Wikipedia talk:WikiProject Elements/Archive 28

Element infoboxes: remove isotope 'decay energy' column and some more
About the main.


 * I propose to remove the data (column) "Decay energy (MeV)" from the isotopes table in all these infoboxes. Recently, this was discussed in Archive 25. I have the impression this removal got support.


 * If agree, please take a look at sodium, which has these unique 2+2+1 decay routes by energy. it is What to do? Just cut out the energy-column results in an incomplete info structure. I'd say, reduce to plain three decay routes, omitting (or footnoting) the *-excited state difference in the decay products.
 * It is perfectly normal that alpha or beta decay leaves the daughter possibly in an excited state, to de-excite via gamma emission; so I think simply writing "β+" as the decay mode and 22Ne as the daughter without regard to excitation should be fine, as the main article isotopes of sodium do. This happens for most elements; the Na infobox just goes into too much detail about it. Double sharp (talk) 14:56, 21 May 2017 (UTC)
 * If it is this obvious, then please change Infobox sodium. I have no opinion on the sodium isotopes article/infobox/#section. -DePiep (talk) 15:21, 21 May 2017 (UTC)
 * Nevertheless I can't remove it and remain correct unless we remove the column for decay energy, because there are multiple possible values for it depending on which state of the daughter is reached. (Yes, this means that a significant number of our infoboxes are technically incomplete about this, which makes removing this column just a bit more urgent. ^_-☆) Double sharp (talk) 23:51, 21 May 2017 (UTC)
 * OK. Later then. -DePiep (talk) 10:03, 22 May 2017 (UTC)


 * Also I propose to: do not add uncertainties in this isotopes table. See rutherfordium.


 * I propose to unlink all half-life periods. See einsteinium. I don't think this link adds any relevant info about half-life (and, in case of linking to second, not even irrelevant info). If the relative time-scale is relevant, the columnheader link 'half-life' should be made to serve this.
 * I have no objections to either of these. Double sharp (talk) 23:52, 21 May 2017 (UTC)


 * Note: These proposals are only for the main element infoboxes. Not the new isotope boxes (which can be more detailed; to be decided by themselves), and not in the #Isotopes section.
 * Technically, this can be done. The almost similar isotopes- (new) and element-tables (old) are more and more diverging, to make such improvements possible. -DePiep (talk) 14:49, 21 May 2017 (UTC)


 * New proposal: Remove daughters too? To consider: mentioned to me that the Archive 25 discussion also is about remove the daughters (decay products) column. Is that consensus? I'm blank on this; while 'deductable' so redundant, it looks like a good illustration (decay process endpoint). !Votes? -DePiep (talk) 10:01, 22 May 2017 (UTC)
 * I think it's all right to remove the daughters: we after all do not have to re-explain how all the decay modes work in every element article, thanks to the power of hypertext. Double sharp (talk) 10:07, 22 May 2017 (UTC)

I actually think it would be cool to remove decay energies from the isobox as well for one reason: we rarely, if ever at all, talk about them. We can leave it for our isotope subarticles.--R8R (talk) 14:29, 22 May 2017 (UTC) I think this suggestion went unnoticed; I want to make sure you've seen it. Any ideas?--R8R (talk) 14:45, 28 May 2017 (UTC)
 * Yeah, we can leave them for the isotope subarticles. One thing I find rather odd is that the main article infoboxes currently give them, but the big tables in the "isotopes of X" articles do not. When I have more time I might go through those lists and add them. Double sharp (talk) 14:59, 28 May 2017 (UTC)
 * This talk only is about the main infobox. For simplicity, other topics are skipped and best be discussed in a new thread. OTOH, I'd like to read more opinions about removing the decay product from the main infobox. ? -DePiep (talk) 15:23, 28 May 2017 (UTC)
 * Yes, I think removing the decay product from the main infobox is a good idea for two reasons: a) it is easily deduced from the decay type and the space here is limited (so this is no showcase place for illustrations); and b) the decay product is (normally) no longer the element the infobox is meant to describe.--R8R (talk) 08:48, 29 May 2017 (UTC)
 * I don't have a very strong opinion about it, but removing the daughters is fine with me. Double sharp (talk) 10:12, 29 May 2017 (UTC)
 * I've come to the conclusion I want the decay products (daughter isotopes) to stay in the table. For me, a semi-layman reader, it makes the decay process complete (start-process-result). Until now, I did not even know the product can be deduced from isotope and mode, and I don't know if I could do so without doing a dedicated studies. While, OTOH, the infobox is perfectly right to have this (not-independent, deductable) information presented as a repetition.  -DePiep (talk) 10:11, 30 May 2017 (UTC)
 * I only insist we remove the products because you insist we should have the minimum information in the infobox (like you propose to remove the actually useful but apparently not infoboxy enough links to neighboring elements) and I try to follow because I like consistency and you have a point there.
 * Repetition is meant to educate. That's good but an infobox is not meant to educate; it is a quick reference that summarizes (how many times has this word been refered to in this discussion) the information and not educates the reader, for which we have the entire article. We can name the products if we want to in the text. If this was some vital isotope information, I would say we can have it, but there are few cases where this is actually important (and even then it's probably too techincal to be bothered with in an infobox). And again, infoboxes on particular elements should not educate readers on general laws of radioactive decay.--R8R (talk) 12:39, 1 June 2017 (UTC)
 * . In the infobox, we don't just mention the main isotopes by name (a list of "lead-204, lead-206, ..." would do then). We also describe them, mentioning half-life/stable and decay mode. I only extend this description with one more quantity (the product) to keep. The main argument is that it makes the info (decay process) complete. That is not 'educational' in purpose (though helpful in an encyclopedia it is), but good information presentation. Not as much repetition, because the info is not there and can only be deduced by the specialised calculation (I myself only found recently, let alone I could perform it). As I wrote before, adding the decay product makes the process complete with no blanks to find out for the reader. With the product, the process is described naturally complete: start - time - process - result. "Too technical" I could say about the decay modes, if at all. And it's not the general process, it's the individual isotope's process. (Theoretically, because describing the isotope, we even could consider adding the N number as in A = Z + N). -DePiep (talk) 09:03, 2 June 2017 (UTC)
 * BTW, you propose removal this from element infoboxes only, or from isoboxes as well? -DePiep (talk) 09:09, 2 June 2017 (UTC)
 * In any case, decay products should not be wikilinked. -DePiep (talk) 20:50, 3 June 2017 (UTC)

Edits in the element infobox

 * I see consensus on these, for Infobox element:
 * Remove column "Decay energy"
 * Remove wikilinks in half-life
 * Remove uncertainty
 * All good edits so far. Thank you.--R8R (talk) 15:14, 31 May 2017 (UTC)

I will make these edits. No consensus yet on removing column decay products (daughters). This can be concluded later without prejudice (we're free to decide any way). -DePiep (talk) 21:29, 30 May 2017 (UTC)
 * ✅. Also put words in table header. -DePiep (talk) 22:41, 30 May 2017 (UTC)

Isobox: main principles


We should discuss what the principles for infobox making should be and how this would affect the isobox.

First of all, I believe we should not use columns we don't talk about in the text. This excludes the current column of "decay energy." We never talk about decay energies.

Second, what isotopes do we discuss? All that are mentioned in the text -- does that sound reasonable? To justify a separate infobox for isotopes only, that would have to be significantly more that what we discuss in the main infobox, and since even secondary infobox shouldn't discuss what the article does not, we should limit the number of isotopes in the main infobox. Can we agree on that?--R8R (talk) 09:31, 29 May 2017 (UTC)
 * I find all of this perfectly reasonable, so yes, we can. ^_^ Double sharp (talk) 10:12, 29 May 2017 (UTC)
 * Then comes the following question. Look at lead, which uses this principle: its isotope table in the main infobox is super redundant as only less than a half of it is useful unique information. We would have basically the same amount of information if we had just a regular linear entry "Natural isotopes: 204 (relative abundance 1.4%) 206 (24.1%) 207 (22.1%) 208 (52.4%)". We could also fit half-life values and decay modes inside these parentheses (like "238 (99.274%, α, t1/2=4.468 y)"), even though this would be rare, around a third of our elements, most not so commonly viewed. How about we do this then?--R8R (talk) 11:16, 29 May 2017 (UTC)
 * I think that the potential problem with scanning a long list, like for example we would get for tin or xenon, is that in the middle it is not so very clear which abundance belongs to which isotope to a skimming reader. If we put line breaks between each isotope, I think it would be fine to scrap the table. Double sharp (talk) 11:42, 29 May 2017 (UTC)
 * Yes, I've had line breaks in mind. Very well then.--R8R (talk) 11:45, 29 May 2017 (UTC)
 * Well, I find your proposal perfectly reasonable with that clarification! ^_^ Double sharp (talk) 11:51, 29 May 2017 (UTC)
 * re "isotope table in the main infobox is super redundant": of course it is redundant! That's the purpose of the infobox. Please take a look at that link to get a primary idea of what the infobox is and should be. It is a bit tiresome having to explain this again and again.
 * Then, changing the isotopes table into linear summary. I don't know for which reason or benefit that would be. Listing isotopes (btw, "204" is not the way to identify or name an isotope) with multiple values in brackets: that is screaming for a table, a table as it now is. So I object. -DePiep (talk) 13:44, 29 May 2017 (UTC)
 * I really don't see you point re redundancy. An infobox is a panel, usually in the top right of an article, next to the lead section (in the desktop view), or at the very top of an article (in mobile view), that summarizes key features of the page's subject. How is redundancy meant to be here? I read the whole page past that phrase, sure, and still, I don't see how.
 * Again, look at lead. Less than a half of that space taken is useful and the rest is just making the thing look more complex than it should be (=>distracting the reader from wanting to read it, which is awful for an infobox). A majority of elements will suffer this, even more so view-wise. Only few will have to have multiple elements in parentheses (around fifteen or twenty) compared to ~70--75 that will simply have a visually less overloaded infobox. That is not to mention, we're going to be just alright with a good parenthesized comment like in "238 (99.274%; α decays, half-life 4.468 y)". Please take another look at it.--R8R (talk) 15:02, 29 May 2017 (UTC)
 * re wikt:Redundancy. You write "super redundant" as an objection. I claim that the infobox invites to repeat info already in the article. For example an infobox should "summarize". So in this, saying info is redundant (repeated, unnecessary, overflowing with) is not an argument against it. (We already assumed that the energy column could be removed: is not the issue here).
 * Perhaps there's a misunderstanding here. I perfectly realize that this info is already present in the article and that's how it should be. What I don't like, though, is how this information can be concluded from what is written in the infobox itself! That's the problem, not an occurrence of this in the text.
 * Then, your proposal too write the data in linear (listing) form, not tabular as it is now (that would be a bulleted/unbulleted list then). It is not clear what problem you are trying to solve. In general I get the impression you are just trying to save space (size reduction) at the cost of clarity & overview. That is a bad approach. The information is perfectly inviting for tabular presentation. These are the issues with your proposal: 1. Shortcutting isotope name to "204" is unacceptable. 2. Having to repeat the symbols like "t 1/2" either per isotope or, the horror, the Reader having to reconstruct it mentally. 3. The word 'abundance' is missing completely. 4. No solution yet for multiple decay modes (currently done by splitting the row). 5. In tabular form, the easy comparing of isotopes is provided (see for example to compare halflives). -DePiep (talk) 10:03, 30 May 2017 (UTC)
 * No, my main point here is not size reduction. My main point is this: we should only present the important information in the infobox; anything else is a distraction, which is terrible for an infobox. Now, as for your points: 1. Yeah, okay, we can have 204Pb. This is not the point. 2. I've already suggested we can write "half-life"; I wonder if you actually missed it. 3. Again, wording can be improved. We can look for exact wordings after we agree on the concept. 4. No problem with this: we can simply list them by symbol. separating decay modes by commas. "22Na (trace; β, ε decays, half-life 2.602 y)" 5. The very point of the whole idea is that there are very few radioactive isotopes among these important ones to be featured in the main infobox. If it wasn't for that, I wouldn't suggest the idea in the first place. (Again, the whole idea is that a vast majority of these super important isotopes (the stable ones) have two useful columns now: isotope and relative abundance, which don't even take half the table. So there is a lot of visual distraction and clearly non-infobox information now.) Practically, this is not something to be worried about.--R8R (talk) 16:49, 31 May 2017 (UTC)
 * only present the important information: sure. And present it complete & clear, I might add. But when doing just a list, either one has to repeat each quantity name (abundance, half-life, decay mode, decay product [I'd keep]) in each, or the reader has to figure this out from a weird code soup:
 * "22Na (abundance: trace; decays:β+→γ, ε→γ, β+; half-life: 2.602 y; product: 22Ne)"
 * "22Na (trace; β+→γ, ε→γ, β+; 2.602 y; 22Ne)"
 * So it's either repetition of all those labels, or leaving them out to guess. While the table form already solves it in the columnheaders, and by its structure (same data in same place). The super-structure is even more impressing: all infoboxes use the same table. I even don't mind the current solution for stable ones. -DePiep (talk) 13:38, 1 June 2017 (UTC)


 * Re the main proposals mentioned.
 * I think the infobox for sodium is too much detailed right now. There are two decay modes of sodium-22: beta plus and epsilon. The fact that the daughter then can undergo de-excitation is not related to sodium. So it should be "β+, ε" only. I'll change that regardless of the result of this discussion.
 * Good. This was already mentioned above, "OK. Later then. -DePiep (talk) 10:03, 22 May 2017 (UTC)" [i.e. could be done after energy column was removed; left to do in isobox. DePiep]. Sodium is the only element that had/has this "(2 2 1)" decay form. -DePiep (talk) 08:37, 2 June 2017 (UTC)
 * "I even don't mind" You're the number one person in this project who I expect to go strict and follow rules and constructive arguments rather than personal liking. The long lines "is stable" are a distraction, both visually (more filled space = more stuff to look at) and informationally ("is stable" is more or less okay, even though no imformation would imply the same; "with 124 neutrons" reads like they had to fill the space with something because it's been implied already. I've had this feeling since the first year here). I propose we remove it.--R8R (talk) 15:24, 1 June 2017 (UTC)
 * my "I even don't mind" is to say: I have no absolute knife to decide in/out. Because, since we decided to add some description to the isotopes (not just their names), the N number might as well be in too. I find it very clarifying within the isotopes business, it reduces mental questions. But that is not a killing argument. It is not some personal belief, but based on the question of what info to provide. (In other places I may have a stronger argument for/against. But I'm not that zealous to remove a plural-s to reduce data). -DePiep (talk) 11:39, 2 June 2017 (UTC)
 * It's clearer now, thank you.
 * Still, I see we could use a good help at it here. The big infobox was maybe a good idea at its creation: it helped standardize the element articles, for instance. Now the project is long established and we might take a good look at its details, which you're doing; I can't thank you enough for that. While I think there is content to add to the infobox, I'd have lots of parameters go away because we should stick to the essentials. (I hope I'll get back to it in a few weeks/months.) In particular, the decay products are not okay with the purpose of the infobox, which I imagine as a list of the essential information because there's a lot to cover in an element. I don't want the decay products in the main infobox because they can be deduced (by simple rules, as simple as basic arithmetic), so they're clearly not essential. We may, by the way, have this information in the isobox, which is absolutely fine to have more information that the table in the main infobox, so they won't be gone.--R8R (talk) 15:10, 4 June 2017 (UTC)
 * Yes, time to cut back to the essentials. BTW, is about both infoboxes. Maybe condense this into an overview, and pingvite the others. -DePiep (talk) 19:33, 4 June 2017 (UTC)
 * By now, I agree on removing "with n neutrons" in stable isotopes. -DePiep (talk) 12:34, 8 June 2017 (UTC)

Arbitrary break

 * I can support removing the 'Decay energy" column completely from all (as is about to be done in the main element infoboxes, see #above). For ease of (technical) editing, I'll speedy-propose this, below. -DePiep (talk) 10:47, 30 May 2017 (UTC)
 * Re "List all isotopes mentioned in the text". That would be in the article section #Isotopes then? This question arises: the isobox formally is the top infobox in the separate isotopes-articles: Isotopes of lead. And it can be nicely reused in Lead#Isotopes section. However, it is not a given fact that these two places require the very same table. IOW, how would this play out for (all) isotopes articles? My guide would be: the isobox should fit the isotope article always. If and when the Article section needs a different listing (isotopes table), it could & should have a new, third table.(So far, we have not thought about restructuring/rewriting isotopes pages. For example, isotopes of lead to reflect a high-quality Lead section, which would make the single infobox more to the point). I want to make clear: there should be no requirement that the isobox must be reused in the #Isotopes section.  -DePiep (talk) 10:47, 30 May 2017 (UTC)
 * Re "Limit the number of isotopes in the main element infobox". I fully agree. In the element article, isotopes are less important overall.
 * So with this, we have three or four lists: 1. The big table in isotopes article (sense of completeness), 2. Isobox used in both isotopes article and article#Isotopes section (or, 2b, in #Isotopes a new, special table), 3. Element infobox small list. As always: all tables should be supported by their body text.
 * Technical note: I am planning to rebuild the tables to make this editing (data & presentation) more easy. -DePiep (talk) 10:47, 30 May 2017 (UTC)


 * New question: I see that infobox lead isotopes has half-lives with an uncertainty now: "36.1(2) min". First impression: too much detailed. Isn't this just for the big complete table? -DePiep (talk) 10:52, 30 May 2017 (UTC)
 * Expanding: we could create the general rule that this isobox has yes/no precision, and values (half-life, abundance) are limited to 3 (or 4) significant figures (number possibly related to the other isotopes in the table: keep the meaningful sigfigs). -DePiep (talk) 07:18, 31 May 2017 (UTC)


 * Intermediate proposal: we could remove column Decay energy completely, and remove links in half-life (as was done in the main infobox). Doubts? -DePiep (talk) 07:18, 31 May 2017 (UTC)

Roundup of the isobox proposals
A roundup of the proposals, and their status, so far. All apply to Infobox element isotopes. [note 1]=already done in main Infobox element. -DePiep (talk) 13:54, 8 June 2017 (UTC)

Consensus: No consensus:
 * Remove column "Decay energy" [note 1]
 * Remove links in half-life [note 1]
 * Remove uncertainty in half-life [note 1]
 * For stable isotopes, remove the "with x neutrons" text. (Leaving: "is stable").
 * Do not use table form, write linear list like:
 * 22Na (abundance: trace; decays:β+→γ, ε→γ, β+; half-life: 2.602 y; product: 22Ne)

22Na (trace; β+→γ, ε→γ, β+; 2.602 y; 22Ne)"
 * No consensus.


 * Remove column "Decay product"
 * No consensus.


 * Add columns for specific isotope types (spin, in decay chain, ...)
 * Separate thread, later.


 * Which isotopes to list?
 * Evolution, converging. (Note: does not alter table setup).

-DePiep (talk) 13:54, 8 June 2017 (UTC)


 * Further discussion
 * re Decay products: asking and others: keep or remove? See discussion above, scattered (describe isotope more complete vs. remove redundant info). (While we're at it: keep/remove from main infobox element?) -DePiep (talk) 13:59, 8 June 2017 (UTC)
 * I don't mind if they stay or leave, since we already have the decay modes. Double sharp (talk) 14:32, 8 June 2017 (UTC)
 * R8R and I were hoping you could give the decisive nod/shake ;-). -DePiep (talk) 15:03, 8 June 2017 (UTC)
 * One column I think may be important enough to put in the main infobox is nuclear spin, which is important for NMR: I usually want to mention this briefly in pretty much every element article (e.g. nitrogen, chlorine, among those I've written), and not including it means I have to put it in the main text separately for each useful isotope. But this is a little borderline, and is not that useful for most unstable isotopes (long-lived ones like 235U and 239Pu being an exception), so I'm not going to strongly ask for it. Double sharp (talk) 14:32, 8 June 2017 (UTC)
 * This section is about removals only. Adding spin etc. proposal is in preparation :-) -DePiep (talk) 15:03, 8 June 2017 (UTC)

Proposal: new columns and tables for isobox
See Template talk:Infobox element isotopes#. -DePiep (talk) 15:21, 14 June 2017 (UTC)

Reclassifying the nonmetals, Continuation
This is a continuation of a discussion, broken here to allow an archival of the previous parts of the discussion.

began this discussion, which went through two iterations (Mk. 1 and Mk. 2) before reaching this point. I have repeated his background section to provide some context. YBG (talk) 04:02, 3 July 2017 (UTC)
 * Those prior discussions have been archived at WT:WikiProject Elements/Archive 27 YBG (talk) 05:21, 3 July 2017 (UTC)

Background
We currently colour code non-metals as polyatomic, diatomic, or noble gas.

Ever since we adopted these categories I've wondered (partly prompted by R8R) if that was the right decision, and if there was a better alternative.

This has been hard. Categorisation of nonmetals in the literature—aside from the halogens and the noble gases—is shabby. Metalloids complicate the situation. Some authors recognise such a category; others don't. The one that don't have to divvy them up between the metals or the nonmetals.

Authors usually throw up their hands and simply look at the leftover nonmetals—the ones other than the halogens and the noble gases—on a group-by-group basis. So you might have separate sections in a book on, say, hydrogen; carbon; nitrogen and phosphorus; and oxygen, sulfur and selenium. And some or all of the metalloids might get added to the applicable sections.

Otherwise there is the question of what to call these leftover nonmetals. The category names I've seen in the literature are "biogen", "CHONPS", "organogen" or "other". The first three of these categories tend to get tripped up by what to do with selenium. The last category—other nonmetals—is the "I-give-up-it's-too hard-I-need-to-get-published-so-I'll-treat-them-as-leftovers" category.

For our own classification scheme the "halogen" category is unavailable since we count astatine as a metalloid (as we should—it's either that or a post-transition metal).

Alternative proposal, Mk. 3
Here I've tried to accommodate the views of Double sharp, YBG and R8R, by retaining "metalloid" as a super-category, and replacing "intermediate nonmetal" with "reactive nonmetal": Winners all round, I hope. Sandbh (talk) 04:30, 14 June 2017 (UTC)
 * Double sharp, I expect you may be more comfortable with the retention of "metalloid" as a top-tier category.
 * YBG, the proposal picks up your "reactive nonmetals" suggestion; Double sharp, this name perhaps better accommodates the schizoid nature of nitrogen.
 * R8R, the "corrosive nonmetal" category that you like is still there.
 * For me, I can still use this categorisation scheme to discuss the progression in metallic to nonmetallic character, and the symmetry in this pattern (given the generally nonmetallic chemistry of metalloids, regardless of their top-tier status).
 * It still feels a bit odd to me, though; O and the halogens are surely reactive and corrosive, no? You can't be the latter without already being the former. "Moderately reactive" works a bit better, but the trouble is that it is inherently subjective (you'd get cases like N; this is mostly again the stability of the bonding – bulk Si is very unreactive, but liquid Si is very reactive)! So I'm not sure which way to go on this one. Double sharp (talk) 05:41, 14 June 2017 (UTC)
 * Yes, certainly O, F, Cl, Br and I are reactive and corrosive. I'm attempting to differentiate between the two kinds of non-noble nonmetals i.e. the ones that are corrosive and the ones that are "merely" reactive compared to the noble gases, without having to use subjective qualifiers like "moderately" or "strongly". The alternative would be to use "reactive nonmetal" and "reactive, corrosive nonmetal", which is redundancy gone rabid. At this level of detail, and in the context of periodic table categories, I don't think it's too much too presume that folks will be able to grasp the difference between reactive and corrosive. Sandbh (talk) 07:04, 14 June 2017 (UTC)
 * It still feels vaguely problematic to me because this would be the only case where you have to mentally iron away the redundancy; all the other categories are disjoint. We can quibble about the borders, but you don't get cases where an element is placed twice; we can argue about whether mercury is a transition metal or a post-transition metal, but no one thinks it is both. I suppose it is merely a symptom of the difficulty of categorisation in the p-block, which makes me think that it is little short of a miracle that the current monatomic/diatomic/polyatomic divide corresponds so well with chemical and physical divisions without creating this sort of redundancy. Let's bear in mind that the periodic table is about the first thing people with no clue about chemistry see when they begin to study the subject: so I think it is important not to force the reader to fill in any gaps there.
 * (A little exception might be that it is possible to regard lanthanum and actinium as both group 3 transition metals and the prototypes of the lanthanide and actinide series, or lutetium and lawrencium if you prefer. But somehow the colouring of Sc and Y and the rectangular shape of the transition metals block makes the implication more plausible, and honestly La and Ac are more usefully considered lanthanides and actinides than transition metals. I recognise that this isn't perfect for Sc and Y, but at least they are physically good transition metals if chemically poor ones. Double sharp (talk) 07:20, 14 June 2017 (UTC))

A small follow on comment about this proposal. After Double sharp queried if the halogens were not also reactive, as well as being corrosive, I began to doubt my own proposal. They certainly are reactive as well as corrosive but it looks like the corrosive nonmetals are then a subset of the reactive nonmetals, hence I became struck with a mild case of category overlap.

Upon reflection I suspect my concern was misguided. The English language is full of words used to denote differing degrees of something. For example: hot–scorching; happy–rapturous; heavy–crushing; destructive–catastrophic; attractive–bedazzling; authoritative–indisputable; complex–tortuous. While each of the second of these terms can be thought of as a subset of the first, the distinction between the terms in each pair is evident. So I now tend to think that "reactive–corrosive" is a reasonable distinction. Sandbh (talk) 02:42, 19 June 2017 (UTC)
 * There is of course a distinction, but how does it solve the problem? If we look towards the average reader who knows nothing about chemistry, which you usually appeal to, he first does not know that "alkali metal" and "alkaline earth metal" are specialised single terms, and are not really modifiers applied to the term "metal". Perhaps he might even suspect that the second is a subset of the first, but he would be rapidly disabused of this notion. Then he gets through all the metal categories, and he learns that all of these terms have a specialised meaning and don't overlap. With this organisation of the nonmetals, what's there to stop him from thinking that there is some actual definition of "intermediate nonmetal", or that this isn't some specialised standard use of "reactive" and "corrosive"? (After all, "react" already has a special meaning in chemistry!)
 * Certainly we can say that the responsible reader should not parrot Wikipedia wholly for things like this, and should at least read the fine print. The problem is that at the level of introductory middle-school chemistry the fine print is incomprehensible, and only the pretty colours will get noticed. Double sharp (talk) 03:36, 19 June 2017 (UTC)
 * I am sorry to say that a similar problem occurs with the alkali metals, alkaline earth metals, and the lanthanides, when the general reader discovers that the Ln are in fact "alkaline" metals, and wonders why there is no reference to this in their category name, and why the Ln includes La, when "lanthanide" means "like lanthanum" etc.


 * When he gets through the metal categories he soon begins to realise that the categories are by no means sharp, as evidenced by questions around where the transition metals start and end.


 * The general reader may assume that there is some actual definition of "intermediate nonmetal" but will soon find that this is not the case just as there is no formal definition of post-transition metal or metalloid, nor is their any standard use of polyatomic or diatomic nonmetals to distinguish between the categories and properties of nonmetals in the way we have done.


 * In introductory middle-school chemistry I hope that the student will compare the pretty colours with the way the periodic table is presented in their text book, and that this prompts some good questions to their teachers, and some 'aha' moments. (I can recall ever seeing only one text book that features a Wikipedia style periodic table).


 * One day the general reader, who has now developed a keen interest in chemistry as a result of their middle school experiences reads King's Inorganic chemistry of main group elements and sees that the Ln are in fact treated by King as main group elements analogous to the alkali metals and alkaline earth metals, and thinks, "I was right all along to wonder why the Ln weren't instead called the alkaline rare earths :)


 * I hope that some of these kinds of issues become a bit less issue-like and more NBD-like (= no big deal), once I post the overview guide, which I intend to do next. Sandbh (talk) 04:26, 19 June 2017 (UTC)
 * The category names are not very good, I know. But you quickly learn to accept them as referring to something else and having non-indicative names ("alkaline rare earths" is funny; first of all they are not rare and second of all Yb and Lu form amphoteric hydroxides, dissolving in hot concentrated NaOH to form ytterbates and lutetates, the latter with a terribly non-indicative spelling). As for the transition metals, I think the ending is pretty clear-cut at group 11; group 3 depends on whether you are looking at it as a physicist (in which case, yes) or a chemist (in which case, no).
 * It generally seems that the group-based categories are much more well-defined than the metallicity-related categories, and this I suspect is because the concept of what a metal is is difficult to pin down. Perhaps we want to say that caesium is the most metallic of the metals. Very well, but Cs+ is so large that it does not really form a well-defined aqueous cation; there is no clearly defined primary solvation shell. Furthermore, from the physical perspective, who ever heard of a metal that is liquid at room temperature and is softer than talc? So maybe we go to a more physical definition and consider gold. Well, it's certainly malleable and ductile beyond compare, but not only does it not form aqua cations at all, but its electrode potential is also ludicrously highly positive for a metal. I suspect that the general idea of what a metal is comes from that staunch workhorse iron; but while indeed the properties of Fe are characteristically metallic, not many elements are very good at being like iron, not even its older sisters Ru and Os. (Its cousins Co and Ni are much better, but then in the +2 oxidation state rather than the +3; in general they are less willing to switch oxidation states than iron itself.) I nevertheless hope that a definition can be found that is better than "I know it when I see it" and "just look at Fe", and I hope that this discussion helps us come up with one, or better, realise that a good one was staring at us in the face all along! ^_^
 * Regarding prompting good questions and "aha" moments, I am not sure a simple colour coding can do this without a lengthly backup full of descriptive chemistry, like you did for example at metalloid and post-transition metal (incidentally, can we perhaps expect transition metal? ^_^). That would certainly do the trick, for those who are interested and do not yet know of all those descriptive inorganic chemistry tomes. Double sharp (talk) 10:30, 19 June 2017 (UTC)

Where things are up to
I saw Parcly Taxel's  question to Double sharp about keeping track of where things were up to, and Double sharp's response, near the end of this section.

Here is my preferred summary of the situation.

Our current nonmetal categories are…


 * Polyatomic nonmetal &#10022; Diatomic nonmetal &#10022; Noble gas

My first proposal was to change this to…


 * Weak nonmetal (metalloid) &#10022; Intermediate nonmetal &#10022; Corrosive nonmetal &#10022; Noble gas

I subsequently suggested this could instead be…


 * Metalloid (weak nonmetal) &#10022; Intermediate nonmetal &#10022; Corrosive nonmetal &#10022; Noble gas

I regard these proposals as dead.

My second proposal was to change to…
 * Metalloid &#10022; Intermediate nonmetal &#10022; Corrosive nonmetal &#10022; Noble gas

The elements most commonly designated as metalloids are B, Si, Ge, As, Sb, and Te. Per Double sharp's observation, the term metalloid gives us a warning that while these elements might look like metals, they don't act like metals. Indeed, the literature (here, 1894–2016) records that their chemistry is generally nonmetallic. Even so they are the most metallic of the nonmetals.

The intermediate nonmetals are H, C, N, P, S, and Se. They are neither as metallic as the metalloids nor as chemically active as the corrosive nonmetals. This is a better explanation of the intermediate nature of these nonmetals.

The corrosive nonmetals are O, F, Cl, Br and I. Their category name speaks for itself.

I believe I've addressed all substantial feedback provided in response to this proposal.

This proposal is still on the table.

My third proposal suggested changing to…
 * Reactive nonmetal &#10022; Corrosive nonmetal &#10022; Noble gas

I am now less interested in this proposal however I wouldn't quite call it dead, yet.

There has been some limited discussion about a two-category approach…
 * Reactive nonmetal &#10022; Noble gas

In summary There is no perfect solution. Our current scheme has weaknesses. My second proposal has weaknesses. Any scheme can be marked down on this basis. But these weaknesses do not matter provided the classification scheme "is beneficial to economy of description, to structuring of knowledge and to our understanding, and hard cases constitute a minority".

Weaknesses of both schemes aside, the second proposal is better than the current division because it is a purer solution, one that is more consistent:
 * with the rest of our categorisation scheme;
 * how the elements are described in the literature;
 * with the progression in metallic to nonmetallic character as you go from left to right across the periodic table, and
 * the way chemists think, which is in terms of where each nonmetal approximately lies on a composite scale (or spectrum) of nonmetallic "liveliness" or "intensity" rather than in terms of polyatomic or diatomic nonmetals.

Tomorrow I'll see if there is some kind of conclusion I can add to this summary.

I'm patiently waiting for any further thoughts from Double sharp and R8R in response to "Response by Sandbh to Double sharp's comments of 03:51, 12 June 2017" and "Response by Sandbh to R8R's comments of 17:03, 12 June 2017". --- Sandbh (talk) 13:33, 14 June 2017 (UTC)
 * We've now both responded (and I'm sorry for setting you some homework on the chemistry of Cr, Se, Te, and U ^_-☆). Double sharp (talk) 05:25, 15 June 2017 (UTC)
 * well I've lost my patience. Things are up to nothing, by now. Nearing 50.000 words, and nowhere a summary or overview or set of conclusion in sight. After three months of discussion, you simply enter a second (Mk. 2) proposal, without any digestion of the enormous subsection wrt the chemical and physical arguments. Scattered around are three 'summary of comment' for each contributor. When I tried to organise the topics I was reverted, because "the flow of words" may not be broken into something logical and legible apparently. Astonishing that this discussion cannot be organised into something useful by you all, wilfully even.
 * To be clear, since and as long as no argument building and developing has been presented, I oppose every change. -DePiep (talk) 14:30, 14 June 2017 (UTC)


 * I oppose the alternative proposals as well (thanks ). How corrosive does a "corrosive nonmetal" have to be anyway? Polyatomic and diatomic classifications are completely objective, whereas the proposals introduce subjectivity. I do not foresee a systematic review of the literature to determine nonmetal groupings like we did for metalloids. So let's just stick with what is live now and archive the black monster we've wrote. [[User:Parcly Taxel|Parcly ]] [[User talk:Parcly Taxel|Taxel ]] 16:57, 14 June 2017 (UTC)
 * To be clear, : I do not oppose because of content arguments (as you mention). I oppose because the whole discussion is a useless diarrhea, and no one gets to organise it. -DePiep (talk) 20:52, 14 June 2017 (UTC)

please take a deep breath and have another look at it.

The discussion is developing. The second and third proposals by Sandbh are reactions to the reception he got. That's alright and good. That's how discussions go. I would not refer to those mini-discussions you've mentioned for the sole reason they are in many ways self-repetitive or secondary compared to the main points.

As for myself, I see that now is a second wave of discussion, fresh and reignited. It is going well.

Not to say I came to this talk just a few days ago and made my own suggestion and there is some early indication it might go reasonably well here. I assume that Sandbh mentioned "limited discussion" because my proposal because it came up so recently. Sandbh is trying to make his points, I'm trying to make mine. DS is complementing this discussion by participation a lot. I am trying to get their reception and see more about about Sandbh's proposal; I've asked a question just today. I have and am expecting we will later proceed to the decision-making. The discussion is not staggered yet. So please don't give up.

(Also, I, too, genuinely don't see the problem in having long talk pages or "unstructured" discussions. I can navigate throughout it by the means Ctrl+F even if nothing else.)--R8R (talk) 17:30, 14 June 2017 (UTC)
 * re please take a deep breath and have another look at it. I did, more than once. And was reverted by Double sharp, some days ago. Does the current TOC really help you or anyone else?
 * re Mk. 2, 3 are "reactions to the reception he got": why do you have to explain that to me? Why was that not clarified by Sandbh right away? Where is the summary & conclusion of all arguments?
 * How many 'summaries of R8R' are there? How many cherry pickings needed? 50.000 words. And I am supposed to search and find deduct the right arguments myself? -DePiep (talk) 21:18, 14 June 2017 (UTC)
 * I note: this is WP:ELEMENTS, and . Our bests. This explains why I aim for and expect higher quality. Both in talkpage and in articles. If I trespass acceptable communication lines, correct me. Also, please allow me to be sharp. -DePiep (talk) 21:48, 14 June 2017 (UTC)

I'll post an overview guide to the arguments.

Allowing for my break, I believe the discussion has been going for 7½ weeks rather than three months (although it may feel like that).

I posted the Mk 2 proposal in an attempt to accommodate the previously expressed views of Double sharp, YBG, and R8R, and bring the discussion to some kind of conclusion. In retrospect this may have been a naive action on my part and it may have benefited from more background summarising where the Mk 1 proposal had gotten up to. So I've been metaphorically knocked down; that's OK; I'll get up again.

I did not follow what happened with your attempt to organise the topics because I thought you would do a good a job as anyone else. Nor did I follow what happened with the reversion. If someone edits and someone else reverts I tend to leave the people involved to sort things out.

In an effort to move forward I'll look at the history of your organisation work and the subsequent reversion, as part of producing the overview guide to the arguments. I expect the guide will include links to the relevant posts, and it might be better organised by subject than chronology but I won't know until I look back at what happened and try and write the overview guide.
 * . In this, I don't exactly expect you to recap the discussion to base your Mk. proposals. That would be weighing arguments too early. Given the size and wideness of the talk, right before posting your Mk.2, we could use a resume of all relevant arguments, open or finished, pro or against. (As it is now, an indiscriminate unstructured list of points that has something for everyone and that no one will read. I tried this TOC that was reverted, so I'm waiting for other talk-structuring proposals). I also find that there are two main areas: categorisation the meta, and chemical/physical properties of elements as a base for any categorisation. WRT the categorisation I can recognise arguments and ideas in there (after shifting), but an overview there is not. But wrt the properties, with which I am very unfamiliar at this level, I cannot grasp what is relevant, and how they do influence the proposal if at all. Any conclusions? The Mk. 1 overview I ask for should summarize all relevant arguments for these. (With that summary, by intention, the Comments section could be moved to Archive). Following that overview (in TOC & time), your Mk. 2 proposal would be perfectly in place. That would also solve the so far unrelated, extensive descriptions Double sharp adds without connection to the proposals.
 * The more stronger option would be to archive all, and restart a discussion (probably picking up relevant arguments). This sounds extreme, but no doubt this will happen anyway if the discussion does not yield anything for being driven into the mud by careless structure. An other route would be that you write a concept-article in your sandbox and have that one commented (seems like you have assumed that approach here?). I note that even in this "Where things are up to" discussion-overview, in creep individual content arguments ("Only S is a polyatomic nonmetal"), defending a single proposal. -DePiep (talk) 11:22, 16 June 2017 (UTC)
 * As I've said above, my sandbox is open. Please add any relevant arguments there. (I've added a few I've found already, and I'll be reviewing the sources for Sarawak in the meantime.) [[User:Parcly Taxel|Parcly ]] [[User talk:Parcly Taxel|Taxel ]] 12:12, 16 June 2017 (UTC)
 * I've drafted an overview summary which is very short. The discussion essentially boils down to four or so themes. I need to sleep and work out how to conclude it (i.e. the summary). I probably need to sandbox a proposed nonmetal article so that project members can see the substance of the proposal rather than just its outline. Sandbh (talk) 12:34, 16 June 2017 (UTC)
 * That sounds interesting. Time is not an issue, nor is sleeping. -DePiep (talk) 12:43, 16 June 2017 (UTC)

as far as I can see the literature refers to O, F, Cl, Br and I as corrosive.

There are some aspects of the diatomic/polyatomic distinction that are a little bit subjective.

The existence of ozone, a polyatomic nonmetal, has bugged me. Sure oxygen is a diatomic nonmetal but you would hardly call it a borderline diatomic nonmetal, so why does ozone exist? I'd expect this sort of thing for an iodine-like nonmetal, not oxygen.

Only S is a polyatomic nonmetal in the sense that one can point to an S8 molecule. There are no discrete molecule-equivalents for C, P and Se in their most thermodynamically stable forms.

Iodine has some polyatomic character. There is evidence for significant intermolecular coupling between the individual iodine molecules (each iodine atom forming weak bond with its two next nearest neighbours, as well as a stronger bond with its molecular partner) implying a bulk coordination number of 1+2, rather than 1. There is a note along these lines in the metalloid article. The funkiness of the crystalline structure of iodine is further borne out by its similarity to that of gallium. Sandbh (talk) 03:34, 15 June 2017 (UTC)

Note that the polyatomic nonmetals we have have each atom strongly (covalently) bonded to more than one other atom in their thermodynamically most stable forms, therefore justifying the polyatomic designation. The diatomic nonmetals have each atom strongly bonded to exactly one other atom in said form. The status quo is justified.

I've already typed out a summary on my sandbox (permalink). I still believe that if we had to use an alternative and ditch the present categorisation, the only viable alternative would be to just highlight each group (H, 14, 15, 16, 17, 18) in a colour gradient and call them all nonmetals, which admittedly has a beautiful presentation (to my eyes; I've rode the MTR and some of its posters use a similar effect at the bottom). You are free to edit my sandbox with any information I might have missed, but please keep things very short. [[User:Parcly Taxel|Parcly ]] [[User talk:Parcly Taxel|Taxel ]] 03:56, 15 June 2017 (UTC)
 * You know, I kind of like the idea of colouring groups separately and ignoring the distinction between nonmetals, metalloids, and metals. After all, in the series {Si, Ge, Sn, Pb}, {P, As, Sb, Bi}, and {S, Se, Te, Po}, there is not a whole lot of difference between each element and the one below it. If you compare the extreme members of each series there is quite clearly a difference, but that's because there are two trends both conspiring to increase metallicity: the dominant oxidation state drops by two and the atomic radius keeps going up. (The metallicity of Po is significantly lower than that of Pb and Bi, because the oxidation state involved is now +4 instead of +2 or +3, and the PoIV cation is very prone to complete hydrolysis to PoO32−; I should address some borderline elements like Sn and Po in the series too.) It may make the colours a little redundant though since the groups are obviously shown even without colours as vertical columns. Double sharp (talk) 04:37, 15 June 2017 (UTC)


 * I did not mean to ignore the distinction between metals, metalloids and nonmetals. My MTR proposal would only colour the nonmetals in a gradient – the metalloids and metals having been defined well in the past.


 * But by your new principle above that a category must be definable without making reference to another category, the alternative proposal(s) are ruled out on the basis that "corrosive", "intermediate" and "reactive" all cannot be properly defined on their own; they require some context. Good job! [[User:Parcly Taxel|Parcly ]] [[User talk:Parcly Taxel|Taxel ]] 06:13, 15 June 2017 (UTC)
 * Well, I think this talk is still serving a useful purpose: if not for my coming up with evidence for my stand, I do not think I would be working for a silicon GA in my sandbox: the chemistry is pretty daunting. I also do not think I would have added the principle of metallicity increasing with higher atomic size and lower oxidation state. I also respect Sandbh's thoughts greatly, so I would want to hear what he thinks of my comparison of metallicity with atomic size, and perhaps this does shed some light on the chemistry of uranium. Even if there is no change, the discussion lets us learn more about chemistry, which translates to better articles, doesn't it? ^_-☆ Double sharp (talk) 07:05, 15 June 2017 (UTC)
 * Indeed. The entire conversation should not be archived with a number, it should receive its own sub-page. And the references will most certainly be useful for writing articles about our period 3 nonmetals, which really are a disjointed mess. (The status quo versus MTR debate can be left for later, as the latter is just an idea now.) [[User:Parcly Taxel|Parcly ]] [[User talk:Parcly Taxel|Taxel ]] 07:28, 15 June 2017 (UTC)
 * I've been planning Si, P, and S since late 2016; it looks like they might finally bear fruit now, in mid-2017. Mg and Al would complete the row; I think R8R wants to do Al, but I don't think he'll mind a collab. After all, we're trying to demonstrate the viability of the "GA-break-FA" scheme with Pb and Th: so maybe I could write a stopgap GA for Al and Au first, awaiting eventual improvement to FA. It will be at least better than keeping them all in the current not-so-great state.
 * There are also a number of leftovers, like Ca, As, and Sn. And one thing I want to do before the decade is out is the the rest of the group articles (at least the main groups); it is slightly absurd that I started this idea in 2011, went happily to do group 1, and then we are still here six and a half years later because the tale grew in the telling. At the very least doing the halogens would make a GT now. Double sharp (talk) 08:53, 15 June 2017 (UTC)

I think I am nearly up to date with my responses. I am going to attempt to do the overview summary, and then respond to Parcly Taxel's  good work. Sandbh (talk) 05:17, 16 June 2017 (UTC)
 * It seems that our discussion has been primarily organized as a sequential discussion of proposal after proposal: (a) proposal 1 and why it is good (b) proposal 2 and why it is good and better than proposal 1 (c) proposal 3 and why it is better than the others and so on.
 * This largely ignores all of the good work we did in our mega-discussion that resulted in the diatomic/polyatomic division. IMHO, we should have a significant bias toward the status quo, primarily as a recognition of all of the work in that mega-discussion. Consequently, I think our discussion should proceed as follows
 * I presume we have a consensus to treat the noble gasses as a separate subcategory of nonmetals.
 * If this is not the case, please ignore the rest of this outline
 * If this is the case, continue to the next step
 * Is there a compelling reason to abandon the polyatomic/diatomic division?
 * If we reach a consensus of "NO", skip to the final step.
 * If we reach a consensus of "YES", continue to the next step.
 * Is there a compelling reason to subdivide the non-noble-gas nonmetals?
 * If we reach a consensus of "NO", skip to the final step.
 * If we reach a consensus of "YES", continue to the next step.
 * What is the best way to subdivide the non-noble nonmetals?
 * Given the decision about subdividing the nonmetals, does it make any sense to consider the metalloids to be a subcategory of the nonmetals?
 * I believe our discussion has faltered because we have started at step 4 instead of step 1. What do you folks think about this scheme? I suggest that use this outline to organize his summary and that we attempt to reach a consensus on each of these questions in order. YBG (talk) 14:12, 17 June 2017 (UTC)
 * I'll leave you to sort this out, but we have no compelling reason to abandon either the poly/di classification or the metalloids as a top-level group. I need a break. [[User:Parcly Taxel|Parcly ]] [[User talk:Parcly Taxel|Taxel ]] 14:35, 17 June 2017 (UTC)
 * Yes, keep NG as a separate category. I expect the sandbox revised nonmetal article will enable the project to assess the merits of 2, 3, 4, and 5. Sandbh (talk) 08:25, 19 June 2017 (UTC)
 * Yes, absolutely keep the noble gases. It would be perverse to give up such a homogeneous category just because of how oganesson might behave. Double sharp (talk) 10:38, 19 June 2017 (UTC)


 * If only could be brought to join the discussion. Instead of unloading ten-thousands of words for others to sort out, no head or tail. A 530.000 bytes page and no improvement in sigh. How can you people ever produce an FA? -DePiep (talk) 10:58, 19 June 2017 (UTC)
 * I think I'll defer to Sandbh's judgement of my contributions, since it is after all his proposal, and my notes on descriptive chemistry of some elements help improve it by pointing out edge cases. As for FAs, it is very simple: it is a lot easier to write a structured page on something you already know about beforehand than something you do not. Double sharp (talk) 11:01, 19 June 2017 (UTC)
 * I'm here to salute your contributions. And also to ask: how are they related to the proposal? Your contributions being 'great' (dixit Sandbh), what is their relation to the Proposal? Thousands of words you add, and no conclusion at all. -DePiep (talk) 23:40, 19 June 2017 (UTC)
 * The proposal includes the classification of metalloids as a type of nonmetal, on account of their descriptive chemistry. The thousands of words give an overview of the descriptive chemistry of the transition metals and show that almost all of them are about as nonmetallic as the metalloids chemically, and yet no one wants to call them metalloids or nonmetals. So they help to critique this element of the proposal.
 * If you like, Sandbh in his overview gives one of the important questions raised by the proposal as "metalloids as nonmetals: is this kosher?". His proposal argues for "yes", and my contribution is trying to make the case for a "no". Double sharp (talk) 23:47, 19 June 2017 (UTC)

Proposal recap table

 * &#10022; = "Metalloid" retained as super category

Argument themes
The thread has four broad themes, as listed in the following sections.

The first four themes are each accompanied by a short analysis. The other sections arise more from the thread as a whole.

Category organising principle
DePiep said that the most important thing is the category organising principle; category names are a secondary consideration.

Analysis: This insightful contribution received little specific attention although there are general references to the concept of metallic character, or the like, running through the thread. Since we have never spelt out our organising principle, and so we are all on the same page, what follows is my go at it.

Broadly, our organising principle is to categorise elements having regard to metallic or non-metallic character, and cross-cutting similarities in chemistry. Here, the character of an element is a composite of such chemical properties as activity, reactivity, electronegativity, acid-base character, and cationic/anionic tendencies, as influenced by relevant physical properties. Cross-cutting similarities in chemistry refers to e.g. the similarities in chemical properties that distinguish most of the transition metals.

This organising principle is generally applied at a level of abstraction above the group level. It results in a categorisation scheme—consistent with the literature—that shows a progressive change going across the periodic table: from strongly metallic (groups 1-2, Ln, An) to weakly metallic (post-transition metals), and weakly nonmetallic (metalloids) to strongly nonmetallic (corrosive nonmetals). The progression transcends any within-category anomalies, and is accompanied by bells and whistles in the form of the noble metals, which make up a subset of the transition metal category, and the noble gases.

Boundaries
Some project members have queried the objectiveness and sharpness of the proposal's boundaries compared to the current scheme.

Analysis: Polyatomic and diatomic are quite objective. I say "quite" because "diatomic" iodine is characterised by a significant degree of bonding between the molecules in its layers, "to form intermolecular, many-centre…[sigma] bonds, spread through the layer and populated with delocalised electrons, as reflected in the properties of iodine (lustre, colour, moderate electrical conductivity" and gallium-like structure). One can address this by tightening the criterion so that it's based only on strong polyatomic or diatomic bonds but this does not tell the fully story in the case of iodine.

The distinction between (a) corrosive; and (b) non-corrosive i.e. intermediate nonmetals, is objective.

There are boundary overlaps in both schemes, just as there are boundary overlaps across our other categories. Our nonmetal and periodic table articles acknowledge this.

The most egregious boundary overlap in the current scheme would be sulfur's classification as a polyatomic nonmetal. It certainly is a polyatomic nonmetal but you have to look very hard to find the associated properties.

There is no comparably standout overlap in the current scheme. N might be a candidate however I have posted at length that this is, ultimately, not borne out in the literature. The divide between Se and Te seems arbitrary however this an outcome of the wisdom of the masses in deciding Te is a metalloid whereas Se is four times less likely to be classified as such. Double sharp and Sandbh have both further discussed why this divide is where it is. H might be another candidate for an egregious misfit however this will likely be the case wherever you put it. If you overlook some of the unique aspects of H chemistry, it is no more than a mildly electro-active nonmetal compared to the corrosive nonmetals.

Having regard to their overall nonmetallic character and their position in nonmetal activity series published by different authors, H and N are more like C, P, S and Se than they are O, F, Cl, Br, and I. They each have their idiosyncrasies but such non-standard behaviour is nothing new within each of our categories. An analogous situation occurs among the metalloids. If you drill down, you can find differences between the chemistry of the B ( metalloid) and the chemistry of Te (also a metalloid). But if you take a Mendeleev-style broad-brush approach,^ as highlighted by DePeip, there are enough higher level similarities to warrant assigning B and Te to the metalloid category, consistent with our organising principle.


 * ^ in which Mendeleev ignored atomic weight

Category names: Meaning and status
(a) Will "intermediate nonmetals" be meaningful enough? (b) Will such a name attract undue status?

Analysis:

(a) The term "intermediate nonmetal" will be no less meaningful to the general reader than the term "transition metal" or any of our other category names. "Intermediate" is plainer English than "transition". An analogy can be made with the concept of a strong metal (such as steel) and a weak metal (such as aluminium).

(b) I feel that this will not happen just as "other nonmetal" and "polyatomic nonmetal" have not become standard terms. Yes, I do see "poor metal" from time to time in unreliable on-line sources of no consequence but this term has been around since before Wikipedia and hasn't become thought of as a standard term. As noted, terms such as "somewhat", "moderately" or "mild" have been used to describe nonmetals in the literature, and the idea of progressive change in metallic to nonmetallic character, encompassing strongly metallic, weekly metallic, weakly nonmetallic, and strongly nonmetallic etc, is well documented and consistent with our organising principle. A descriptive phrase such as "intermediate nonmetal" is nothing new. There will be no "intermediate nonmetal" article, only a link to the nonmetal article. The nonmetal article will explain the sorry state of nonmetal category names, a bit like it does now, and that "intermediate nonmetal" is a descriptive phrase only.

Metalloids as nonmetals. Is this kosher?
This theme applies only to proposal Mk. 2.

Analysis: It is common knowledge that metalloids look like metals but behave like nonmetals. Our periodic table legend, the one with the super category at the end that says "Unknown chemical properties", obliges us to show metalloids as a subclass of nonmetals since the legend is ultimately organised by chemical properties, consistent with our organising principle.

There is a cognitive dissonance in the literature since metalloids are often, but not always, referred to as a third category of element yet their descriptive chemistry is generally nonmetallic.

Ontologically, Real Essentialism (Oderberg 2007, p. 97) argues that anything not a metal is therefore a nonmetal, and that this includes semi-metals (i.e. metalloids).

Thus, much of the literature is inconsistent. We can note this in the metalloid article but we need to make a decision on how to deal with it.

Accuracy, Science, Pragmatism, Educational value, and Professionalism (these five are watching me again as I type this) say we should treat metalloids as a subclass of nonmetals, consistent with the general chemistry at hand.

Other pertinent or peripheral topics
Project members discussed or mentioned the following:
 * the nature of At and Rn
 * what constitutes a metal/metallic behaviour/metalloid
 * colouring H on its own
 * chemistry trumps physical
 * the metal-normative focus of the periodic table
 * dividing the nonmetals into reactive and noble nonmetals
 * focussing too much on the elements in their free states
 * the division of the table into groups as a complement to categories
 * how nonmetals are described in the literature
 * defining a category without referring to another category
 * surpassing IUPAC standards/knocking ourselves out
 * getting rid of the colours.

Claimed advantages of the proposal/s

 * The new categories are self-evident in the way the elements involved are described, as a whole, in the literature, and they and their names are more consistent with our organising principle.
 * The resulting nonmetal categories facilitate a more symmetrical comparison with the metals.
 * The sequence of categories from left to right is more consistent with the way chemists think about nonmetals in terms of where each nonmetal approximately lies on a composite scale (or spectrum) of nonmetallic "liveliness" or "intensity" rather than in terms of polyatomic or diatomic nonmetalsa—at one end of such a scale lie the corrosive nonmetals; at the other end are found the weakly nonmetallic metalloids.
 * It is easier to grasp the sequence of metallic character going across the periodic table.
 * If project members will give me the benefit of the doubt, pending the production of a revised nonmetal sandbox article, then I claim that the proposed scheme will do a better job at weaving the cacophony of nonmetallic properties and anomalies shown by the nonmetals into a cohesive category ensemble, boundary overlaps and all.

Points of acknowledgement, or agreement with at least Sandbh, as interpreted by Sandbh

 * R8R likes the corrosive nonmetals category.
 * Double sharp acknowledges the flaws in the current scheme and that—at least etymologically—metalloids are nonmetals; and likes simple category names.
 * YBG suggested the current scheme seems to be on its way out, on the grounds that polyatomic and diatomic overemphasise physical properties.
 * DePiep says category names and meanings are secondary (that's what links are for); suggested our poly-di scheme was not a true chemistry categorisation; and, as mentioned, noted Mendeleev's broad-brush approach in working out the periodic table (ignoring atomic weight).
 * Parcly Taxel noted the "thread" that relates all of the elements in the reactive/intermediate category as similar.

What's the hold up?
The following factors represent Sandbh's limited judgement, are not all evident to the same degree, and do not necessarily apply to any particular project member/s:
 * Perceptions that the proposal falls short of a perfect solution
 * Magnifying the imperfections;^ neglecting the claimed advantages
 * Focussing on anomalies rather than taking the balcony view
 * Losing sight of (a) the fact that Wikipedia articles usually improve incrementally, and (b) the messy reality of the periodic table
 * Fear of being too bold on an issue that most chemists would not lose any sleep over and would probably welcome i.e. presenting the neglected nonmetal-category-garden based on the way nonmetals are generally described in the literature, anomalies and all—Mendeleev would never have produced his table if he worried about all of the anomalies
 * Looking at the proposal from the perspective of an experienced reader rather than a general reader
 * Incredible tension between expectations for a plain English category name v the need to avoid neologisms
 * Absence of a sandboxed proposed nonmetal article.


 * ^ I do this too and found myself somewhat foolishly dissing proposal Mk. 3 for this reason

What next?
I intend to sandbox a proposed nonmetal article so that project members can see the substance of the proposal rather than just its outline. The thread has given me much to draw upon, including the excellent pure chemistry observations by Double sharp. Thank heavens for our project and the wisdom of our members.

Crazy things I've called, or thought of calling, the other nonmetals
-- Sandbh (talk) 05:19, 19 June 2017 (UTC)

Comments by DePiep
I suggest to add to : the inability or lack of want to structure and organise this discussion. No effort has been made (nor proposed) to condense the arguments, draw an intermediate conclusion, bring in a logical reasoning. So here we are, a 500.000 byte Talk and growing, no limits in sight. The sheer volume and unstructured sectioning renders the talk a continuous reiteration of (some) arguments. -DePiep (talk) 11:08, 19 June 2017 (UTC)
 * I beg to differ.
 * The structure pervading the discussion is given by the four themes.
 * It is impractical IMO to restructure the discussion since there a bits of themes all over the place and being able to trace thought as it develops over time would become quite challenging.
 * The four themes largely eliminate the need to reiterate arguments.
 * Matters to ponder include the accuracy of the organising principle; the ever-presence of boundary overlaps regardless of the scheme; category names and the general reader; and the bugger's muddle that is the treatment of the status of metalloids in the literature.
 * One intermediate conclusion is that different project members appear to agree with different aspects of the proposal/s.
 * Blame me for the other intermediate conclusion that we need a sandbox so we can see how the whole thing comes together, including whether the claimed advantages stack up, before we can make decisions about these ponderous matters.
 * I expect we'll soon be able to archive the thread up to section 2.12 -- Sandbh (talk) 12:59, 19 June 2017 (UTC)
 * I would personally say 2.10, since some things from 2.11 and 2.12 are still actively being addressed. Double sharp (talk) 13:43, 19 June 2017 (UTC)
 * Cannot 'discuss' a 500.000 size 'talk'. (says it all: even here you re-gorge arguments).-DePiep (talk) 21:11, 19 June 2017 (UTC)

DePiep (talk) 00:26, 20 June 2017 (UTC)
 * Final call: why cannot you both FA-making editors make this a sound discussion,  ? Already R8R and YBG have bee chased away. -DePiep (talk) 21:49, 19 June 2017 (UTC)
 * I understand that R8R has RL commitments and that he contributes when he is able to, and is still following the discussion. Sandbh (talk) 02:06, 20 June 2017 (UTC)


 * Today, this page is 557k. This 'discussion' is suffocating all other WT:ELEMENTS talks. By size reasons only, I am excluded. -DePiep (talk) 21:23, 2 July 2017 (UTC)

Comments by YBG

 * I have not been "chased away". Rather, I find I have less wiki-time now than during the mega discussion. DePiep does have a valid point that the discussion is overly large. I respect DP's attempt to improve the discussion by refactoring it, but I also respect the push back that resisted the refactoring of others comments. I appreciate the time that S&BH is putting in to summarizing the discussion. I presume that the end of that effort will be a short, concise and self-contained summary. As soon as the other participants agree that their points of view have been adequately captured, then the previous discussion can be archived off leaving us with a more concise and organized discussion. This works well for me with my wikitime limitations.(Clarifying superscript text added YBG (talk) 17:41, 20 June 2017 (UTC))
 * My one suggestion to S&BH would be to make sure that the summary is self-contained. For example, just reading the summary, it seems that some context is missing on the specifics of the proposal(s) being discussed. To keep things concise, I think we don't need any but the barest details of earlier proposals that S&BH has abandoned without anyone else picking them up. It might do well to completely eliminate all reference to such vestigal, unless of course there is any concern that someone coming to the discussion later in the game might think "But what about X", where X is one of these vestigal proposals.
 * One other thing. While it might be interesting to read the sandbox nonmetals article, I'm concerned about the implication that such an article would replace the current nonmetal article. I for one would object to removing all reference to the poly/di/monatomic division. Perhaps we should have articles classification of nonmetals by reactivity and classification of nonmetals by atomicity where atomicity is my attempt the property whose variables are "polyatomic", "diatomic", and "monatomic". (If there is no doubt a perfectly good term already for this, please let me know.)
 * A much better "touchstone" for decision making would be the presence of sandbox articles for the proposed categories. I note that much water has passed under the proverbial bridge since our mega-discussion, but polyatomic nonmetal and diatomic nonmetal are just redirects and monatomic nonmetal is not even a redirect. I think any new categorization should do better at fleshing out the category articles than that. I'm willing to hear arguments to the contrary, but my instinct is that we really should have such category articles in the pipeline before we institute a new categorization scheme.
 * I would still like to see the discussion organized on a step-by-step basis along the lines I set out above
 * I presume we have a consensus to treat the noble gasses as a separate subcategory of nonmetals.
 * If this is not the case, please ignore the rest of this outline
 * If this is the case, continue to the next step
 * Is there a compelling reason to abandon the polyatomic/diatomic division?
 * If we reach a consensus of "NO", skip to the final step.
 * If we reach a consensus of "YES", continue to the next step.
 * Is there a compelling reason to subdivide the non-noble-gas nonmetals?
 * If we reach a consensus of "NO", skip to the final step.
 * If we reach a consensus of "YES", continue to the next step.
 * What is the best way to subdivide the non-noble nonmetals?
 * Given the decision about subdividing the nonmetals(i.e., a consensus of (a) "NO" from #2, (b) "NO" from #3, or (c) a particular subcategorization from #4), does it make any sense to consider the metalloids to be a subcategory of the nonmetals? after original edit, I added the clarifying superscript text YBG (talk) 05:25, 20 June 2017 (UTC)
 * Under this discussion scheme, I don't think we should proceed to step N+1 until we have a consensus at step N.
 * We clearly have consensus on #1 (Keep NG). Regarding #2, I originally thought there was a consensus to scrap mono/di/poly, but so far the only categorical statement is a NO by.
 * If we were to follow this scheme, we should not proceed to #3 until we have consensus on #2.
 * Now I recognize that this is merely my preference for how the discussion ought to proceed. But because of my lack of bandwidth, I really don't have enough skin in the game to insist on my preferences. So I am content to follow the discussion as best as I can.
 * Thanks to one and all for their contributions to this little corner of WP.: YBG (talk) 23:09, 19 June 2017 (UTC)
 * See: arguing again, not organising the discussion. -DePiep (talk) 00:26, 20 June 2017 (UTC)
 * I do believe you have gone too far in chastising the other participants and harming the very arguments/discussions you were trying to organise. I suggest you sit out. I myself have had enough of all this wall of black and am only waiting to improve the silicon/phosphorus/sulfur articles in the future. Parcly   Taxel  01:22, 20 June 2017 (UTC)

Hi, The nonmetal sandbox article would not replace the current nonmetal article. The sequence of events would be (1) complete sandbox; (2) seek views of project members; (3) subject to views of project members, post a formal proposal at Template talk:Periodic table; (4) seek consensus. Sandbh (talk) 02:02, 20 June 2017 (UTC)
 * Question prompted by Sandbh's response: Do you think that your new categories would have "enough within-group similarity and enough between-group dissimilarity so that each group could be the subject of a separate encyclopedia article"? If not, is the categorization enough to justify violating rule #3? If so, would it be too much to ask for a sandbox of the article before we decide to implement a new categorization scheme? YBG (talk) 05:12, 20 June 2017 (UTC)
 * Given the dissimilarity of the nonmetals, I am not too sure that this can be done very well; the main difference I tend to find is between the 2p elements and the ones below them, and then the tendency towards metallicity creeps slowly enough that that there's not much dissimilarity between the groups. Then you have some funny cases, since GeIV is more similar to SiIV and GeII is more similar to SnII (I need to check how the pnictogens and chalcogens split and get to know them a little better). I can see a metal-metalloid-nonmetal trichotomy on the basis of which properties dominate (or neither, for the middle case), but splitting the p-block nonmetals seems a daunting task. The current grouping does at least cappture the main distinction in the hyperelectronic elements; carbon is a bit different and fits reasonably with phosphorus and selenium. Sulfur is however a troublesome intermediate case, forming discrete octasulfur molecules instead of giant covalent lattice structure, though at least this makes a significant amount of sense. So I find the idea of using structure as a guide reasonable; even if it does not work directly from chemistry, it correlates with it well. Whereas putting N and P together makes me wonder just how wide the net is being cast. Double sharp (talk) 05:51, 20 June 2017 (UTC)
 * If the proposed recategorization will be into categories that we don't ever expect to merit a WP article, then I say, let's not recategorize nonmetals in that way.
 * If polyatomic nonmetal and diatomic nonmetal really are in fact WP:Redirects with possibilities, then mark them with R with possibilities and expand them into articles ASAP and keep the current categorization (i.e., #2 NO).
 * But if we don't ever expect them to become WP articles in their own right, then I say let's scrap the current nonmetal categorization (#2 YES) in favor of just keeping H,C,N,O,F,P,S,Cl,Se,Br,I all together (#3 NO) so that we divide nonmetals into two mutually exclusive and jointly exhaustive subcategories: noble gases and chemically active nonmetals (or some other synonym). YBG (talk) 06:16, 20 June 2017 (UTC)

The Well of Wisdom

 * Re addressing your fine criteria of clarity, unambiguity, and meaningful, please see my earlier analysis here. (Yes, I believe the proposals have the potential to meet all three criteria, noting Double sharp's concerns).


 * I think I treat the expressions polyatomic nonmetal, diatomic nonmetal, intermediate nonmetal, reactive nonmetal, and corrosive nonmetal as descriptive phrases rather than formal terminology in the sense that there is no well-documented concept in the literature of treating any of them as taxonomic categories with multiple shared characteristics.


 * As such they don't warrant articles.


 * But they do provide a convenient way of labelling, structuring, describing, and explaining knowledge about the nonmetals in a way that is more (as per my proposals) or less (current scheme) consistent with the well-documented progression in metallic to nonmetallic character going across the PT, including fundamental notions of weaker, stronger, and almost inactive nonmetals


 * The nonmetal article is the place where this broader context can be provided, accompanied by the descriptive phrases in question, as I have attempted to show in the current version of this article.


 * As noted, I intend to provide a sandbox of a revised proposed nonmetal article, for consideration by project members, before seeking consensus for implementation. No consensus = no implementation.


 * I was recently in the US and among other things visited the Beat Museum, in San Francisco, where I purchased a Jack Kerouac fridge magnet, that says, "I have nothing to offer anybody except my own confusion". I feel like Jack now. Having said that—¡Ay, caramba! Pass me the headache tablets!—I think I am happy with my answer. Sandbh (talk) 00:43, 21 June 2017 (UTC)


 * So you are saying these terms are merely descriptive phrases and not RS-recognized formal terms. Consequently, they might reasonably be the subject of a subsection of, but it would be over-the-top to expand them out into a full-blown encyclopedia articles. (As an aside, I now finally think I understand what you are intending to sandbox!)
 * But aren't you acknowledging that these categories don't fully meet the meaningfulness criterion?
 * 3. Meaningful. The categories should have significance more than just dividing for the sake of dividing. There should be enough within-group similarity and enough between-group dissimilarity so that each group could be the subject of a separate encyclopedia article.
 * If the categories aren't significant enough to merit separate encyclopedia articles, then IMHO, we are precariously close to "dividing for the sake of dividing".
 * Now I would be a bit more lenient using categories defined by mere descriptive phrases if there were sharp dividing lines between the two categories. Thus, I may be willing to accept the diatomic/polyatomic distinction. But when a property takes on values that vary continuously without huge defining gaps, then I'd say we are "dividing for the sake of dividing".
 * One could, I suppose, separate Hydrogen from the rest of the group because it is SO different in so many ways - rather like Donald Bradman, but IMHO having a subcategory of one is sufficiently problematic that it should be avoided if at all possible. YBG (talk) 03:08, 21 June 2017 (UTC)

Yes, the category names are merely descriptive phrases based on those used in the literature and not formal taxonomic nomenclature found in the same literature. It would nevertheless be reasonable to mention them in an article which provides the context for their application.

I forgot about that part of your criterion that says, "...could be the subject of a separate encyclopedia article."

In my view this part of the criterion is not necessarily essential. We should not be prevented from using descriptive phrases to label, structure, describe, and explain knowledge in a way that is consistent with the literature.

I feel that we are NOT precariously close to "dividing for the sake of dividing". The fundamental notion of weaker, stronger, and almost inactive nonmetals, as a result of the progression in metallic to nonmetallic character going across the PT, is well documented in the literature. The simple binary distinction between corrosive nonmetals and the rest of the non-noble nonmetals is sharp. The correlation with high electronegativity, oxidising power, simple anion formation, and activity, is marked.

The gap between the polyatomic nonmetals and diatomic nonmetals is distinct but partly superficial. As I have observed, the validity of the gap when it comes to assigning sulfur to the polyatomic nonmetals, and iodine to the diatomic nonmetals, in terms of measuring non-metallicity, is problematic.

Sir Donald Bradman is indisputably, and by a wide margin, the greatest batsman of all time. He sits in a category on his own. Overall, Bradman might be to F as F is to I.

As a cricketer I would describe H as a mediocre batsman with an unorthodox style, some aspects of which are unique. Sandbh (talk) 10:27, 21 June 2017 (UTC)
 * The distinction would likely look sharper if nitrogen was not there sitting in the middle of it. As you have noted, its high electronegativity does not always count for much, but I can think of two striking cases where it does:
 * Nitrogen compounds (e.g. NH3) show very significant hydrogen bonding. If it was around the same level of strength as C or P, one would not expect to see this. In this respect it seems that N is a notch ahead of even Cl.
 * Like O and F, but markedly unlike the nonmetals in the third row and below (Cl and Br taking an intermediate position), it is very difficult for nitrogen to adopt a central position in a multicentre bond, because it attracts the electrons too closely towards itself. This is famously not a problem for Be and B, and even C is okay with it in carbocation rearrangements.
 * Incidentally I will note that iodine, which tends to find a great deal of happiness in positive oxidation states, seems to be halfway already to being a "weak nonmetal". It succeeds in slaying gold, yes, but (1) it is not as good as doing it as cyanide and (2) that's due to the formation of a complex, presumably [AuI4]− (I'm on a train and writing it from memory). In your sandbox you give a table of electrode potentials: I wonder what the authors were thinking of by including fluorine, ozone, and perbromates but not hyponitrous acid, when none of the four will get saved by the overpotential effect! Yes, N only shows its true strength when O and F are also around and it is pushed into a positive oxidation state, but then the same is also true of bromine and iodine in aqueous solution, isn't it?
 * So if iodine is in a strange position, perhaps that is a good thing; and sulfur seems intermediate as it should be between the diatomic structures of H, N, O, and the halogens and the giant covalent structures of C, P, and Se (and white phosphorus similarly looks intermediate between the diatomics and S). And I still think that the "too-reactive" first approximation for the behaviour N given by the current scheme is better than the "not-as-reactive-as-it-really-is" one given by the proposed one for the above reasons. Double sharp (talk) 10:55, 21 June 2017 (UTC)

As noted, I think generalisations are usually more important than anomalies, as per how Mendeleev ignored atomic weight. I cannot do anything about the fact that the literature describes (appropriately, in my view) the non-noble nonmetals in terms of weaker and stronger overall nonmetallic character, rather than their polyatomic and diatomic structures, and that chemists think in terms of the former rather than the latter, and that the rest of our categories correspond more closely to overall metallic or nonmetallic character, rather than structure.

On nitrogen, and its idiosyncratic nature, I have previously said all I need to say for now, pending further research, here.

Oh! I'll look again at the table in my sandbox as to why ozone is included, since I would't have regarded ozone as a stable species in aqueous solution. Sandbh (talk) 12:28, 21 June 2017 (UTC)
 * Peculiar. Inorganic chemistry (Wulfsberg 2000, p. 247) shows ozone as a thermodynamically stable species in aqueous solution "at the potential indicated" i.e. 2.08–3 V. Below 2.08 V the stable species is O2. Does that sound plausible? Sandbh (talk) 12:40, 21 June 2017 (UTC)
 * It is more stable in alkaline solutions: the half-life of ozone in 1 M NaOH is about 2 minutes, but that in 5 M NaOH is 40 minutes and that in 20 M NaOH is 83 hours (Greenwood and Earnshaw, p. 609). Mind you, in alkaline solution the standard reduction potential is only +1.246 V, which is really not that impressive given that it is actually approximately equal to that for the reduction of O2 to H2O.
 * BTW I can think of another problem with the table, which is that Xe would show up as a strong nonmetal. Since iodine is calculated from periodic acid, iodate, and iodine, I'd compute Xe from perxenic acid and xenon trioxide, getting (2.42 + 2.12)/2 = 2.27, greater than oxygen(!). (It is significantly milder in alkaline solutions, but since everything was calculated from the acidic values, this should be fairer). Meanwhile XeF2 ends up at +2.32 V, at the same ballpark; putting it in hardly affects the average (now +2.29 V). And even more impressively: for Kr we only have one figure, that of KrF2, because no krypton oxyacids or oxides are known. That gives it an electrode potential of +3.27 V, even higher than fluorine. Clearly something needs clarification if Kr and Xe are not to be called strong nonmetals by this measure. I realise that the Kr species is obviously not very stable in aqueous solution, but the Xe species seem no worse than elemental fluorine (which is included), and it makes the non-inclusion of hyponitrous acid look all the stranger; including it gives N a value of +1.39, just above Cl as I would have expected. (I also think the inclusion of perbromate, despite its unwillingness to exist, explains why Br is strangely placed above Cl; as you can see, I find the inclusion criteria rather unsatisfactory!) Double sharp (talk) 15:02, 21 June 2017 (UTC)
 * I hope you saw my email about ozone. Xe has an average standard reduction potential of 2.19, placing it ahead of oxygen. The noble gases also have ionisation energies and electronegativities that go off the top of the scale yet no one regards them as being "strong" nonmetals. Rather, they seem to be regarded as about what you would expect to find between group 17 and group 1, somewhat like Dr Edward Teller shaking hands with his antimatter equivalent. If metallic/nonmetallic character is a composite of multiple factors then the noble gases appear to represent an example where indicative factors like e.g. electronegativity, ionisation energy, and standard reduction potential get swamped by one or more other (counter)factors. I don't know how you would count electron configuration as a factor here but that seems to play role. Sandbh (talk) 23:43, 21 June 2017 (UTC)
 * I find that group 18 is far more allied to group 17, to be honest; all of its interesting chemistry is hypervalent, as is all of that of the halogens that isn't that of the halide anions. Whereas you will not find any alkali metal chemistry in oxidation states +3 and above (except I guess maybe in eka-francium, but that's not important). In general, noble gas chemistry is rather like halogen chemistry in even rather than odd oxidation states.
 * Actually, it is interesting to ask why xenon for instance is not considered a "strong" nonmetal. Most of its physical properties are the epitome of nonmetallicity, and the chemistry is pretty much what you would expect in each oxidation state extrapolating from tellurium and iodine. I think the main reason is probably that its main oxidation state is 0, which is frankly probably the single distinguishing property of this strange group. At the very least it tells us that the properties we like to refer to are not so useful for group 18, because they only work when the elements are in compounds, which is not something they particularly like doing. (Even nitrogen appears quite happy to be in compounds like the famous ammonia and nitric acid, for example.) But this raises another question, because if reduction potentials can be swamped like this for group 18, then perhaps they can be swamped elsewhere. And it does not address my concern about the choice of species being considered; I shall look at your email on ozone as soon as I can (later today), but it seems to me that the inclusion of perbromate for example is still fairly questionable. Double sharp (talk) 05:20, 22 June 2017 (UTC)
 * AFAIK the perbromate ion is a stable species in aqueous solution (mind you, I haven't checked the literature) so I'm not sure why its inclusion is questionable. I'll wait to see what you think about that ozone e-mail. Sandbh (talk) 08:14, 22 June 2017 (UTC)
 * I find it a little questionable because perbromate suffers the same problem you point out with many nitrogen species: it is a great oxidising agent on paper, but it is extremely sluggish, almost as inert as perchlorate, due to kinetic factors. So if the electronegativity of N is taken as misleading, than so is the redox potential of perbromate: it certainly requires extreme oxidising agents for its chemical production, like F2 or XeF2, but once you make it it is rather disappointing. Double sharp (talk) 08:26, 22 June 2017 (UTC)
 * P.S. Interesting thing about ozone I see in the email (from Prof. Wulfsberg, at that!). Still, given that it is unstable at more normal voltages I still wonder why things like it and F2 are included. Double sharp (talk) 14:24, 22 June 2017 (UTC)
 * We could remove ozone and perbromate from the average standard reduction potentials (although I'm not saying that such exclusions are necessarily justified). Even so, and once removed, there is a still a reasonable gap to nitrogen, so I'm not sure anything has changed. I presume F2 is included only for comparative purposes. Here is a quote that may shed some light on the subject:
 * "Very little of the chemistry of the group 15 elements in that of simple ions…nearly all the chemistry of the group…involves covalently bonded compounds. The thermochemical basis of the chemistry of such species is much harder to establish than that of ionic compounds. In addition, they are much more likely to be kinetically inert, both to substitution reactions…and to oxidation or reduction when these processes involved making or breaking covalent bonds, as well as the transfer of electrons.' --- Housecroft CE & Sharpe AG 2008, Inorganic chemistry, 3rd ed., Prentice Hall, Harlow"
 * Sandbh (talk) 03:41, 23 June 2017 (UTC)

I can't believe I forgot about such a famous compound, but: why isn't nitrous oxide there? IIRC it is stable in water and the standard reduction potential of the N2O/N2 couple is +1.77 V, enough to push nitrogen to 1.1 and make it equal to iodine. Double sharp (talk) 10:00, 23 June 2017 (UTC)
 * The problem is that the nitrate ion reduces to nitrogen, and so does nitrous oxide. I don't see how you can count both on the same redox predominance diagram. Wulfberg's selection of the nitrate to nitrogen reduction potential may be on the the grounds of the low solubility of nitrous oxide (1.5g/L = 0.15%) whereas I understand nitrates are much more soluble. I'm wildly guessing though. Sandbh (talk) 13:05, 23 June 2017 (UTC)
 * Curious: Greenwood and Earnshaw (pp. 434–7) do not list a reduction of nitrate to nitrogen at all, giving only potentials for the couples /N2O4, /HNO2, and /NO...and neither do we on either Nitrogen or Standard electrode potential (data page). Double sharp (talk) 14:21, 23 June 2017 (UTC)
 * The aqueous chemistry of the elements, (Schweitzer & Pesterfield 2010, p. 202) says, "Many reactions of nitrogen and its compounds are controlled primarily by kinetics, and hence conclusions drawn from thermodynamically based E–pH diagrams are likely not to be in accord with experiment. Taking this into account, nitrogen has not been introduced into [its E–pH] diagram, and even though nitrogen may be liberated in certain aqueous reactions, the oxidation or reduction of nitrogen in water is kinetically hindered." Sandbh (talk) 23:17, 23 June 2017 (UTC)
 * The same is true of and  too, though. Double sharp (talk) 09:36, 26 June 2017 (UTC)
 * For agree; for  only if we are talking about a dilute solution. In any event, whereas such behaviour is generally characteristic of N and its compounds, as per Schweitzer & Pesterfield, and the blockquote from Housecroft & Sharpe (a few paragraphs ago), this not the case with Cl and Br generally. Sandbh (talk) 03:29, 27 June 2017 (UTC)
 * It does seem to be true for Cl in its most stable oxidation states of −1 and +7, though. What seems to be going on is that Cl is fairly small, compared to Br and I, and thus perchlorate vs periodate is kind of the same situation as CCl4 vs SiCl4. (The Cl vs Br difference is minor. I think the 3d contraction dampens it by making Br not much larger than Cl, but also keeps it alive because Br is much less happy to be in the +7 state than Cl.) Nitrogen, being of course even smaller than chlorine, could very plausibly be in this situation, which explains the kinetic hindrance when it is surrounded by more electronegative O and F atoms (which we must remember are not very much smaller than N). So I get the feeling that N is in about the same situation as H in group 1, where the quantitative differences are vast enough that a few turn qualitative because of overly small size, but with no other group looking as appealing and giving the right prediction. (Physically speaking nitrogen is hardly an intermediate nonmetal in the sense of carbon or phosphorus, after all.) Double sharp (talk) 06:53, 28 June 2017 (UTC)

Re compelling need to subcategorize nonmetals
Is there a compelling need to subcategorize the non-noble nonmetals? There are only eleven of them (H,C,N,O,F,P,S,Cl,Se,Br,I), right? It seems obvious that the seven noble gases should be in a separate category, but do we really need two separate categories for the rest of the nonmetals? If memory serves, most of this discussion has revolved around the advantages or disadvantages of particular subcategorizations. I can recall only one argument about the advantages of subcategorization in general, namely, that it would better show that gradual nature of the metallicity trend across the table. (Please remind me if I have forgotten another argument.) Frankly, that seems to me to be tenuous basis to justify subcategorization. If that is our primary reason, why not agree on a single metric for metallishness and then color each element a particular color accordingly along a more-or-less continuous spectrum? YBG (talk) 05:51, 23 June 2017 (UTC)
 * Sanbdh's opinion
 * Yes, there is.


 * We are obliged to follow or accommodate the literature.


 * The theme of a metallic to nonmetallic progression across the periodic table pervades the literature, including the stark contrast between the alkali metals and the halogens/noble gases etc, as set out below in the average view section, items 2, 3, 4, and 5. One result is a "stronger-noble-weaker" trichotomy of the nonmetals (ignoring the metalloids, for now) encompassing the halogens, noble gases, and the rest of the nonmetals.


 * There is no single representative measure of metallicity or nonmetallicity at the individual element level. It is only feasible to attempt to reach agreement at the broader category level, having regard to the way the literature has attempted to categorise the elements. Sandbh (talk) 11:27, 23 June 2017 (UTC)
 * (By the way, my suggestion to color each element separately is almost entirely tongue-in-cheek).
 * OK, granted, we ought to follow or accomodate the literature. The difficulty here, I as understand it, is that the literature does not give us a single answer to our question; otherwise, we would simply follow it and be done.
 * At any rate, I would appreciate hearing others chime in on my question: is there a compelling reason to subdivide the eleven non-noble nonmetals? YBG (talk) 05:42, 26 June 2017 (UTC)
 * There is a single answer to our subdivision question: halogens, noble gases, other nonmetals. We cannot adopt this completely since we made a decision to count astatine as a metalloid, and to have something better than "other nonmetals". But we can get quite close (much closer than our poly-di scheme). Sandbh (talk) 06:33, 26 June 2017 (UTC)


 * R8R's opinion
 * I rather argue that no, there is not. Just as well, I will point at the literature, which we indeed are obliged to follow.
 * From what I have seen so far, there has indeed been a strong theme of the strong and the not so strong nonmetals. Yet the idea of this not being a criterion for classification (which we are aiming for) is just as common. From the major books off the top of my head, Wilberg and G&E both recognize the concepts and neither uses them as the line for the categorization they'd use; I genuinely don't know a single source that would use this sort of classification as the main classification. I believe that this is a general observation rather than the ultimate classification and shouldn't be used for the latter purposes.--R8R (talk) 10:48, 26 June 2017 (UTC)
 * I partly agree.


 * Most general chemistry books start with (a) the broad descriptions, and then go into (b) the detail later on. The two go hand in hand: (a) gives the context to (b). The broad descriptions are in terms of the contrast between the alkaline metals and the halogens/noble gases etc. The detail later on, when they get to the nonmetals, is group by group. I presume "group by group" is what R8R means by the ultimate classification.


 * We follow this line too: our main table shows the broad categories.^ Our big table adds the group names.


 * ^ The Encyclopedia Britannica article on the Periodic table of the elements leads with a similar table, in which the nonmetals are categorised as other nonmetals, halogens, and noble gases.


 * Both cases represent approaches to classification i.e. the structuring of knowledge based on shared characteristics. Sandbh (talk) 03:59, 27 June 2017 (UTC)


 * Sorry, I've missed this one.
 * Fair enough. As for broad vs. detail: yes, I agree. That, however, still leaves me thinking that "this is a general observation rather than the ultimate classification and shouldn't be used for the latter purposes." Compare with the rest of categories we have: alkali metals/AEMs are clearly defined; transition metals/lanthanides are almost as clear, having uncertainties only at the very borders but they still have grasp defining concepts behind them (incomplete d/f shells; and, for example, whether the complete d shell of group 12 still counts is an issue that still revolves around this grasp concept). In contrast, weak/strong nonmetals are, by definition, fuzzy concepts and require something else to complete the definition (phosphorus is weak/intermediate only because we can compare it to fluorine and chlorine and see the difference). Which, to my understanding, is a clear negative answer to the question in the beginning of this section: "Is there a compelling need to subcategorize the non-noble nonmetals?". (Of course, this only shows why we don't need a subcategorization into strong and weak nonmetals or the like but IIRC there hasn't been any sign of a discussion on a basically different principle.)--R8R (talk) 16:04, 2 July 2017 (UTC)

YBG's rephrasing of the question
If you'll allow me a little bit of literary license, I will anthropomorphize the nonmetals and rephrase my question. Instead of asking "Is there a compelling need to subcategorize the non-noble nonmetals" let us ask "Have any subgroups of the non-noble nonmetals (NNNM) banded together and demanded of us, 'Hey, we're special! We need to be colored differently from those (ugh) other nonmetals! We're special! We DEMAND to have a separate encyclopedia article written about us!"
 * This calls to mind my 2012 edit summary which resulted in 's comment preserved at

Well, OK, I didn't really need to use anthropomorphism. I could have said, "Do the literature indicate - yea demand - that any subcategories of the NNNM rise to the significance as to require treatment as a separate subcategory?" Of course, one of the marks of such an indication is the presence of enough material in the literature so that we could create a separate encyclopedia article. (Sorry to harp on this,, but I really think it is a key point.) Now as to 's statement that the best subcategorization is noble-gasses/halogens/others, I would say that this criterion is only partly satisfied. There is certainly enough info for an article about the halogens (in fact, we already have one), but I don't think there is enough info about the others (H,C,N,O,P,S,Se) to merit an article about them collectively. The very fact that there isn't a particularly good name for them is suggestive (though certainly not determinative). So let me take a stand with in favor of NOT subcategorizing the NNNM. This does not eliminate the need for WP to reflect the what the literature has to say, nor does it mean that WP cannot reflect a distinction (a) between the halogens and the NHNNNM (non-halogen-non-noble-nonmetals) or (b) between the corrosive NM and the reactive-but-not-corrosive NM. It merely says that we have made an editorial choice to use something other than our categorization scheme to reflect these distinctions. YBG (talk) 03:34, 3 July 2017 (UTC)
 * Sandbh's response
 * Ultimately, we have no compelling need to do anything except, I suppose, do our best at summarising what is known about the elements in some kind of organised manner.


 * The average view of the literature when it comes to nonmetals is that there are noble gases, halogens, and these other nonmetals.


 * We cannot reflect this in our categorisation scheme since we decided quite a while ago to colour astatine as a metalloid.


 * The nearest we can get to the literature is a one-move adjustment in which O is moved out of the other nonmetals, and is placed with F, Cl, Br, and I, consistent with the resulting five elements collectively representing the most chemically active nonmetals. Cue quotes from the literature.


 * Our current categorisation scheme, in contrast, involved a three-move adjustment that saw H, N, and O transferred out of the other nonmetals, and co-located with F, Cl, Br and I to form a new diatomic nonmetal category.


 * Looking back, that was a "seismic" move that resulted in a major misalignment with the literature.


 * Respectfully, it seems to me that arguments about an ultimate classification, a compelling need, and article worthiness distract us from our goal of doing our best at summarising what is known about the elements in an organised, internally consistent, and literature-consistent manner. Sandbh (talk) 06:32, 3 July 2017 (UTC)


 * Double sharp's suggestion: At as a halogen
 * Question: what's wrong with calling At a halogen, while making reference to its being more metallic than you might expect? We already have no problem calling Bi a pnictogen and Po a chalcogen. Doesn't "halogen", at least according to IUPAC and most of the chemistry community, just mean exactly "elements of group 17"? If astatine is the one reason we cannot simply go back and use the old scheme, then perhaps we are putting too much weight on elements like At and Ts at the expense of those that actually matter more.
 * Given how iodine appears to be happier to form cationic species in ultratrace solutions in about the same way that astatine does, and how minuscule adjustments to the calculations suggests different structures of bulk astatine (coming down to whether it is molecular or metallic at all), perhaps it is not so bad an idea to call astatine a halogen and leave its physical metallicity in doubt, and realign ourselves with the literature in this way. Double sharp (talk) 07:49, 3 July 2017 (UTC)
 * YBG's response
 * 1. Actually, we DO call At a halogen, by virtue of putting it in the same column as the other halogens and by including it in the halogen article. So, to the extent that the halogens are special, that is already represented in our PT by virtue of them being in the same column.
 * 2. The question is, do we need to indicate their specialness in a second way, by assigning them a special color?
 * 3. As far as compelling need and article worthiness, I would say that these questions are what protect us from giving undue weight and undue emphasis to things that are not given corresponding emphasis in the literature.
 * 4. Being assigned a separate color on our periodic table color scheme is a very weighty form of emphasis and if the literature have not used this as a form of classification and have not provided us with enough information to form a separate article, then perhaps we should not award this level of emphasis. YBG (talk) 17:22, 3 July 2017 (UTC)
 * Well, when we say that it seems obvious to keep He, Ne, Ar, Kr, Xe, and Rn together, that is of course partially because most of the literature treats them as such. Since most of the literature additionally treats F, Cl, Br, and I together, and invariably appends any small discussions of At to that context, it seems to me that there is justification for calling out the halogens like we do the noble gases. Perhaps the situation may be a little different in groups 13 through 16 where division along the lines of metallicity has also been done. Double sharp (talk) 23:24, 3 July 2017 (UTC)

YBG's useful table
Is the following table useful?
 * ^ The term "halogen nonmetals" indicates that At is excluded and only nonmetallic halogens are in the subcategory.
 * That is rather clever if, potentially, a little confusing. It may just be clever enough to outweigh this potential confusion, however. Sandbh (talk) 04:56, 4 July 2017 (UTC)

If not, I'll gladly revert. YBG (talk) 02:01, 4 July 2017 (UTC)


 * Sandbh's response
 * I think that not showing a tripartite colouring of the nonmetals would simply obscure knowledge that is set out in the literature in the form words, pictures, descriptive phrases, labels, colours, symbols, broad, fuzzy, or specific concepts, or a mixture of all of these.


 * Our job is to codify this knowledge.


 * Here are a few more specific comments in response to YBG's contribution of 17:22, 3 July 2017 (I numbered the applicable paragraphs for ease of response).


 * 1. Agree. That is the beauty of our current approach, which follows the top-down approach set out in the literature. First there is the table organised by progression in metallic to nonmetallic character. Then the big table overlays the groups, including the pnictogens, chalcogens, and halogens.


 * 2. No. The first table is for metallic and nonmetallic character. The second table overlays the groups, as per the literature.


 * 3–4. The metallic or nonmetallic character of the elements (and the periodicity of their properties, as overlying or underlying themes) pervades the literature. I am sorry that the focus of the literature when it comes to the forgotten post-transition metals, the metalloids, and the other nonmetals, is not as good as that which applies to our other categories. I nevertheless feel we should aspire to represent the lay of the periodic table landscape in the most helpful literature-consistent manner possible. Colours are a wonderful way of doing this, by representing, abstracting or codifying what is said in words. In the case of the weaker nonmetals there is enough information to form a separate article but no compelling need to do so, as opposed to mentioning them in a broader article. Sandbh (talk) 02:22, 4 July 2017 (UTC)
 * YBG's 1st request for clarification
 * I notice that you numbered my paragraphs 1-4 but your responses are only numbered 1-3. I'm not sure what to make of the discrepancy. YBG (talk) 03:59, 4 July 2017 (UTC)
 * I mean to say 3–4, rather than 3. Sandbh (talk) 04:48, 4 July 2017 (UTC)
 * I too am not entirely sure which points are responding to which of YBG's points. And it seems to me from the majority of large descriptive inorganic chemistry texts that the starting point is the periodic table (and hence the groups), before showing how these imply more or less metallicity for the individual elements involved. Double sharp (talk) 04:14, 4 July 2017 (UTC)
 * This a chicken or the egg phenomenon, I reckon. AFAIK the great majority of authors discuss metals, nonmetals, and trends in metallic character, and the contrast between the alkali metals with the halogens/noble gases. Then they drill down into the groups, including the relatively obscure ones like groups 13, 14 and the pnictogens. Sure, the PT is in the background of the first discussion just as is the case with out PT article. Sandbh (talk) 05:14, 4 July 2017 (UTC)
 * Exactly; the PT is introduced first. This allows one to use electronic structure as a theoretical basis, especially valence electron configuration, and derive from them plausible interpretations of the chemical and physical properties we see. Even the idea of comparing the alkali metals, halogens, and noble gases derives from keeping in mind the common configurations of ns1, ns2np5, and ns2np6 (or 1s2) respectively. So we're still starting from the groups as a basis, aren't we? Double sharp (talk) 12:05, 4 July 2017 (UTC)
 * Yes, superficially. In most general chemistry books you get the "superficial" overview section at the front, then the physical chemistry, then the descriptive group-by-group chemistry when the groups and the "Aha!" of the PT come to life. Sandbh (talk) 07:08, 5 July 2017 (UTC)
 * So we start from the groups and then we end with the groups. Surely that if anything would show their importance over metallicity considerations? Double sharp (talk) 14:13, 5 July 2017 (UTC)

Parcly Taxel's suggestion: Eleven undivided nonmetals ,  I do not see any need to split the non-noble nonmetals, as  explained in the quote I put up in the grand summary above. Leaving the group of eleven nonmetals undivided allows us to do this for colouring the periodic table: which is far neater due to the "of which" construction. (Astatine remains a metalloid.) Then in the nonmetal article we should have So the halogens are given some importance, but not too much as to beg the astatine question. Sounds like the best compromise to me. Parcly  Taxel  05:03, 4 July 2017 (UTC)
 * Nonmetals (one colour)
 * of which noble gases (another colour)
 * seven sections detailing the seven non-group 17/18 nonmetals individually
 * an eighth section on the four halogen nonmetals, detailing each element individually
 * a ninth section for the noble gases, discussing them as a group
 * a tenth section on minor groupings (SPONCH, etc.)
 * Double sharp's response
 * I am still wondering: if we can call out the noble gases, why can't we also call out the halogens? They are one of the most homogeneous groups, along with groups 1, 2, and 18. Given how the chemistry of tracer amounts of radioiodine looks a lot more metallic and significantly different from what we see in bulk iodine, I suspect the same may have happened in astatine: given the huge change that happens to the results when spin-orbit effects are concluded, I am inclined to agree with this particular quantum chemist that we need a better study of condensed At. I would think it more reasonable to call it a halogen and leave its amount of metallicity in doubt. Double sharp (talk) 12:19, 4 July 2017 (UTC)
 * R8R's response
 * For the record: there is no saying that "halogen" is an illegitimate or insignificant category. We may well refer to this concept in articles on particular elements. (However, comparison to noble gases is not comparison of two alikes. Noble gases stand out far more. Perhaps, a conparison with alkali metals would be fair to a first approximation.)
 * As for why: because there is no name for the rest of the nonmetals. "Other nonmetals" is meaningless on its own and thus not legitimate. To my understanding, that is enough reasoning to consider it settled. Moreover, I'd say that this group alikeness does not trump metallicity; is this alikeness so important if it does not prevent such a shift in metallicity anyway? Besides, we already have one uncertainty in our coloring (we have 38 transition metals) and there is hardly any justification for introducing a second one. Internal consistency should be maintained (if we have the category "metalloid" and we agree that At is one, we should color it as one).--R8R (talk) 13:02, 4 July 2017 (UTC)
 * What this tends to suggest to me is that we already have too many colours and no matter what we do we will do something wrong; and since the groups are the most important thing anyway perhaps we should indeed get rid of all of them. Mind you, I think that not acknowledging the similarities of {F, Cl, Br, I} because At might act funky sometimes (never mind that we don't even know if it would in bulk) is a bit silly given the almost total lack of importance of At compared to the other four. It is kind of like not wanting to use the term "noble gas" because Og might not be one.
 * The more I read about At, anyway, the more I think it would be best considered to be of unknown metallicity, so that the only thing you can say about it is "it's in group 17, so it should continue the trend down from iodine in its chemistry, but we don't really know". Meanwhile, I'd rather let the consequences of the similarity speak for themselves in F, Cl, Br, and I.
 * By the way, I am not sure there is a good name for "nonmetals that are not noble gases" either that means anything by itself while remaining correct. "Reactive nonmetals" is a bit odd given how N2 and C (diamond, slightly less so for graphite) are not terribly reactive at standard conditions; "typical nonmetal" trips on H; and "non-noble nonmetal" is so bad it starts to become funny. That's why I think the best option is just to go by the groups like everybody does. Double sharp (talk) 23:39, 4 July 2017 (UTC)

YBG's 2nd request for clarification I'm not sure I follow your response to point number 2. I said:
 * 2. The question is, do we need to indicate their specialness in a second way, by assigning them a special color?

You responded:
 * 2. No. The first table is for metallic and nonmetallic character. The second table overlays the groups, as per the literature.

I have two questions about this: First, did you really mean to answer my question "No"? That would seem to be saying that "No, we do not need to indicate [the halogen's] specialness in a second way, by assigning them a special color". But that seems to run counter to the rest of your discussion. Secondly, it would help me if I better understood what you mean by "The first table" and "the second table". The gallery shows what I think you mean, but perhaps you could verify or clarify. Thanks! YBG (talk) 16:46, 4 July 2017 (UTC)
 * Sandbh's response
 * Yes, I meant to answer "no" to your question no 2. The first table is the one that first appears in the periodic table article i.e. the one you have labelled "Is this 10-color one "the second table"? The halogens are not assigned a special colour here. The second table is the one that appears second in the periodic table article, and has the overlay of the group names including the halogens. This is what I mean by a top-down approach. Sandbh (talk) 22:59, 4 July 2017 (UTC)
 * OK, so the first table is File:Periodic_Table_Chart.png and the second table is Periodic table, right? But I notice that groups are shown in both, although only by numbers in the first and by numbers and names in the second. YBG (talk) 07:53, 5 July 2017 (UTC)
 * Given that groups 1–12 are all pretty well called out even in the "metallicity" table (even with separate categories for groups 1 and 2!), it strikes me that this is not so clear-cut. I also would question the value of a "metallicity-first" view here, as above. Double sharp (talk) 23:42, 4 July 2017 (UTC)
 * , yes you have that right. Yes, while the group names are overlaid in the second periodic table appearing in our periodic table article, the bells and whistles don't make an appearance until the third table, in the Groups section. Sandbh (talk) 12:41, 5 July 2017 (UTC)
 * I was referring to the way the categories of "alkali metal", "alkaline earth metal", "lanthanides/actinides", and "transition metals" clearly follow groups 1, 2, (3), and 3–12 in the table. These in particular are very obviously group-based, perhaps even more so than metallicity-based. The noble gases (group 18) are similarly also called out as a group; in fact more elements are called out by groups than are not. Double sharp (talk) 14:58, 5 July 2017 (UTC)
 * Yes, agree! Since its inception in ca. 2002 the Wikipedia table has been based on a hybrid of individual group names, and groupings (Ln, An, TM, post-TM, metalloids, and two variable groupings of nonmetals). Until very recently our article on Names for sets of chemical elements had a note to this effect: "This English Wikipedia uses the below hybrid system in its periodic table subgroupings. These groupings are called category [sic] in this Wikipedia (without any claim to universality)." Those words or their predecessors were extant from 2008, per Mav, to 2017. Perhaps the biggest single contributor to the hybrid nature of the table was the felt need to accommodate the metalloids as a marker between the metals and the nonmetals, which left us with the challenge of how to deal with the other metals aka poor or post-transition metals, and their erstwhile counterparts, the other nonmetals. It did, however, result in beautiful, and rich, representation of the progression in metallic to nonmetallic character going across the table from left to right, as enthusiastically noted in the literature. Sandbh (talk) 23:50, 5 July 2017 (UTC)
 * But, as I have said before, the opening with "alkali metals, alkaline earth metals, lanthanides and actinides, transition metals" tells us nothing about metallicity and only about electronic structure. We never get told that metallicity is physically all right in the transition metals but chemically has long since fled, because we never get told that metallicity goes up when the oxidation state goes down: that could only be implied by the groups (and colouring VIA and VIB together, for example). So our categorisation has a very strange divide: up to group 12 it is about groups, which it is again in group 18, but in the middle we have a sudden break into metallicity. It is a hybrid indeed, but I am not sure that it is one for a good reason, given that Ta is chemically about as unimpressive as a metal as Sb. Double sharp (talk) 03:33, 6 July 2017 (UTC)
 * Hi I agree, in spirit! Pragmatically, I'm sorry that at the level of abstraction we are dealing with in our rainbow table, we cannot go into the amount of detail you have in mind. Per De Piep's astute observation, Mendeleev would have still been working on his table until his death, unpublished, had he concerned himself with all of its anomalies. I feel like we can't see the forest for the trees. It's a fab table for its high-level purpose of tracing the broad contours and flow of the period table, and transcending its anomalies, which I suspect is what the founders of our table had in mind all those years ago when they first colour coded it.


 * Re Ta and Sb, and as you know, metallic character is not purely based on chemical behaviour. Attempting to define metallic character based purely on either physical properties (like was done in days of yore) or chemical properties (like was done in early in the post-days of yore, until it started producing nonsensical results like concluding W was a metalloid) does not work. One needs to consider physical and chemical properties.


 * You also agreed with me that this was a good approximation: "In summary, and riding roughshod over some of the nuances and subtleties noted by Double sharp and DePiep, it appears that the s- and f- metals are chemically strong but physically weaker; the d- metals are chemically strong to weaker (or even noble) and physically the strongest; and that the p-block metals are chemically weak and physically weak.


 * For sure you can drill down to the individual element level and find anomalies, but these are not relevant at the level of detail that our rainbow table attempts to cover. Sandbh (talk) 05:09, 6 July 2017 (UTC)
 * It nevertheless occurs to me that you could get a better guide to this if you followed the groups. Then you could guess that elements in groups I, II, and III would be the best metals chemically, and that things would get worse progressively in groups IV, V, VI, and VII (VIII being a bit of an anomaly). Meanwhile you could guess that the metals would be physically strengthened across the A subgroups and weakened across the B subgroups.
 * Mendeleev's great idea to break out of the impasse was by looking at groups and valency: why shouldn't we go back to it, since it appears to work so well? Double sharp (talk) 05:26, 6 July 2017 (UTC)


 * re the 11 elements classification. Allow me to add a third criterium (next to: element's properties (Sandbh, Double sharp) and category requirements (YBG). I add: if it is not worth a Wikipedia article, it cannot be a category. (Example of bad: polyatomic nonmetal, diatomic nonmetal). -DePiep (talk) 21:17, 6 July 2017 (UTC)

The need to divide the 11 NNNM - A summary
Coming back to the question at hand, should we subdivide the nonnoble nonmetals (NNNM), which I believe everyone understands to be referring to our color metalicity scheme. Here is hopefully a reasonably accurate representation of editors' response to this question.
 * 16:04, 2 July 2017 (UTC)  In contrast, weak/strong nonmetals are, by definition, fuzzy concepts and requires something else to complete the definition (phosphorus is weak/intermediate only because we can compare it to fluorine and chlorine and see the difference). Which, to my understanding, is a clear negative answer to the question in the beginning of this section: "Is there a compelling need to subcategorize the non-noble nonmetals?"
 * 05:03, 4 July 2017 (UTC)  I do not see any need to split the non-noble nonmetals, as @R8R: explained in the quote I put up in the grand summary above. Leaving the group of eleven nonmetals undivided allows us to do this for colouring the periodic table: ...
 * 03:34, 3 July 2017 (UTC)  Let me take a stand with R8R in favor of NOT subcategorizing the NNNM. This does not eliminate the need for WP to reflect the what the literature has to say, nor does it mean that WP cannot reflect a distinction (a) between the halogens and the NHNNNM (non-halogen-non-noble-nonmetals) or (b) between the corrosive NM and the reactive-but-not-corrosive NM. It merely says that we have made an editorial choice to use something other than our categorization scheme to reflect these distinctions.
 * 22:59, 4 July 2017 (UTC)  I meant to answer "no" to your question no 2. [Do we need to indicate [the halogen's]] specialness in a second way [other than by group], by assigning them a special color?] (but I note that elsewhere Sandbh argues strongly for subdividing the NNNM into strong/weak where strong is halogen NM + O)
 * 12:19, 4 July 2017 (UTC)  If we can call out the noble gases, why can't we also call out the halogens? They are one of the most homogeneous groups...

This seems to say that three of us are clearly in favor of coloring all 11 the same and two of us are not willing to do so. This barest of majorities hardly counts, first because consensus is not determined by votes, and second I believe most of us would agree with DePiep who refers to S&BH and DS as "two of our best" here at WP:ELEM

Does this seem to be a fair summary of where we each stand? YBG (talk) 08:24, 5 July 2017 (UTC)

Given the respect I have for and, IMHO there are only two ways we could come to a sustainable consensus: Does this seem right? YBG (talk) 08:24, 5 July 2017 (UTC)
 * 1) One of them is convinced to come over to the 11-undivided NNNM position
 * 2) Both of them agree on a single proposal that the rest of us can live with even if it not be our first choice

Another related item, what would you call the NNNM? It is mentioned above that "reactive nonmetals" seems ill chosen considering the unreactiveness of graphite and diamonds, for example. But to me an element that exists in so many different compounds, as carbon does, can well be considered reactive, even if some of its elemental forms are highly stable. And keep in mind that the reason for going with weak/strong instead of reactive/corrosive nomenclature was not the unreactiveness of some of the elements, but that corrosive was a special type of reactive. YBG (talk) 08:24, 5 July 2017 (UTC)

And another related item: If we call out the noble gasses, why can't we call out the halogens? Of course, we can, but the question is whether we should or not. By my way of thinking, the conceptual gap (H,C,N,P,O,S,Se) vs. (F,Cl,Br,I) is not nearly as stark as the one between (11 NNNM) and (noble gases). This is clearly true from an historical perspective, the noble gases were discovered much later and were first called "inert gases". But my understanding is that even now that the noble gases are known to not be inert, their reticence to form compounds still sets them far apart from the rest of the nonmetals. YBG (talk) 08:24, 5 July 2017 (UTC)

Double sharp responses
 * Well, yes, I was partially posing the question rhetorically. Depending on what point you are trying to make it can make more sense to group F with N and O instead of Cl, Br, and I. The halogens do happen to be a model example of group trends, but once again the divide between F and {Cl, Br, I} is quite evident, just to a slightly lesser extent than what is seen in the pnictogens and chalcogens. Double sharp (talk) 08:30, 5 July 2017 (UTC)

I think you are pretty close to having your consensus as of today, since a few hours ago I wrote: "within the nonmetals we have no vast difference like the astronomers have with the terrestrials vs the giant planets, or even {Jupiter, Saturn} vs {Uranus, Neptune}. That makes me think that we should just leave well enough alone and not categorise the nonmetals, if we keep using metallicity as the guide in the p-block."

The question is just how to name that single category. Double sharp (talk) 08:30, 5 July 2017 (UTC)

Sandbh response YBG, it's a fair summary.

I have a philosophical passion for breaking down complex things into digestible chunks, and a strong philosophical aversion to obfuscating knowledge. I seek to use formal terminology where available and descriptive phrasing for any gaps.

For me personally, a set of 11-undivided pre-noble nonmetals would obfuscate ("To darken or obscure to the mind or intellectual perception" – Oxford English Dictionary) knowledge whereas a division structures and clarifies knowledge, and facilitates understanding, in a 1 + 1 = 3 manner.

If anybody in addition to Double sharp wants to see what planet I'm apparently living on with my constant references to a literature-based dichotomy between less nonmetallic and more nonmetallic nonmetals (or, as R8R put it, "strong and the not so strong nonmetals"), you can read all about it in the parsing the nonmetals section of my sandbox.

Double sharp and I have been incidentally discussing some aspects of this planet on my talk page. Sandbh (talk) 01:41, 6 July 2017 (UTC)
 * I read your sandbox. Certainly a division based on standard reduction potentials into what I call oxidative (O, F, Cl, Br, I) and reductive (H, C, N, P, S, Se) nonmetals makes sense in dividing the pre-noble nonmetals (as I've called the NNNMs on YBG's prompt – see below), but we could as well carve out the halogens or (as we currently do) diatomic elements. There are simply too many correct ways to subdivide the pre-noble nonmetals and each division would be of too little importance to merit two full-blown articles. In contrast, we can have one article for all the eleven pre-noble nonmetals, with different sections highlighting the oxidative/reductive, halogen/non-halogen, polyatomic/diatomic, biogen/non-biogen and other further divisions. Since I believe the oxidative/reductive division is most sound, though, it should go first in that unified article.
 * if you were looking for some name to replace "reactive", "corrosive", "weak", "strong", "intermediate" and all those subjective terms, I've found it here: oxidative nonmetals and reductive nonmetals. Parcly   Taxel  02:50, 6 July 2017 (UTC)
 * Oh, and another thing: lead has been promoted to FA status. With that, I feel I can take a break from Wikipedia until at least 14 August, on which date I'll be starting mathematics at university. See ya! Parcly   Taxel  03:02, 6 July 2017 (UTC)
 * I agree with the point, though not very much with the names: surely S is also rather oxidative as a silver-slayer? What I would do with the nonmetal article is to discuss all 11 of them together, then mentiom the noble gases as a special case, and then mention some ways the literature has split the eleven. There are just too many reasonable ways to do it and there isn't really much in the way of an intermediate set when you consider that we cannot even agree in the literature over which elements are nonmetals, as shown by Sandbh even being able to make a case for calling the metalloids as such. (And here's someone who apparently considers Bi and Po to be nonmetals! He also adds N and removes I from his list of strong nonmetals, and points out that while O is a good oxidising agent it is often kinetically hindered.) So I think not splitting is better. Double sharp (talk) 03:55, 6 July 2017 (UTC)
 * ✅ Parcly   Taxel  04:12, 6 July 2017 (UTC)
 * Small world, Parcly Taxel. I'll be returning to university at the end of August and doing mathematics. Sandbh (talk) 23:32, 6 July 2017 (UTC)

Needing something else to refer to?
Parcly Taxel, in your sandbox you commented

"By a new principle formulated by Double sharp that a category must be definable without making reference to another category, the alternative is also ruled out because none of the new terms is unambiguous without making reference to another term."

R8R, you made the following contribution, as listed by Parcly Taxel in the Grand Summary:

"In contrast, weak/strong nonmetals are, by definition, fuzzy concepts and requires something else to complete the definition (phosphorus is weak/intermediate only because we can compare it to fluorine and chlorine and see the difference). Which, to my understanding, is a clear negative answer to the question in the beginning of this section: 'Is there a compelling need to subcategorise the non-noble nonmetals?'"

The above quote is referred to again by Parcly Taxel in the "Parcly Taxel's suggestion: Eleven undivided nonmetals" section, and reiterated by R8R in the "Re compelling need to subcategorise nonmetals section"

Parcly Taxel, R8R, it seems to me that your concerns about needing something else to refer to apply equally to "transition metal" (transition between what and what?); "post-transition metal" (what is it these things are post- to?) and "metalloid" (meaning resembling a metal, hence referring to something else). Is that so? Sandbh (talk) 04:44, 5 July 2017 (UTC)
 * We may define all the groups essentially on their own, without referencing other categories, but only as a process:


 * 1) Define the transition metals as the d-block.
 * 2) Define the lanthanides, actinides, alkali metals and alkaline earth metals as sixth-period f-block, seventh-period f-block, group 1 minus hydrogen and group 2 respectively.
 * 3) Define the metalloids as we have done at lists of metalloids – the average view of several objective and subjective criteria.
 * 4) To the left of the metalloids but to the right of the transition metals are the post-transition metals. These are too metallic to fit the metalloid criteria, but still lie within the p-block.
 * 5) To the right of the metalloids, and including hydrogen, are the nonmetals. These are too non-metallic to fit the metalloid criteria, but still lie within the p-block.
 * 6) The noble gases within the nonmetals are obviously group 18, and the halogen nonmetals are group 17.
 * These categories, then, may be fuzzy at the borders, but at least their interiors are well-defined (i.e. I could pick out a representative element). The problem with "weak", "strong", "corrosive", "intermediate", "other" and similar descriptors is that they're completely fuzzy, as has mentioned: there are no representatives. This is what I interpreted 's principle to be and is akin to defining the boundary between Europe and Asia: while Orsk's continent may be disputed, Moscow's is not.  Parcly   Taxel  05:14, 5 July 2017 (UTC)

Like we have stated many times, in words like transition metal, transition does not have its usual English meaning. Therefore concerns that it is undefinable without referencing anything else are not really valid: one does not ask "transitional between what and what" except as an exercise in etymology. It simply means "those elements that use their d-electrons as valence electrons". And that's all there is to it. To say that titanium is a transition metal, you do not have to compare it to calcium and zinc. You just note that it can form Ti3+ with the 3d1 configuration. We don't look for the knees of polygons either just because that's what γόνυ originally meant in Ancient Greek, and I doubt Euclid or Archimedes would have done so either. So why do it with "transition metal"?

Similarly, the other categories spring up directly from considering electron configurations. The exceptions are the ones relating to metallicity, which is one reason why I think it would be better to scrap those and just divide the elements in that region by groups from 12 to 18. Already we have the silly problem that {Se, Te, Po} as a relatively homogeneous triad with a clear trend go into three different categories. The current treatment obscures this sort of thing, because thhe average reader is drawn to the colours, and may think that the groups stop having meaning there. (I distinctly remember talking to Sandbh about an alternative periodic table whose author claimed that putting C and Pb in the same group was a failure of periodicity.) It would be helpful to get rid of that misconception early.

I find it vaguely distasteful to define a category's borders by what other people have done. Firstly, it pushes back the problem: just what are the defining attributes of this category? If we cannot define this and appeal to others, then surely we can end up with something as nebulous as heavy metals, for which the definition seems to be "I know it when I see it", and the term has become almost meaningless? Secondly, different people may choose different boundaries depending on their focus. Referring to the IAU definition of planet again, someone interested in dyanmics would prefer to distinguish {planets} from {dwarf planets + small Solar System bodies}, while someone intereted in planetary geology would prefer to distinguish {planets + dwarf planets} from {small Solar System bodies}. (This is of course a simplification, not counting the rounded moons, and trying to avoid hard cases like the partially differentiated Pallas and Vesta, but you get the idea.) Similarly, perhaps someone interested in general inorganic chemistry might prefer to distinguish {metals} from {metalloids + nonmetals}, but an organometallic chemist would probably advocate distinguishing {metals + metalloids} from {nonmetals}, and no doubt their mental maps of the categories will be different. (For example, I am not sure I think of Te or At as being on a par with the other metalloids.)

This is where I think the IAU definition of a planet has something important to teach us: in such cases, we need an intermediate category for things that tick a consistent subset of the boxes for the "best" category, whether that be "planet" or "metal". And that consistent subset then becomes a definition that requires no outside reference to another category.

We surely can define what a metalloid ought to be, and we surely can define what a nonmetal ought to be, and have them be largely distinct. Unfortunately, unlike in astronomy (where Mercury and Mars are far from being dwarf planets and Pluto and Eris are far from being major planets), we have a lot of hard edge cases like selenium or astatine, and we have more of a continuum. That alone would suggest avoiding the issue by using the parallel categorisation of groups as primary. But furthermore this affects the small categories too: within the nonmetals we have no vast difference like the astronomers have with the terrestrials vs the giant planets, or even {Jupiter, Saturn} vs {Uranus, Neptune}. That makes me think that we should just leave well enough alone and not categorise the nonmetals, if we keep using metallicity as the guide in the p-block.

P.S. The thing that worries me most about what we call astatine is that it reminds me uncomfortably of what we used to think of Pluto. It was reasonable enough to call it a planet at its discovery; it was still vaguely reasonable to say it without some doubt until its moon Charon was discovered; and even after that, it took the discovery of other big KBOs like Quaoar, Orcus, and Sedna to make a reclassification reasonable enough to gain momentum, and it took the discovery of the heavier Eris to bring matters to a breaking point. (I am still a little bit disappointed about it, by the way: I think it is perfectly fair that minor planet 134340 is not considered a major planet any longer, but I think Pluto was a great name, and it is now a great shame that we cannot use it for the hypothetical Planet Nine, especially since it may have been exiled to its current abode in the "underworld" by Jupiter!) I think we may be making a similar error in classifying astatine based on what little we know: I suspect that if we only knew of iodine from tracer studies, we would be calling it a metalloid too. But this is a separate topic. Double sharp (talk) 06:28, 5 July 2017 (UTC)


 * looking through your six groups I see you've defined the post-transition metals by referring to the transition metals, and the nonmetals by referring to the metalloids. On this basis, and following your process, a seventh grouping can be defined as follows:


 * "7. To the left of the halogens, and including hydrogen, but to the right of the metalloids are the other nonmetals. They are too nonmetallic to fit the metalloid criteria but still lie within the p-block."


 * In other words, I don't follow the logic of your response. Sandbh (talk) 06:37, 5 July 2017 (UTC)
 * I get his point, but I don't think he's doing it well either: his references to the location of groups by saying what lies to their left or right makes that look more important than it is, since I know that phosphorus is a nonmetal because its physical and chemical properties are those of one, not because it lies to the right of the metalloid line. I think we would need a definition of metals, metalloids and nonmetals that looked more like that of planets, dwarf planets, and small Solar System bodies to do it the way I would like it, though. Double sharp (talk) 06:51, 5 July 2017 (UTC)
 * You mean we should take an approach similar to the heatmap we have for heavy metals to classify metals, metalloids and nonmetals? (I am still supportive of not classifying the nonmetals at all, since the halogens and noble gases are already groups.) Parcly   Taxel  07:20, 5 July 2017 (UTC)
 * No, that is the opposite of what I meant. I was referring to "heavy metals" as a nebulous, meaningless term above, after all. I am saying that we should seek to define metallicity by a small set of physical and chemical properties and not by scattergun-like choosing vague ideas from the literature and making lists of elements according to how often or not they are singled out. Double sharp (talk) 08:11, 5 July 2017 (UTC)
 * I got your point about "transition metal" but was thinking about it from the pov of (a) the general reader, whom we write for, and (b) what I wrote about earlier re killing ourselves tying to surpass the weaknesses of terms such as these including post-transition metal, and metalloid, and forgetting about the value of plain English. Sandbh (talk) 07:28, 5 July 2017 (UTC)
 * And as we said many times, I would consider this quite irrelevant because if this general reader wants to learn anything about chemistry, he needs to learn what "transition metal" means in chemistry, just as he has to relearn new meanings of words like "mole", "equilibrium", "concentration", "analysis", and "coordination complex". The same is not true of something like "intermediate nonmetal". Paradoxically, the fact that we cannot use plain English for something like "alkali metal", "transition metal", or "actinide" seems to indicate that these categories have more rigour from inside chemistry, and to learn chemistry you must learn those
 * P.S. I sometimes wonder if the general reader today actually has the patience to read through everything we write instead of just searching for brief nuggets for his or her science presentation in middle school. Certainly the Article Feedback Tool of a few years ago did little to convince me otherwise. As a result I tend to think of my audience here not so much as the general reader, but as the general interested reader, who is willing to at least put in some minimal effort to understand what is going on. Learning basic terminology is surely the first step towards that. Double sharp (talk) 08:11, 5 July 2017 (UTC)

Hi I see no need to make life for the general reader, whether casual, or interested, any harder than it needs to be, hence my ongoing references to the value of plain English, where no technical nomenclature is in common use.

It may well be enough for the general reader to see that transition metals are situated between the alkaline earth metals and the post-transition metals.

And yes, if the general or interested reader wants to know more about the alkaline earth metals, the transition metals, or the post-transition metals, they will have to dig deeper.

That said, when it comes to attempting to a better job of mapping the nonmetals I feel that we become entangled in undue concerns about whether or not the general reader will understand the meaning of plain English terms like "intermediate nonmetal" ("Hmm, I see they are located between the metalloids and the corrosive nonmetals, in a manner analogous to the way the transition metals are located between the alkaline earth metals and the post-transition metals—that's interesting…) or "corrosive nonmetal", and the great burden such descriptive phrases would apparently impose on these readers.

If the curiosity of the general or interested reader is piqued enough to click on the link to intermediate nonmetal they will be helpfully taken to the relevant section of the nonmetal article which will explain (a) the descriptive rather than formal status of this phraseology, and (b) the nature of the intermediate nonmetals, including their relationship to the rest of the nonmetals, akin to what we do now in distinguishing the properties of the polyatomic nonmetals from the diatomic nonmetal, and vice versa.

For the general or interested reader, the same process applies to any of our other more formal category terms.

Meanwhile, the detached chemist and related science professional, appreciates the effort that has been made in attempting to describe the pre-noble nonmetals consistent with strong literature conceptions of (as R8R put it, "strong and the not so strong nonmetals") without any pre-tense at introducing neologisms, which is effectively what we tried to do with the current poly-di scheme.

I sincerely hope I've spoken plainly, and without getting anyone's nose out of joint. Bravo! the general reader, whether casual or interested ^__^  Sandbh (talk) 00:54, 6 July 2017 (UTC)
 * The trouble, to me, is that intermediate nonmetal feels so much more like a neologism than polyatomic nonmetal. I think chemists would not bat an eye at the latter; it might be a little odd to categorise them this way, but they know what "polyatomic" means, and they know it is correct. At the most they might quibble about ozone. But they do not know what "intermediate" means in this context, which is understandable because we just made it up; and since we're not considering the metalloids to be nonmetals now it really starts to be odd. How can you have an "intermediate" category at the edges like that?
 * Once again I agree that there is a notion that some of the chemically active nonmetals (hey, that sounds perfectly natural!) are stronger than others. No doubt they are, but different authorities would put the gap differently. We have argued about nitrogen; I could argue about iodine; and I could even note that while oxygen is a strong oxidising agent it is kinetically hindered so that the oxidation takes a lot of time (or it needs a flame or a spark to start going quickly, like in the case of Mg). Only F, Cl, and Br seem beyond reproach. The IAU-inspired solution I guess would be to have a middle ground and put N, O, and I in it, but at this point I would ask: with only eleven nonmetals, instead of all those scores of dwarf planet candidates and minor planets, aren't we going a bit too far? Can't we just let groups and electronegativity tell the story? Double sharp (talk) 04:05, 6 July 2017 (UTC)

1. Well, at the end of the day, I only go by what the literature says.

2. Fundamentally (according to the literature):


 * a. Metalloids generally behave chemically like nonmetals;
 * b. There is a consistent theme of strong nonmetals and less strong nonmetals; and
 * c. There is a firm distinction, as per my sandbox, between the two kinds of nonmetal, the twin hallmarks of which are (i) the widely acknowledged recognition of O and halogen nonmetals as the most chemically active nonmetals; and (ii) their corrosive nature.

3. I should ask some other editors about the neologism question.

4. A chemist would grasp the concept of an intermediate nonmetal in two shakes.

5. Arguments around the edges, by their nature, are peripheral rather than fundamental.

6. I should ask some published chemists about how they see these things.

7. If we were to supersede the poly-di scheme, I guess my preference would be: 's active nonmetals + noble gases + keep showing At as a metalloid.

8. There is no need to call them chemically active nonmetals, just as there is no need to call the noble gases chemically noble gases. I have seen the non-noble nonmetals collectively called either "active nonmetals" or "reactive nonmetals" in periodic tables in two chemistry books.

9. Please give me 24 hours to consider if I really meant to say all of this. Don't let that stop you from posting a response in the meantime. Sandbh (talk) 07:35, 6 July 2017 (UTC)
 * Well, I'll graciously accept your invitation! ^_-☆
 * 1. I agree that the literature notes that some nonmetals are stronger than others. Where I am in doubt is whether they are consistent about which they are, save for I guess fluorine, chlorine, and bromine. There is also the issue that "strong" is a bit hand-wavy as a term and may be interpreted in different ways.
 * 2. Metalloids certainly behave chemically like nonmetals. However so do quite a few of the transition metals, and I think you could make a similar case for SnIV and perhaps Bi and Po. (The rest has been responded to in 1.)
 * 3. I agree.
 * 4. The concept may be clear; it may not be so clear that one half of the "intermediate" is being more nonmetallic than the metalloids; and the reason for elevating this to such an important status as a separate colour may be questioned.
 * 5. Indeed; but when nearly the whole category is edges (e.g. N, O, I), one might start to question its viability.
 * 6. Yes, please do; I'd love to see their responses.
 * 7. OK, but...
 * 8. ...there is similarly, I think, a reason why we call the heaviest elements out as having "unknown chemical properties". Evidently chemistry is being thought of as primary in some respect here, and even if the idea of a physically active nonmetal is nonsense, it is good to clarify things if it only adds one word like this.
 * 9. No problem! Double sharp (talk) 08:14, 6 July 2017 (UTC)

I'm OK with everything I wrote above. On item 1 I've added a few more cites to my sandbox. My contention would be that this represents at least the average view if not the majority view. You've prompted me to think more about fundamental v peripheral issues, and I'll post more about that somewhere. On item 8 I tend to agree that chemically active is better than active. Sandbh (talk) 12:06, 7 July 2017 (UTC)

Mark 4 + Mark 5
Summary I now believe that both the Mark 2 and 3 proposals are dead, and in their place are two new proposals:
 * Mark 4 is 's proposal to completely scrap the metal/metalloid/nonmetal groupings, leaving the very structure of the periodic table to suggest similarities and differences. (no visual? -DePiep (talk) 21:26, 6 July 2017 (UTC))
 * Mark 5 is my "eleven undivided nonmetals" proposal, but with the metal/metalloid/nonmetal groupings defined like the planet/dwarf planet/SSSB groupings. Note that I do not give the nonmetals excluding the noble gases any special name, instead leaving it as the result of the set-theoretic operation $$NM\setminus NG$$ (via the "of which" construction). Parcly   Taxel  08:28, 5 July 2017 (UTC)
 * Well, Mark 4 is kind of only a half-serious suggestion, but I pose it as food for thought anyway. Since if we want to keep the colours, perhaps we should think about the reasons for doing so, other than "it looks pretty with colours". ^_^
 * As for your version of Mark 5: okay, but how do you do that without creating an "Other nonmetals" in the legend of ?
 * I would be tempted to go for "Typical nonmetal", incidentally, except that I am not sure if we can really consider H to be one. But I wouldn't be convinced unless I saw them as common terms in the literature: for "reactive nonmetal", I'd like to see the literature that considers C and N as such first before I would support it. Double sharp (talk) 08:34, 5 July 2017 (UTC)

Legend alternatives for the Mark 5 categorization
Here follows three alternative legends for the Mark 5 categorization. YBG (talk) 23:02, 6 July 2017 (UTC)
 * My idea, visually (Mark 5a):


 * (The "of which Noble gas" cell can either be to the right of the "Nonmetal" cell, or below it. This accommodates a grouping of all the cells left of "Metalloid" under a "Metal" cell.) Parcly   Taxel  08:51, 5 July 2017 (UTC)
 * So you would propose removing the top row saying "Metals, Metalloids, Nonmetals" as super-categories? Double sharp (talk) 09:45, 5 July 2017 (UTC)


 * As I explained, a "Metal" super-category can be accommodated with my proposal, as follows, Mark 5b:


 * This second colour scheme has the advantage of giving metals, metalloids and nonmetals three distinct colours, stressing both the coarser and finer classifications. I do not propose removing the super-categories, but we currently do it in the default periodic table navbox, and my realisation of the eleven undivided nonmetals proposal checks out both ways. (No super-categories are used on the French, German, Japanese, Chinese or Dutch Wikipedias.) Parcly   Taxel  10:04, 5 July 2017 (UTC)
 * Hey, I like this idea! If we removed the background colour from the "Metals" cell in the legend, then we could even do away with the words "of which". It's certainly a very neat way to avoid using that dreaded word "other". Double sharp (talk) 10:16, 5 July 2017 (UTC)

Perhaps an idea Mark 5c: Double sharp (talk) 15:07, 5 July 2017 (UTC)
 * with the reservation that we currently use singular forms for the category names in the navbox (I'm fine if you want to use plurals though). Parcly   Taxel  15:38, 5 July 2017 (UTC)
 * I have changed the numbering to "5a, 5b, 5c", as these are all essentially a set of one-off proposals that were quickly rejected when something better came along (active nonmetals / chemically active nonmetals). Double sharp (talk) 23:37, 6 July 2017 (UTC)

Category name alternatives for the Mark 5 categorization
Here the discussion shifts to the question of category naming YBG (talk) 23:02, 6 July 2017 (UTC)


 * I think that even with the "of which" construction, any of these legends are confusing. The current legends (and all previous ones) show two different complete classifications into mutually exclusive and jointly exhaustive categories: 3+ categories using the top row and N+ categories using the bottom row (where the + indicates the unknown category). I find the above proposals so ugly that I'd almost be willing to resurrect "other nonmetals". Yes, that bad.
 * If reactive nonmetals fails the sniff test, what about active nonmetals or chemically active nonmetals? The latter seems most precise; the only negative I can think of is that it is longer than any of our other category names. At any rate, we should put more thought into what to call the NNNM. YBG (talk) 16:33, 5 July 2017 (UTC)
 * I've got it. We have
 * post-transition metals after transition metals
 * ??? before noble gases
 * and this suggests pre-noble nonmetals as the name for the NNNMs. Each element in this group can realistically gain electrons to reach a noble gas electron configuration, which also fits the pre-noble moniker. Parcly   Taxel  17:19, 5 July 2017 (UTC)
 * Which would be nice if anyone actually called them that. I think YBG's chemically active nonmetals is the best we have at the moment, but we should link it as chemically active nonmetals to avoid the idea that this is a single term like "transition metal". Double sharp (talk) 03:25, 6 July 2017 (UTC)
 * Yes, "chemically active nonmetals" sounds completely doable and feasible. We need to break out of the mindset that all the words in a category must be part of the link. Parcly   Taxel  04:17, 6 July 2017 (UTC)
 * ++ from me for "chemically active nonmetals." To be precise, I think two words would be certainly better than three ("active nonmetals," how about that?) but I will easily accept the three-words version.--R8R (talk) 11:24, 6 July 2017 (UTC)
 * Post-metalloid nonmetals, anyone? YBG (talk) 15:08, 6 July 2017 (UTC)
 * Or could we call the Halogen NM, with or with O, "hyperactive nonmetals"? ADHD nometals? Then could we call the rest the couch potato nonmetals? Or would the noble gases rise up and say that term should be reserved for them? YBG (talk) 15:18, 6 July 2017 (UTC)
 * Negative for Mark 7: the bottom row does not accomodate the 11 NNNMs at all. -DePiep (talk) 23:04, 6 July 2017 (UTC)
 * The same is true for both of the two-row legends, which is why I believe they should not be considered. And the one-row option seems to indicate that noble gases are not nonmetals. So I believe all three legends options should be abandoned. But we are still left with the original Mark 5, which categorizes the elements as follows
 * Metals
 * Alkali metals
 * Alkaline earth metals
 * Lanthanides
 * Actinides
 * Transition metals
 * Post-transition metals
 * Metalloids
 * Nonmetals
 * As-yet-unnamed 11-element category of nonmetals
 * Noble gases
 * Unknown chemical properties
 * YBG (talk) 23:16, 6 July 2017 (UTC)
 * I don't know if I like it but the "as-yet-unnamed 11-element category of nonmetals" are rather indisputably molecular nonmetals, made up of polyatomic and diatomic varieties, are they not? There are some references in the literature that use this terminology: "Many of the nonmetals exist as diatomic molecules rather than individual atoms...Not all molecular nonmetals are diatomic however..." (Basic Concepts of Chemistry, Malone & Dolter 2008, p. 115) ♦  "The melting and boiling points of the atomic and molecular nonmetals are given in Table 4-3." (Chemical principles, properties, and reactions, Henold & Walmsley 1984, p. 117)  ♦  In Chemical Bonds: An Introduction to Atomic and Molecular Structure (Gray 1994, p. 192) there is a periodic table classifying the nonmetals as network nonmetals (B, C, Si) and molecular nonmetals. Unfortunately the situation is clouded since the noble gases are also counted by Gray as molecular nonmetals. He appears to regard the van der Waals forces holding the noble gases as being sufficiently molecular to warrant counting them as molecular in nature. I suspect this would be a minority view by a very wide margin, however. Sandbh (talk) 06:47, 7 July 2017 (UTC)
 * That has possibilities. Not that it counts for much, but it does provides a bit of continuity with our current polyatomic and diatomic names. YBG (talk) 07:35, 7 July 2017 (UTC)
 * Unfortunately it appears that this is a subfield difference: our article on molecule says "In the kinetic theory of gases, the term molecule is often used for any gaseous particle regardless of its composition. According to this definition, noble gas atoms are considered molecules as they are in fact monoatomic molecules." Double sharp (talk) 07:58, 7 July 2017 (UTC)

Coming back to the question of how to link the category name, what about "chemically active nonmetals", where the target of the link would be a high-level survey article that could stand in parallel to noble gas chemistry. YBG (talk) 07:35, 7 July 2017 (UTC)

The leading name contenders
Here is a list of the potential names that have been suggested and received at least some positive feedback from another editor. (I haven't included my "reactive nonmetals" or Parcly Taxel's "pre-noble metals" because I couldn't remember any positive response to either. But feel free to add either if you wish.) I could live with any of these. YBG (talk) 07:35, 7 July 2017 (UTC)
 * Active nonmetals
 * Chemically active nonmetals
 * Molecular nonmetals
 * I suppose chemists tend to think of nonmetals in terms of their electroactive character, working back from the inside corner trio of O-F-Cl or maybe just the halogen nonmetals, so it would probably be nice to have some consistency on the right hand side of the table and refer to active or chemically active nonmetals, as analogs of the whatever-active metals on the left hand side of the table. I don't think the molecular nature of the molecular nonmetals is the first thing that comes to the minds of chemists.


 * I further suppose that chemists, when one refers to "nonmetals", tend to presume that one is referring to nonmetals other than noble gases, unless one has specifically referred to the noble gases. Which makes the reason for calling the active nonmetals "active" or "chemically active" nonmetals seem somewhat strange (but it would help the general reader). Sandbh (talk) 08:23, 7 July 2017 (UTC)

The average view
Double sharp mentioned earlier that, "we are an encyclopaedia, trying to sum up the average views of most authors..." Most other project members have mentioned or alluded to the same kind of thing.

I guess the average of the views of most authors goes something like this:


 * 1. Nonmetals, at the outset, are characterised by a lack of metallic properties (as per Coin845's "metal-normative" observation)


 * 2. The alkali metals and the halogens provide the most distinct contrast between metals and nonmetals


 * 3. The most reactive metals are found towards the bottom left of the PT; the most reactive nonmetals are found in the upper right hand corner just inside the noble gases


 * 4. Apart from the halogens and the noble gases, there are these other nonmetals


 * 5. Apart from the alkali metals (and their cousins the alkaline earths), there are the transition metals, the Ln, the An, and the rest of the metals


 * 6. The metals and the nonmetals meet in the p-block, as a by-product of the general progression in metallic to nonmetallic character going across the PT


 * 7. The elements in this vicinity are sometimes, but not always, called metalloids or semimetals or semiconductors, and sometimes, but not always, regarded as a third category of element.


 * 8. Metalloids:


 * (a) look like metals but behave like nonmetals;
 * (b) are thought to be semiconductors, and this is mostly explained in terms of their electron band structure or the fact that they don't conduct electricity as well as true metals but not as badly as nonmetals;^
 * (c) are thought to have amphoteric oxides.^^


 * ^ most authors get this wrong or confuse the two ideas
 * ^^ this is strictly correct only for Ge, Sb and Te; usually but not alway regarded as correct for As; and the minority view for B and Si

Does this look about right? Sandbh (talk) 02:24, 20 June 2017 (UTC)
 * It certainly does to me! Double sharp (talk) 03:03, 20 June 2017 (UTC)
 * My simple-minded expectation of 8(a) caused me to expect that metalloids would have physical properties akin to metals and chemical properties akin to nonmetals. This is only borne out ever so slightly in . I would have expected a much greater contrast in Metals-metalloids-nonmetals: compare, overview. Am I missing something? YBG (talk) 05:12, 26 June 2017 (UTC)
 * I think it's mostly true as a simplification, but if you just look a little bit harder you will see metallic properties even there as I showed above for B and Ge (I would finish it, except that I'm trying to find more convincing ones for Si). And the trouble is that they are not the only elements that are physically like metals and chemically like nonmetals: quite a lot of the d-block also fits that criterion. Take ion formation: metals are listed as "tend to form cations". Yes, but only in low oxidation states. When the oxidation states are forced to be high, as they are in the early part of the d-block (e.g. vanadium), the metalloid description fits better. And in many organometallic compounds the d-block metals are perfectly happy to exist in negative oxidation states. As for true sulfate salts, includes a very telling gap running through the middle of the d-block, and while VII and VIII are perfectly happy to form true sulfates, VIV can only do so as VOSO4.
 * The trichotomy given in the article probably looks more clear-cut than it really is because only the main-group metals are really being seriously considered, and they have low oxidation states: if you stay there and move to higher and higher oxidation states, you move into metalloid and nonmetal territory. But if you do this in the transition metal block, you stay there while you move to higher and higher oxidation states physically, but the chemistry becomes more and more nonmetallic as well. Double sharp (talk) 09:44, 26 June 2017 (UTC)
 * The phrase "look like metals" is a bit of a give away. The name "metalloid" is a bit of a give away too, as long as you know that "metalloid" means 'resembling a metal' i.e. rather than actually being a metal. The comparison of the properties of metals, metalloids, and nonmetals table is a bit obscure, because in the "Elemental chemistry" section of the table, the Overall behaviour row doesn't really do justice to the situation, at least not for the metalloid entry. Sandbh (talk) 12:35, 26 June 2017 (UTC)