Talk:Roman numerals/Archive 9

Archive 5

←

Archive 7

Archive 8

Archive 9

Archive 10

Typo in last edit summary

Reinstating a paragraph break between two paragraphs I remarked that 'if the paragraphs were the tiniest bit more distinct we'd need a new sentence'. Of course I meant a new section. In fact it would be better to 'break the section' here than reunite these paragraphs - although each paragraph is brief, there is a quite distinct change of topic, and each reads much more clearly for the line br[e]ak. --Soundofmusicals (talk) 11:21, 6 June 2020 (UTC) [corrected typo in (own) post]

Is that all, Soundofmusicals? has our discussion on my proposed content run its course? Xcalibur (talk) 15:31, 10 June 2020 (UTC)

So far as I am concerned - long ago! --Soundofmusicals (talk) 03:38, 11 June 2020 (UTC)

Latest rendition of ruleset

Here's what I've got currently. the LURNC (Lee K Seitz) source has been particularly helpful, and I just added a new rule to fill in a gap (I'm surprised no one pointed it out). Now, I'm almost certain my ruleset is complete.

Rules-based section

Standard form

The modern era has seen the emergence of a standardized orthography for Roman numerals, which permits only one permutation for any given value. This system may be described as a decimal pattern, as above, but also as a logical set of rules. While exceptions can be made (notably IIII instead of IV on clockfaces), the modern convention is widely recognized and adhered to, and may be prescribed by the following ruleset:

Basic Rules

1. Powers of ten are dealt with separately, ordered from greatest to smallest and from left to right, with numerals either added or subtracted.
2. Repeated 'tens' numerals (I, X, C, M) are added together, with up to three (3) permitted in sequence (subtraction allows an additional non-sequential repetition once per power).
3. 'Fives' numerals (V, L, D) may only be added once per power (not repeated nor subtracted).
4. Numerals placed to the right of a greater value are added, while those placed to the left are subtracted; if placed between two larger numerals, the 'tens' value is subtracted (i.e. subtraction takes precedence).
5. Only one (1) 'tens' numeral may be subtracted from a single numeral per power, and this must be by 1/5 or 1/10 (i.e. the next lower 'tens' numeral).
6. Addition must be less than subtraction for any given numeral (i.e. the added value to the right must be less than the subtractor).
7. Any subtracted 'tens' numeral must either be first, or be preceded by a numeral at least ten times (10x) greater.

The above describes the basic pattern of integers from 1 - 3,999.

Examples

Following the above rules:
90 must be XC. It cannot be LXXXX because that would break rule 2 (up to three in sequence), and it cannot be LXL because that would break rule 3 ('fives' not repeated).
45 must be XLV. It cannot be VL because that would break rule 3 ('fives' not subtracted).
99 must be XCIX. It cannot be IC because that would break rule 5 (subtract by one fifth or tenth).
18 must be XVIII. It cannot be IIXX or IXIX because these would break rule 5 (subtract once per power).
19 must be XIX. It cannot be IXX because that would break either rule 5 (subtract by one fifth or tenth) or rule 6 (subtract>add).
10 must be X. It cannot be IXI because that would break rule 6 (subtract>add).
14 must be XIV. It cannot be IXV because that would break rule 6 (subtract>add), and it cannot be VIX because that would break rule 7 (at least 10x).

1894 is MDCCCXCIV. Consider how this agrees with the rules: powers of ten are arranged properly (rule 1), C is used three times sequentially and a fourth time non-sequentially (rule 2), 'fives' numerals D and V are added once each (rule 3), X and I are subtracted by being placed to the left of larger values while the rest are added (rule 4), they are subtracted once per power by one tenth and one fifth respectively (rule 5), IV is less than X (rule 6), and C is 10x X and 100x I (rule 7).

Fractions & Vinculums

The following rules extend the system further:

• Fractions may only be added once to the right of all numerals (not subtracted), without redundancy, and are duodecimal.
• Beginning at 6/12, the S (semi) is used, followed by marks (unciae) for additional quantities up to 11/12.
• Placing a bar (vinculum) over a numeral multiplies it x1000.
• The vinculum is used for values of 4,000 and greater (it is not used up to 3,999), and may be iterated up to three (3) times.
• Starting at 4,000, and at the next two powers of a thousand thereafter, a new vinculum is added, with the pattern beginning at IV.
• Once vinculums are employed, the higher power is modified in preference to the lower (eg IV, not MV).
• Vinculums are always contiguous and are placed leftmost (ie no broken overlines or lower powers to the left per basic rule 1).

Thus, 2 2/3 would be rendered as IIS••, while 6,986 would be (VI)CMLXXXVI. Note that the symbol V is used more than once in that example, this is only permissible with the vinculum setting them apart. Also note that while M is used, 7,000 would be (VII).

(I), (II), and (III) are not used, with M, MM, MMM being preferred under 4,000. SS is never used in place of I (ss is sometimes seen as a variant of semi, but this is strictly pharmaceutical notation).

With fractions and vinculums employed, the lowest possible value is • or 1/12, while the highest is (((MMMCMXCIX)))((CMXCIX))(CMXCIX)CMXCIX or 3,999,999,999,999 (within standard usage). Four trillion, rendered as (((MMMM))), would be non-standard since it breaks basic rule 2, although the four M's can be used as a variant form, and it is more aesthetically pleasing as a maximum figure. Another alternative is ((((IV)))), but since the general principle is three sequential iterations, this should also apply to vinculums. The reason for the three-rule is that the eye can readily distinguish between one, two, or three marks, while four or greater become illegible. Thus, 4 M's are preferable as a variant form.

admittedly, the intro is a bit redundant, but the text can always be reworked. the extended rules would be a nice addition, but I don't insist on their inclusion -- the now 7 basic rules are enough. I could add the legal/illegal examples for each rule if that is preferred. for the latest rule, the example would be: 1400 is MCD, not DCM. Xcalibur (talk) 17:57, 26 May 2020 (UTC)

fixed the intro, and now on subpage: Talk:Roman_numerals/Rules Xcalibur (talk) 18:03, 27 May 2020 (UTC)

Still vastly vastly redundant. You have described "sorted from largest to smallest" several times. And all the stuff about subtractive could be replaced by "the digit for 4 is IV, ..." as no other subtractives are allowed. And almost everything else is covered by "use the shortest pattern". Furthermore any claim that this has something to do with programming is bogus, a program could correctly encode a number by repeatedly choosing the largest "digit" less or equal to the number, writing it down, and subtracting it, as long as "IV" is considered a "digit" with value 4. In any case again I strongly object to these "rules" unless you actually replace the existing description. There should never be "the same information shown twice" in any Wikipedia article, because all that happens is the poor user wastes a lot of time comparing them and trying to figure out what the hell the difference is so that it had to be described twice. Don't do that.Spitzak (talk) 19:01, 26 May 2020 (UTC)

there are redundancies, especially because the 'powers of ten' rule overlaps with others, but it's not to the extent you claim. how else would you define the system by rules, while making it explicit and understood, and while disallowing all illegal combinations? Your 3 rules are logical enough, but don't spell things out as mine do, and don't follow the RS. I reiterate, my content is directly compiled from the sources, which also use the dual approach of prescriptive/descriptive. I don't see a problem with this, and I don't think it would confuse readers so much as provide greater illumination. In fact, I just rewrote the intro to take that into account. Lastly, my objective here is to provide quality, well-sourced content to WP. Why must that be difficult? Xcalibur (talk) 19:40, 26 May 2020 (UTC)

If any of the rules "overlap" any of the others then the rules are obviously "redundant:", which implies they need to clipped - what's very much worse, they are contradictory, and some, even with your "examples" are so unclear that even you can not offer an intelligible explanation. Your whole "intro" duplicates the first paragraph of the "description section - so that there is no need for it at all. Your stubborn inclusion of matter (Roman fractions and "barred" large numbers) even after you seem to admit they have nothing to do with the case is also intensle irritating. These things have been pointed out, repeatedly, by at least three other editors, and your response has been just to restate your case in more or less the same words, and lay down an identical or almost identical "alternative" set of rules that fails to address any of these problems. It is this behaviour that is "difficult". --Soundofmusicals (talk) 08:57, 28 May 2020 (UTC)

there are redundancies, but that's ok, as long as it's not excessive. the only rule I could really do away with is the first (powers of ten), but I keep it because it provides clarity. the rules are not contradictory, and not difficult for anyone familiar with the subject matter; on the contrary, their logic is impervious and their style refined (thanks to multiple revisions and constructive criticism), and I can explain each one if you want me to. I can always edit the intro further if you'd like. as for fractions/vinculums, again I'm willing to compromise and leave that stuff out if you'd like. I've made significant fixes and improvements so that it's essentially a finished product. overall, the 'problems' you object to don't really exist, or are minor and can be easily remedied. I think the rules would make a constructive addition, and it's unfortunate that you continue to blockade my efforts. Xcalibur (talk) 09:32, 28 May 2020 (UTC)

To each "point" made above - yes they are, or in other words saying over and over that something is not so doesn't make it not so! --Soundofmusicals (talk) 09:40, 28 May 2020 (UTC)

I agree that repeating something doesn't make it true. there seems to be a fundamental disagreement or misunderstanding, or maybe a philosophical objection to a prescriptive approach. to answer one of your earlier criticisms, the subtract>add rule makes perfect sense, it's the reason why strings like IXV or IXI are not legal. IC and IIX are not permitted by the subtraction rule, neither is IXIX (because you can only subtract once per power). this stuff really is self-explanatory, and based on Reliable Sources; I don't think the average reader would have trouble with it, as I don't underestimate their ability. you seem intelligent enough, even making useful constructive criticisms, so I'm not sure what the obstacle is? Xcalibur (talk) 09:56, 28 May 2020 (UTC)

You have had more than ample opportunity to share your point of view on this one, and nobody agrees with you. This has happened to all of us at some time or other - usually (on sober reflection) because we are wrong. In any case this discussion really has moved to what version of the table we are going to use, and how it should be incorporated into the "description" section. Watch this space! --Soundofmusicals (talk) 10:42, 28 May 2020 (UTC)

Alternatively, maybe I'm right, and you (& other editors) are wrong. at least, you can't rule out the possibility (no pun intended). I could potentially go through other channels, although I'm not sure if that would be viable. I'll let you know if/when I do. In any case, you have some sort of deep-seated objection to this, and in spite of my valiant efforts, I haven't been able to get through to you, and there it stands. Xcalibur (talk) 11:01, 28 May 2020 (UTC)

References

Indo-arabian numbers

I would like to propose to replace the words "arabian numbers" with the more accurate "indo-arabian numbers". It has long been established that the modern western numbers used with the decimal number system are in fact Indian numbers, brought to the west from India (see links below). I believe, the indo-arabian numbers came via islamic Spain (caliphate) to Europe. The messenger (Arabic) is not the creator (Indian).

https://www.britannica.com/topic/Hindu-Arabic-numerals

https://en.wikipedia.org/wiki/Hindu%E2%80%93Arabic_numeral_system — Preceding unsigned comment added by Faustdownunder (talk • contribs) 00:32, 24 July 2020 (UTC)

NO! Those are ARABIC DIGITS. They look like this: 0123456789, not like other glyphs. "Hindu-Arabic Numeral System" means base 10 which is written in a large variety of digits, and was in fact invented long before the digits being talked about. Trying to change Arabic numerals to "Hindu Arabic" or whatever is insulting to Hindus as it makes it sound like they did not invent the system long long before the digits were developed. Just please stop you are not serving your "cause".Spitzak (talk) 02:14, 24 July 2020 (UTC)

"rn" template

This is currently used to display Roman numerals in a slightly enlarged Roman font. It is up for deletion (so that in this article, for instance, the default display for roman numerals would be standard sized non-serif. (IMHO a retrograde step, but that is not the point here). The point is that this article becomes illegible with the template itself modified to indicate that it is under discussion. I have taken the liberty of restoring the template to its former (operative) form. Please forgive me if this step was unnecessary, and the "messed-up" form of the article did not display that way to casual users or something like that! --Soundofmusicals (talk) 06:53, 9 July 2020 (UTC)

Incidentally - the discussion in question is here--Soundofmusicals (talk) 07:33, 9 July 2020 (UTC)

There can be some opposition to the template, I used it on William II of England and was immediately reverted. The discussion seems to be against it. Martin of Sheffield (talk) 07:58, 5 August 2020 (UTC)

Paraphrasing from what I wrote at the TfD, when discussing Roman numerals, characters like I and X pop up frequently and are confusing without specific styling ({{rn}} is used 387 times at Roman numerals). However, at Pope Stephen II (and many similar articles), it's pretty obvious that II is an id, and most of those fluent in English would know it's Roman numerals. Special styling in that case is not needed and may be distracting. That is, I agree that {{rn}} is only desirable in an article such as Roman numerals. Johnuniq (talk) 09:59, 5 August 2020 (UTC)

Description and Standard form sections

Ignoring the rather rude edit summary we do have a real attempt at a resolution of these sections. Which is basically what we have been waiting for for years! To outline my the structure of the two section as edited - and I might suggest that any improvements keep to the point by leaving this structure intact

The first paragraph of the description section compares the basic structure of Roman numerals with the system that largely replaced them in everyday use.

The second paragraph points out that Roam numerals - even in modern use, are not altogether consistent.

The third paragraph points that the use of a consistent notation is not just desirable but actually necessary in some cases.

The following section describes the Standard form (although since we have said that there strictly ISN'T a "standard" I would really have rather kept the quotes. The intoduction to the table is kept to a minimum - basically just a statement that the table is a substitute for a set of rules.

The examples that follow the table might still be a bit "bloated" - if no one else does I may trim it a little. --Soundofmusicals (talk) 06:15, 9 August 2020 (UTC)

I had no intention of being rude, just honest. until now, my focus has been on adding the rules, which are completely neglected by the article. since that proposal is stalled, I turned my attention to the existing content, and found it lacking. it starts with a sentence (borrowed from my content) which doesn't apply -- the rules narrow it down to a single permutation per value, a basic overview doesn't do that. then the next line essentially contradicts this: is there a single set of permutations, or does it not attempt to endorse or refute various combinations? further on, it claims that Roman Numerals have digits, which they don't at all. RN are an additive/subtractive system using a set of fixed values, a more rudimentary design which pre-dates the invention of place value notation in early medieval India. of course, we can overlay digits onto RN (as the table does), but this is a modern contrivance which doesn't describe the system as it really functions. it's essentially imposing an alien structure onto RN to make it easier for us to follow (since we're native to the alien structure). the rules I've documented are actually much more relevant to how RN operate. the table, while useful, is no replacement.

Reducing this convention to a set of 'rules" that is at once comprehensive, unambiguous and easily understood is highly problematic no it isn't, because I've already done this. btw, I checked the Reddy/Khan source already used in the article, and it has a set of rules for RN overlapping my own! yet you still won't let me make a constructive addition. Xcalibur (talk) 18:09, 9 August 2020 (UTC)

The letters IVXLCDM certainly are "digits", in that they are single symbols that represent numerical values. And I think we have pretty good proof right here in the talk page that "rules" are highly problematic.Spitzak (talk) 20:40, 9 August 2020 (UTC)

no they aren't. they function more like building blocks, which are combined in different ways to generate values via addition/subtraction. it's a fundamentally different, more rudimentary system as opposed to place value notation. for example, RN 1-9 are as follows:

1, 1+1, 1+1+1, 5-1, 5, 5+1, 5+1+1, 5+1+1+1, 10-1.

while we can line these up with digits 1-9, that's a modern contrivance, and not the original design.

The word "digits" does not mean place-value notation. It means a symbol that has a numeric value.Spitzak (talk) 23:33, 9 August 2020 (UTC)

IDONTLIKEIT and misunderstandings are not proof. the rules I compiled are supported by many RS, while the arguments made against them are just opinion and conjecture. that's especially the case now that I've incorporated many constructive criticisms and refined the content. Xcalibur (talk) 21:11, 9 August 2020 (UTC)

For goodness sake, we have got past all that, I hope. Your rules never "worked", and no one ever agreed with you that they did. When "sources" come up with all kinds of rubbish and contradict each other like mad then invoking a particular one is not enough - we are forced to use our own common sense to a degree that would not be appropriate elsewhere. There were never a set of rules that shaped the system - the rules - just like the table - are a way of describing "the system". The table is obviously better in every way at doing just this and we don't (and never did) need both! In any case, what impressed me no end about your latest edits (and why I kept as much as I could) is the apparent acceptance of consensus on one point - and moving on to a general effort to improve the article. Do hope I wasn't wrong there. --Soundofmusicals (talk) 23:12, 9 August 2020 (UTC)

I have taken the liberty of re-reverting to my "new version" although I am still working on it in an attempt to get to something you might find less objectionable. The first paragrah of the "Standard" section (or something else conveying the same message) is essential - to define exactly what we are doing here. --Soundofmusicals (talk) 23:30, 9 August 2020 (UTC)

Still at it! One thing you seem to be missing, by the way, is that it is important that any description of the system doesn't attempt to "refute' or "endorse" (i.e. lay down the law). Roman numerals as they are actually used is far more complicated (one reason why "rulesets" are doomed to failure! --Soundofmusicals (talk) 23:46, 9 August 2020 (UTC)

yes they do, and they're better than ever. a couple of editors disagreeing doesn't change that fact. you can't pull up one RS refuting my ruleset, which is now backed by 8 RS (+1 already in the article); nor can you find any gaps in its logic. the sources do not fundamentally contradict, they may use different structures, they may vary in comprehensiveness, but all are in agreement and converge on the same rules. in other words, my ruleset is compiled from the RS, and I used common sense in pulling it all together, just as you said. your denial of the rules has no source except conjecture. and yes, we do need both. the table and rules complement one another. there's no reason to leave out relevant content, especially when RS already used refer to rules!

as I said, those first two sentences essentially contradict. claiming that rules are 'problematic' is baseless personal opinion. RNs don't really correspond except by a convenient, contrived method. but more importantly, there's no need to keep emphasizing that the rules and/or chart are not set in stone, because the description already says so, and we have an entire section on variant forms. Xcalibur (talk) 00:44, 10 August 2020 (UTC)

"Matching", rather than "sorting" is really closer to what the table does, for instance we "match" 3,000 with MMM. The aim here is not what any one editor thinks, but hammering out a version we can all agree to (this is what consensus means). No need to mention "rules" at all if we're not including them. (We're also not including "my" original power by power description of the "pattern", which would still be my first choice. --Soundofmusicals (talk) 21:43, 15 August 2020 (UTC)

my point was that, if you apply place-value to RN, you're looking at it through a foreign lens. it may be natural and convenient to us, but it's unrelated to the original additive/subtractive design. however, you could put it either way, so I made a slight adjustment to the other wording.

No need to mention "rules" at all if we're not including them. the rules are referred to in RS already used. moreover, their exclusion is temporary.

(We're also not including "my" original power by power description of the "pattern", which would still be my first choice. I thought that was fine actually, and could probably coexist with the table, although that would be redundant. Xcalibur (talk) 04:14, 16 August 2020 (UTC)

Actually you have a point if you mean that the table (or for that matter any of the various sets of "rules") wouldn't have meant very much to the Romans (or the Etruscans) - they didn't use place keeping zeros because they were indeed a totally alien concept. On the other hand in this section we're talking about the current convention - for much of the history of RNs they have co-existed with Arabic numbers - gradually losing their status as the usual way of writing numbers, and in the meantime becoming standardised. Incidentally - the proper section for stuff about the "original design" would be in a new section under "Origins of the system" - I'm serious - you could probably even slap in something like your rules there. For a meaningful description, however you really need to relate everything to the numbers that our readers (we hope) already use and understand. --Soundofmusicals (talk) 05:50, 16 August 2020 (UTC)

the rules took shape in the modern era. historical usage was full of variants: for example, one source mentioned IXCS for 89 1/2, which would be considerably out of bounds today. thus, rules can't be categorized under an 'origins' section, rather they should be separate but adjacent to the existing description. I agree that superimposing digits is one effective way to describe RNs, and we should do that, but that's not the only way by any means. my rules are much more consistent with the inherent additive/subtractive structure which, for all their standardization, is still retained by Roman Numerals. Xcalibur (talk) 06:18, 16 August 2020 (UTC)

Actually READ the section I referred to as a suitable home for (among other stuff) the "inherent additive/subtractive structure" (it actually may need renaming, I am talking about the content of the section rather than its heading). There is now a gap between "Earky Roman numerals" and "Use in Middle ages and Renaissance" where a really good section about "inherent additive/subtractive structure" could fit very well indeed. It would have to be really good though. --Soundofmusicals (talk) 00:02, 17 August 2020 (UTC)

additive/subtractive applies to RN at all stages, and I've already covered it in the new description. there is a gap between those sections, but it should be filled by a discussion of classical/imperial usage, which is mentioned a few times under variant forms. Xcalibur (talk) 03:06, 17 August 2020 (UTC)

RfC: Orthography/logic/rules for Roman Numerals

I'd like to add a section analyzing the conventional structure of Roman Numerals, using a logical set of rules & examples, based on the RS. The section in question can be seen here: Talk:Roman_numerals/Rules should this content be added to the article? Xcalibur (talk) 21:18, 25 July 2020 (UTC)

Since you've raised an RfC the correct thing is to preserve the status quo ante until such time as the RfC is decided differently. For this reason I have reversed your edit. Martin of Sheffield (talk) 21:46, 25 July 2020 (UTC)

Ok, that's fair enough. Xcalibur (talk) 22:03, 25 July 2020 (UTC)

As an addendum, a few other editors have raised various objections to my addition, which can be seen in the lengthy discussion above. Xcalibur (talk) 02:32, 28 July 2020 (UTC)

Survey

Put *'''Yes''' or *'''No''' with a brief reason and signature.

• Yes the content is compiled from RS, including a scholarly journal, and fills a gap in the article's coverage. the 'dual approach' of descriptive & prescriptive is also seen in the sources. in addition, I've provided examples/explanations to clarify. my proposed expansion is consistent with numerous Reliable Sources, which is my strongest argument in favor. Xcalibur (talk) 04:09, 13 August 2020 (UTC)

Comments

This has been discussed above. You then raised this at the dispute resolution noticeboard where an independent editor closed it with the comment:

Closed. The discussion is being repetitive. The filing editor can either post a Request for Comments or accept that there is a rough consensus against them.

— Robert McClenon

This is actually starting to become disruptive, may I suggest a read of WP:DROP before pursuing this any further. Martin of Sheffield (talk) 21:46, 25 July 2020 (UTC)

I simply took the next step, as advised on the noticeboard. if no progress can be made, I'll accept that, but I'd like to try out the available options first. I can just as easily argue that this is a case of status quo stonewalling. Also, I gave this a decent window of time before using other options, as you mentioned on there. Xcalibur (talk) 22:13, 25 July 2020 (UTC)

@ Bigdan201/Xcalibur (still only one of you - dishonest at best to imply otherwise). This (especially your latest re-reversion) is called disruptive editing. It is destructive of the project of Wikipedia, if only because of the constant "cleaning" work it gives others, and will very probably eventually lead to your being banned. "We" have adhered to the principle of "assumption of good faith" in your case hoping against all hope that you really ARE acting in "good faith" and will eventually see the point. Your alleged "improvements" have not attracted one iota of support - either from the several editors working on the article, or anyone else. There is a reason for this, and it is the obvious one. If you cannot accept that you are conclusively "outvoted" in this instance, and that the clear consensus is absolutely against you, I really am at a loss to know what I can possibly suggest. -Soundofmusicals (talk) 09:41, 27 July 2020 (UTC)

Yes, I'm one person. I have no intention of being disruptive, and I don't think I have been. I put up a new version because of substantial changes made (which directly addressed a major point of criticism); I abide by it being reverted in order to follow protocol during RfC. I discussed matters at length and tried to address all points, only for you to stop discussion on June 11. this forced me to seek out other venues, and even then I waited for weeks in order to give a break to all involved. I'm willing to accept consensus of course, but in this case, it's a consensus of just a few editors, led primarily by you. You can't argue from a lack of involvement, see Warnock's_dilemma and WP:SILENCE. for now, we'll see if the RfC makes a difference. Xcalibur (talk) 10:41, 27 July 2020 (UTC)

• Your effective use of two user names borders on sockpuppetry, even though there is only one user account involved. Please cease and desist, as this is disruptive.
• You did not "put up a new version" but made a major edit to the article when you had agreed that it was "fair enough" to leave the article itself intact until a consensus to the contrary was reached.
• To call the this version "new", in any real sense of the word is risible. Most, if not all the problems raised with the text you wish to add to the article continue to be ignored.
• Consensus can be (and very often is) reached by just two editors. The number of editors is actually pretty immaterial. I have said this before, but you can't deny the validity of the current consensus until someone agrees with you. --Soundofmusicals (talk) 14:11, 27 July 2020 (UTC)

it's a handle I often use, so I updated my WP account to reflect that awhile back. there's no rule against that, and I haven't acted in a misleading manner (which is the only time it would be an issue).

I was responding to Martin's request to follow protocol. there's nothing wrong with attempting to be BOLD again after I've made such a significant change to content.

but it is new, and it directly resolves one of the main critiques, namely that my content is difficult for a beginner. and so, I added a sub-section of examples and explanations to address this. that was the only valid complaint, and now it should be resolved; the rest have been proven false or irrelevant.

of course, a consensus can be a few or many. however, greater numbers and variety make for a stronger consensus, and this one is a bit lacking. my concern is that there's not enough attention or involvement in this dispute, which is why I seek other views. Xcalibur (talk) 01:24, 28 July 2020 (UTC)

@Soundofmusicals: It's not sockpuppetry. It is a bit irritating when a person's user name (what appears on history pages) and signature are different, but it is an accepted procedure adopted by quite a few editors for various reasons. Let's just focus on the issue and get the RfC over.

@Bigdan201/Xcalibur: An RfC should post a precise question without waffle. Obviously other editors object, otherwise the RfC would not be happening. An RfC does not consider whether content is "permitted". I assume the question is "Should the content at Talk:Roman numerals/Rules be added to the article in a "Standard orthography" section?". If so, that's what the question should state. If really needed you can comment about other editors in the comments section. Johnuniq (talk) 02:17, 28 July 2020 (UTC)

Received and understood. I've made the appropriate changes. Xcalibur (talk) 02:36, 28 July 2020 (UTC)

@Bigdan201 An even better question for others' comments might centre around whether a valid consensus has been established. This goes past the opinion of a single dissenting voice. And I did say borders on sockpuppetry. In particular, for people coming fresh to this "discussion", it is likely to create the impression that there are at least two people on "your side". It really would be less confusing if you picked one "wiki name". Statements like "that was the only valid complaint, and now it should be resolved; the rest have been proven false or irrelevant" go well past being "irritating". Quite apart from anything else, you have acknowledged the justice of a good many "constructive criticisms" without attempting to remedy them. The rules (after all these years!) still take more or less the same form and suffer from the same drawbacks. And, even more to the point, they remain every bit as redundant. --Soundofmusicals (talk) 04:06, 28 July 2020 (UTC)

I've made no pretense of being anyone other than Bigdan201, so that's not an issue.

but I have taken constructive criticism into account, and made many fixes and adjustments, including the addition of a new section to provide examples and explanation, which should address the 'difficulty' complaint. As for the rest, I'll cover them briefly:

"there is no modern convention" yes there is, and the RS prove it. "the sources are vague and contradictory" false, the RS define a rules-based logic with total consistency. "fractions/vinculums are not conventional" also false, every source I've consulted defines vinculums, and many discuss fractions as well. "descriptive and proscriptive shouldn't go together" they complement each other just fine, as seen in my RS. "we don't need two different explanations side-by-side" yes we do, because this provides more comprehensive coverage and greater insight into the topic; a basic description of the pattern combined with logical analysis derived from the RS is just what we need. That should answer most of the points you've raised, Soundofmusicals. as for Spitzak: his highly condensed approach is interesting, but not supported by the RS as mine is, which makes it Original Research; it also lacks the detail and comprehensiveness of my ruleset. as for inefficiency, describing a decimal system in 7 rules certainly qualifies as efficient.

that should provide a useful summary of our (admittedly long-winded) discussions. finally, this is about making a constructive, well-sourced contribution in good faith, which you don't seem to appreciate. Xcalibur (talk) 07:21, 28 July 2020 (UTC)

You are still refusing to address complaints, just claiming they don't count. Number one is: if this is a better description as you claim, IT MUST REPLACE THE CURRENT ONE. No edit that produces two descriptions of how to write Roman Numerals is acceptable. Furthermore you keep claiming your "rules" are backed by references, but I read your references and they all have about 3 rules. They also don't describe fractions or the bars and if you think those should be discussed, PUT THEM IN THE SECTIONS THAT ALREADY EXIST!!!Spitzak (talk) 16:25, 28 July 2020 (UTC)

no, it shouldn't replace the basic overview, it should be added alongside; there should be two sections (descriptive & prescriptive) because they provide different insights into the topic, and more importantly, the dual approach of descriptive/prescriptive is used by my sources. in fact, my sources each have 5-7 rules, even if they're bundled together into a few sections, so 7 rules is an accurate summary. and yes, they do cover fractions & vinculums (perhaps you skimmed over the content) so there's nothing wrong with mentioning them (in addition to the in-depth sections). Xcalibur (talk) 23:52, 28 July 2020 (UTC)

btw, added another source: [1] eta: I now have 7 sources, which should be plenty (any more is venturing into OVERCITE territory) Xcalibur (talk) 13:11, 4 August 2020 (UTC)

Debating here will not achieve anything. Anyone interested should add their brief view in the Survey section above. Johnuniq (talk) 05:02, 29 July 2020 (UTC)

More "rule" bs

There are no "official rules". The fact that you (or somebody you reference) can write a set of statements that if all obeyed will restrict the possible patterns to the subset shown by the table does not mean it "is governed by a set of rules". At best you can say "it CAN BE governed by a set of rules". Please just stop this it is getting annoying.Spitzak (talk) 22:03, 23 August 2020 (UTC)

rules can be broken by variant forms, but they are in fact governed by a set of rules. this is according to numerous RS, including the Reddy/Khan source already used in the article under Standard Form, and a scholarly journal on mathematics. the sources don't say, you can use rules if you like, they say there ARE rules. we're supposed to follow RS, or so I thought. also, Roman Numerals don't have digits, and they're not designed as a parallel to place-value notation. they have a different, more rudimentary system, which may be compared to place-value notation for convenience, but that's not inherent in them. and I don't want to annoy you at all, I'm trying to improve the article. Xcalibur (talk) 11:43, 24 August 2020 (UTC)

ps, there's no contradiction between RNs being decimal and not using place-value. they are in fact base-ten, since the fixed symbols scale up by powers of ten, rather than some other quantity. but it's base-ten in a different structure. Xcalibur (talk) 11:48, 24 August 2020 (UTC)

Roman numerals are not base 10, but bi-quinary. The symbols do not "scale up by powers of ten" but alternatively by fives and twos. They are a tally derived system and as with many tally systems tend to use simple marks for 1-4 and then distinguish 5 before repeating the sequence. If you are counting up a flock of goats the easiest way in the world is to raise one finger (digitus) for each of the first four goats. For the fifth, raise your thumb and close the fingers. 6-9 are likewise done by raising fingers. For ten, close the fist and raise one finger on the other hand. That way you can count 99 goats, actually a superior system to the Arabic numbers. I doubt that any wandering herdsman cared a brass farthing for a ruleset, they just did the accounting as part of everyday life. You keep banging on about the rules governing Roman Numerals but there aren't any. You are pushing an analysis of one variant of an existing system, not a definition. When a Roman mason or medieval scribe used a variant they were not breaking rules, there weren't any.

Consensus has been against including this, you are not improving the article merely annoying other editors, so please WP:JUSTDROPIT. B/X – if you are really interested in rules have a read of Letter case#Case styles, it will make your text easier to understand. Martin of Sheffield (talk) 12:47, 24 August 2020 (UTC)

Roman Numerals are base ten, and also bi-quinary. that is to say, they use ten as the main base, with 2 and 5 as the sub-radices. as for the rest, of course RNs were flexible in history, that's well established, with many variant forms from the ancient/medieval eras. but a standard did emerge in the modern era, and it can be described with rules or a table, that's why there's one variant given in the table. rules governing Roman Numerals but there aren't any. all the RS say there are rules. why do you keep denying well-established facts? Xcalibur (talk) 13:20, 24 August 2020 (UTC)

btw, it occurred to me that you may have assumed that this is about the 'rules' section, but it's not. leaving aside that my substantial, well-sourced work was blockaded, this is about my more recent edits to the existing content. under 'standard form', I mentioned that RNs are governed by rules, but can also be described as a decimal pattern for convenience. this is true, and the RS already used in that passage (Reddy/Khan) lists rules. even if you're opposed to covering the rules, there's no need to balk at a brief mention. Xcalibur (talk) 15:33, 24 August 2020 (UTC)

@B/X, given that the section heading is "More rules bs" and that the first two posts are about rules, then it's pretty clear your rules obsession is being discussed yet again. Martin of Sheffield (talk) 22:11, 24 August 2020 (UTC)

yes, but this time it's about a brief mention of rules under 'standard form', not the section. that's an important distinction. Xcalibur (talk) 22:56, 24 August 2020 (UTC)

A big problem with the "rules" described is that they are making modern assumptions that are not supported by evidence and directly contradict the "decimal pattern" in the very first paragraph. In particular claiming that "a lower symbol before a higher one subtracts". This allows "IC" for 99 which is no longer decimal because it does not have an 'XC' in it that is used for other numbers with 90 in the second digit. Of course you can add more "rules" to exclude the resulting patterns but that is where your "rules" quickly become unreadable verses the "these 6 arrangements are allowed" used by the current text.

As for the current mention of "rules": it is a mathematical fact that any finite set of patterns (and any countable infinite set) can be described by a grammar or "rules". So the sentence saying this "can be described by a set of rules" actually conveys zero information.Spitzak (talk) 00:05, 25 August 2020 (UTC)

they don't contradict. of course you use multiple rules in tandem, that's what all my sources do, which tend to list 5-7 rules each. there's nothing 'unreadable' about a short prescriptive list. I can describe the RN standard just as unambiguously as the chart using 7 rules. it's true that any pattern can be defined with rules, but it's still worth mentioning, especially since the rules are straightforward and numerous RS are consistent on the matter. my rules aren't made up, they're all derived from sources which are being blithely ignored. Xcalibur (talk) 02:03, 25 August 2020 (UTC)

Here is a right-hand regular grammar for the described Roman numerals, if you want "rules". This is nice because it is regular, most breakdowns are not (or they assume "IV" is a terminal). This is some kind of compressed notation, showing all the options as individual rules and making it not match the empty string will greatly expand the number of expressions. I believe each rule can be assigned a numeric value so that when summed they give the value of the number (though some of them will have unexpected values like -2):

   NOEMPTY(
       ['M'['M'['M']]]
       ['C'['D'|'M'|'C'['C']] | 'D'['C'['C'['C']]]]
       ['X'['L'|'C'|'X'['X']] | 'L'['X'['X'['X']]]]
       ['I'['V'|'X'|'I'['I']] | 'V'['I'['I'['I']]]]
   )

Spitzak (talk) 16:35, 25 August 2020 (UTC)

once again, an interesting approach, but not one I've seen in the sources. I didn't just make up those rules (that would be Original Research), I compiled them from various RS. yet for some reason, you won't let me make a constructive addition. also, RN are both decimal/base ten and non place-value, and there's no contradiction there either. Xcalibur (talk) 21:17, 25 August 2020 (UTC)

My complaint about "decimal", which I have tried to fix again, is that it says "it is basically a decimal system" (which I agree) and then in the same sentence after a comma says a bunch of reasons why this is *different* from decimal. It makes no sense, that should be a different sentence, and it would be nice to add at least one statement in support of it being decimal.

As for the rules, any edit that ends up describing how to write Roman numerals twice is unacceptable. It you think your rules are so great you MUST replace the current description. I also don't like the rules after doing further research, as they state things that there is no proof that the Romans thought of, and then add more rules to exclude these from the resulting pattern set. There are FAR too many "rules" trying to restrict subtractive notation to 6 patterns.Spitzak (talk) 00:28, 26 August 2020 (UTC)

I can see re-writing it a bit, but there's no need for an overhaul. there's really no conflict: RNs are decimal, and they function quite differently from place-value, both of these are true. in theory, you could use RNs for a different base, so that if they were octal, you'd have V for 4, X for 8, L for 32, C for 64, etc. claiming that RNs use combinations for 'digits' is superimposing place-value onto a different system. it's a convenient way to sort it, but it's not inherent; RNs are inherently about adding/subtracting set values.

I see no problem with describing Roman Numerals twice in different ways, in fact that's what some of my RS do. and the rules are not excessive -- it's possible to reduce RNs to a single pattern using just 7 rules, as I've already done. Xcalibur (talk) 01:23, 26 August 2020 (UTC)

I restored my description to the article, with an alteration to the first sentence which will hopefully clear up the issue. it should be emphasized that RNs don't have digits, that they're built on a different (more rudimentary) principle, and applying digits is a mere contrivance. btw, I approve of the change to 'translated' under standard form, although 'converted' would also work. Xcalibur (talk) 21:18, 26 August 2020 (UTC)

just to clarify, digits are not numbers 0-9, they are places within place-value notation. and place-value is closely tied to the concept of zero, so it follows that number systems that don't use place-value don't use a zero. Xcalibur (talk) 20:14, 27 August 2020 (UTC)

First paragraph of "description"

This first paragraph needs to cover these points:

1. Arabic numbers

That is what they are called in English - this may well be unfair on the Indian mathematicians who actually invented place value notation - but it refers to the actual symbols we use in the west - in fact it is the title of the Wikipedia article!

2. Decimal or "base 10" number system

Reinstating long standing consensus that this is essentially so. The "contention" is really over the definition of "decimal". The comparison with Arabic numbers really doesn't make sense unless this is mentioned.

3. Place-keeping' zeros enable a digit to represent different powers of ten

It won't be self-evident to every reader that this is the essential defining feature of place value notation. In any case doesn't hurt to remind people, and make the comparison clearer.

4. Omit value judgements about RNs being more "rudimentary"

NPOV (even if almost everyone would agree). It IS a value judgement, and adds nothing to clarity of information we want to convey.

5. Set of symbols representing fixed values

Clearer if we state this in its own sentence rather than portmanteau it into the next?

6. Correspondence or equivalence of combinations of symbols with digits

Unless this is clear nothing makes sense!

___Soundofmusicals (talk) 01:13, 15 September 2020 (UTC)

Q as a Roman numeral

Q "has been used" - apparently for several different numbers! As noted elsewhere in the article it was apparently sometimes "redundant with D - abbreviating quingenti, Latin for 500". Confusingly, the Latin words for 5,000, 50,000, 500,000 etc. also begin with a "Q", although how relevant this is - if we discount M as "abbreviating" mille and C as "abbreviating" centum - may be questionable! Usefulness and notability are not the same thing, but...

• The "use" of Q as a redundant alternative for D is definitely, well, redundant. It is (just) notable that it was used in this way in Medieval times.
• As a numeral for 5,000 Q could be "used" (although rather less than usefully) to extend the current numeral system to allow an extended (if not complete) "thousands" place - M, MM, MMM, MQ, QM, QMM, QMMM (for a numeral for 9,000 we need one for 10,000). A suggestion to this effect might bring out the people who suppose we actual NEED a way to express large numbers in RNs, and would in any case be uncited speculation, not to mention OR.
• As a numeral for 50,000 or 500,000 etc. - I would like to read the text of the citation that has been offered to "verify" this one (alas, it is not available as on-line text so someone needs to track down a print copy) - but any actual USE of Q in this way seems awfully speculative. It wouldn't fit into any known "systematic" use of RNs.
  

Perhaps a new section called "Q as a Roman numeral"? --Soundofmusicals (talk) 04:04, 19 June 2020 (UTC)

Adriano Cappelli, Lexicon Abbreviaturarum, available online here. Page 417 shows the use of a Q with an elongated tail for 500,000. SweetPotatoGolem (talk) 21:53, 19 September 2020 (UTC)

The two entries of interest are at the bottom of the page. Q is given as 500 but Q with a bar over it is 500,000. A Q with a horizontally elongated tail is shown as an alternative to Q-bar, it looks like a lazy scribal variant. Martin of Sheffield (talk) 22:09, 19 September 2020 (UTC)

A modest proposal

for quite awhile, I've been trying to add a rules-based analysis to the article, based on RS, only for a few other editors to reject this. my concern here is that 'consensus' is just a few regulars, with a lack of wider input. while I could try relisting the RfC and advertising further, I have another idea. how about I add the content, on the condition that no one who has participated in discussion so far (Soundofmusicals, Spitzak, et al) reverts it? then we'll see how stable it is. surely, if my content is the dumpster fire you claim it is, there's no way it could stay up, especially when it's a very visible change to a well-trafficked article such as this. I realize this is unorthodox, but I think it's a worthwhile experiment. what do you say? Xcalibur (talk) 09:38, 6 September 2020 (UTC)

For "quite a while" read "several years". If Spitzak, yours truly, or any one of the many other editors who have NOT approved this quite unnecessary, totally useless, and plain confusing nonsense fail to step straight in and revert anything of the kind I suspect there is always someone else who will. Try to accept that not every idea any one of has has is a good one, and this is a real stinker. Reverting any such edit you make will in fact be a very much more useful change than any of the myriad (if repeated) edits you have made to this article. This is extremely sad, as there is a lot I'm sure you are capable of doing that WOULD be an improvement to this articles (and mo doubt others). There - can't say you are not forewarned that no one is buying the latest permutation of the old story. --Soundofmusicals (talk) 12:14, 6 September 2020 (UTC)

The full force of the above applies - at least as far as I am concerned - but after all this time I persist in giving you the credit for being in good faith, and sincerely I do want to avoid hurting you (or anyone). No article I have ever edited (and I suspect this is pretty typical for other Wikipedia editors) has the precise contents, form, and wording that I would personally consider the "optimum". No article in Wikipedia is the property of one editor - every article is in fact more or less of a compromise between differing points of view (hopefully producing better articles than any of us could have written on our own). The only way of "pushing through" a drastic and universally unpopular change to an article, especially a "well-thumbed" one, is to convince your peers. No sneaky tricks, no repeated re-reverts, just changing the consensus. Lots of very good guideline articles to read if you think I'm making this up! You have had a very good trot on this one - no one has formally called you out for disruptive editing (or even edit-warring) as they very well might have. Reams of kindly (kindly-meant, anyway) advice has been spent - greatly wasting the time of other sincere, well-meant editors whose only apparent fault is disagreeing with your goodself. With the kindest will in the world it is high time you simply gave up (we have all had to do this, in similar case). --Soundofmusicals (talk) 12:43, 6 September 2020 (UTC)

thanks for your response. this began 2 years ago, and I picked it up again after a wikibreak. yes, content disputes are decided by consensus (with no way of appealing to experts, unfortunately). I don't want to waste anyone's time & energy, or my own for that matter; I've put a fair amount of effort into developing the rules section, incorporating many of your constructive criticisms (eg folding rules together, clarifying rule 4, moving 'powers of ten' to the top, adding examples, etc). perhaps there's nothing to be done, but I'd like to try first. If Spitzak, yours truly, or any one of the many other editors who have NOT approved this quite unnecessary, totally useless, and plain confusing nonsense fail to step straight in and revert anything of the kind I suspect there is always someone else who will. that's what I want to find out! if I leave this up, will the article be stable? will there be someone else to oppose the change? I think it's a worthwhile, if unorthodox experiment. keep in mind, it may not work out in my favor, perhaps the consensus against me will only strengthen, then I'd have no choice but to concede. Xcalibur (talk) 06:15, 7 September 2020 (UTC)

Can you confirm that I have understood your proposal correctly? You are proposing making a change, and asking all those who have contributed to the article to do nothing and see if some random bystander disagrees with the great B/X, and just to put us in our place every other editor (apparently bar yourself) is by your definition not an expert. Unorthodox would be a very mild way of putting this. Frankly I'm amazed at the forbearance of Soundofmusicals who has remained calm and polite throughout and who certainly has expertise, even if disagreeing with you disqualifies him from being an expert in your personal eyes. Martin of Sheffield (talk) 21:37, 7 September 2020 (UTC)

First of all, I'm not an expert, and I don't claim to be. my mention of expertise was a comment on WP as a whole, where arbitration is only for behavior, and content disputes are decided by simple consensus (with a few avenues to assist in this). consensus is often good enough, but sometimes it goes astray, and it would be helpful if there were (hypothetically) a panel of experts I could appeal to. of course, too much reliance on expertise and credentialism invites its own problems, so I'm not in Larry Sanger's camp. still, a balanced element of expertise would be helpful, but all this is going off on a tangent.

my concern is that there hasn't been enough attention brought to my content. what better way to solve this than to plunk it down in the article and see what other, uninvolved parties have to say? admittedly this is an innovation of mine, but I think it's worthwhile, and certainly wouldn't do any harm. and as I said, there's no guarantee it would turn out in my favor. Xcalibur (talk) 10:56, 8 September 2020 (UTC)

• an update to all concerned: I've done an experimental restoration of the content. it occurred to me that this lack of cooperation may be because you haven't checked the latest version, and are judging it by past iterations. the section is quite different now, and resolves the most relevant complaints (ie that it's cryptic and hard to follow, so I added examples to illustrate). I've reinstated it to make the new & improved version easily visible. I humbly ask that you read through to see all the differences before taking action. hopefully you'll see that it's much more viable now. Xcalibur (talk) 20:02, 14 September 2020 (UTC)

You failed to REPLACE the description with your new one. Will you just stop wasting everybody's time when even you seem to acknowledge that your description is not usable and that another one is required.Spitzak (talk) 21:16, 14 September 2020 (UTC)

The "modest proposal" was that if your "humble" reinstatement of your "rules" wasn't reverted by an editor not not part of the current consensus then it should be allowed to stand - well, now it seems to have been more or less instantly reverted by by a "new" editor - so all your "experiment" has done is confirm the consensus. Is this just one more person who is obviously wrong for disagreeing with you, or will you accept the result of your "experiment"? --Soundofmusicals (talk) 22:20, 14 September 2020 (UTC)

I primarily wanted to bring attention to the latest version, which differs markedly from past iterations (changes to the rules, extra sources, a new section of examples to help explain). in fact this was productive, because the latest objection is that my sources are insufficient. a fair criticism, and one that can hopefully be remedied. Xcalibur (talk) 14:17, 16 September 2020 (UTC)

first of all, @Spitzak: I have no idea why you insist upon replacing existing content with the rules. each section serves a particular purpose, they are not redundant. we need a basic description that covers decimal, fixed symbols, additive/subtractive instead of place-value, etc., then there's the table that shows the set of legal numerals, and then there's rules-based analysis of why that set is legal. each one represents a different sort of content. I shouldn't replace the description with the rules, any more than I should replace the description with the table. there's room for all of these. Xcalibur (talk) 01:32, 17 September 2020 (UTC)

At least the discussion is moving forward. Mathglot stated in an edit summary: I see no support for the assertion "The modern era has seen the emergence of a standardized orthography for roman numerals" in any of your eight sources, seven of which are blogs or equivalents, and one reliable source (Shaw, 1938) which quotes "The Tutor's Assistant" (Walkingame, 1845) thus hardly a modern standard. This appears to be OR with cherry-picked sources. Shaw is the strongest of my RS, but not the only one. as for it referencing an 1845 work, that doesn't disqualify it from being modern -- I see 'modern' in this context as referring to the past 2 centuries (19th and 20th). if the issue is my assertion that a standard emerged in the modern era, I can always rewrite the lede and leave that out.

here are my sources: admittedly LURNC/Seitz is blog-like, but it's excellent and widely duplicated, idk if that counts for anything, but it should. nmsu.edu is an academic resource, which qualifies it as an RS. Paul Lewis is blog-like, factmonster and edugain may be RS (not sure), mytecbits is blog-like. thus, my sources vary in reliability, but are stronger than you admit. they're certainly not cherry-picked, as I didn't leave out any dissenting views -- I searched all sorts of sources, and they all converged on the same ruleset. my content is not OR, because it's entirely drawn from the sources; we're allowed some freedom in structuring and formatting information, as long as it's derived from RS. finally, if what you're asking for is stronger RS, I've come up with two more: an educational resource and a scholarly journal. I could also add the Reddy/Khan RS used in the article, a pharmacy textbook which begins with rules. surely this should be sufficient? Xcalibur (talk) 01:52, 17 September 2020 (UTC)

What you need is ONE or perhaps two non-blog sources that give a set of rules - preferably identical to yours - or at least broadly similar. A much shorter list than yours obviously would not count. Multiple sources, mostly blogs, some of which list rules but none of which resemble your rules in particular is not "citation" at all - but as I have pointed out more than once an "invocation". Please read WP:RS, WP:CS and even WP:REFB (there are other guidelines etc. you can probably find for yourself which would give you a better idea. Failing this, Mathglot's edit summary is, alas, pretty accurate - all of your "rules" edits are indeed :cherry-picked" (where they are not pure O.R. --Soundofmusicals (talk) 09:46, 18 September 2020 (UTC)

none of this is an issue. I now have 2 scholarly journals, 3 academic sources, and 2 reasonably good sources; thus my sources are mostly RS with a few blogs thrown in. each and every one lists rules which are broadly similar to mine, and every clause of every rule I put up is from a source. they're not "much shorter", only more compounded. keep in mind, we're allowed a bit of freedom in formatting and structuring content, we don't have to copy & paste from sources, in fact paraphrasing is encouraged (as long as it's accurate, which this is). none of this is cherry-picked, as I didn't selectively leave anything out; and none of it is OR, because every bit is sourced. I encourage you to read and analyze the sources I've provided, instead of just skimming, and you'll see my point. Xcalibur (talk) 16:57, 18 September 2020 (UTC)

There is a world of difference between "allowing a bit of freedom in formatting and structuring content" and formulating your own own more or less unsupported OR - especially while constantly bleating about how closely you are "following RS" (to be followed with a justification for not following it closely at all). A blunderbus of every source you can lay your hand on - most of which have little or no relevance to the point you are trying to make, is just not how we do references in Wikipedia (or anywhere else that I am aware of). Just one reputable source with something MUCH more like your "set of rules" is necessary - even if "every bit WERE sourced" - far from the case, and anyway it's up to your to specify this, this is in fact a classic description of "cherry picking". And none of this bears on the questions of redundancy, introduction of unnecessary confusion, and above all blatant disregard for consensus (if that isn't "ownership" I don't know what is!). --Soundofmusicals (talk) 00:57, 19 September 2020 (UTC)

in response to Spitzak's edits: throughout this whole discussion (with Soundofmusicals telling me to get off his lawn), no one suggested splitting up my section and incorporating the content into existing sections. I'd be totally fine with that, if it's done properly. unfortunately, you decided to be innovative and put up your own set of rules, which have two problems: 1. the rule which states 'the shortest pattern must be used' is not to be found in any of the numerous sources I've provided; unlike my rules, it's definitely OR. 2. while your rules cover alot of ground, they're not foolproof as mine are. for example, 14 is XIV. but why are IXV and VIX illegal? after all, they're all 3 characters in length. this is answered by my rules, specifically rules 6 & 7 (subtract>add, at least 10x). there should not be any gaps in logic.

perhaps just adding the ruleset under standard form would work. I still think examples are a good idea however, since they resolve one of the most common critiques, that my content is advanced and tricky to follow. Xcalibur (talk) 17:08, 18 September 2020 (UTC)

VIX would not work because the V would be subtracted.

You are right that IXV would be allowed by the text I wrote. I was thinking of changing it to instead say "only !XC can be subtracted, can only be followed by one larger digit, which must be immediately after it and can only be 5x or 10x the subtracted value". This forces there to only be one (since IIX has an I not followed by a larger digit) and also prevents IXV (because there are two larger digits after the I). In any case, the crazy amount of text being used to restrict the output to six 2-letter patterns is why people are mostly objecting to these "rules". They don't really work.

Your rules are still ridiculous redundant. You don't need to state that the powers are written from highest to lowest, since any other arrangement would cause some of the digits to be subtracted. And I greatly prefer "shortest" as a rule rather than rules describing the results, which get confusing when saying how many times a symbol can be repeated when one of them could also be subtracted. Also I can only repeat 3 times, not 4 as you state.

Spitzak (talk) 18:56, 18 September 2020 (UTC)

maybe, but I still think rule 7 (at least 10x) helps clarify. rule 6 (subtract>add) is indispensable -- in a single stroke it rules out not only combos like IXV or IXX, but also redundant combos like IXI or IXIII. you're correct in stating that rule 1 (powers of ten) is redundant; while I could leave it out, I'd rather keep it simply because it helps clarify and explain the rest. the rest of the rules have to stay to keep the logic complete and consistent. yes, you can repeat up to 3 times sequentially, but subtraction allows 4 non-sequential repetitions, eg XXXIX has 4 Xs but is still legal; I was careful to distinguish between sequential and non-sequential. as I said, 'use the shortest pattern' is an interesting way of summarizing the logic, but I haven't seen this published anywhere.

In any case, the crazy amount of text being used to restrict the output to six 2-letter patterns is why people are mostly objecting to these "rules". They don't really work. naturally, if you don't seal every gap efficiently, you'll run into issues, which is probably why the editors here are wary of rules. however, I think I've finally pulled it off, I've completely subsumed the underlying logic of the standard set of RNs (1-3,999) within 7 rules, with the help of RS of course. this should qualify as efficient. Xcalibur (talk) 23:30, 18 September 2020 (UTC)

References

Standard notation

There are theoretically dozens of ways we could demonstrate the standard RN notation. My own preference was a 'place by place' illustration of the common pattern - another editor arranged this as an even simpler and clearer table which makes essentially the same point. Attempts to list "rules" have proved highly problematic - at best they are less than comprehensive, and have caused misunderstandings. This is the current consensus, accepted by everyone except one editor. Please accept it. --Soundofmusicals (talk) 22:02, 12 October 2020 (UTC)

I'm not trying to add my content, I'm only adding a brief reference to rules under the Standard Form section. this is sensible, especially since the Reddy/Khan source used there lists rules in the first two pages! this aversion to any mention of the fact that there are rules/logic given in RS, as if it's some sort of taboo, really is strange. a brief mention is not much to ask, and shouldn't cause issues. Xcalibur (talk) 22:45, 12 October 2020 (UTC)

All patterns can be described by a set of "rules" so your statement is meaningless. If we want "rules" we should at least use standard notations, such as LALR parser grammar, or something.Spitzak (talk) 02:09, 13 October 2020 (UTC)

yes, patterns can be described by rules, that doesn't make this meaningless at all. using LALR parsing or similar, aside from probably being OR, is too technical for the layman reader. my content, on the other hand, contains rules that are straightforward and derived from numerous RS. but this isn't about my comprehensive, foolproof ruleset that was wrongfully excluded. rather, you're balking at a single passing reference, which is supported by the Reddy/Khan source already used, from which I quote: In the usage of Roman Numerals, the following set of rules applies:. we're supposed to follow the RS, or so I thought. Xcalibur (talk) 22:32, 13 October 2020 (UTC)

Enough already! --Soundofmusicals (talk) 05:02, 17 October 2020 (UTC)

again, it's a single passing reference, directly supported by the RS already used there. I don't see what the problem could possibly be, unless it's sheer spite. Xcalibur (talk) 05:06, 17 October 2020 (UTC)

But RNs are very simply NOT governed by rules - in fact they're not actually "governed" at all. The problem is that if you want a change in this VERY long standing consensus you'll have to have it formally changed. Constant and repeated disruptive editing is not the way. We STILL assume good faith in your case - why I am starting to wonder.
Even if they were "based on RS", which yours never have been - and incidentally quite rightly so (!) - since the sets of rules in the sources are much more limited than yours in several respects, rules are very simply not the way to go here. The very purpose of rules is to proscribe rather than describe. We do not, especially at this point, and under the heading "description", want a list of what you are not "allowed" to do - but a concise and comprehensive description of what is (usually) done.
This needs to be followed (as it is in the current text) by a further description of the many other things that are sometimes done (or used to be done) instead of what is usually done - under the heading "variant forms". Between them, these two sections cover everything meaningful that a set of rules could possibly convey. The answer to anyone who wants our permission to do something else again is NOT "you're not allowed to do that" but "go ahead and do that if you like, but it's not the way it's usually done".
So basically, however good our rules were they are not required. (Nor, just quietly, is a mention of them - although no one, I fear, is naive enough to fall for that one).
But, if that were not enough, your rules don't even adequately describe everything that IS done, nor do they adequately proscribe everything that (usually) isn't. In other words, they don't work. Nor, perhaps, do anyone else's, but that's not a lot of help. Over the years several people (NOT just me) have raised various difficulties with certain of your rules - and you, to be fair have at times made valiant attempts to resolve the various difficulties. At one stage, for instance, you went so far as to invent a unique jargon ("ones" and "fives") for two classes of RN symbol, (I,X,C and M) and (V,L and D). Fair enough, although to properly explain this to a complete tyro is only marginally less fiddly than a full "descriptive" explanation of the whole system. But jargon only works when it is widely known, and has an agreed meaning.
--Soundofmusicals (talk) 10:13, 17 October 2020 (UTC)

I don't believe a single brief reference (supported by an RS) qualifies as disruptive. my rules are in fact based on RS. some are more comprehensive, some less so, but they converge on the ruleset I put up. every clause is based on a source, and Seitz in particular had lots of coverage and was a major influence on my work. the RS are stronger than ever, with scholarly journals, .edu sites, the pharmacy textbook, and so on. take another look, you'll see this is no issue. I agree that we should have a basic intro, a description, and a section on variants. however, I think there's also room for documenting rules/analysis which is commonly seen in the RS, especially because they directly answer common questions about RNs. there's no need for rules to replace another section, just as there's no need for the table of RNs to replace the description either -- they serve different purposes. I agree that the rules are not hard and fast, rather they analyze the common convention as seen in Constitutional Amendments, Superbowls, prefaces of books, and so on. you CAN use variant forms, but it's discouraged.

I did indeed make many revisions to my work, some of which were in direct response to your constructive criticisms. it took awhile for me to create a finished product, to seal all gaps and iron out inefficiencies, but I've finally done so (with your help, in fact). the 'jargon' isn't really a problem, the sources refer to 'powers of five' and 'powers of ten', and I simply shortened this to 'fives' and 'tens'. I also rewrote the rules and created a new section of examples to illustrate, all to resolve the common complaint that my work was too difficult for the layman. But, if that were not enough, your rules don't even adequately describe everything that IS done, nor do they adequately proscribe everything that (usually) isn't. In other words, they don't work. this is simply incorrect. the current form of my ruleset does work, it describes all that is done, and proscribes all that typically isn't. earlier on, there were gaps (eg I had to expand rule 4) but now it's complete and polished. for reference, here's the section as it currently stands: Talk:Roman_numerals/Rules. I invite you, or anyone else reading this, to find a single weakness in its logic or style, because I don't think that's possible. Xcalibur (talk) 03:09, 18 October 2020 (UTC)

Two reversions I have NOT made (for the time being at least)

If only in the name of peace, and to extract a (necessary) renunciation of further nonsense about "rules" - two of the efforts of the same editor that we have left stet do need to be reconsidered, at least.
The first is the passage in the lead which describes Arabic numerals as being more "efficient" than Roman ones. The original text at this point described them as being more "convenient" (which they are). In fact the "convenience" of Arabic numbers is very much more to the point than their "efficiency", especially as an explanation of why we use them. A small point this - but the change to "efficient" stems from an unexplained and unjustified revert. Perhaps when the dust sets over the "rules" bit?
The other, more pressing, thing we need to fix is the current version of the first paragraph of the "Description" section. The apparent meaning (when one has teased it out) is pretty well spot on, but the wording is obviously very poor from the point of view of clarity and comprehensibility. Not something that a reader coming "fresh" to the subject will get much out of. --Soundofmusicals (talk) 13:37, 6 September 2020 (UTC)

but why are hindu-arabic numerals more 'convenient'? they're only convenient because we're thoroughly accustomed to using them, so that they're our 'native tongue' in mathematics. they're actually built on a more complex concept than Roman Numerals -- place-value notation as opposed to additive/subtractive. for those used to RNs, they did not lack convenience, which is why they were used by Medieval Europe centuries after the fall of Rome. while convenience is debatable, efficiency is a matter of fact. it can be proven that hindu-arabic numerals and their place-value are a more effective, powerful, and efficient system for processing numbers. thus, the change to the wording.

and there's nothing wrong with the first paragraph under Description. it describes how Roman Numerals really work, with each statement linked to the next, and a segue into the next paragraph. maybe it could be elaborated, but otherwise there's no issue there. Xcalibur (talk) 00:10, 7 September 2020 (UTC)

Of course you are perfectly right that the "convenience" of Arabic numerals has a lot to do with how familiar we are with them, but as you hint yourself, for centuries the situation was reversed and Roman numbers were the more familiar, so that if that were all there was to it we'd never have adopted Arabic ones. The point, is what is the very best word in this context? Given that Arabic numbers are indeed more "effective", "powerful", and "efficient" - why are these things important? Might it just be because they are in the last analysis more "convenient"? That word sort of sums up the others, doesn't it? At least in this context. So it does seem a shame to have changed it, for no good reason. Not that it's such a major thing we couldn't bear it at a pinch. Worth a lot more than that to get you off our backs about so-called "rules".
Much more to the point, I never said that the first paragraph doesn't cover most of the bases so far as the meaning we want to convey. Point is it conveys it rather badly "from the point of view of clarity and comprehensibility". It may even need simplification rather than elaboration but it DOES need to be clear and comprehensible. It IS clear and comprehensible, I hear you say. The very intelligent person I tried it out on couldn't understand a word. The "only numbers she has ever used" are Arabic ones, mind you (until I pointed out the case of Henry VIII) - but I have a feeling it would stop the average reader in their tracks, with an impression that RNs are very much more difficult than they really are. Not what we want, at least in an encyclopedia article... --Soundofmusicals (talk) 03:04, 7 September 2020 (UTC)

I'd say the word 'efficient' sums up the qualities of hindu-arabic numerals. there's nothing inconvenient about RNs, for centuries they were good enough for most common uses. however, they lack the sophistication and capability of hindu-arabic, place-value numbers. our number system is mathematically stronger, a sense which I capture with 'efficient'.

it seems clear and concise enough to me, especially since it follows after the lede, which does its own part in explaining the topic. the description should cover the following: RNs use fixed symbols, these are added/subtracted, this is different from place-value notation, the fixed symbols scale up by powers of ten instead of using place value, they form various additive/subtractive combinations, and this allows flexibility. if there's a better way of capturing those points, I'd be interested. Xcalibur (talk) 03:29, 7 September 2020 (UTC)

"Efficient" in this context is actually very vague indeed - certainly far from the best word available - but as I have said a number of times, all of can't have all our own way - so I'll leave this for someone else to fix.
Unfortunately "concise" and "clear" are two quite different things. While we want to be concise, especially in a long article like this one, clarity is much more important again. Anyway, that paragraph definitely needs fixing, so that it not only means something, but what it means is crystal clear. Just so I can't be accused of "owning" this one I'll reinstate a better version (one of Spitzak's perhaps) when I get around to it. Or, better still, someone else will do it for us. --Soundofmusicals (talk) 04:27, 8 September 2020 (UTC)

'efficient' isn't vague, it's accurate. of course, the word could just as easily be 'sophisticated' or 'advanced'. 'convenient' on the other hand doesn't cover the difference.

if you or someone else would like to rewrite it, then by all means, go ahead. the main thing is to describe the difference between additive/subtractive and place-value notation, and not conflate the two. Xcalibur (talk) 11:01, 8 September 2020 (UTC)

Actually "efficient" is a pretty poor choice. It would be easy to claim that Roman Numerals are 3x as "efficient" at representing 100 (as "C" which is 1 character, not 3). I would certainly revert that back to "convenient".Spitzak (talk) 18:59, 18 September 2020 (UTC)

Agree with Spitzak. Roman numerals are more efficient for finger counting and tallying, which is what they were developed from. Qualitative judgements like "better" or "more efficient" are not universally applicable. Martin of Sheffield (talk) 20:10, 18 September 2020 (UTC)

Couldn't agree more - why I find the term "vague" (and for that matter, in this context, unencyclopedic) is that there are two different basic senses of the word - the mathematical/scientific - with quite a precise meaning - and the "common" which is actually very ambiguous and vague indeed. I think we have a very good case here for changing (either back to "convenient" or to another word that fits the meaning here) if only on the grounds of long standing and overwhelming consensus. Just don't ask me to do the changing, since I am on the block here for being stubborn! --Soundofmusicals (talk) 23:42, 18 September 2020 (UTC)

I was thinking of cases like 3,888, which is certainly more efficient than MMMDCCCLXXXVIII. but you're right, while hindu-arabic are more efficient in some ways, RNs have the advantage in other contexts (as mentioned above). in light of this, I've changed it to 'sophisticated', which is far less debatable than other qualities. hindu-arabic numerals and place-value are in fact a more sophisticated, advanced system than RNs, hopefully we can agree on that. Xcalibur (talk) 23:58, 18 September 2020 (UTC)

"Sophisticated" is at least an apparently sincere attempt to find an other term - but unfortunately it is even more ambiguous (just look it up in a good dictionary) and even more NPOS! While I am wilting under accusations of "owning" the article (talk about pot calling kettle black!!) I will not be the one to make this relatively minor change, but what the [VERY naughty word expunged to protect my image as a nice old man] is wrong with "convenient"? --Soundofmusicals (talk) 00:31, 19 September 2020 (UTC)

I already answered this above. 'convenient' is debatable, and if RNs are more efficient for tallying or writing 100 or 1000, it follows that they're more convenient for those uses as well. we want a word that captures the consistent advantage of using hindu-arabic, and 'sophisticated' is a good candidate, as place-value is undeniably a more advanced system than additive/subtractive. Xcalibur (talk) 01:05, 19 September 2020 (UTC)

I really don't know why you feel constrained to start so many o your talk posts with a direct terminological inexactitude. You quite specifically avoid answering either of my objections to "sophisticated" - firstly that it is ambiguous (viz. the dictionary) and secondly - assuming you take the meaning of "advanced" - it is very NPOV! Not up to us to make value judgements about RNs, however universal or obvious they might seem. "Convenient" at least states a single more or less value free fact. --Soundofmusicals (talk) 01:39, 19 September 2020 (UTC)

convenience is not a fact. RNs are not inconvenient, in fact (as others pointed out) for certain uses such as tallying, they're more convenient. 'convenience' is the ambiguous concept here, since much of it is subjective. sophistication/advancement is the value-free fact, since it's based on the respective mathematical structures of the two systems. Xcalibur (talk) 02:37, 19 September 2020 (UTC)

Again - your remarks are not exactly logical - if "sephistication" IS a fact - how is "convenience' not? This is an encyclopedia, and value judgements are something we do (rightly) try very hard to avoid, or at least mitigate, even in milder cases than this one. As I have said several times - not the biggest deal in the world - but all other things being equal it is a shame, at least, to go against consensus, especially such long standing consensus, even on a minor point like this. --Soundofmusicals (talk) 06:08, 19 September 2020 (UTC)

RNs are convenient enough, if they weren't, they would not have persisted in common use for centuries after antiquity. if anything, the switching cost of using a new number system would've been inconvenient. the reason why we switched over, and why hindu-arabic numerals dominate worldwide, is because they're objectively superior to traditional systems. that's what I'm referencing with my word choice. Xcalibur (talk) 10:09, 19 September 2020 (UTC)

"objectively superior to traditional systems" – no unless you qualify this with "generally" or similar phrasing. Remember that one exception breaks a generalisation. Let's do a quick gedankenexperiment. Hold up the fingers for 47 sheep counted into the fold. Tally five more (one at a time), then write down the result. No cheating, do it all on your fingers. It's a lot easier to use (LH: four fingers up, RH: thumb and two fingers) = XXXXVII count then (LH:thumb, RH two fingers) = LII. Note I used the original (additive only) technique. It's an "objectively superior" system in this instance, ergo you cannot make the generalisations. Martin of Sheffield (talk) 14:17, 19 September 2020 (UTC)

yes, I should've clarified that hindu-arabic numerals are superior in most contexts. there are exceptions such as tallying, but for the most part, place-value is more effective and versatile. Xcalibur (talk) 20:11, 19 September 2020 (UTC)

This conversation seems to have gotten a bit heated, hope it's okay I chip in with some thoughts: this question of whether Roman numerals were "less efficient" or "less sophisticated" than Arabic numerals is very characteristic of how moderns have historically approached ancient numeracy, an in fact it's incredibly common to see this claim in some older scholarship. I'd say that it is fair to say that Roman numerals are "less sophisticated" than Arabic numerals, if we hold them to the same standards, and expect that both systems are supposed to perform same functions. We use Arabic numerals as calculation tools, for which they are very efficient, but it is maybe a bit misguided to compare Roman numerals against this standard. Currently, I think, the article does not adequately reflect the fact that the Roman numerals were NEVER intended as calculation tools. Although some moderns have tried to develop inventive methods for calculating with Roman numerals (e.g. Detlefsen, Erlandson et al. (1976) 'Computation with Roman numerals', in Archive for History of Exact Sciences 15.2: 141-148), we have absolutely no evidence that any Roman ever tried to calculate (addition, subtraction, multiplication, division...) with Roman numerals. Romans calculated with abaci and counting pebbles, or simple addition/subtraction with fingers and tally marks. Roman numerals were developed only to be mnemonic and communicative tools. For this, they are at least just as efficient and good as Arabic numerals? Somebody might even say that the (standard) Roman numerals have some advances, since they are fairly "visual" representations of numbers and take advantage of the fact that human and animal cognition has the innate ability to apprehend small numbers up to four (even pigeons have this ability: BrysBaert, 2005, Number Recognition in Different Formats in Handbook of Mathematical Cognition, p. 24). So, reading and using at least smallish Roman numerals might have required less education and arcane knowledge than Arabic numerals. In summary: I'm okay with saying that Arabic numerals are more "sophisticated" than Roman numerals for the purposes of calculation, but there should be some comment to make clear that Roman numerals weren't ever used for calculation. Mythoplokos (talk) 12:54, 8 November 2020 (UTC)

Welcome to this discussion! Our current use of the far from ideal "sophisticated" here is an attempt at a compromise on this relatively trivial matter, to which the users concerned can at least be reconciled. BUT what do you think of our original word, which was "convenient"? Perhaps something like "oonvenient, at least for the purposes of calculation"? "Efficient" is of course plain wrong (as a word to convey our precise meaning here) - "sophisticated" is very little better - in many respects Arabic numerals are not a "sophisticated" form of Roman numerals (albeit they share a common "decimal" base, but a fundamentally new system, based on a feature (the place-keeping zero) essentially alien to Roman-era Europe, and originating in the mathematical insights of South Asia. --Soundofmusicals (talk) 15:04, 8 November 2020 (UTC)

Thank you, glad to be here! I am new to contributing to Wikipedia, so please everyone be patient :D I know the word-choice might be a bit of a trivial question, but I think this issue addresses the wider (and interesting!) questions of social contexts and cognition re: numeracy. Traditionally, Roman numerals have been much 'derided' for the supposed lack of sophistication in modern scholarship, scholars used to even hold that the failure to develop a positional system somehow stopped Romans from advancing further in sciences (even when they clearly did make huge leaps in e.g. engineering). Current scholarship is basically in the process of abandoning this efficiency/sophistication approach. The move to Arabic numerals was very slow - people were aware of them in Europe already during the 10th c., but the movement towards them intensified between the 12th and 17th cs. - and is directly related to how people took their time to embrace a completely new way of calculating and conceptualising numbers, and replace the abacus. There were actually multiple attempts during the Medieval period to develop ciphered-positional systems for Latin numerals, so that they could be used to calculate with the same rationale as Arabic numerals, but none of them took hold. So, IDK how to efficiently reflect this in the introduction. Don't know if 'convenient' alone quite captures it (although it is better than "sophisticated"!), I really do think that this is non-calculating element is quite an important element in the history and use of Roman numerals. Maybe I would rewrite the paragraph slightly in the lines of: "The use of Roman numerals continued long after the decline of the Roman Empire. From the 14th century [?] on, Roman numerals began to be replaced in most contexts by Arabic numerals, which had better potential for calculation: Roman numerals were originally developed and used only for communication. However, this process was gradual, and the use of Roman numerals persists in some minor applications to this day." Mythoplokos (talk) 17:18, 8 November 2020 (UTC)

A lot of good ideas there - but I think we probably shouldn't get bogged down with too much of this sort of thing in the "lede" - which is supposed to be a purely introductory section. I have actually cut the Gordian Knot here - hope everyone will at least agree that no adjective at all here is better than an inappropriate one. This by no means bars us from inserting some of the detail suggested by Mythoplokos at an appropriate juncture in the body of the article! -Soundofmusicals (talk) 02:09, 10 November 2020 (UTC)

this is fine, I have no objection to removing the adjective. Xcalibur (talk) 04:01, 10 November 2020 (UTC)

Section on rules/analysis (for reference)

Since the discussion has stalled out for the time being, with no apparent way forward, I've decided to document the relevant content in this section. this way, it's convenient and accessible for any interested parties, with no need to go spelunking through the talk page.

here's my finished rules section, which analyzes the conventional style of Roman Numerals by collating numerous RS. it also provides examples to illustrate the rules, and extends coverage to fractions & vinculums. [2]

Rules-based section

Standard form

The modern era has seen the emergence of a standardized orthography for Roman numerals, which permits only one permutation for any given value. This system may be described as a decimal pattern, as above, but also as a logical set of rules. While exceptions can be made (notably IIII instead of IV on clockfaces), the modern convention is widely recognized and adhered to, and may be prescribed by the following ruleset:

Basic Rules

1. Powers of ten are dealt with separately, ordered from greatest to smallest and from left to right, with numerals either added or subtracted.
2. Repeated 'tens' numerals (I, X, C, M) are added together, with up to three (3) permitted in sequence (subtraction allows an additional non-sequential repetition once per power).
3. 'Fives' numerals (V, L, D) may only be added once per power (not repeated nor subtracted).
4. Numerals placed to the right of a greater value are added, while those placed to the left are subtracted; if placed between two larger numerals, the 'tens' value is subtracted (i.e. subtraction takes precedence).
5. Only one (1) 'tens' numeral may be subtracted from a single numeral per power, and this must be by 1/5 or 1/10 (i.e. the next lower 'tens' numeral).
6. Addition must be less than subtraction for any given numeral (i.e. the added value to the right must be less than the subtractor).
7. Any subtracted 'tens' numeral must either be first, or be preceded by a numeral at least ten times (10x) greater.

The above describes the basic pattern of integers from 1 - 3,999.

Examples

Following the above rules:
90 must be XC. It cannot be LXXXX because that would break rule 2 (up to three in sequence), and it cannot be LXL because that would break rule 3 ('fives' not repeated).
45 must be XLV. It cannot be VL because that would break rule 3 ('fives' not subtracted).
99 must be XCIX. It cannot be IC because that would break rule 5 (subtract by one fifth or tenth).
18 must be XVIII. It cannot be IIXX or IXIX because these would break rule 5 (subtract once per power).
19 must be XIX. It cannot be IXX because that would break either rule 5 (subtract by one fifth or tenth) or rule 6 (subtract>add).
10 must be X. It cannot be IXI because that would break rule 6 (subtract>add).
14 must be XIV. It cannot be IXV because that would break rule 6 (subtract>add), and it cannot be VIX because that would break rule 7 (at least 10x).

1894 is MDCCCXCIV. Consider how this agrees with the rules: powers of ten are arranged properly (rule 1), C is used three times sequentially and a fourth time non-sequentially (rule 2), 'fives' numerals D and V are added once each (rule 3), X and I are subtracted by being placed to the left of larger values while the rest are added (rule 4), they are subtracted once per power by one tenth and one fifth respectively (rule 5), IV is less than X (rule 6), and C is 10x X and 100x I (rule 7).

Fractions & Vinculums

The following rules extend the system further:

• Fractions may only be added once to the right of all numerals (not subtracted), without redundancy, and are duodecimal.
• Beginning at 6/12, the S (semi) is used, followed by marks (unciae) for additional quantities up to 11/12.
• Placing a bar (vinculum) over a numeral multiplies it x1000.
• The vinculum is used for values of 4,000 and greater (it is not used up to 3,999), and may be iterated up to three (3) times.
• Starting at 4,000, and at the next two powers of a thousand thereafter, a new vinculum is added, with the pattern beginning at IV.
• Once vinculums are employed, the higher power is modified in preference to the lower (eg IV, not MV).
• Vinculums are always contiguous and are placed leftmost (ie no broken overlines or lower powers to the left per basic rule 1).

Thus, 2 2/3 would be rendered as IIS••, while 6,986 would be (VI)CMLXXXVI. Note that the symbol V is used more than once in that example, this is only permissible with the vinculum setting them apart. Also note that while M is used, 7,000 would be (VII).

(I), (II), and (III) are not used, with M, MM, MMM being preferred under 4,000. SS is never used in place of I (ss is sometimes seen as a variant of semi, but this is strictly pharmaceutical notation).

With fractions and vinculums employed, the lowest possible value is • or 1/12, while the highest is (((MMMCMXCIX)))((CMXCIX))(CMXCIX)CMXCIX or 3,999,999,999,999 (within standard usage). Four trillion, rendered as (((MMMM))), would be non-standard since it breaks basic rule 2, although the four M's can be used as a variant form, and it is more aesthetically pleasing as a maximum figure. Another alternative is ((((IV)))), but since the general principle is three sequential iterations, this should also apply to vinculums. The reason for the three-rule is that the eye can readily distinguish between one, two, or three marks, while four or greater become illegible. Thus, 4 M's are preferable as a variant form.

additionally, I tried to insert a passing reference to rules under the 'Standard form' section (since the Reddy/Khan RS used there documents rules in its first 2 pages), only to trigger the dispute again. Xcalibur (talk) 07:15, 19 November 2020 (UTC)

References

Renewal of old argument between Bigdan201|Xcalibur and Soundofmusicals

The next thread was getting very off topic - I have taken the liberty of moving the posts in question here, in an attempt to get back to its purpose - which is not to please everyone but to establish a consensus. First post here is duplicated from next thread for continuity.

• "How to write Roman Numerals" must not be duplicated. If you think you have a better method of showing how to write them, it must replace the currrent description. This also means that your talk about fractions and viculums must be merged or replace the existing sections on these.
• IMHO your "rules" are unreadable and require considerable mental effort to figure out what patterns are allowed. "Write the shortest sequence that adds up to the correct value" would work far better, and if you insist about 4 of your "rules" could be listed as side-effects of following this. And you go into far too much unreadable convolutions trying to restrict the subtractive patterns, rather than just list the six allowed.
• Lots of references are not very convincing when it is obvious they are all cut & pasted from each other

Spitzak (talk) 20:46, 2 November 2020 (UTC)

must not be duplicated/must replace -- wrong. I'm not sure why you keep insisting on this, there's no reason we can't have both, this isn't a zero-sum game. the basic description & rules describe the system in different ways and are not interchangeable. would you say that the chart of RNs must replace the description, or that it "can't stand on its own"? of course not, we have both because they serve different purposes. my rules cover analysis/logic, the basic description covers other points (outlined by Soundofmusicals above: Talk:Roman_numerals#First_paragraph_of_"description"). fractions/vinculums -- I'd be fine with merging that content into the respective section, I included it alongside the basic rules simply for convenience.

difficulty -- I think they're fairly straightforward, but since this is a common critique, I added a subsection full of examples to illustrate how the rules work, which should resolve the issue. shortest sequence -- that's not as comprehensive or detailed as my version, moreover, it's not reflected in the RS, which makes it OR (as I've mentioned earlier). every clause of my rules is derived from the RS. restrict subtractive patterns/list them instead -- that's missing the point. I could just say that only 4s and 9s (to a power of ten) are permitted, but that's not explaining *why* that's the case. among the most common questions about Roman Numerals are why combinations like IIX, VL, or IC are not legal; my rule 5 answers this explicitly.

Lots of references are not very convincing when it is obvious they are all cut & pasted from each other -- Wrong, that's obviously not the case. I explicitly avoided using any duplicate sources (a number of sites copied the excellent LURNC/Seitz source which was a significant influence, but I only used the one ref for that). the reason my RS/sources are independent yet similar is due to convergence, as they're all describing the same system in a similar manner. Xcalibur (talk) 01:22, 4 November 2020 (UTC)

I'd like to add one more statement. the criticisms offered so far would be valid concerns, if there was a single grain of truth to them, which there isn't. if I didn't have the RS, or if the rules didn't cover every combination coherently, or if they were redundant with the existing description, then in any of those cases it shouldn't be published. but I do have the RS, they do clearly address every possible question on orthography, and their scope is quite different than the basic description. if you can prove me wrong on any of these points, and if I can't address or resolve the issue, then I'll concede, but that can't be done. Xcalibur (talk) 23:31, 5 November 2020 (UTC)

Repeated and ever more emphatic assertions do not an argument make. Everyone here accepts (I hope) that you sincerely believe you are right. Alas - this does not mean you are right - or, even if you were, that "what you want to share" is what wikipedia is about. Philosophical unity is unattainable, and (fortunately) unnecessary. Progress towards an ideal on-line encyclopedia is impeded rather than advanced by stubborn opposition to consensus. If you are a musical theatre fan you may notice that "we" call the stage works of Gilbert and Sullivan "operas". This (still, years after the question was settled) irritates me intensely. I wanted to call them what they are, which is "operettas". I had to bow to consensus, as you do here. Nobody else agrees with you that a set of rules would be an improvement here. Nobody. Accept it. --Soundofmusicals (talk) 00:57, 6 November 2020 (UTC)

I'm sorry but your continued insistence that for some reason your rules do not duplicate the information in the current section, that your viculum and fraction stuff does not duplicate information in existing sections, and that your examples do not duplicate information in the existing examples, is showing an unbelievable stubborness. Just because "I say so" and you use words like "perscriptive" or whatever does not make anything you say true. Just shut up. Or if you really really think "rules" are better, REPLACE THE EXISTING TEXT!!! Or go awaylSpitzak (talk) 18:01, 6 November 2020 (UTC)

First of all, there's no need for hostility. to the best of my ability, I've participated in these discussions in good faith, and I'm willing to acknowledge any valid criticism. for example, I'm fine with moving the fractions/vinculum content down to that section and merging it; that's never been a problem, I only included it next to the rules out of convenience.

your primary critique seems to be that my rules are redundant, i.e. stating the same information as the intro in a different format; if that were true, then certainly it shouldn't go up, but it's demonstrably false. to elaborate, the basic intro covers: fixed symbols, the use of addition/subtraction instead of place-value, decimal base, etc. which are not covered by my rules. likewise, the rules logically dictate how the symbols may be added/subtracted, which is not covered by the intro. therefore, they each have a different scope and are not redundant or interchangeable.

as for the examples, there again is a difference, since I show how particular combinations are correct or incorrect, which the current examples don't do (correct me if I'm wrong).

if you can prove me wrong, or find a flaw in my work, I'll listen, and if I can't address it or correct it, then I'll concede. I'm trying to share my knowledge and make an improvement, but unfortunately, I'm locked out by gate-keeping. Xcalibur (talk) 18:25, 6 November 2020 (UTC)

All this has no relevance to the thread in which it was posted so I have moved it here. Please let us keep all "off-topic" things (especially bleats about hard-done by we are and how we wish the other(s) would go away) HERE in future. If we must keep telling each other to shut up, at least keep it out of a thread started by an administrator for a specific purpose (getting a consensus).

One apposite remark, I have done a bit of work on our existing "viculum and fraction stuff" [sic] that, while it is not a straight "merge" of any proposed addition does attempt to work these up into a more comprehensive form that might render their (duplicated) placement elsewhere redundant. --Soundofmusicals (talk) 21:38, 8 November 2020 (UTC)

on second thought, much of the above was unproductive, so I've taken the liberty of deleting some comments (including my own) and only leaving in the most relevant points. if anyone objects to this, feel free to revert. Xcalibur (talk) 07:53, 20 November 2020 (UTC)

Proposals regarding standard form and rules

Please do not post to this thread anything not of strict relevance to this topic. --Soundofmusicals (talk) 01:03, 6 November 2020 (UTC)

There are ongoing disagreements regarding (1) Roman numerals#Standard form; and (2) whether additional rules should be added. I propose that we handle these by dealing with (1), then considering (2) soon after. For (1), a discussion might be sufficient but if a clear consensus does not emerge, I intend starting an RfC to settle the matter for now. For (2), I'm unsure whether a discussion would be conclusive so it is likely that I would start an RfC after a discussion drafted a precise question for an RfC to settle (2).

Re (1): I had some minor reservations about the wording and propose the changes shown in diff (these were drafted by Soundofmusicals and myself). The changes are:

• Simplify the first sentence.
• Remove "nowadays" as contrary to style and "mainly" as unnecessary and hard to reference.
• Simplify the last sentence to remove points that are unnecessary and hard to reference.

@Soundofmusicals: I made some changes in the diff above. Are you happy with them? Views from others would be greatly appreciated—are the changes in the above diff ok? Should anything else be changed? Johnuniq (talk) 00:56, 20 October 2020 (UTC)

Yes - this is definitely the best version yet suggested for this section, succinct and comprehensive - IMO any further changes would at best be unnecessary nit-picking. --Soundofmusicals (talk) 01:20, 20 October 2020 (UTC)

Comment: The suggestion still mentions "ambiguities inherent in Roman numerals". I can't see any: an ambiguity means there is a putative expression in RN whose value is not clear. The closest I could come would be a MS fantasy in which SIX could be either 8-1/2 or 9-1/2. It does seem to me that the currently argued about "ruleset" is absurdly long-winded, but also that the claim that there "are no rules" displays an ignorance of the (mathematical rather than legal) meaning of "rules". Imaginatorium (talk) 03:57, 20 October 2020 (UTC)

It's an attempt to say, for example, that 4 is IV, not IIII. The ambiguity is about how to write a particular number so that (with design exceptions such as clock faces) numbers are written the same way by different people. Let's leave discussing rules for the moment. Johnuniq (talk) 04:15, 20 October 2020 (UTC)

the current text is fine. however, the Reddy/Khan source used there documents rules on the first 2 pages, so I thought it would make sense to add a clause mentioning rules, something like "the convention is governed by rules, and is illustrated by the following table". ultimately it's not necessary, but really, I wasn't expecting you to balk at a brief mention supported by an RS.

to the new arrival who described my ruleset as "absurdly long-winded", it really isn't. it consists of 7 statements, which is fairly succinct; I doubt it could be much reduced while still retaining comprehensiveness and clarity. keep in mind, much of the section's length is from extra content: fractions/vinculums are there because I wanted to develop the concept further; the examples are there to explain and illustrate, because a common criticism was that my content is too advanced for a beginner. Xcalibur (talk) 07:08, 20 October 2020 (UTC)

We have a whole section ("Variant forms") on variant and ambiguous Roman numerals. They are not specifically treated in this section (and neither are fractions) because these are simply not part of the "standard" - nor, for that matter, are any of the systems by which RNs were historically extended to cover large numbers. Like a number of other topics they do have their own sections. Nothing to stop us discussing the content, and even the placement, of these sections in their turn - but for the moment we need to stay on topic. --Soundofmusicals (talk) 19:59, 20 October 2020 (UTC)

I didn't mean to go off-topic, I was only trying to address the comment above, by pointing out that the length of my section is mostly due to optional content. to respond to your edit summary: there's no conflict between my rules and the rules in any given source. Xcalibur (talk) 00:57, 25 October 2020 (UTC)

With all due respect everyone knows by now (or can easily discover by looking up old posts of yours) exactly what you think about this - more or less same goes for me of course. Nothing to be gained by contined exchanges between you and me. The topic of this thread is quite clear (just read Johnuniq's first post) - If you have something genuinely new to say that is on topic then by all means say it - otherwise you and I really need to back off for a week or two to give anyone else who might be interested a chance. --Soundofmusicals (talk) 05:54, 25 October 2020 (UTC)

For the time being we can accept the absence of the "ambiguity" bit - but I would personally have preferred something that makes it clear that the table links "Arabic" digits with RN equivalents on a one-to-one basis. Or is this nit-picking? The editor who could not find any ambiguities apparently missed the whole of the "Variant forms" section, which lists a number of "ambiguities" - not only Arabic digits that have two (or more) possible Roman numeral equivalents, ('IIII' and 'IV' of course, but also many others) but also Roman numerals that can be "interpreted" as the equivalents of different Arabic numbers! (e.g. "IIXX" (just one "variant" form with at least two possible meanings, 18, and 22). --Soundofmusicals (talk) 06:02, 27 October 2020 (UTC)

Perfection is desirable of course, but it often leads to opaque text because it tries to make a statement that is 100% correct under all conditions. I'm not wedded to the idea of removing ambiguities but I doubt that it is helpful at the start of that section. If ambiguity needs a mention, it might be later with the examples you mention. Johnuniq (talk) 06:40, 27 October 2020 (UTC)

Yes I was coming round something to that opinion myself. I think the "Variant forms" section is fine in itself - was just a little concerned that anyone could have missed it so entirely and wondered if we needed to call some attention to it. On the other hand - if "ambiguities" was not the way we used to describe variants (and I have a feeling it wasn't...) then perhaps it is no loss. Oh well. --Soundofmusicals (talk) 18:57, 27 October 2020 (UTC)

Things that must be addressed if this is ever to be considered:

• "How to write Roman Numerals" must not be duplicated. If you think you have a better method of showing how to write them, it must replace the currrent description. This also means that your talk about fractions and viculums must be merged or replace the existing sections on these.
• IMHO your "rules" are unreadable and require considerable mental effort to figure out what patterns are allowed. "Write the shortest sequence that adds up to the correct value" would work far better, and if you insist about 4 of your "rules" could be listed as side-effects of following this. And you go into far too much unreadable convolutions trying to restrict the subtractive patterns, rather than just list the six allowed.
• Lots of references are not very convincing when it is obvious they are all cut & pasted from each other

Spitzak (talk) 20:46, 2 November 2020 (UTC)

I responded to Spitzak's post in the section above. Xcalibur (talk) 04:00, 10 November 2020 (UTC)

Bigdan201 has changed the article yet again! I don't like to be the one to revert this, at least not right away. One would really like to assume "good faith" and accept that Bigdan is genuine about a "compromise" in the sense of letting a mention of hypothetical "rules" stand for the rules themselves, rather than an attempt to wedge them right back in! If this is really the case, then by all means let's test that this meets a consensus of concerned editors. I do feel, however, that if we mention "rules" at all we need to include a meaningful comment indicating why we don't include them. Something on the lines of:

• Some sources suggest more or less elaborate "rules" which attempt to define Roman numeral notation in terms of what is (or is not) "permitted" - the following table, on the other hand, shows the simple pattern behind the usual, or "standard" form, and is a simpler and clearer way of conveying this information.

If any of this is basically objected to by Bigdan - then I find it very hard to accept that a "compromise" is actually intended. --Soundofmusicals (talk) 15:55, 18 November 2020 (UTC)

No, there is no need to say there are "rules" or that they are somehow different from the current description. The current description and the "rules" both are instructions on how to produce a limited subset of the possible arrangement of the letters. It is a basic fact that if you make *any* subset of patterns there is going to be a "set of rules" that restricts output to that subset. Therefore saying "there are rules" is totally meaningless. And transcribing between decimal and roman numerals using this table is precisely how any modern user would read/write roman numerals. They certanly in no way try to apply Dan's "rules". Please don't mention it at all. It is misleading readers into thinking maybe there are two sets of patterns, or that there is some government set of laws about Roman numerals.Spitzak (talk) 19:13, 18 November 2020 (UTC)

FTR, I reverted because between the two options:

Roman numerals typically use a conventional notation, as shown in the following table:

and

Roman numerals are typically written with a conventional notation. The convention may be described by rules, or shown as a pattern, as in the following table:

I can't see what the reader gains in the second version. Because they mean basically the same thing. So, it's just adding extra words for the sake of adding extra words, without giving any greater understanding. It's better to be concise. Kahastok talk 20:34, 18 November 2020 (UTC)

usually, being concise is better. I expanded the text to insert a mention of rules, which are in the sources and relevant to the topic. the table is fine, and transcribing between our numbers and RNs is a convenient method, but it's a contrivance not based on the actual structure of RNs. among the most common questions that arise are why combinations like VL, IC, and IIX aren't permitted. the table shows you which permutations are correct, but it doesn't tell you why that is. sure, you can say "subtractives are only powers of 4 and 9" but that's still just describing, not analyzing; you could still ask why IIVI or VVC aren't allowed, in spite of them meeting the condition of 4 or 9 x 10^. only the rules I linked to a couple sections above achieve a proper analysis. and the rules do work, no one can refute them. It is misleading readers into thinking maybe there are two sets of patterns, or that there is some government set of laws about Roman numerals. it doesn't say that at all, it says the same convention can be described in two different ways.

with all that said, experience has shown that my correspondents aren't willing to read or listen to anything I say, so that no progress can be made in debate. even if I somehow got my content up, certain others would try to "improve" it and end up mangling it; this already happened when parts of my text were cut&pasted into the article without any regard for style, structure or context (I mean no offense by this, but it's true). so there's nothing to be done here. Xcalibur (talk) 06:57, 19 November 2020 (UTC)

I'd like to tie up some loose ends in regards to this. Soundofmusicals made a number of edits I took issue with, but I haven't had the chance to address and criticize them, so I'll do so now.

the usual modern standardized orthography this is an awkward mouthful of a phrase, and one that I didn't use. the furthest I went was 'standardized orthography'. even that's a little rich, and I can see an argument for simplifying it to 'standard form'; any more verbiage is excessive. I also don't know why you kept trying to include permits only one permutation for any given value, this only makes sense in the context of rules/logic (which narrow down possibilities until there's one left), without that it's oddly out of place. then there's this: To resolve the ambiguities and inconsistencies inherent in Roman numerals, a conventional orthography, permitting only one permutation for any given value, is widely recognized and adhered to. you cut&pasted my text and clumped it together, resulting in a mess of a run-on sentence. compare this to the original lede:

The modern era has seen the emergence of a standardized orthography for Roman numerals, which permits only one permutation for any given value. This system may be described as a decimal pattern, as above, but also as a logical set of rules. While exceptions can be made (notably IIII instead of IV on clockfaces), the modern convention is widely recognized and adhered to, and may be prescribed by the following ruleset: yes it's technical writing, but it has a much more natural, logical flow to it. I suppose if you wanted to simplify it, you could use something like: The modern era has seen the emergence of a standard form for Roman numerals, allowing one combination per value. ... the modern convention is widely followed, and obeys the following ruleset: I wouldn't have a problem with simplifying the text, if that were the issue.

last but not least, there's this edit: [3] in which you formatted the basic description as 'rules'. I still find this mystifying months later. they're not rules to begin with, and tacking numbers onto lines of prose doesn't make sense. that was a nonsensical attempt at compromise, no offense.

the overall impression I get from these edits is that you don't really understand where I was going with the rules/logic. the point isn't numbered lines or exact wording, it's analysis based on sources. Spitzak doesn't seem to get that either, but for different reasons. I've tried my best to explain, but to no avail. None of this is meant in the spirit of negativity, I just wanted to offer constructive criticism, and wrap up remaining issues. After all, Soundofmusicals' constructive crit was quite useful, and helped me achieve a finished product. Xcalibur (talk) 08:46, 4 December 2020 (UTC)

None of the above has any relevance to the current article - as you rightly point out, many of my proposed changes and edits have not passed the test of time - in many cases having been modified or deleted by my own hand. To put this another way - none of the phrases you complain about is actually still part of the article. The understanding of our long sought consensus is that previous versions are indeed "history" - no "cleaning up" is now necessary. --Soundofmusicals (talk) 17:33, 4 December 2020 (UTC)

I understand, I just wanted to 'set the record straight' since I didn't weigh in at the time. Xcalibur (talk) 00:01, 5 December 2020 (UTC)

Semi-protected edit request on 8 November 2020

The section 1.5.2 for "Vinculum" currently implies that there is some debate over whether the vinculum was used already in ancient Roman times, when there absolutely isn't: the vinculum to mark thousands was very common way to indicate thousands in Latin inscriptions of at least the Imperial era. I would completely delete the David Smith reference for this confusion - this is an outdated general overview to the history of mathematics from the 1950's from someone who isn't a specialist in Roman numerals, and not the best source to use in this article and of no relevance to current specialist scholarship. I would edit the paragraph to make clear that the vinculum was standard imperial practice, and reference some epigraphic examples. I would also add a comment that the practice might cause some confusion, because Romans (and Greeks) also often used bars over numbers in inscriptions and papyri to highlight letters that were numbers, without any numerical significance.

The *vinculum* continued to be used in the Medieval times, but there is currently no reference specifically for this claim - from the excellent book by S. Chrisomalis, he notes that it was in Medieval period called *titulus*.

So, change:

• "Another system was the vinculum, in which conventional Roman numerals were multiplied by 1,000 by adding a "bar" or "overline".[45] Although mathematical historian David Eugene Smith disputes that this was part of ancient Roman usage,[46] the notation was certainly in use in the Middle Ages. Although modern usage is largely hypothetical it is certainly easier for a modern user to decode than the Apostrophus [...]"

to

• "Another system was the vinculum, in which conventional Roman numerals were multiplied by 1,000 by adding a "bar" or "overline".[45] It was a common alternative to the apostrophic ↀ during the Imperial era: both systems were in simultaneous use around the Roman world (M for '1000' was not in use until the Medieval period).[REFERENCES1] The use of vinculum for multiples of 1,000 can be observed, for example, on the milestones erected by Roman soldiers along the Antonine Wall in the mid-2nd century AD.[REFERENCES2]. There is some scope for confusion when an overline is meant to denote multiples of 1,000, and when not. The Greeks and Romans often overlined letters acting as numerals to highlight them from the general body of the text, without any numerical significance. This stylistic convention was, for example, also in use in the inscriptions of the Antonine Wall,[REFERENCES3] and the reader is required to decipher the intended meaning of the overline from the context. The vinculum for marking 1,000s continued to be in use in the Middle Ages, though it became to be known more commonly as titulus.[REFERENCES4] Although modern usage is largely hypothetical it is certainly easier for a modern user to decode than the Apostrophus [...]"

Where [REFERENCES1] S. Chrisomalis, 2010, Numerical Notation: A Comparative History, p. 109-112.; A. E. Gordon, 1988, Illustrated Introduction to Latin Epigraphy, pp. 122-123.

[REFERECES2] e.g.: https://romaninscriptionsofbritain.org/inscriptions/2196, https://romaninscriptionsofbritain.org/inscriptions/2208

[REFERENCES3]: https://romaninscriptionsofbritain.org/inscriptions/2193, https://romaninscriptionsofbritain.org/inscriptions/2171

[REFERENCES4]: S. Chrisomalis, 2010, Numerical Notation: A Comparative History, p. 119. Mythoplokos (talk) 13:43, 8 November 2020 (UTC)

 Not done for now: This sounds quite unencyclopedic, especially the use of multiple examples. Certain words such as it is certainly easier may be considered as an opinion, where encyclopaedias should not exhibit opinions. I suggest rewriting this in a more encyclopaedic tone — Yours, Berrely • ^Talk∕_Contribs 15:36, 8 November 2020 (UTC)

Thanks Berrely; I can try to edit the section to be more "encyclopedic", my first attempts at trying to contribute. I thought of presenting some specific examples as reference, since this is just something anyone worked with Roman numeral practices KNOWS since this is really such an obviously common way to denote multiples of thousands - but maybe I'll use some pages from epigraphy handbooks or something to back up instead. As it stands the section for vinculum is just plain incorrect and misleading, and on a second look the lower part in the same section seems to imply that "boxing" numerals for multiples of 100,000 was a Medieval practice, when, again, this was already widely in use during Roman times. Before I try to rewrite, would you (or someone else) mind elaborating where I used "certain words" that make this section sound too much like an opinion, so I can avoid them? I didn't use the phrase it is certainly easier in the text anywhere! Mythoplokos (talk) 16:11, 8 November 2020 (UTC)

DONE!!I hope that my reworking of your edit (basically excellent - it fills the section out nicely!) meets full satisfaction! I have regularised your references into standard "wiki" format and added a few parameters - but otherwise left things pretty much "as was" (any mistakes that I missed will still be there!) so you may want to go back to it when you have been granted full privileges.

The "unencyclopedic" text another editor objected to was already there - I have recast the passage in question.

And welcome to the Wikipedia editing team! From the look of this one we can expect some very useful edits from you in the future. --Soundofmusicals (talk) 18:04, 8 November 2020 (UTC)

 Already done Melmann 22:35, 8 November 2020 (UTC)

Semi-protected edit request on 15 November 2020

Please remove

The use of a final "j" is still used in medical prescriptions to prevent tampering with or misinterpretation of a number after it is written.

and add

Into the early twentieth century, a final "j" was still used in medical prescriptions to prevent tampering with or misinterpretation of a number after it was written.

This sentence has two sources, one dating from 1906 and the other from 1918. Sources more than 100 years old aren't good for "is still used" statements. 2601:5C6:8081:35C0:2C8A:AB39:C6BE:2DEB (talk) 16:52, 15 November 2020 (UTC)

 Not done: I've added sources for the occasional modern use of j. Danski454 (talk) 17:15, 15 November 2020 (UTC)

Thanks! That's even better. But Danski454 (talk), could you remove the old sources, leaving the sentence referenced with your new source only? The two old sources still aren't good references here. 2601:5C6:8081:35C0:2C8A:AB39:C6BE:2DEB (talk) 17:42, 15 November 2020 (UTC)

I've actually changed the text as requested - I've left just one source here - although its well "dated" so is the information, so the old source is fine (in this context). What we don't need is four sources in a block! --Soundofmusicals (talk) 22:16, 15 November 2020 (UTC)

Semi-protected edit request on 19 November 2020

In Section 4.1 of this article, the 8th example of Roman Numerals used in specific disciplines says "In music, Roman Numerals are used in several contexts." The second of the three following bullet points, "In music theory, the diatonic functions are identified using Roman Numerals (See: Roman Numeral Analysis)" is misleading-- in analysis of classical, jazz, and other contemporary music, there are countless examples of analysis of non-diatonic harmony using Roman Numerals (where, in most cases, Roman Numerals are the only common and well-accepted analysis of said chords). To name a couple of very common examples, secondary dominants are denoted using Roman Numerals, (V7/V, V7/II, V7/VI, V7/III, V7/IV, etc.), as are modal interchange chords (IV-7, bVII7, #IV-7b5, bII∆7, to name a few). Of course, these obviously use a combination of Roman and Arabic numerals, but diatonic harmonies do as well-- a few examples include VI-7, V7, II-9, IV∆7(#11).

TL;DR Please change "In music theory, the diatonic functions are identified using Roman Numerals (See: Roman Numeral Analysis)" to "In Roman Numeral Analysis, harmonic function is identified using Roman Numerals".

Reason: "Music theory" is vague-- Roman Numerals are specifically used in an area of music theory which is itself named after Roman Numerals-- and "diatonic functions" is misleading. Awh.1998 (talk) 21:59, 19 November 2020 (UTC)

Done: Although this is part of an article about Roman numerals rather than music - nothing wrong with the requested change provided we keep the link to Roman numeral analysis to provide the required detail --Soundofmusicals (talk) 03:45, 26 November 2020 (UTC)

"Last gasp" of old controversy (at least one hopes?)

If one reference is truly insufficient for the point that "some sources like rules" then do we really need five? And one of them a bare url? Reddy was never a good source for this point anyway, its "rules" are tied to the (obsolete!) application of rns to pharmaceutical calculations. We've had it there for a while, but then I honestly feel that for all the good it is currently doing it ought to go! Overall, two references are plenty. --Soundofmusicals (talk) 07:46, 7 December 2020 (UTC)

I moved Reddy/Khan down to 'specific disciplines' under 'modern use'. I was meaning to flesh out the simple url, it was included because it's another strong RS (alongside Shaw). however, it's not a big deal, and 2 refs should be fine, especially when they're 2 of my best, Shaw & Seitz (while LURNC isn't considered very reliable under our policies, it's a thorough source that heavily influenced my work). one other matter -- the sentence "these are are only subtractives in standard use" shouldn't be italicized. there's no need to spoonfeed the reader, and as I said, it goes against typical style. at most you should italicize important phrases (as the current description does), not a whole sentence. Xcalibur (talk) 08:51, 7 December 2020 (UTC)

Perhaps italicising the last words of the offending sentence (as I have now done) rather than the whole thing may bring it into line with considerations of style without losing too much force? We do have another source for the "pharmaceutical" bit - but if we must include this one as well it certainly fits better there than where it was. --Soundofmusicals (talk) 13:06, 7 December 2020 (UTC)

Reddy/Khan fits in either place, because it covers pharmaceutical notation as well as general rules. however, since you prioritize reducing reference clutter, I'm willing to cooperate. as for italics, Wikipedia:Manual_of_Style#Italics is more relevant. there's simply no need to emphasize that last sentence, especially since the table already shows the standard subtractives. as for the deleted text, of course logical analysis is more complex than a simple description, that's why I wanted both. if I simply named the subtractives, then I could just as well put "look it up in the table, if it's there it's legal". that's not explaining the mechanics of how the standard form works. I still haven't seen any valid critiques, but it's obvious that I'll never get through to the regulars here, so I won't keep rehashing. Xcalibur (talk) 03:36, 8 December 2020 (UTC)

Standard form inflation

For reasons I missed, the Roman numerals#Standard form introduction has expanded from:

A conventional notation is used to write Roman numerals, as shown in the following table:^[ref]

to:

Various sources suggest more or less elaborate rules, which define Roman numeral notation in terms of what is (or is not) permitted;^[ref1][ref2] the following table, on the other hand, shows the simple pattern behind the standard form, and is a simpler and clearer way of conveying this information:

Is there any reason the new text should be retained? What is it saying that is missing from the earlier version? Is the suggestion that retrospective sanctions would be applied to anyone who broke the rules carefully documented in the two references? This article should stick to simple and encyclopedic information, namely the earlier version. Apart from the vacuousness of the new text, "Various" is a weasel word, and "more or less" is from a personal essay, and I count four unsourced claims in the new text, aka original research. Johnuniq (talk) 08:53, 10 December 2020 (UTC)

It seems to me that the new text is largely pointless waffle, with several clear flaws and no obvious advantage. I support going back to the initial version you describe. Kahastok talk 17:42, 10 December 2020 (UTC)

I agree but there is a problem with a contributer who thinks it is really important that the fact that any subset of patters can be described by a "list of rules" be mentioned here, rather than in some page about pattern grammars.Spitzak (talk) 18:28, 10 December 2020 (UTC)

The idea, of course, was basically to deflect Bigdan from a prolonged and highly disruptive campaign to introduce a set of so-called "rules". He claimed that a suitable "mention" of rules would lead to his acceptance of the obvious consensus on this point. (Although, alas, the obsession remains.) And yet I do feel that the additional matter - especially the references - is by no means a complete waste of space, if only for the hypothetical reader who may miss the rules featured in some print and online "sources" - and justify our preference for a simple table of values instead. The alternative may be getting Bigdan permanently banned from editing this article - but this sort of thing is really not what Wikipedia is about, is it? --Soundofmusicals (talk) 19:57, 10 December 2020 (UTC)

No, a ban/block is not going to happen, not unless things seriously escalate. Life at Wikipedia is a lot simpler than would appear from this topic. What has to happen is this:

1. Editors discuss what wording should occur in a section.
2. If there is consensus, the edit is made.
3. If someone edits against that consensus, their edit is reverted with an edit summary pointing to the consensus on talk.
4. If the disputed edit is repeated, another editor repeats (3). Someone might put {{subst:uw-3rr|Roman numerals}} ~~~~ on the editor's talk.
5. If the disputed edit is repeated, someone like me reports the problem at WP:3RR.

That's it, there is no problem. The only reason this has dragged on for months is that people won't commit to consensus text. So long as editors continue to pick at the wording, there can be no objection if others continue to pick at the wording. If you want to spend the rest of your life arguing here, carry on. If you don't, stick to the script above. By the way, the wording in my post above comes from 26 October 2020 permalink. It's not perfect but it's better than other proposals. Can we agree what wording should be in the article? Can we agree to follow standard operating procedure? Johnuniq (talk) 23:04, 10 December 2020 (UTC)

I for one am ok with all of this but who is going to do it? With respect, all Bigdan's efforts have been directed at getting his own way - all mine have been towards establishing what is obviously the majority view here. So who is going to actually restore (and defend) the "pre-bellum" status quo? (Step 3. of Johnuniq's procedure as outlined above). I am fed up with having this treated as if it were still "an ordinary dispute" between two editors. It went past that point long, long ago. --Soundofmusicals (talk) 06:28, 11 December 2020 (UTC)

You might like to do one revert but should leave more than that to others. If others want to do further reverts, that would demonstrate a real consensus. If they don't, the dispute will never end. You must stop referring to another editor during a discussion at an article talk page. I assure you that this is a dime-a-dozen dispute as far as Wikipedia is concerned, and a noticeboard would not assist because if two editors are editing the same section without edit warring, there is no problem as far as outsiders would be concerned. Johnuniq (talk) 10:06, 11 December 2020 (UTC)

Sorry, but while so many editors express a preference (which I share) for another wording, but refrain from reverting to that version then I can't see us getting anywhere. Not for me to "kick off" this procedure, however, in view of past history, where I have been accused of being one half of a two-editor dispute. This particular thread - from beginning to end - shows that this is far from the case. --Soundofmusicals (talk) 10:43, 11 December 2020 (UTC)

there are in fact logical rules, and they do govern how RNs are written. I put forth my best efforts in creating a section documenting this, with almost a dozen sources (this can be seen on the /rules subpage, linked under 'section on rules/analysis'). however, the regulars here refuse to let me contribute and improve a C-quality, high-priority article. in light of that, I've tried for a very modest compromise by inserting a brief reference to rules. the exact wording is negotiable, and if you find the current version too long-winded, you could try my earlier edit: Roman numerals typically use a conventional notation, as described by rules or the following table:. nowhere have I mentioned 'sanctions', that's a strawman. as to Kahastok, with all due respect, I don't think you're aware of the larger context here; the topic of rules/logic is neglected due to gatekeeping of content, and I'm trying to add a brief mention as compromise. I thought I was being reasonable, and wasn't expecting the conflict to flare up again when I'm not even adding the main content. I also don't see how my contributions or discussions have been disruptive, on the contrary, I've remained civil and addressed points as best I can, only to hit a complete roadblock. Xcalibur (talk) 11:14, 11 December 2020 (UTC)

Bluntly, no. There is no useful distinction between a section describing the "rules" of Roman numerals on one hand, and a section describing how Roman numerals work on the other. There is no reason why the section describing how Roman Numerals work has to be described in terms of "rules", and there is no benefit to the reader in trying to shoehorn the word "rules" into such a section. The reader wants to understand how Roman numerals work. They aren't interested in random philosophical debates on the nature of "rules".

I would also note that, in general, a claim like I also don't see how my contributions or discussions have been disruptive, on the contrary, I've remained civil and addressed points as best I can... is a logical nonsense. It is perfectly possible for an editor to be both civil and disruptive at the same time. For example, an editor who persistently refuses to drop the stick could easily remain civil while doing so. Kahastok talk 17:36, 11 December 2020 (UTC)

again, I'd attribute this to you stumbling into the dispute, without following the lengthy context. of course the reader wants to understand how RNs work, that's why I tried to add another section on rules/logic, to complement the existing explanation. I'm not sure if you even know what I mean by rules in the first place: have you read my section, or the two refs that were linked? maybe then you'd realize the difference between a basic explanation vs analysis of how it works; they're two different things. the current article doesn't actually explain how RNs work, or why combos like IIX or VL are not considered standard, it just lists the correct permutations and provides examples, without telling us why that is. as for the last point, that may be true, but I've tried to be cooperative, and the latest flare-up is over what I consider a very generous compromise. Xcalibur (talk) 18:08, 11 December 2020 (UTC)

So basically what you're saying is that my view is invalid because you don't think I've spent long enough discussing it? That's not how this works.

You seem to think that the only way anyone might disagree with what you say is because they do not understand what you say. This is not the case. Sometimes people understand perfectly well, but do not agree with your view.

I've seen your section. What I saw added little if anything to the article, because it did little more than duplicate the existing content. That is not to say that the status quo cannot be improved. It can. But these apparently-endless discussions are not improving it. As the example at the beginning of this discussion amply demonstrates. Kahastok talk 18:20, 11 December 2020 (UTC)

once again, the text is negotiable, and that version was (mostly) not my work. no, it's not because you haven't discussed it, it's because you evidently haven't been reading or comprehending what you read. this is further proven by this statement: because it did little more than duplicate the existing content.. this is demonstrably false, and reveals that you don't understand the material in question. finally, I'm not trying to add the section back in, I'm trying to add a brief reference, which shouldn't be too much to ask. Xcalibur (talk) 18:30, 11 December 2020 (UTC)

@Xcalibur Oh Dear!!! so much for me leaving a pointless "discussion"! to quote you - I'm not trying to add the section back in, I'm trying to add a brief reference, which shouldn't be too much to ask. - You nearly had me in with this one. Several editors, of whom our new friend is just the latest, have rightly pointed out that adding a "rules" section to the article considerably degrades its usefulness, with not one dissenting voice (sorry, just one, namely your goodself). But I (silly me) took your remark I'm trying to add a brief reference, which shouldn't be too much to ask. as a basically sincere statement of your intent here (assumption of good faith and all that) - and in lieu of simply reverting your "brief reference" I expanded it a bit into a form where it took note of something of the anti-"rules" sentiment on this page. I even ran this past you and got an "OK with this" response in return. For a few days we seemed well on the way to settling the "rules" business - people were even coming in with questions about other parts of the article, which I was happily addressing. And then you had to charge back in, and stirred up this new hornet's nest! I am going to give you one more shot with the brief three-liner you agreed to - but I will not stand in the way of returning to a section wholly purged of your influence - as "they" may well do - regardless of me! --Soundofmusicals (talk) 19:48, 11 December 2020 (UTC)

I've given up on including the section, because it's a brick wall. yes, I'm fine with a compromise that briefly mentions the rules, my position hasn't changed on that. I was trying to make a few adjustments to the text, that's all. the rest of my commentary was in response to points being made, as always. Xcalibur (talk) 20:24, 11 December 2020 (UTC)

Would everyone please focus on what text should be in the article! Compromise if you like, but we now have evidence of how that will work out. If you enjoyed the experience, keep doing the same thing. Otherwise, take my advice and decide on some text and stick to it. Johnuniq (talk) 22:28, 11 December 2020 (UTC)

Thanks for this. It seems to me that four editors have endorsed the simple, short text in your original comment, with one opponent. We can discuss variations to see if we can get the extra person on board. But it seems to me that we can reasonably consider that text to have consensus per WP:NOTUNANIMOUS.

While it may seem premature in this thread to talk about WP:NOTUNANIMOUS already, I think that it is appropriate given that this discussion seems have been going on for literally years. Kahastok talk 22:51, 11 December 2020 (UTC)

@Xcalibur OK on your last one. All I was after was a compromise that you would agree on - so we could all get back to better and more productive things. If you will harden your apparent promise above - and lay off the "few adjustments" i.e. just leave the "compromise text" which is actually the result of a good deal of thought on my part and let the whole matter rest - THEN I think there is a good chance (can't promise for any of the others though) that all this kerfuffle will cease - you will have got your wish of a brief mention of the rules - and even a couple of references where the (very hypothetical) reader who is interested can look at an example or two. Mucking around further will only result in a reversion to the original sentence and all your effort will be wasted - in fact this may even happen anyway - there are a good many "bricks in the wall" (other editors) who may have other ideas. In fact see above, which went in while I was typing this! --Soundofmusicals (talk) 23:03, 11 December 2020 (UTC)

calling this consensus is jumping the gun. more importantly, we need to reference the rules, otherwise the section as it stands is baseless OR. there is currently no RS for the table of numerals, or for the statement: A conventional notation is used to write Roman numerals. the sources I've read don't refer to a convention, they refer to rules. Xcalibur (talk) 02:28, 12 December 2020 (UTC)

You have been pushing this for over two years. If this is "jumping the gun" on consensus, then there are no circumstances in which WP:NOTUNANIMOUS applies.

Your arguments on the substance are unconvincing. You claim, the sources I've read don't refer to a convention. Yet the source you added to the article here does refer to a convention in its second sentence. And even if it hadn't used the word "convention", there is nothing in WP:NOR that says that we have to use exactly the same words as our sources.

This change is not an improvement to the article. There is nothing else here that leads me to believe that a wider consensus is possible. I thus see no reason not to proceed with the consensus we already have. Kahastok talk 11:48, 12 December 2020 (UTC)

it's more like several months, with a long break in between. this started in late 2018, and I was away during 2019. the latest controversy over adding a brief mention has been about a few months iirc, and this discussion is a few days old. the proposed section, vs a single passing mention, are two different issues.

I'm not in opposition to the shorter text, it's just that there needs to be a reference to rules. I'm also not sure where Soundofmusicals stands, he's the author of the longer text, not I.

yes, it uses the word, but the content is focused on a rules-based approach. point is, as far as the RS support a convention for RN, it's based on rules; that includes the Reddy/Khan source currently used. thus, we need to mention rules to 'anchor' the section, otherwise it's baseless OR. currently, there's no ref supporting the statement or the table of numerals. in fact, some editors here have seemingly denied that there's a convention, which is rather contradictory.

there is nothing in WP:NOR that says that we have to use exactly the same words as our sources. this is what I've been trying to say for awhile.

there's no firm consensus for the reasons I stated. my edit is an improvement, again for the reasons stated. at the very least, a single passing reference should not be a problem. Xcalibur (talk) 12:59, 12 December 2020 (UTC)

@Xcalibur I have supported a reasonable compromise with YOUR text (against very long standing consensus) in the vain hope this might cause you to accept that you are a minority of one on this question and move on. To have my efforts described as "I'm also not sure where Soundofmusicals stands, he's the author of the longer text, not I." is even more breathtakingly mendacious than anything else you have come up with over the years (yes, years are indeed composed of months, but it has been years. And in all this time you have not been able to interest, much less convince one other editor. If this is not consensus against you I don't know what is. Flat lies do not become true by repetition, to be blunt. I warned you in my last post what would happen if you persisted - it has, and quicker than I expected. Sorry, but that's it. --Soundofmusicals (talk) 20:54, 12 December 2020 (UTC)

the only thing that's happening is a "consensus" being railroaded, possibly out of spite.

there are two separate issues here: the proposed section, and the proposed reference. they should not be lumped together. as I said, I first proposed the section in late 2018, left it on the shelf for over a year, then picked it up again this year, so the discussion has been short of a year (not counting the long break). after numerous revisions and debates, I hit a dead end, so I made a second proposal -- a single brief mention in the text, which I thought would be quite reasonable.

the key point is, the proposed section is not relevant to this thread at all! this has more to do with the proposed reference, which has only been on the table for a few months. thus, this is not a case of STICK (as it would be if I were still insisting on the section). moreover, the reference needs to be there. without it, I could make a case for deleting the 'standard form' section as unsourced conjecture. however, I don't want to do that, I'd rather anchor it with a brief reference to the rules, woven into the preferred version of the text. Xcalibur (talk) 21:55, 13 December 2020 (UTC)

You've been filibustering on this point, by your own admission, for months. We have a consensus for this text. Not unanimity, but that's OK. Consensus does not need to be unanimous. If the consensus is being "railroaded", then that is entirely in line with both policy and standard practice on Wikipedia.

And I thought you'd said you'd been civil? The suggestion possibly out of spite is a personal attack. We've not interacted before, so I don't know where you think this "spite" might come from, but there we are. That one's on you. Kahastok talk 22:42, 13 December 2020 (UTC)

no, not filibustering, once again you're being misleading. there are two different points, and progress was made throughout the debate until it hit a dead end. compare the first draft of the ruleset to the finished product, and you'll see a major difference. even in the latter phases it was constructive, eg my modest proposal led to new criticism, which led to improving my sourcing. more significantly, all that applies to the first point (the section) which is not relevant here. the second point (a brief mention) has only been discussed for a few months. you're pushing this narrative where I've been arguing pointlessly for 2 solid years, which is not true (not even I would take it that far).

it wasn't a personal attack, but it was a violation of ASSUMEGOODFAITH. yes, that's on me, but there are reasons. first, I can't think of why else there'd be such strong opposition to a brief mention, and without discussing it properly. second, as I said above, you keep misrepresenting my position and what's happening (eg I never tried to insert "philosophical debates on rules" into the article at all). third, while I can't pin it down exactly, something about your posts suggests hostility.

with that said, I'm not necessarily opposed to the agreed-upon lede. it is in fact concise and to the point, and I can see why you'd want to keep out any extra verbiage (I wonder why no one brought this up, it's perfectly valid). I'm willing to come around on this, perhaps we can come to terms... Xcalibur (talk) 15:10, 14 December 2020 (UTC)

Reverted edit: "numerals" vs. "numeral system"

Hi,

an edit of mine was reverted:

 https://en.wikipedia.org/w/index.php?title=Roman_numerals&oldid=prev&diff=993119695

The point of the edit was to distinguish between "numeral" and "numeral system", which are two different things (you wouldn't say, e.g., that XV is a "numeral system"). The author of the revert seemed to focus on the fact that the entry for "numeral" is actually a disambiguation page, which isn't the main point. If there aren't significant objections here, I'd reinstate the edit. Alternatively, we can avoid the link, changing the introductory sentence to:

 Roman numerals are numerals of a numeral system that originated in ancient Rome...

or

 Roman numerals are numerals of a system that originated in ancient Rome...

(I believe the link to numeral is useful, though, because that page defines the term).—Gennaro Prota^•Talk 08:28, 16 December 2020 (UTC)

We could theoretically have named this article "The Roman numeral system" - but this simply isn't what the subject of the article is usually called. While it may seem illogical - the primary subject of this article refers to the system rather than all or any of the numerals as such. Many examples of particular numerals inevitably occur in any description/discussion of the system, but as you might say yourself that "isn't the main point". --Soundofmusicals (talk) 09:50, 16 December 2020 (UTC)

on rules, consensus, and RS

for those of you getting fed up with this, I beg your pardon. it's no easier for myself, that's why I'm trying to finally get this settled.

the current consensus, which I now accept, is that the text under standard form should be as follows: A conventional notation is used to write Roman numerals, as shown in the following table:

as an aside, I still think the discussion was rushed. all you had to do was point out that this is more concise and to the point, and you could've gotten me on board right away. but no matter, I've come around on my own.

but the point is, the consensus is for that text under Standard Form. it says nothing about Description, and it's not a moratorium on referencing rules elsewhere. if you want to create a consensus against mentioning rules, perhaps we can do an RfC for that. in the meantime, there's a bit of a problem here: the ref used for the consensus text refers to rules, in fact that's primarily what it has to say; it doesn't say there's a 'standard form' or 'conventional notation', it says there are rules. this, on top of the fact that multiple other RS state that there are rules, is contradictory. and this ties into a larger problem, namely that Standard Form and other parts of the article, as they currently stand, are unsourced conjecture. there's no RS supporting the 'standard form' claim or the table of RNs. this is a defect in the article, which should override partisan conflicts. this defect could easily be resolved by inserting a brief reference to rules, which is what the standard form/conventional notation is built on (as the sources state). a very brief reference worked into the text should not be too great a demand, especially not when there are larger issues at stake than personal preference.

as another aside, JohnUniq stated the following in an edit summary: rv: consensus is against false suggestion that the Romans published rules. as I said in my own summary, no one said that rules were published in classical antiquity. I didn't say it, the sources don't say it, no one in the world said this; it's a ridiculous strawman. additionally, the consensus is not against mentioning rules (yet). I and the sources state that there are rules governing the standard form, which took shape in the past 200/CC years (not 2000/MM). more to the point, if you disagree and believe that there are no rules, then it follows that there's no standard form/convention either, and that large chunks of the article should be deleted as unsourced conjecture.

to summarize: I'm trying to add a brief reference to rules, to support article content which is currently not supported. this is not a violation of the current consensus, which is only concerned with the Standard Form lede. if you want a consensus against rules, you may proceed with that, but that would contradict both RS and existing content. I hope my fellow editors can come around to supporting my reasonable compromise, without getting bogged down by partisanship. Xcalibur (talk) 19:39, 21 December 2020 (UTC)

References

If you have an actionable proposal based on policies to improve the article, please make it. Otherwise, please use another website to post opinions on life and rules. Editors are not required to debate forever and any future discussion should focus on precise proposals without extra TLDR wording. Johnuniq (talk) 22:43, 21 December 2020 (UTC)

I agree.

But I would note the context. I would note that this broad push has been going on continuously for months. Given this, the claim that we need to restart the consensus process on this change from scratch - barely a week after this substantially identical change was rejected by clear consensus - is WP:FILIBUSTERing. Kahastok talk 23:24, 21 December 2020 (UTC)

the suggestion to add a brief reference dates back a few months. this is not the same issue as the proposed section, which consensus decided against. they're two different issues, as I already said. also, I'm not sure if you used the correct diffs? either way, the consensus was on the lede text, not on mentioning rules, which is a key difference. also, the post made above is entirely about improving the article, nothing to do with "opinions on life and rules".

my proposal is to let this edit stand: [4] because it's a brief mention, it's relevant, and provides support to an entire chunk of the article, which is otherwise unsourced conjecture. the proposed text for the end of Description is: it is desirable to follow the standard form, which is defined by consistent rules. Xcalibur (talk) 02:03, 22 December 2020 (UTC)

References

In summary, the proposal is to change the last sentence at Roman numerals#Description from:

On the other hand ... it is desirable to strictly follow the usual standard form described below.

to:

On the other hand ... it is desirable to follow the standard form, which is defined by consistent rules.

The proposal adds three references, none of which appear to verify the text. The home.hiwaay.net page is "How Roman Numerals Work" with "This is the convention I was taught in school"—that does not satisfy WP:RS. The Google books link for "Essential Math and Calculations for Pharmacy Technicians" (238 pages) has no page number. The JSTOR link gives a brief outline of how to write Roman numerals, with nothing verifying the proposed text. The proposed text is not useful and is not verified. A better proposal would be to remove the "On the other hand ..." sentence altogether. Its reference regarding copyright ([5]) is interesting but only states the obvious, namely that making a mistake in a copyright notice might invalidate the copyright. Johnuniq (talk) 02:08, 22 December 2020 (UTC)

I can fix the page # issue, presumably that wasn't included because the relevant content is right there on the first two pages. I agree that Seitz doesn't satisfy RS, but I wanted to include it alongside RS because its content is high quality. if you want more RS, I've got plenty more. I'm not sure I agree that it doesn't verify the text, I'm mentioning that there are rules, and the sources agree with this. perhaps I could change the wording? keep in mind, the table of RNs and the claims that there's a standard form are unsupported, which is why this matters.

btw, I realize that this discussion (if one includes the first proposal) has been excessive; one reason for that is because I kept responding to every point being made. I'll try to keep it shorter from now on, since I'd like to conclude this. Xcalibur (talk) 02:23, 22 December 2020 (UTC)

I repeat, there's currently no consensus for proscribing any mention of rules from the article. I'm proposing a brief mention, backed by sources/RS, which provides support to article content which is otherwise unsupported. Xcalibur (talk) 19:36, 22 December 2020 (UTC)

I don't know about that, at the moment I see three people opposing this and one in favour. At the very least, there is no consensus to include your text, and so you shouldn't be trying to edit war it in. I do actually think there's probably a good non-unanimous consensus against mentioning rules in this article at this time.

I view this proposal as being only trivially different from the previous proposal. If anything, it's slightly worse, because the mention of "rules" is more obviously contrived. I agree with the above the best solution here is just to remove the sentence completely because it's not adding any useful information.

Reality is, this entire section could do with significant improvement. The trouble is, that improvement cannot happen on a wider scale while the article is being held hostage by a philosophical debate on "rules". Kahastok talk 21:36, 22 December 2020 (UTC)

Strongly oppose any mention of "rules" that does not make it clear that existing "rules" are inherently unhelpful, and at best an ambiguous and misleading substitute for our table. Moving this from one section to another, and/or marginally rewording does not affect its relevance to this very old consensus (it goes back at the very least to 2018 and has only ever been opposed by one editor). Also oppose deletion of mention of U.S. copyright law - we should keep this, as it is significant as the only actual case where the precise form of a Roman numeral has a specific legal consequence. --Soundofmusicals (talk) 22:28, 22 December 2020 (UTC)

if the consensus is against any mention of rules in the text, that should've been made clearer. if that's the case, I have to abide by it. but that presents another problem, which is that the 'standard form' content & table is completely baseless. my plan was to insert a mention of rules, which are supported by RS and are consistent with the table; thus providing support for the existing content. since this isn't possible, I'd advise coming up with another means of supporting the 'standard form' section, since it's conjecture at this point; it may be conjecture I agree with, but it's still unsourced conjecture.

as an aside, there's no need to assume the worst. it seems that JohnUniq didn't read any of the original post, and presumed that I was going on a pointless rant. that's not the case, all this is concerned with the article. I'm also not going to debate about rules any further -- I tried to do so, answering every point, only for it to go nowhere except towards accusations of FILIBUSTERing. no more. Xcalibur (talk) 23:36, 22 December 2020 (UTC)

From one side a single editor persisting with patent filibustering (for over two years) - on the other only occasionally less than saintly patience and forbearance from quite a few editors (I'll refrain from enumerating them at this point). You have had ample opportunity to push your obsession and I really don't think you have any just cause to complain about the way you have been treated here. --Soundofmusicals (talk) 03:44, 23 December 2020 (UTC)

not really complaining, just speaking my mind. and that's a false narrative -- as I said, I brought up a rough draft in late 2018, then dropped it for over a year. I revisited this in May of this year, making many revisions and discussing further, only for it to stall out in June. I gave everyone a break, then proceeded with dispute resolution in July. when that stalled out too, I switched gears towards making other changes, and putting in a brief reference instead of a whole section. the last attempt to include the section was in September, and I responded to posts on it afterwards. add all that up, and it's a handful of months (not counting breaks). I agree that it's been excessive, due to my tendency to answer every point being made, but I haven't really acted in bad faith (eg the 'renewal of old argument' was in response to Spitzak). all I've said is verifiable. I did what I was supposed to do, but now that consensus has proscribed mentioning rules, it's come to a close. it wasn't entirely a waste of time though, I was able to improve and polish my work with your help (in fact I just now polished the wording of rule 2). Xcalibur (talk) 04:19, 23 December 2020 (UTC)