Chetvorno^TALK 06:50, 9 November 2023 (UTC)[reply]

  For an article about an equation, it is important to include the equation in the lede. The lede should be a capsule version of the article, as accessible as possible, but also useful to likely readers of the article (which can include experts, graduate students and other researchers, etc, in more technical areas). Tito Omburo (talk) 12:31, 10 November 2023 (UTC)[reply]

• "It is essential to use the correct terminology (i.e. jargon) in the introduction"
  Okay, but a brief plain-language explanation of jargon terms should also be included. --Chetvorno^TALK 06:50, 9 November 2023 (UTC)[reply]

  Sometimes that would take many pages of explanation and background. Tito Omburo (talk) 12:31, 10 November 2023 (UTC)[reply]

Studies show that bad readability in technical articles is Wikipedia's biggest problem when compared to competitors (Britannica, etc.). Seriously, folks, is this so hard to understand that our articles should not be hard to read and understand? Writing plainly is not dumbing things down. Academese and like don't make thinks smarter, just dumber and obsfurcated. _{Piotr Konieczny aka Prokonsul Piotrus| reply here} 03:42, 14 November 2023 (UTC)[reply]

Which study are you thinking of? Can you be specific? –jacobolus (t) 05:48, 14 November 2023 (UTC)[reply]

@Jacobolus CTRL+F readability in my own article at wikiversity:WikiJournal_of_Medicine/Where_experts_and_amateurs_meet:_the_ideological_hobby_of_medical_volunteering_on_Wikipedia, where I cite some relevant works. _{Piotr Konieczny aka Prokonsul Piotrus| reply here} 06:03, 14 November 2023 (UTC)[reply]

These references seem to be about some quantified "reading level" scores (based on syllables per word, proportion of uncommon words, grammatical structure, etc.), rather than subjective assessments of the effectiveness or legibility of the articles for any particular audience. Is that what you were thinking of / referring to before, when you called this Wikipedia's biggest problem compared to competitors? –jacobolus (t) 09:16, 15 November 2023 (UTC)[reply]

Much of what I have had to clean out of science articles was incomprehensible not because it was full of jargon, but because it was written by somebody who didn't have a clue. They got their "knowledge" out of a YouTube video, which got it from a website, which got it from a magazine, which got it from a coffee-table book. Jargon and equations would be a net advance over that: a presentation that requires a couple years of college to understand would still benefit a nonzero number of readers, rather than benefiting nobody at all — or worse yet, actively harming the lay audience by instilling the illusion of understanding.

Nobody wants to write articles that are "hard to read and understand". We should of course aim to cover each topic in a way that is as widely comprehensible as feasible... but what is feasible varies widely. XOR'easter (talk) 13:39, 14 November 2023 (UTC)[reply]

I agree that nobody wants to write articles that are "hard to read and understand", but sometimes humans want to show off. I've worked hard for that jargon; I've been told that this jargon is how True™ Mathematicians talk to other professional mathematicians; I want the world to know that I'm a True™ Mathematician, too, so I've just got to use the True™ Mathematician jargon. Using the kind of language that Prof. Matt Math actually uses in front on my own class would be mortally embarrassing, because he explains everything.

It reminds me of the prepubescent boys in choir. They're desperately waiting for their voices to crack and telling the choir director, in their high little-kid voices, that they can't possibly sing notes in the vocal range they use for everyday speech. Sometimes it's about wanting to feel important and knowledgeable and grown up, rather than anything at all about the effects on other people. WhatamIdoing (talk) 20:31, 14 November 2023 (UTC)[reply]

I think this is a good point to bring up from time to time. It's not easy to distill one's own expertise into a simple explanation for lay-people. One example in overcoming that, I think we've managed to make a small improvement to the introduction of Mandelbrot set with a couple lay-person friendly sentences, when it used to start very technically. I'm a fan of taking the WP:IMPERFECT approach on such issues in general, working to improve things as we can. —siroχo 08:33, 9 November 2023 (UTC)[reply]

One sentence of context-setting is good. But two things stand out. First, there is a vast chasm between being able to set the context for a topic in one sentence, and making the topic accessible to a "general audience" (which is often impossible). Secondly, the Mandelbrot set (or similar article) is a topic for which there is a superabundance of sources aimed at a general audience. But this is not the case for most concepts in the mathematical sciences. I agree that we need to temper our expectations about what is possible, and in particular nip the "everything can be made accessible" fiction in the bud. Tito Omburo (talk) 12:41, 10 November 2023 (UTC)[reply]

+1, some topics are popular among the general public and have and deserve a more lay person explanation, but most math topics are not like that. Galobtter (talk) 01:07, 15 November 2023 (UTC)[reply]

Let's look at Bijection. I assert that the third sentence of the lede, with minor modification, should be the first sentence. As follows:

"In mathematics, a bijection is a relation between two sets, such that each element of one set is paired with exactly one element of the other set, and vice-versa."

Is that an accurate statement?

Consider the linked terms of jargon currently in the first sentence that are eliminated (they can be introduced later):

function, injective, surjective, codomain, domain of the function.

My revision would retain "set" as a linked term, but the word is recognizable to everyone as an expression for collection of things. Unless my reworking of this lede is completely wrong, I assert that ledes of all mathematical articles can be similarly simplified by removal of much opaque jargon from the first one or two sentences and the substitution of plain English, which is very often found later in the same article. DonFB (talk) 21:59, 10 November 2023 (UTC)[reply]

Whether you call a bijection a "function" or a "relation between sets" doesn't really make much difference in how understandable the concept is, in my opinion. I'll agree that "injective (one-to-one) and surjective (onto)" can probably be deferred by a few sentences. This article would also benefit by putting a couple examples in the lead section, by explaining how a function (or pairing) can fail to be a bijection, and by giving short inline definitions of some of the other terms used. –jacobolus (t) 23:03, 10 November 2023 (UTC)[reply]

My purpose here was to show that an opening sentence with what I believe to be unnecessary jargon can be written to be instantly understandable to everyone. I'm not a mathematician (probably obvious), so I would defer to those who are in matters of accuracy, while urging and trying to demonstrate that more accessible language can be used. In this example, I used "relation between", which seems even more apparent in its meaning than "function". I looked up function and find its opening paragraph to be...difficult. The second paragraph begins with what I've understood the word to mean, although it uses "originally":

"Functions were originally the idealization of how a varying quantity depends on another quantity."

For my money, that's how the article should start, (with some adjustments), unless more recent standadization has made it truly obsolete. I offer this as another example of how I think it is possible to use plain English at the very start of an article and defer the techno/jargon/lingo/formula to a position later in the article, or the lede, when that's necessary. DonFB (talk) 00:00, 11 November 2023 (UTC)[reply]

My impression is that people most commonly think of a bijection as an invertible function. While it's true it can also be thought of as a pairing between two sets, I feel like it's a bit less obvious to put that version first.

I'm not really sure if "binary relation" is more likely to be understood than "function". The concept of functions is usually discussed at length in high school while the concept of a binary relation is (while not inherently any more complicated) a bit more obscure I would guess.

In the case of "bijection" per se, I don't think the concept of a function as one varying quantity depending on another quantity is really the right concept to think of. The concept of a function as an abstract machine that turns an input into an output is probably more useful here, but the abstract mathematical definition in terms of ordered pairs of set elements is probably also alright. –jacobolus (t) 01:23, 11 November 2023 (UTC)[reply]

Ok, to show a little more of what I'm getting at, consider the statements:

"think of a bijection as an invertible function"

"can also be thought of as a pairing between two sets"

I submit that the second statement is a lot more accessible than the first. Disclosure: I never heard of bijection until today, hopping among various math articles while reading the discussion above. I'm guessing that "mathematicians most commonly think of a bijection as an invertible function", but most people, like me, don't know the term at all. DonFB (talk) 02:22, 11 November 2023 (UTC)[reply]

I took a crack at rewriting the first couple of paragraphs of bijection. Does that help? –jacobolus (t) 23:45, 10 November 2023 (UTC)[reply]

With all due respect, no. I'm sure it's accurate, but I believe my first sentence version is immediately understandable to anyone in comparison to the current version, or your offer. With the caveat, of course, that what I wrote is accurate in the broadest sense. But really, I'm not here to debate particular articles; I'm only trying to show how I think the beginning of articles on abstruse topics can be written in plain English. DonFB (talk) 00:22, 11 November 2023 (UTC)[reply]

This is still not really "plain English" though. The concept of a set in this context is at least a bit tricky (we're talking about potentially infinite collections, e.g. of numbers, lists of numbers, functions, geometric transformations, ...), and a binary relation is concept that is not usually encountered by students until college (and turning it into the undefined term "relation" just makes it ambiguous rather than particularly more accessible IMO). It probably would be worth trying a few versions on e.g. high school students to see which ones were easiest for them to decipher while giving a correct understanding of the concept.

You could well be right that first describing a bijection as a pairing between set elements would be gentler than describing it as a function. –jacobolus (t) 01:38, 11 November 2023 (UTC)[reply]

@DonFB okay, I tried again. How about this version? –jacobolus (t) 02:26, 11 November 2023 (UTC)[reply]

That (almost) works for me. I would drop "binary pairing" from the first sentence and introduce it later. No need for a speed bump in the first sentence; it doesn't help. Also, some garble seems to exist: "element of each set is paired..." Not "the either set." DonFB (talk) 02:41, 11 November 2023 (UTC)[reply]

I'm also a little confused by the text example (1,2 etc); does not look consistent with the graphical example. DonFB (talk) 02:47, 11 November 2023 (UTC)[reply]

If considered as a pairing per se, the graphical example could be likewise written as {(1, D), (2, B), (3, C), (4, A)}. Thinking of it as a function, you might instead write {1 ↦ D, 2 ↦ B, 3 ↦ C, 4 ↦ A} or {f(1) = D, f(2) = B, f(3) = C, f(4) = A}. –jacobolus (t) 03:26, 11 November 2023 (UTC)[reply]

I put in a ↦ notation instead, but looking at it I'm not sure it's any better. One issue with it is that general binary relations don't necessarily have a single output for any input. That is, this notation works for bijections but not "pairings" in general. It's also not as common, which might leave some people confused. –jacobolus (t) 03:50, 11 November 2023 (UTC)[reply]

Articles covering math topics up through early undergrad level

@XOR'easter I wonder if there's a good venue for coordinating on "top-notch coverage of math topics taught in high school and the first couple years of college". I'd also like to see this happen, but it could be nice to organize some group effort rather than individual contributors trying to take each one as a one-off, perhaps something along the lines of the old "collaboration of the month" or the like. I think even aiming for green checks or gold stars shouldn't be the top priority, but just getting all of the most common pages to a consistently decent "B class" kind of standard, with special focus on the lead sections. I feel like the math Wikiproject isn't quite the venue (or rather, it plausibly could be, but I feel people might get annoyed if the traffic there increased a lot discussing a wide set of such articles).

I personally have been trying to put in the research to further improving (or write where they don't yet exist) the articles about spherical geometry and its history, spherical trigonometry, the most common map projections, and more generally some other kinds of geometry topics that could (but typically don't) appear in school. But while these topics have plenty of applications to industry (GIS, attitude control, kinematics, etc.) they're at least somewhat obscure compared to stuff right in the main line of the school curriculum. I wouldn't mind spending some time looking up sources, drawing or redrawing diagrams, etc. for high-value articles. I wonder if there's a good way to e.g. get a page view summary of all articles currently in the math wikiproject and pick out a few that are important and high traffic but currently mediocre. –jacobolus (t) 05:21, 11 November 2023 (UTC)[reply]

To get a sense of where things stand, I grabbed the list of level-3 "Vital articles" in mathematics and looked up their 30-day page view counts. The winner? 0, with 241,262. XOR'easter (talk) 21:01, 11 November 2023 (UTC)[reply]

Thanks. I have a medium-term interest in cleaning up in particular polygon, triangle, line (geometry), circle, angle, area, trigonometry, and complex number, among various other articles not on your list (most especially sphere, spherical geometry, etc.). If anyone wants to collaborate on any of these, I've been chipping away here and there, figuring out how to draw decent diagrams, gathering sources about several, doing my basic reading, .... –jacobolus (t) 01:36, 12 November 2023 (UTC)[reply]

I'm introducing a new heading just because the previous discussions are now a bit...unwieldy. So far, I've avoided getting into the weeds of Policy/Guideline, because it's tedious to look up that stuff, and because I think we're all pretty familiar with the rules. However, I feel compelled to quote from policy now, because it directly contradicts some of what's been stated above, and what one can find in many articles. Before quoting, I'll add that I'm aware that Policy is, in theory, not top-down, but bottom up. Somewhere in the site (again, too tedious to find it, but I've seen it), one can find a statement that the reason rules are not hard and fast ("Wikipedia has no firm rules"--Five Pillars) is because they develop from common practices; that's why they can change. So maybe the practices we see in many articles on obscure subject matter (and many not so obscure) are pressuring the community to reconsider the policy text I'm about to quote. But my view is that the policy should be followed:

From: What Wikipedia is Not:

Section heading:

"Wikipedia is not a manual, guidebook, textbook, or scientific journal"

"...Wikipedia articles should not read like:"

6) "Textbooks and annotated texts: The purpose of Wikipedia is to summarize accepted knowledge, not to teach subject matter. Articles should not read like textbooks, with leading questions and systematic problem solutions as examples. These belong on our sister projects, such as Wikibooks, Wikisource, and Wikiversity. However, examples intended to inform rather than to instruct, may be appropriate for inclusion in Wikipedia articles."

(Take special note of the following--Don):

7) "Scientific journals: A Wikipedia article should not be presented on the assumption that the reader is well-versed in the topic's field...Introductory language in the lead (and sometimes the initial sections) of the article should be written in plain terms and concepts that can be understood by any literate reader of Wikipedia without any knowledge in the given field before advancing to more detailed explanations of the topic...."

(Italics in the original in #6)

DonFB (talk) 06:29, 15 November 2023 (UTC)[reply]

The complaints above are not about encyclopedia article that seem like textbook chapters (they don't generally have "leading questions and systematic problem solutions", nor are they especially like textbooks lacking those features). They also are not about articles reading like research journal papers per se. While these are somewhat related issues, I don't think these policies are closely applicable to the concerns under discussion.–jacobolus (t) 07:30, 15 November 2023 (UTC)[reply]

The incomprehensible hyper-specialized math articles that people get so upset about are bad in part because they perfectly comply with point 6: they're not written like textbooks. Those "leading questions" we're not supposed to include can help lead a reader into a subject, providing motivation while giving a notion of the topic's contours before delving into the nitty-gritty. Those "systematic problem solutions" we're not supposed to include can make abstractions more concrete and give readers a means to double-check that they're following what's going on. Now, there are reasons why we shouldn't include those things. I tried to explain some of those reasons here. But if you really don't like the state of Wikipedia's technical coverage, maybe you shouldn't like that policy! XOR'easter (talk) 07:34, 15 November 2023 (UTC)[reply]

What XOReaster says. 67.198.37.16 (talk)

I read through the argument above, about the lede to the article on bijections, and I'd like to point out that bijections are not relations, and relations are not functions. The issue here is that the word "relation" has a specific technical meaning in mathematics, and using it in the wrong context results in non-sense. This is one (of many) reasons why writing adequate accessible intros is difficult.

There's also a converse issue: After many decades of doing physics and math, I understand certain topics well enough to write broad, accurate non-technical intros. I've done so on a number of occasions, and I'm proud of that work. I've sometimes had this work blanked, because (I kid you not) "it's not rigorous enough". From what I can tell, these were college students cramming for a mid-term exam; they'd replace the text with incorrect formulas copied from chapter one of some "intro to..." textbook. It's hard to please everybody all the time, especially when the topic is popular. 67.198.37.16 (talk) 07:40, 15 November 2023 (UTC)[reply]

bijections are not relations, and relations are not functions – a function is a special case of a binary relation, as those concepts are commonly defined. You can find a wide variety of reliable sources (e.g. textbook chapters from reputable mathematicians and publishers) making this claim; I don't think it's especially controversial. –jacobolus (t) 08:10, 15 November 2023 (UTC)[reply]

Sigh. Any and every book on model theory goes to great pains to distinguish relations from functions. Topology and set theory and order theory are conventionally defined in terms of relations, not functions. Functions really come into thier own only in analysis and calculus. But whatever, this is not the place to argue. I agree w/98% of your other discussion comments and general outlook, and I have no particular desire to litigate the concept of a relation (mathematics) vis-a-vis a function (mathematics). I only wanted to point out that laymans terms and formal terms often clash, and that this is a problem for writing article intros and surveys. 67.198.37.16 (talk) 09:30, 15 November 2023 (UTC)[reply]

Would you mind starting a conversation at talk:bijection? I'm sure you are more of an expert about this than I am. –jacobolus (t) 15:04, 15 November 2023 (UTC)[reply]

Point 7 is vague, appealing to the subjective standard of "plain terms and concepts" and the illusory general "literate reader". Actually spelling out what vague, aspirational language in a policy means requires a guideline to supplement it, which is just what this guideline is. (If you really want to wiki-lawyer, the policy only talks about "Introductory language", so anyone could riposte that a sufficiently technical article contains no "introductory language" and thus that item is beside the point.) I would actually say that point 7 is poorly constructed. The real reason that Wikipedia doesn't function as a scientific journal is WP:NOR, not any issue of writing style. Honestly, it almost seems like it was written by somebody who didn't quite know what scientific journals contain. XOR'easter (talk) 07:53, 15 November 2023 (UTC)[reply]

From what I've seen on the site, I think it's true that technical articles don't give "leading questions and systematic problem solutions as examples". What I've noted, however, from the discussions above is that some editors seem preoccupied with matching their content with the education level of the presumed readers. That's a direct violation of #7 in the policy text I quoted.

I'm surprised to hear the opinion that "Point 7 is vague". Rather, it is explicit. Nor is there any reason to describe the idea of "literate reader" as "illusory". Think of any person with more than a primary school education. Any such person can readily see the difference between "plain terms and concepts" and the scientific and technical jargon that appears in many articles. The hypothetical idea that a "sufficiently technical article contains no 'introductory language'" seems like a suggestion to simply overthrow the standard practice of introducing a subject. So far, the responses seem to wish to invalidate the policy, which perhaps is understandable since it directly contradicts some of the assertions in the discussions above. DonFB (talk) 08:42, 15 November 2023 (UTC)[reply]

Can you provide an explicit example of an article that you think violates either 6 or 7? (I'm finding this discussion nebulous; there is some general unease, but I've never actually seen an article that violated these. Well, I have, but they were always proded and speedy-deleted.) 67.198.37.16 (talk) 09:38, 15 November 2023 (UTC)[reply]

Thanks DonFB. I don't see that Point 7 is vague at all. The introduction should be written in "plain terms and concepts" understandable by "any literate reader of Wikipedia without any knowledge in the given field" before "advancing to more detailed explanations". This policy removes any remaining rationalization that it's okay to write the introduction just for specialists. The last clause allows the introduction to include technical definitions and jargon, but they must be accompanied by plain language explanations.--Chetvorno^TALK 13:31, 15 November 2023 (UTC)[reply]

67.198.37.16 Although there are examples of incomprehensible WP introductions throughout technical fields, I will confine attention to mathematics. The problem here is not finding examples of articles that violate #7, but finding articles that don't. Cohomotopy set and Diffeomorphism have as introduction a single lead sentence of solid jargon. And of course there is Galois cohomology. None of these have a single sentence in the introduction that attempts to explain the jargon terms in plain language.

Compare these to the introduction to Homeomorphism. The first paragraph is a jargon-ridden technical introduction that will only be understandable to mathematicians. That's okay, because the 2nd paragraph is an attempt to explain homeomorphism in plain language for general readers. It is illustrated by a GIF that shows the famous coffee-cup-transforming-into-a-donut, as an example of homeomorphism. Any of these introductions could be made more compliant with #7 by including general-interest information like who invented the concept, when, why it is important, any important mathematical or physics problems it is used in, and any practical uses. Context is an important part of understanding. --Chetvorno^TALK 13:31, 15 November 2023 (UTC)[reply]

Can you explain what is wrong with any of those ledes? They certainly are not "jargon-filled". They are written in plain, simple terms, using clear and direct language. All four are eminently readable and well-structured articles. My knee-jerk reaction is they're "C-class" or "B-class". Pretty much anyone reading these will sail right on through; none of this is hard. Take Cohomotopy set for example. It's almost exemplary: the overview introduces the basic notation, chapter one type stuff that anyone/everyone would normally know. It's standard algebraic topology. What's the first concept you're tripping over? Why are we having this discussion here, instead of the article page? 67.198.37.16 (talk) 17:31, 15 November 2023 (UTC)[reply]

I looked more carefully at Cohomotopy set. It's a start-class article. The properties are relatively shallow factoids; there's nothing deep here. I suppose the article could be expanded; I'm not sure how. It's providing a basic background definition; that definition would be used in other articles, where assorted theorems and deep results are presented and explained. Think of it as a handy-dandy cheat-sheet one might keep at one's elbow. Whether or not this article should be gummed up with more detailed explanations is a hard call; depends on how it would be written. 67.198.37.16 (talk) 18:08, 15 November 2023 (UTC)[reply]

Do you think "any literate reader of Wikipedia without any knowledge in mathematics" knows what a contravariant functor is? How about a pointed topological space? --Chetvorno^TALK 18:42, 15 November 2023 (UTC)[reply]

Do you think any literate reader of russian without any knowledge of Lithuanian can read Lithuanian Wikipedia? 67.198.37.16 (talk) 19:28, 15 November 2023 (UTC)[reply]

So you're acknowledging that the articles read like a foreign language to English speakers, correct? DonFB (talk) 19:50, 15 November 2023 (UTC)[reply]

Right. And of course the policy doesn't require the English Wikipedia to be understandable in other languages. --Chetvorno^TALK 20:21, 15 November 2023 (UTC)[reply]

Nor is there any reason to describe the idea of "literate reader" as "illusory". Think of any person with more than a primary school education. You're replacing a vague standard with a different, much more specific one. That doesn't make the actual language of the policy any less vague. XOR'easter (talk) 20:25, 15 November 2023 (UTC)[reply]

If the policy keeps us from covering technical material, then it's the policy that must change. The literal wording of NOTPAPERS has mostly been politely ignored because it's just not workable; the only way of complying with it would be to remove tons of valuable content from WP. I hope and trust that no one here actually wants that. --Trovatore (talk) 19:34, 15 November 2023 (UTC)[reply]

As I said opening this section, "maybe the practices we see in many articles on obscure subject matter (and many not so obscure) are pressuring the community to reconsider the policy text I'm about to quote." But for any topic that can reasonably be considered non-obscure, the current policy prescription should apply. Otherwise, why bother? And If articles on difficult concepts are written only for those who already understand them, who is the encyclopedia serving? DonFB (talk) 20:07, 15 November 2023 (UTC)[reply]

The policy you are quoting is "Wikipedia is an encyclopedic reference, not an instruction manual, guidebook, or textbook. Wikipedia articles should not read like [...] Scientific journals"

That is not the question we are dealing with here, which is rather: what compromises should we make so that encyclopedia articles serve all of their plausible audiences as best as is practical. Notice that outside Wikipedia there are encyclopedia articles of a wide variety of levels from the most basic aimed at young children with lots of glossy photos, up through specialist encyclopedias written at a very advanced level. The level of difficulty involved is distinct from whether they read like research papers.

Unfortunately the textual elaboration of this policy bullet point is itself poorly worded and somewhat off topic, probably written by someone who wasn't thinking clearly about the fundamental differences between an encyclopedia article vs. a paper. It's confusing to editors reading it for guidance and should probably be changed. –jacobolus (t) 20:24, 15 November 2023 (UTC)[reply]

Regarding bullet #7: I agree that the comparison to science journal is not too apt. However, the longstanding policy text which follows is explicit in prescribing plain language for ledes. DonFB (talk) 21:03, 15 November 2023 (UTC)[reply]

The policy text following there is off topic for the page "what wikipedia is not", and should be clarified and moved somewhere else. –jacobolus (t) 22:16, 15 November 2023 (UTC)[reply]

There are different threads here, so I'll bullet my responses just to keep them clear (not meant to be aggressive).

• "[F]or any topic that can reasonably be considered non-obscure" — that presumably excludes a lot of the articles we've been talking about, then. We're no longer talking about cohomotopy set or Galois cohomology, right? But we could still be talking about bijection. In that case, fair enough — I agree that there's a considerable number of math articles that could have their intros formulated in understandable language, but don't, and that this is a problem that should be addressed.
• On the "written for those who already understand them" point — this is a recurring complaint, but in fact most of these articles are not written only for those who already understand them. They're written also for those who already understand them, as makes sense for a reference work; this is where you come to look stuff up that maybe you already knew but want to make sure you get it right. But they're also written for those who don't yet understand them but have the background necessary to understand them. Or at least a lot of them are; this is hard to know when reading articles for which we don't have the necessary background (and that's all of us, at least sometimes). --Trovatore (talk) 20:26, 15 November 2023 (UTC)[reply]

Even a start-class article on a bizarre math topic that effectively presumes the reader already learned the subject in school can still benefit somebody. For example, they not infrequently mention where the idea was first published or point to a textbook that covers it. These kinds of details do actually benefit people who already have a grasp on the idea but need an extra datum. Should those articles benefit more people than they currently do? Sure. But the value they already provide is more than nothing. XOR'easter (talk) 20:33, 15 November 2023 (UTC)[reply]

Yes, the articles have some value; I don't disagree, nor do I suggest elimination. But editors should make the effort--in accordance with policy--to introduce the bizarre in normal language, rather than simply falling back on the jargon and equations they're already intimate with, but read like a foreign language to others. DonFB (talk) 21:19, 15 November 2023 (UTC)[reply]

Who decides what is "normal language"? You? Me? And what happens when there is no "normal language"? Do we violate WP:NOR to invent some?

The snippet of policy you point to provides no useful guidance for resolving any situation that is actually difficult. It is confused about what scientific journals are and why Wikipedia can't be one. It's worse than every other item in the list at connecting the bold part to the supposed rationale. Including it in WP:NOT makes suboptimal material — like niche mathematics articles that are too densely written — sound as bad as fundamentally unencyclopedic material, like paid advertising and political soapboxing. As written, it's not fit for purpose. XOR'easter (talk) 21:35, 15 November 2023 (UTC)[reply]

Like the rest of Wikipedia, the answer would seem to be "consensus", which is what we're attempting to hash out right now. It seems pretty impossible to look to any standard other than "trying your best as a speaker of language yourself", I don't see how this is any different from any other matter of writing or style guideline onsite. Remsense聊 21:41, 15 November 2023 (UTC)[reply]

It's rather worse than most things that style guides try to provide answers for: at face value, the call to do it everywhere would violate WP:NOR, a bedrock principle. XOR'easter (talk) 21:55, 15 November 2023 (UTC)[reply]

Do what everywhere? Explain unfamiliar concepts using non-specialist language? Does not strike me as an insurmountable challenge. DonFB (talk) 22:05, 15 November 2023 (UTC)[reply]

Sometimes it isn't. Sometimes it is. XOR'easter (talk) 22:08, 15 November 2023 (UTC)[reply]

Well, that seems a little defeatist. Policy tells us to use non-specialist language in intros. That's what we should consistently endeavor to do. DonFB (talk) 22:15, 15 November 2023 (UTC)[reply]

It's not defeatist; it's just reality. It's not about trying harder. It's not about being better writers. Those things can help, but they can never help enough to be able to say anything worth saying about some of these things in truly non-specialist language. Less specialist language yes; non-specialist, no.

But there are certainly cases where we can get closer. Someone called out diffeomorphism — that one actually is something you could say something intuitive about in the first sentence or two, that most people could get some idea about. It wouldn't be enough to fill up the whole intro, though — you are going to have to use specialist language before the TOC. --Trovatore (talk) 22:21, 15 November 2023 (UTC)[reply]

Consensus decides. Let's differentiate: there is a "normal" language in math; in astronomy; in biology; nuclear physics; and on. The disciplines use specialized terms normal to them, but not to everyone. The normal I refer to is the language people use every day apart from specialized or arcane fields of study. This page is filled with such normal language. Use it as a guide. DonFB (talk) 21:55, 15 November 2023 (UTC)[reply]

The difference between everyday language and specialized language is not so clear-cut as that. One spills over into the other; the same word can have connotations in one that are irrelevant to or in contradiction with its use in the other. What kind of word is energy, technical or everyday? How about infinity, heritability, linear, exponential, and spike protein? XOR'easter (talk) 22:04, 15 November 2023 (UTC)[reply]

Sure, that's not news. All the more reason for editors to carefully weigh their words, while making the effort not to default to less accessible lingo in introductions. DonFB (talk) 22:24, 15 November 2023 (UTC)[reply]

It's not possible to use only lay conversational English in a specialized math article. These are extremely abstract and esoteric objects which are many steps removed anything encountered in daily life. To fully unpack each word all the way to concepts recognized by a high school student would in some cases require dozens to hundreds of pages of elaboration, and wouldn't really be that helpful anyway.

What we can try to do is make the article as accessible as practical by e.g. providing examples, pictures, historical context, and analogies, and by trying to expand obscure jargon words in terms of more commonly known jargon words which are couple levels below the level of the subject. –jacobolus (t) 22:29, 15 November 2023 (UTC)[reply]

Thanks to XOReaster for cleaning up my mess — I should work through and understand that merge-conflict tool sometime. When it pops up I don't know exactly what to do. --Trovatore (talk) 20:32, 15 November 2023 (UTC) [reply]

Can I say that having a conversation on this point is always going to be incomplete without concrete examples of what we're talking about? As with the rest of wiki norms, these things are implemented on a per-article basis, and if we're going to craft a better, more helpful policy, we should be doing so based on specific examples so that we don't just argue over the connotations of various adjectives. Are there any articles people would cite as an example either for or against their particular view of what the policy should reflect? Remsense聊 21:57, 15 November 2023 (UTC)[reply]

It appears my analogy was mis-understood. All I'm saying is that perhaps the ledes of the articles in Lithuanian Wikipedia should be re-written in russian. This would allow ordinary people, without specialist training in Lithuanian, to at least be able to understand what the article is about. This is a win-win situation. Pretty much everyone in the world can read russian; these articles would become accessible to a broad, generalist audience. Wikipedia would be vastly improved by this one simple step. 67.198.37.16 (talk) 20:48, 15 November 2023 (UTC)[reply]

Let's try to leave out the sarcasm. I think we have similar goals. We all want articles to be as understandable as possible, and I hope no one wants to remove technical content. --Trovatore (talk) 20:51, 15 November 2023 (UTC)[reply]

This is not sarcasm. The problem with the russian proposal is that it is chauvinistic: russians believe that the world should adapt to them, instead of russia adapting to the world. This is the chauvinism I see above. Instead of learning enough math to understand some article, the proposal is that the lede should be re-written so that non-mathematicians can understand it. The chauvinistic attitude that "literate readers" should understand them. The reality is, of course, that the readers are illiterate; they have not taken the effort to understand the language of mathematics. And yes, mathematics is a foreign language. There's no way to make it not-foreign. 67.198.37.16 (talk) 21:20, 15 November 2023 (UTC)[reply]

The argument you're making is simply not an argument at all. Instead, it is a rhetorical sleight of hand that breaks down when you try to apply rigor to it. Granted—in a sense, mathematics is its own language. It has its own writing system of vocabulary and syntax. Certain mathematical papers may even manage to be written entirely in it. However, you are plainly making a false equivalence: it is asinine to deliberately overlook that Wikipedia is still being written in a human language, in the context of an existing body of literature in said human language, in this case English.

Please stick to making a case instead that it is simply too difficult to make certain technical concepts readable enough for a given standard, because this is an infinitely more compelling argument to make. Remsense聊 21:31, 15 November 2023 (UTC)[reply]

The above conversation already singled out cohomotopy set and Galois cohomology. I believe that they are reasonably well-written and clear, and quite understandable to anyone who has gone through a college course on algebraic topology and Galois theory. I don't see any particular way of making them comprehensible to anyone with only high-school math, and yet, it seems like that is the standard being asked for. 67.198.37.16 (talk) 22:33, 15 November 2023 (UTC)[reply]

It is not "chauvinist" (aggressively nationalistic) to wish that articles about math topics which are relevant to e.g. computer science or physics or economics should be at least basically accessible to professional computer scientists or physicists or economists, or that articles about math topics encountered in grad school should have at least their first few sentences comprehensible to people with undergraduate math degrees, or that articles about basic topics used throughout science and engineering should if possible have a few sentences accessible to laypeople or first year college students. Examples about algebraic topology etc. are pretty esoteric, so as a concrete lower level example, it would be great if Eigenvalues and eigenvectors did not require readers to have already gone through two semesters of undergraduate linear algebra coursework to make sense of the first few sections. –jacobolus (t) 22:43, 15 November 2023 (UTC)[reply]

Algebraic topology is not a course that you will magically evade on your way to an undergrad math degree, so its hardly "esoteric". As we are all in general agreement that articles should be understandable, then what are we arguing about? 67.198.37.16 (talk) 23:14, 15 November 2023 (UTC)[reply]

What I mean is, a bunch of the people discussing here are 3+ years of full-time pure math study away from making sense of Galois cohomology, which makes it hard to use as an example that we can try to get on the same page about. –jacobolus (t) 23:20, 15 November 2023 (UTC)[reply]

I think it would be great to discuss concretely what we might do with the eigenvector article, in the context of this discussion. Maybe start a new section for that? The topic is accessible enough that non-mathematicians can potentially contribute, but technical enough to come up against some of the issues we're dealing with. --Trovatore (talk) 23:27, 15 November 2023 (UTC)[reply]

I'm game. DonFB (talk) 23:51, 15 November 2023 (UTC)[reply]

Great idea! I'll look at it. Should we start a new thread? --Chetvorno^TALK 23:55, 15 November 2023 (UTC)[reply]

Our article about Eigenvectors and eigenvalues (henceforth E&E) has some great material included, and does some things well, but I think its structure and language for the first few sections still make it less accessible than it could be for an audience of laypeople or early students who might plausibly want to know what the word nominally means if they encounter it somewhere, which many people are likely to do as this is a very common tool used throughout science and engineering, one of the most important ideas in linear algebra.

There's been some significant effort put into making accessible material in E&E § Overview with some pictures and relatively informal text. However, the lead section and the first following section, E&E § Definition are full of jargon, and even the overview section doesn't in my opinion sufficiently prioritize accessibility.

I think this could be yet better if we made the lead section itself less technical, moved material from the overview section into the lead / first section, deferred a formal definition, and tried to unpack any jargon showing up near the top. In my opinion, this article should try to provide an introductory explanation in the first 2–3 sections accessible to someone who does not know the meanings of the jargon words field, vector, linear transformation, matrix, scalar multiple, principal axis, rigid body, shear mapping, differential operator, matrix decomposition, diagonalize, coordinate vector, differential operator, quadratic form, etc.

I think we could fit a short inline explanation of linear transformations, vectors, matrices, lead with some pictures/examples, and then give some clear idea near the top of why this subject is important. We of course also want undergraduate linear algebra students (and more advanced math/technical students, math historians, etc.) to get the full story they are looking for. But this doesn't have to come at the expense of a very quick and basic accessible explanation. –jacobolus (t) 23:52, 15 November 2023 (UTC)[reply]

Mona Lisa illustration and caption are superb; highly informative and helpful. How about putting that in the intro section. It seems like it would be universally understandable. DonFB (talk) 00:07, 16 November 2023 (UTC)[reply]

Agree. I'd like to mention up front that the intro needs context: history of the idea, etymology, importance, mathematicians who developed it, physical problems it applies to, modern uses. For the many readers who won't be able to comprehend even a simplified explanation, this is what "understanding" the concept consists of. --Chetvorno^TALK 00:19, 16 November 2023 (UTC)[reply]

I like the concept of the Mona Lisa picture, but I think it works better with a less busy picture, on which the arrows would show up more clearly. –jacobolus (t) 00:54, 16 November 2023 (UTC)[reply]

While we're talking about pictures and 2d transformations, the article has a section near the bottom about E&E § Eigenvalues of geometric transformations; I think it would probably be helpful to move this to the first third of the article instead of relegating it to "applications" near the bottom, and maybe use paragraphs rather than a table, ideally also showing a picture of what each transformation does to the same base image. This material is more accessible than parts about characteristic polynomials and so on. –jacobolus (t) 01:12, 16 November 2023 (UTC)[reply]

Here is a kind of schematic that simplifies the syntax, though does not immediately simplify the terminology:

"In linear algebra, an eigenvector (/ˈaɪɡənˌvɛktər/) is a nonzero vector that changes at most by a constant factor when a linear transformation is applied to it."

In short, is it necessary to repeat "linear transformation" twice? Paraphrasing, here's what it seems to say now: "in a transformation this is what happens when you do the transformation". Cutting redundancy can also cut instances of jargon. DonFB (talk) 01:07, 16 November 2023 (UTC)[reply]

I think it is necessary to repeat "linear transformation". An eigenvector is an eigenvector of a transformation, not by itself. XOR'easter (talk) 03:46, 16 November 2023 (UTC)[reply]

Ok, not being a wise-guy here, just want to nail this down. In the Mona Lisa (eg), before it is transformed, are any of the vectors (if they exist) eigenvectors? The article says an eigenvector is a "characteristic vector" of a transformation. That seems to say it is not an eigenvector until the transformation happens, and then the eigenvector becomes "characteristic" of what a transformed vector looks like. Or, to put it another way, the sentence appears to say that a vector (generic) is not an eigenvector until the space is transformed. Is that correct? Stripping away much of the surrounding text, the sentence literally says, "an eigenvector is a vector that changes when that transformation is applied to it." Reduced further, we see "a vector that changes". DonFB (talk) 04:23, 16 November 2023 (UTC)[reply]

Once you have a mathematical description of the transformation, you can find what its eigenvectors are. Asking whether a vector is an eigenvector before the transformation is applied is kind of a meaningless question.

In trying to strip away the surrounding text, you're actually changing the meaning. An eigenvector of a transformation is a vector that changes in at most in a specific, very limited way when that transformation is applied to it. The italicized part is an essential part of the concept, not an optional one. XOR'easter (talk) 04:34, 16 November 2023 (UTC)[reply]

Yes, I acknowledge that stripping away text can change meaning. So instead of asking if a vector is an eigenvector before transformation, is it correct to say: If there is no transformation, then an eigenvector cannot exist in the specified space (or object, or whatever is the correct term). Is it therefore correct to say an eigenvector can only appear after a transformation.? And going out further on a limb, can we say, in a generalized way, "A vector can become an eigenvector when a linear transformation is applied to it"? I do not mean to imply that all vectors in a space will become eigen, as Mona Lisa shows. DonFB (talk) 04:50, 16 November 2023 (UTC)[reply]

The concept of "eigenvector" is only meaningful with respect to a linear transformation (specifically, a linear transformation of some vector space to itself), for example the vector space of displacements in the 2-dimensional Euclidean plane can be transformed by e.g. stretching every displacement two times taller vertically and then skewing them upward proportionally to the horizontal displacement.

It seems weird to me to say that a vector "becomes" an eigenvector. Like, if I give you the nickname "D" I wouldn't be likely to say "DonFB became a nicknameholder"; it't not wrong exactly, but seems strange.

I would instead say that if you start with a linear transformation, you can find out which vectors didn't change direction under that transformation, but were only scaled. –jacobolus (t) 05:07, 16 November 2023 (UTC)[reply]

I was modifying my text, but got a conflict. My change says, "A vector can be recognized as an eigenvector when (or after?) a linear transformation is applied to it. It is a vector which did not change direction, but only changed its size (magnitude?)" [seems incorrect] Also incorporates Jacobolus comment. As may be seen, I'm attempting to distill the concept into unadorned terms, for use at the very beginning of the article. I recognize that "linear transformation" should have some explanation very close by. DonFB (talk) 05:19, 16 November 2023 (UTC)[reply]

I find the current first sentence's "changes at most by a constant factor" to be confusing and a bit misleading.

It's probably clearer to say that an eigenvector is a vector to which applying the transformation is equivalent to scaling by a constant factor. I think it's likely clearer in this case to first start with a (moderately) technical definition, and then add 1–2 paragraphs afterward explicitly defining the terms involved such as linear transformation, vector, and scaling. –jacobolus (t) 06:18, 16 November 2023 (UTC)[reply]

Not to put too fine a point on it, but that whole first sentence is a train wreck.

Try this:

"An eigenvector is a vector that doesn't change direction when a linear transformation is applied."

Is it wrong, or is it simply bare bones, with more details to be added in the ensuing sentences, using equally plain language?

DonFB (talk) 07:34, 16 November 2023 (UTC)[reply]

Keep in mind, in this part of the article (the beginning), we are only introducing the concept to a broad public audience, not emulating the type of textbook language a mathematics student would expect to see. DonFB (talk) 13:02, 16 November 2023 (UTC)[reply]

It's not wrong (and you can find sources saying more or less this precise sentence), and this would be a better "short description" than the current "Vectors that map to their scalar multiples", but (a) it still presumes someone knows what "vector" and "linear transformation" mean, (b) it presumes someone can guess that changing magnitude or flipping backwards is allowed and knows knows what it means for a vector to change directions, and (c) I think some people may find "direction" to be insufficient as a description in some more abstract vector spaces (I think that's fine).

You might want to start be looking at some non-Wikipedia sources. For example, Grant Sanderson's video about the subject. There's some good meta-discussion in Wawro; Watson; Zandieh (2019), "Student understanding of linear combinations of eigenvectors", ZDM, 51: 1111–1123, which e.g. notes that "Hillel (2000) found that instructors often move between geometric, algebraic, and abstract modes of description without explicitly alerting students; although the various ways to think about and symbolize linear algebra ideas are second nature to experts, they often are not within the cognitive reach of students." –jacobolus (t) 14:15, 16 November 2023 (UTC)[reply]

It does presume foreknowledge of vector and linear transformation. I don't think, however, that articles of this type begin with sub-definitions. So, for the ensuing couple of sentences, try this:

"A vector is a way of representing quantities that have both magnitude and direction. A linear transformation is a way of changing or stretching space, possibly rotating or even flipping it, while maintaining continuous structure that does not have holes or gaps."

I would also offer the following as a modification of the first sentence I proposed above:

"An eigenvector is a vector that does not change direction when a linear transformation is applied, but its scale may change, either lengthening or shortening." DonFB (talk) 20:08, 16 November 2023 (UTC)[reply]

This still has the problem that an eigenvector is an eigenvector of a specific linear transformation, not a vector that is somehow immune to change in general. XOR'easter (talk) 20:39, 16 November 2023 (UTC)[reply]

I wonder if putting it in the past tense would address the issue:

"An eigenvector is a vector that did not change direction when a linear transformation was applied, but its scale may have changed, either lengthening or shortening." DonFB (talk) 20:54, 16 November 2023 (UTC)[reply]

That's needlessly complicated, introducing unexplained tense shifts back and forth during the opening sentences. And, if anything, it gets in the way of understanding that the eigenvectors and eigenvalues are determined once the transformation is written down. Really, whatever problems the intro has, I don't think repeating "linear transformation" is among them. XOR'easter (talk) 21:02, 16 November 2023 (UTC)[reply]

I think it should say something along the lines of "An eigenvector is a vector whose direction is fixed by a linear transformation" or "The eigenvectors of a linear transformation are those vectors on which the transformation acts by scaling". Then in a following paragraph, the terms vector, linear transformation, scaling, etc. can be (perhaps informally) defined. –jacobolus (t) 21:15, 16 November 2023 (UTC)[reply]

I rather like the second option. XOR'easter (talk) 21:19, 16 November 2023 (UTC)[reply]

"the transformation acts by scaling" is not clear. It seems appropriate to explicity say length (magnitude?) can change to be greater or lesser/longer or shorter DonFB (talk) 21:27, 16 November 2023 (UTC)[reply]

This would be wikilinked to Scaling (geometry) or Scalar multiplication, and unpacked at greater length in the following paragraph. It could be extended by a few words inline, for instance:

The eigenvectors of a linear transformation are those vectors on which the transformation acts by scaling, without change of orientation; the eigenvalues are the corresponding scaling factors.

A vector is a multi-dimensional quantity with an orientation and a magnitude, often pictured as an arrow. Adding two vectors can be understood geometrically by the parallelogram rule: when two oriented sides of a parallelogram are added together, the result is the parallelogram's diagonal. Multiplying a vector by a scalar represents the geometric transformation of uniform scaling, which stetches, squishes, or possibly reverses a vector along its line through the origin but does not rotate it away from that line. A linear transformation is a function between vector spaces which preserves the operations of vector addition and scalar multiplication; geometrically, this means that a linear transformation maps any regular grid of parallelograms in one vector space to a regular grid of parallelograms in the other.

For a linear transformation from a vector space to itself, applying the transformation to one of its eigenvectors scales that vector by the corresponding eigenvalue. Whenever an arbitrary vector can be rewritten as a sum of eigenvectors, the operation of the linear transformation can be likewise written as a weighted sum of the same eigenvectors, using with the eigenvalues as weights. Because scalar multiplication is a simpler operation than general linear transformations, ...

–jacobolus (t) 22:31, 16 November 2023 (UTC)[reply]

I am certain it will be approved by math specialists, but I think it is much too complicated as an introduction to a lay audience. As I've discussed before, and as multiple rules in Policy and Guideline say, links should not be used in the introduction as a crutch to substitute for an explanation that laypeople can understand. Links in the intro are fine for readily understandable terms where the linked article offers supplementary information, but not because the term itself is obscure or has a highly specialized meaning that is not obvious.

May I be so bold to ask you for this: a list, in plainest English you can manage, of the four or five most important things to understand about eigenvectors.

In the meantime, I offer a new revision of my proposed first sentence:

"In linear algebra, an eigenvector for a linear transformation is a vector that does not change direction when the transformation is applied, but its scale may change, either lengthening or shortening." DonFB (talk) 23:09, 16 November 2023 (UTC)[reply]

Striking my comment above; terms you linked look reasonable; but the overall complexity of the suggested intro remains too great. DonFB (talk) 23:16, 16 November 2023 (UTC)[reply]

The only fundamental thing to understand about eigenvectors and eigenvalues is that they satisfy the equation ${\displaystyle Av=\lambda v,}$ where ${\displaystyle A}$ is a linear transformation, ${\displaystyle v}$ is a vector, and ${\displaystyle \lambda }$ is a scalar. To make sense of that requires explaining what vectors, scalars, and linear transformations are and how they are related to each-other. –jacobolus (t) 23:55, 16 November 2023 (UTC)[reply]

much too complicated as an introduction to a lay audience – quite the opposite, I would expect a very technical audience to be annoyed for wasting too much space defining jargon and unpacking the prerequisite ideas. –jacobolus (t) 23:13, 16 November 2023 (UTC)[reply]

I agree it's quite possible a very technical audience will be annoyed for wasting space. I submit that's not a consideration for writing introductions. The experts can jump to the main body in an instant. The laypeople are the ones who will be annoyed by the unnecessary complexity of the intro, as drafted. DonFB (talk) 23:20, 16 November 2023 (UTC)[reply]

I actually like to read leads of technical topics I'm competent at, to see if they properly summarize the material for a lay audience. It's not an annoyance, especially when the material is done well. If I'm in a hurry for some particular factoid and not interested in the lead, I know how to use the ToC or in-page search to get directly to what I'm looking for.  — SMcCandlish ☏ ¢ 😼  23:32, 16 November 2023 (UTC)[reply]

Sure, if you're knowledgeable in the topic, it can be fun to see if the intro is done well. If you're not well-versed in it, getting through and understanding an unnecessarily techy intro it can be a pain, as innumerable comments over the years have attested. DonFB (talk) 23:49, 16 November 2023 (UTC)[reply]

To clarify, I'm suggesting that a lead that isn't unnecessarily techy is still not an annoyance for readers who could understand a techy version. Some, hunting for something specific, are apt to skip it, but I don't think they'd be annoyed that an accessible intro has been provided, and will understand that it was written by experts not for them.  — SMcCandlish ☏ ¢ 😼  23:54, 16 November 2023 (UTC)[reply]

Agreed. DonFB (talk) 00:08, 17 November 2023 (UTC)[reply]

If the simplification is done well, it's fine. If it isn't, say if it propagates a hoary old confusion by being careless and lazy, it becomes what in technical terms is designated annoying as fuck. XOR'easter (talk) 02:31, 17 November 2023 (UTC)[reply]

What I mean is, my proposal is significantly less technical than anything previously proposed here, and about as non-technical as reasonably possible (to take it further means further unpacking another couple levels of concepts to the length of a book chapter, which obviously doesn't fit in a lead section anywhere). –jacobolus (t) 23:50, 16 November 2023 (UTC)[reply]

Here is text from your draft, slightly edited and with a sentence I added, that I think is helpful and readily grasped by the lay readership:

"A vector is a multi-dimensional quantity with an orientation and a magnitude, often pictured as an arrow. A linear transformation stretches, squishes, or possibly reverses a vector, but does not rotate it. The amount it gets changed is called the eigenvalue." DonFB (talk) 00:02, 17 November 2023 (UTC)[reply]

No, this version is outright incorrect. Linear transformations, as I wrote, are transformations which preserve scalar multiplication and vector addition (or geometrically, fix the origin and transform regular grids of parallelograms into regular grids of parallelograms). It's only the eigenvectors of a linear transformation which are scaled without a change of direction. –jacobolus (t) 01:09, 17 November 2023 (UTC)[reply]

Does this mean a linear transformation is not what "stretches, squishes, or possibly reverses a vector"? DonFB (talk) 01:28, 17 November 2023 (UTC)[reply]

I'm sorry to say this, but your current draft is much less accessible than what exists now in the second paragraph of the intro section. From my perspective, the main problem with the section is the first paragraph, for which I've been attempting to create an improved first sentence.DonFB (talk)

No, a linear transformation only (just) scales the eigenvectors. The only restriction on the transformation of arbitrary other vectors is that the linear transformation preserves scalar multiplication and vector addition. In other words, the transform applied to a scalar multiple of one vector is equal to the scalar multiple of the transformed vector – symbolically ${\displaystyle A(ku)=k\,Au}$ – and the transform applied to the sum of two vectors is equal to the sum of each separately transformed vector – symbolically ${\displaystyle A(u+v)=Au+Av.}$

The part we have so far is clearly insufficient for you as an audience: you don't yet understand the idea. (This is entirely to be expected, and not a personal failing; students usually don't encounter this topic until near the end of a semester-long introductory linear algebra course where they have spent dozens of hours doing practice exercises, and here we have only a few sentences.) This is the problem with taking "accessibility" to just mean "what seems at a glance to make sense"... it's easy to superficially feel like you know what is being said without really following it if you don't have any experience with the prerequisite concepts involved. If we use a word such as "linear transformation" and the audience (in this case you, as a representative uninitiated reader) does not know what that word means, it's not really possible to explain what a linear transformation's eigenvectors are. –jacobolus (t) 02:21, 17 November 2023 (UTC)[reply]

To further your point a little: one can't really get far editing mathematical prose just by pushing the words around. One has to understand the mathematics to make sure that nothing breaks. In this case, we're dealing with a concept that a student might meet in their second year of college or thereabouts. (For example, at MIT they teach this stuff in 18.03, which comes after the two semesters of calculus, 18.01 and 18.02. Caltech does it about the same way.) This means that if we manage to say something meaningful for students in their first year of college math, we're making progress. XOR'easter (talk) 02:53, 17 November 2023 (UTC)[reply]

I agree with the following; it's a fair comment:

one can't really get far editing mathematical prose just by pushing the words around. One has to understand the mathematics to make sure that nothing breaks

Having said that, I offer:

Edited Jacobolus draft:

"The eigenvectors of a linear transformation are those vectors on which the transformation acts by scaling, or sizing, without a change in vector direction. Eigenvalues are the corresponding scaling factors. A vector is a quantity with an orientation and a magnitude, often pictured as an arrow. A vector space is a set whose elements are vectors. A linear transformation maps any regular grid of parallelograms in one vector space to a regular grid of parallelograms in another.

Vectors can be added and multiplied in ways that meet certain requirements. Multiplying a vector represents a geometric transformation which stretches, squishes, or possibly reverses a vector, but does not rotate it. For a linear transformation from a vector space to itself, applying the transformation to one of its eigenvectors scales that vector by the corresponding eigenvalue.

The eigenvectors and eigenvalues of a transformation serve to characterize it, and so they play important roles in all the areas where linear algebra is applied, including geology, navigation, electricity, magnetism, physics, and quantum mechanics." DonFB (talk) 05:32, 17 November 2023 (UTC)[reply]

This proposal suffers a big problem that with three paragraphs it completely fails to summarize the article. (This is a deficiency of the current lede as well, although less so.) Your proposal might be more appropriate for a separate introductory section in the article (e.g., the first few paragraphs of the Overview section). I think this illustrates a big problem with the current guideline: it's not really possible to give a completely non-technical description of everything in the lede, without breaking it. The lede is for summarizing the article that exists, with due weight, not the article we wished existed. Tito Omburo (talk) 10:39, 17 November 2023 (UTC)[reply]

I like your changes, good work! (I'm not a big fan of the phrasing "changes at most by a constant factor" though, which seems confusing.) I agree that we can probably delay further unpacking these concepts until one of the following sections. I'll try to think about how to make some clear illustrative diagrams, with a particular focus on a single lead section diagram. –jacobolus (t) 15:56, 17 November 2023 (UTC)[reply]

(For what it's worth, the proposal above was just for how a lead section could start; I was imagining a few more paragraphs after. But you are probably right that trying to pack enough context into the lead section to make it legible to e.g. high school students in a self contained way is too big a challenge for just the lead section alone.) –jacobolus (t) 16:52, 17 November 2023 (UTC)[reply]

I wonder if "linear transformation whose outputs are fed as inputs to the same inputs (feedback)" can be made any clearer. –jacobolus (t) 17:13, 17 November 2023 (UTC)[reply]

I worked a bit more from Tito Omburo's version. Current state:

In linear algebra, it is often important to know which vectors have their directions unchanged by a linear transformation. An eigenvector (/ˈaɪɡənˌvɛktər/) or characteristic vector is a such a vector. Thus an eigenvector ${\displaystyle \mathbf {v} }$ of a linear transformation ${\displaystyle T}$ is scaled by a constant factor ${\displaystyle \lambda }$ when the linear transformation is applied to it: ${\displaystyle T\mathbf {v} =\lambda \mathbf {v} }$. The corresponding eigenvalue, characteristic value, or characteristic root is the multiplying factor ${\displaystyle \lambda }$.

Geometrically, vectors are multi-dimensional quantities with magnitude and direction, often pictured as arrows. A linear transformation rotates, stretches, or shears the vectors it acts upon. Its eigenvectors are those vectors that are only stretched, with no rotation or shear. The corresponding eigenvalue is the factor by which an eigenvector is stretched or squished. If the eigenvalue is negative, the eigenvector's direction is reversed.

The eigenvectors and eigenvalues of a transformation serve to characterize it, and so they play important roles in all the areas where linear algebra is applied, from geology to quantum mechanics. In particular, it is often the case that a system is represented by a linear transformation whose outputs are fed as inputs to the same inputs (feedback). In such an application, the largest eigenvalue is of particular importance, because it governs the long-term behavior of the system, after many applications of the linear transformation, and the associated eigenvector is the steady state of the system.

This is definitely not entirely lay-accessible, but I think we can further introduce/unpack vectors, vector addition and scalar multiplication, linear transformations, etc. in one of the immediately following sections. It would also probably be good to write some more introductory explanations of the effect of a repeatedly (or continuously) applied transformation – does anyone have recommend sources explaining this clearly? –jacobolus (t) 17:36, 17 November 2023 (UTC)[reply]

I am very satisfied with the recent revisions by multiple editors to the lede section. The effort shows what's possible for an article on a such a specialized topic that not many people in the general public ever heard of. Thanks. DonFB (talk) 18:54, 17 November 2023 (UTC)[reply]

Just even glancing at this, the order of the sections alone would be of great help to improve the article, keeping in mind that our articles in general should move from most general to most specific. Having a definition with a mathematical formula front and center is a problem; instead, putting the Overview first (and eliminating mathematics in that) followed by History then delving into the Definition and other sections would do wonders. Masem (t) 05:12, 16 November 2023 (UTC)[reply]

Perhaps an unpopular opinion, but an article about a mathematics concept should have a definition of the concept very early in the article, and arguably even part of the lede. Otherwise it would be less accessible to many likely readers. And the idea of "eliminating mathematics" from an article about mathematics seems totally misguided. Tito Omburo (talk) 15:37, 16 November 2023 (UTC)[reply]

I've attempted, above, to offer a layman's definition as the first sentence. I don't think anyone is suggesting "eliminating mathematics" from this article, not even me. I do strongly urge, though--in accordance with policy--that the introductory section be written for laymen, and should have very little--if any--math symbology or equations. After the introduction, go hog wild with math; that seems appropriate. DonFB (talk) 20:15, 16 November 2023 (UTC)[reply]

The previous editor suggested "eliminating mathematics". My basic position on the lede is that it should summarize the article, giving due weight as per reliable sources. In a mathematics article, that means that it will have to summarize the mathematical content of the article, not giving undue weight to introductory matters. It should be accessible when possible, but not at the expense of doing what it's supposed to do. A separate introductory section is often appropriate, but lots of introductory stuff often doesn't belong in the lead. Tito Omburo (talk) 00:36, 17 November 2023 (UTC)[reply]

We should continue this conversation on the article talk page, leaving a little note and link here. DonFB (talk) 01:46, 16 November 2023 (UTC)[reply]

The point is not to fix this one article, but to focus the discussion on something concrete. Therefore in my opinion the discussion belongs here. --Trovatore (talk) 03:09, 16 November 2023 (UTC)[reply]

I'm not gonna make a big deal of it, but really, this page is supposed to be talking about the Guideline, not host extended discussions of specific articles. After the initial talk here about Bijection, I continued the interaction with Jacobolus on that talk page. This page could be used to alert editors to new discussions about specific articles, and the discussions could continue at those specific pages. Does not seem quite proper to hijack this page for long talks about individual articles. But I won't say more. DonFB (talk) 03:25, 16 November 2023 (UTC)[reply]

The point is that there are many thousands of articles all of which are equally deficient, broken, lackluster, hard-to-understand and need work in general, to bring them into GA-type status. Everyone understands this. How can policy guide this effort? Early parts of this conversation start with a complaint; paraphrasing: "I don't understand this article, therefore it is a bad article." When this statement was challenged, the reply was "Well, I can prove its a bad article because it violates policy 6 & 7". When this is challenged, an experiment is proposed: "Is it possible, in general, to take some popular article, and bring it into conformance with policy?" As can be seen, this becomes surprisingly difficult. Of course, the conversation can be moved to the article talk page, but that doesn't resolve the earlier policy dispute.

WPMath has something like 20K articles, WPPhysics has a similar number. Throw in WP Astronomy, Chemistry, Engineering, you've got maybe 80K or 120K pages all of which violate policy. How did we end up here? How many math/phys/eng editors are there? Maybe 50K? Maybe 100K? How is it that a hundred thousand editors, all working with the best of intentions, created 100K-odd pages that are unacceptable from a policy standpoint? Maybe, if we can understand how we got to here, we can find a way to do things better. Create a policy that is workable and works to change the attitude of a vast ocean of editors. That's my rah-rah go-get-em speech. Privately, I'm not so sure. We haven't reached agreement as to what the problem is, never mind how to fix it. 67.198.37.16 (talk) 06:23, 16 November 2023 (UTC)[reply]

I don't think these articles "violate policy", and I frankly don't think trying to solve this problem with a top-down "policy" hat on is that useful. What I would say instead is that these articles need love and attention from people who are explicitly trying to make them more accessible. It's hard work and takes prioritization. Each one has to be taken one at a time, takes days (weeks, months) of dedicated effort (finding sources, synthesizing them, writing clear prose, drawing pictures, debating which version is most useful, etc.), and there are indeed a lot of them.

But then if you consider that a wikipedia article about any of these topics is likely to get at least an order of magnitude more readers than the next most popular source, you can see the beneficial impact of even tackling a single article, especially some of the most popular ones. Each time a Wikipedia article is made into a great resource about some topic, a lot of future readers benefit. –jacobolus (t) 06:55, 16 November 2023 (UTC)[reply]

Policy drives behavior, and sometimes bad behavior. Fifteen years ago, there was a Category::Proof that contained all math articles that had embedded proofs (embedded proofs were (are?) considered controversial, so the category was a way of tracking those. After 3-4 years, someone put it up on Categories for deletion. There was no policy to notify WP:M, and so there was no rally to defend. Non-math people voted to delete it. Poof, it was gone. Six months later, it was recreated as a category for all of the recordings released by some rap artist called "Proof (rapper)". Here: Category:Proof Oh, well, now its just a redlink, I see. Well, anyway, if there had been some appropriate policy, then maybe this drive-by shooting of a large category would not have happened. 67.198.37.16 (talk) 07:12, 16 November 2023 (UTC)[reply]

Just to give a flavor of the scope of work, below is the current assessment table. Note that there are 166 GA-level articles. The Eigenvalue article is assessed as B-class, High priority. Based on the theory of statistical sampling, Eigenvalue is typical of B-class articles. It is safe to assume that they're all like this. All of them, or, precisely, all except 166 out of 27K will be hard to understand, and inaccessible to anyone without a formal education in math. In numbers, that's 99.4% of them. Now what? 67.198.37.16 (talk) 06:39, 16 November 2023 (UTC)[reply]

Mathematics articles by quality and importance
Quality	Importance
	Top	High	Mid	Low	NA	???	Total
FA	9	4	12	4		2	31
FL			2	1			3
FM					64		64
GA	15	20	56	84		8	183
B	127	266	538	415		49	1,395
C	52	555	1,022	1,311		629	3,569
Start		66	2,425	6,426		2,258	11,175
Stub			103	5,123		1,111	6,337
List	3	14	73	193	6	65	354
Draft				1	150		151
Category					1,717		1,717
Disambig			5	11	43		59
File					285		285
Portal					2		2
Project		1		25	26		52
Redirect		4	35	470	684		1,193
Template					1,013		1,013
NA					1		1
Assessed	206	930	4,271	14,064	3,991	4,122	27,584
Unassessed				15		345	360
Total	206	930	4,271	14,079	3,991	4,467	27,944
WikiWork factors (?)		ω = 112,724			Ω = 4.97

People often say that our math article should follow the model set by the encyclopedia Britannica. I do not agree with this, but it is instructive to compare our article with theirs:

When studying linear transformations, it is extremely useful to find nonzero vectors whose direction is left unchanged by the transformation. These are called eigenvectors (also known as characteristic vectors). If v is an eigenvector for the linear transformation T, then T(v) = λv for some scalar λ. This scalar is called an eigenvalue. The eigenvalue of greatest absolute value, along with its associated eigenvector, have special significance for many physical applications. This is because whatever process is represented by the linear transformation often acts repeatedly—feeding output from the last transformation back into another transformation—which results in every arbitrary (nonzero) vector converging on the eigenvector associated with the largest eigenvalue, though rescaled by a power of the eigenvalue. In other words, the long-term behaviour of the system is determined by its eigenvectors.

Notably, the second sentence of the article includes an equation, the first sentence uses the forbidden term "linear transformation". Tito Omburo (talk) 00:46, 17 November 2023 (UTC)[reply]

FYI

 – Pointer to relevant discussion elsewhere.

Please see: Wikipedia talk:Manual of Style/Lead section#A wording dispute about technical material.  — SMcCandlish ☏ ¢ 😼  20:17, 16 November 2023 (UTC)[reply]