Talk:Roman numerals

From Wikipedia, the free encyclopedia
Jump to: navigation, search
          This article is of interest to the following WikiProjects:
WikiProject Classical Greece and Rome (Rated C-class, Mid-importance)
WikiProject icon This article is part of the WikiProject for Classical Greece and Rome, a group of contributors who write Wikipedia's Classics articles. If you would like to join the WikiProject or learn how to contribute, please see our project page. If you need assistance from a classicist, please see our talk page.
C-Class article C  This article has been rated as C-Class on the project's quality scale.
 Mid  This article has been rated as Mid-importance on the project's importance scale.
WikiProject Mathematics (Rated C-class, High-priority)
WikiProject Mathematics
This article is within the scope of WikiProject Mathematics, a collaborative effort to improve the coverage of Mathematics on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
Mathematics rating:
C Class
High Priority
 Field: Basics (historical)
One of the 500 most frequently viewed mathematics articles.
WikiProject Writing systems (Rated C-class, Low-importance)
WikiProject icon This article falls within the scope of WikiProject Writing systems, a WikiProject interested in improving the encyclopaedic coverage and content of articles relating to writing systems on Wikipedia. If you would like to help out, you are welcome to drop by the project page and/or leave a query at the project’s talk page.
C-Class article C  This article has been rated as C-Class on the project's quality scale.
 Low  This article has been rated as Low-importance on the project's importance scale.
This article has been mentioned by a media organization:


Unicode chart[edit]

Should the chart be replaced with {{Unicode chart Number Forms}}:

Number Forms[1][2]
Official Unicode Consortium code chart (PDF)
  0 1 2 3 4 5 6 7 8 9 A B C D E F
1.^ As of Unicode version 9.0
2.^ Grey areas indicate non-assigned code points

I don't like how it looks, but that entire group of templates could use some cleanup. —Random832 14:27, 20 September 2007 (UTC)

I see Japanese characters in my web browser, Firefox, when I look at that chart. However, when I view the PDF, I can see that the chart should contain Roman numerals in place of the Japanese characters. This may vary from one browser to another. It would be nice to have an explanation (or a link to one) of how to get the Roman numerals to appear in a browser correctly. --Lance E Sloan (talk) 14:45, 3 January 2008 (UTC)
I also use Firefox but do see the PDF Roman numerals, not Japanese characters. Nevertheless, I still see Chinese/Korean/Japanese characters when they appear in Wikipedia articles. The difference may involve how I initially downloaded Firefox because I knew I wanted to see all characters used throughout the world. — Joe Kress (talk) 05:25, 4 January 2008 (UTC)
I particularly am seeing _some_ special characters, but also a lot that are what I guess pass for FFx's/its unifont's attempt at a "code not found" signal - the four-nibble unicode value inside a rectangular box. EG above, the bottom row of figures in the 3~8 columns (with boxes that show 21/83, 21/84 etc, in 2 rows of 2 each). The special characters in 0, 1 and 2 look just fine, though. This is using UTF-8, the only encoding that seems to give any result that isn't a/ everything on the page turning into a mass of unformatted chinese, b/ complete loss of intelligble special characters, replaced by extended ASCII pairs and triplets. Do I need to download some special font (poor show! maybe they need small graphical icons to go in their place?), or is it a coding error by whoever wrote the table? (talk) 19:34, 8 November 2009 (UTC)

Something seriously wrong[edit]

Hi there,

I came on here to determine the exact definition of roman numerals, and the article contradicts itself as far as I can see. The question is: can V proceed L?

The page says "10^n may not precede any symbol larger than 10^(n+1)". 5 is (approximately) 10^0.7. 10^(1 + 0.7) = 50.12. Therefore 5 cannot proceed any number larger than 50.12. Therefore 5 CAN proceed 50, and VL is ok. However, the list at the bottom has XLV for 45.

Well, then there's the fact that you approximated and 10^(n+1) = 10 * 10^n, so 5 cannot precede numbers larger than 50.
However, I believe that Kwamikagami below sums it up pretty nicely - n is an integer, so it only refers to I, X, C, M, etc.
For example, XD will not equal 490, but LD will not equal 450 either. ZtObOr 01:35, 25 February 2009 (UTC)

Furthermore, I can't see in the article anywhere why VX is not allowed! Of course it shouldn't be (since VX = V). Same with LC, and DM. This should be in here!

Would someone please take the time to clean this up? There should be a nice coherent explanation of what is allowed in roman numerals and what is not! And I should not have to search through the entire article to find it!

Rob (talk) 16:51, 24 February 2009 (UTC)

Because n is an integer, as it stated in the second line. I reworded it. kwami (talk) 18:32, 24 February 2009 (UTC)
I only learned Roman Numerals intuitively. But it seems that only integer powers of ten can be used to "decrement" a decimal digit. I.e. I,X,C and logically eventually M. They are used only to create 4's and 9's.--SportWagon (talk) 19:39, 24 February 2009 (UTC)

Think of as Decimal Digits[edit]

Thinking about the confusion made me realize that I keep Roman Numerals straight by thinking of them as decimal digits (with never any zeroes, but all digits are actually "scaled" by the appropriate power of ten). But when I over-think about that, it seems it should be wrong. Were there historical reasons why Romans would have concepts of decimal digits even though they represented numerals with their own system? That is, would it be wrong to include in the article somewhere the notion that Roman Numerals are not completely distinct from decimal digits? That is, as far as I can tell, a correct Roman Numeral can always be split into its non-zero decimal digits. I.e. XLV splits into "XL" for "40" ("4"), and "V" for "5". The incorrect "VL" would not divide that way. Perhaps "XLIX" versus the incorrect "IL" is an even more pointed example. ("Can we find a citation for that?"). MMVIX is still just plain sloppy. ("VL", in contrast, is at least creative). --SportWagon (talk) 20:00, 24 February 2009 (UTC)

The spoken language was decimal. It makes sense that if you say "two hundreds and thirty (and) five", you would write CC + XXX + V. Also, in speech the rule is "five and forty" (as in German) or "forty five" (as in English), so maybe writing "VL" would have been confused with 55. 9, on the other hand, was just "nine", and there was no "one and ten" for 11, so IX and XI do not have that problem. kwami (talk) 21:50, 24 February 2009 (UTC)
Yes, the current article does imply Romans counted in decimal. For people who get confused, it seems it could be suggested that any valid Roman numeral must use zero or one combination from each of the lines of the table at the end of the Symbols section (written from left to right, lower lines of table first). But that notion could be introduced in at least two places. and there are already notes about duplication in the article, so I just offer the thought here for now.--SportWagon (talk) 22:08, 24 February 2009 (UTC)
Zero? There is no zero. kwami (talk) 23:01, 24 February 2009 (UTC)
Either you jest, or I didn't explain correctly. Take the table I indicate. From the bottom, go up to the first line with values small enough for the number in question. Select the appropriate value from that line. (This will be the first "digit" of your number, suitably scaled). Then go up to successive lines. If the particular digit is zero, then you pick zero combinations from that line, otherwise you must pick exactly one--the one corresponding to the digit in question--and append it to the numeral you are creating. You keep doing that until you get to the top of the table. Thus my "zero or one". The table defines the combinations which can occur in valid (modern) Roman Numerals; it is more restrictive than more general rules would imply.--SportWagon (talk) 00:18, 25 February 2009 (UTC)
I think that's suitably explained now. kwami (talk) 08:21, 25 February 2009 (UTC)

Definitely Not Decimal? Not.[edit]

A revision comment states that the Roman Numeral system is definitely not decimal. Simple inspection would seem to say that the system is decimal (based on powers of 10). The basic symbols all represent either powers of ten, or five times a power of ten. A Roman Numeral can be visualized as its decimal digits, and omitted zero placeholders (which are unnecessary because the magnitude is explicitly indicated by the choice of symbols). It's not like the symbols represent dozens and gross, or other truly non-decimal quantities. True one might be able to construct restrictive definitions of "decimal system" which would exclude Roman Numerals, but that wouldn't seem to qualify as "definitely". Given the obviously decimal nature of the system, I don't see why one would need an explicit source for using the word, especially in one of the deletions.--SportWagon (talk) 18:12, 25 February 2009 (UTC)

Attempting to find web citations found this unfortunate garbage where, in the middle of the page, they seem to encourage very young readers to do things like VL. [1]--SportWagon (talk) 18:36, 25 February 2009 (UTC)
Here's someone who agrees with the "decimal" concept.[2]--SportWagon (talk) 18:36, 25 February 2009 (UTC)
web search does indicate most people contrast "Roman Numerals" to "decimal", however. That is, when they say "decimal" they mean more than what we mean by "decimal".--SportWagon (talk) 18:36, 25 February 2009 (UTC)
Another programmer who agrees with much of what we say. [3].--SportWagon (talk) 18:53, 25 February 2009 (UTC)
A "Math Forum" (not forum in that sense...)[4]. Contains further references.--SportWagon (talk) 18:57, 25 February 2009 (UTC)

I was the one deleting the "decimal" remark: thanks for your comments. I must admit that I have not yet checked the links you give, but if Wikipedia has to be consistent as a whole, this article must agree with Decimal (and with Decimal representation). In those articles, and for what I know in the current use of the word (when talking about numeration systems), a decimal numeration system is before anything else a positional system, that is, one in which each digits gets a meaning depending on where it is in the representation of a number. So in "13" the digit "3" denotes three units, while in "31" it denotes three "tens", and so on. So merely the fact that some of the symbols used in Roman numerals denote powers of ten is not sufficient to qualify it as "decimal". Thanks, Goochelaar (talk) 18:59, 25 February 2009 (UTC)

I'll leave it to kwami to choose some appropriate compromises. It seems like the word we want might have disappeared from the English language as "decimal" has taken on extra implications. Ignoring pedantic interpretations of "decimal", the now omitted paragraph following the table should simplify readers' thinking. (And true, one can just as easily say our attempted use of "decimal" is the pedantic one...)--SportWagon (talk) 19:12, 25 February 2009 (UTC)
Egyptian numerals and Chinese numerals are examples of decimal systems which are certainly not positional because they use different symbols for every power of ten. — Joe Kress (talk) 19:41, 25 February 2009 (UTC)
Declaring that decimal must imply positional seems pedantic because it insists one cannot understand a broader definition of a term in different contexts. However, the way you used "decimal" can also be considered pedantic because it would appear that "the decimal system" and "decimal numbers", even just "decimal" are commonly understood to imply the currently widespread positional system. That is it requires the reader to not make common assumptions. I will keep out of the editing for now.--SportWagon (talk) 20:58, 25 February 2009 (UTC)
"Decimal" means base 10, just as "binary" means base 2. That's all it means. We speak of languages as having decimal, vigessimal, etc. systems, but spoken languages are neither positional nor do they have a zero. Other than the auxiliary base 5 (presumably due to the limits of visual processing of iterated symbols), Roman numerals are analogous to the decimal numbering system of the Latin language. If our decimal article is wrong, then it needs to be corrected. The intro to this article clearly states that Roman numerals are decimal but not positional. Though a check with a dictionary is all that should be needed, I added a ref from Ifrah. As far as compromising with people who do not understand the concepts involved, that would be like stating that whales are fish as a compromise with people don't know the difference. kwami (talk) 21:15, 25 February 2009 (UTC)
Kwamigami, thanks for your good work on this article, but I must disagree. First of all, the reference you give uses the words "a decimal system in which the number 5 is an auxiliary base" not about Roman numerals, but about a previous, archaic, conjectural numbering system an hypothetical herdsman might have developed in order to tally his animals; in fact, it follows "(and the numbers 2 and 5 are alternating bases)". Second, we might as well say that it is a "quinary", or base 5, system. Third, even if we find and agree on a source that describes Roman numerals as decimal under a broader definition of "decimal", we should immediately modify Decimal and Decimal representation (we have to be able to link the word "decimal" in this article to one of those articles).
You are right about languages, but we are not covering Latin language here; only this numeration system as used then and now in several countries with several different languages (in fact, nowhere is told anything about the Latin names for composite numbers, that is, different from 1, 5, 10 etc.).
As this is an encyclopedia, we are forced to be careful about the meaning we attribute to the words we use (even when this looks like pedantry). Goochelaar (talk) 23:45, 25 February 2009 (UTC)
It is nothing close to being a base 5 system. 25, 125, 625 etc. do not fall out as being simple representations the way C and M, and larger powers of ten do. The use of base 5 is merely a means of shortening what are conceptually decimal digits.--SportWagon (talk) 23:55, 25 February 2009 (UTC)
You are onto something here. The use of a representation of 5 is merely a means of shorting what is essentially a unary system. And the use of a representation for 10 is merely a means of shortening what is essentially a unary system. At no point is it decimal or quinary. The fact that symbols exist for V and X actually specifically exclude it from that - V and X are both 10 in decimal and quinary. It's the point at which the system rolls over and starts using other representations. V, X, L, C, etc... are merely convenient representations for large number of I's. By tradition, of course, you put these in order, and you use the representation that takes the smallest amount of space (so no VV or IIIII). But that doesn't make it decimal. Having a representation for 0, 1, 2...9 and then rolling over to 10 at the end makes a system decimal. If I rolled over to my own representation of 10, such as X, or A, I would not have a decimal system. Math in Roman numerals looks nothing like decimal math. It looks like unary math. Because Roman Numerals are unary. (talk) 03:35, 29 September 2011 (UTC)
Yes, Ifrah was speculating on the origins of the system, but what he ended up with was a decimal system "exactly the same as in the Roman system". As for language, my point was that if decimal means base 10 when describing numeral systems in language, it means base 10 when describing numeral systems in writing. I've seen no reason to believe that the word "decimal" changes definition depending on which medium we're discussing. SportWagon is right. There is no *VVV for 15, or *LLLL for 200, which is what Ifrah meant by 5 being "auxiliary". Roman numerals are base 10, therefore decimal. kwami (talk) 00:24, 26 February 2009 (UTC)
Roman numerals aren't decimal. They're unary with extra symbols thrown in so that you don't have to write out ridiculous number of I's, which are themselves stacked on each other in a unary fashion. The fact that these special symbols, which, in actuality, are simply representations for a large amount of unary I's, are usually based on powers of 10 doesn't make the system decimal. If it were base-10, it wouldn't have a representation for 10. No base-10 system has its own representation for 10. 10 is the place where the system rolls over and starts re-using other representations. So, in fact, having its own representation for 10 is one thing that would actually definitively exclude it from being base-10. And having its own representation for five doesn't make five an "auxillary base". Base 5 doesn't have a 5. 5 is 10 in base-5. Roman numerals are, again, base-1. — Preceding unsigned comment added by (talk) 03:22, 29 September 2011 (UTC)
There are two main definitions. The historically primary definition deals with fractions, as decim means a tenth, but since at least 1684 the phrase "decimal fraction" has been used to disambiguate. Roman numerals are not decimal in this sense, since the fractions were duodecimal. But no-one uses Roman fractions anymore; for nearly everyone, we're talking about the non-fractional part. The second definition, per the OED, is "decimal numeration, the numerical system generally prevalent in all ages, of which 10 forms the basis; i.e. in which the units have distinct names up to 10, and the higher numbers are expressed by multiples or powers of 10 with the units added as required." There is ambiguity with decimal referring to decimal point & decimal places, etc., but the intro is clear enough for the reader to follow.

To be precise, Ifrah says that "the successive order of magnitude [used by the hypothetical herdsman] are exactly the same as in Roman system", not the system itself about which he only says that "the graphical forms for the figures ... are closely comparable with those in the archaic Roman and Etruscan systems". More importantly, the paragraph

The number system of the Latin language was decimal. That is, one said "one thousand and two hundreds and thirty [and] four". When writing a Roman number, the thousands, hundreds, tens, and units in the chart above are strung together the way they are spoken: M (one thousand) + CC (two hundreds) + XXX (thirty) + IV (four), for MCCXXXIV. Thus eleven is XI, 32 is XXXII, and 45 is XLV. Note that the subtractive principle is not extended beyond the chart, and *VL is not used for 45, as it does not correspond to the spoken language.

cannot stay as it is. Of course, in Latin one did not say "one thousand etc."; if anything "mille etc." Moreover, we cannot bring Latin language into this, because it would immediately contradict what is being said: "eighteen" is in Latin "duodeviginti", that is, "two-from-twenty", which would suggest such an expression as *IIXX, and similarly for 19, 28, 29 and so on. In order to find a solution acceptable to everybody, I suggest rewriting the former along the lines of

A practical way to write a Roman number is to consider it as if it were written in the modern decimal number system, and string together separately the thousands, hundreds, tens, and units as given in the chart above. So, for instance, 1234 may be thought of as "one thousand and two hundreds and thirty [and] four", obtaining M (one thousand) + CC (two hundreds) + XXX (thirty) + IV (four), for MCCXXXIV. Thus eleven is XI, 32 is XXXII, and 45 is XLV. Note that the subtractive principle is not extended beyond the chart, and *VL is not used for 45.

Would this be acceptable? (The problem would still remain to make this article and those on decimal notation not contradict each other.) Goochelaar (talk) 01:11, 26 February 2009 (UTC)

Yeah, that's better. I'd forgotten about duodeviginti. I'll go ahead and change it. kwami (talk) 02:17, 26 February 2009 (UTC)

The new paragraph is about what I thought could be asserted. The advice also helps for reading Roman Numerals. Why not say "numeral" rather than "number"? When you begin to move towards insisting that decimal comes from "tenth" rather than "ten", you also start to call some explanations of the term "decimal system" into question. Strange, web searches for "decemal" find first references to "decemal point". Oh well. Would some remaining uses of the term "decimal" in this page be better changed to references to "powers of 10", or possibly "base 10"?--SportWagon (talk) 03:21, 26 February 2009 (UTC)

We could probably come up with different wording, but I don't see the point. Decimālis means 'pertaining to decima, which means 'tenth' or 'tithe'. The earliest usage I can find is for writing fractions rather than xx/xxx. However, that is only one use out of several, and when people speak of decimal numeration, it has nothing to do with fractions, just as it has nothing to do with tithing. —kwami (talk) 04:11, 26 February 2009 (UTC)

It seems that it would be more accurate to consider Roman Numerals a unary system with extra symbols thrown in so that you don't have to write out a ridiculous amount of I's to represent larger numbers. But these extra symbols are simply stacked on each other in a base-1 fashion. Decimal implies that it's base-10, which it's clearly not. Having a convenient symbol for 10 doesn't make something base-10. In fact, base 10 actually obviously doesn't even give 10 it's own symbol. That's one thing that sets it apart from more primitive systems. (talk) 03:09, 29 September 2011 (UTC)

XCIX vs. IC[edit]

The section is a fine explanation of a set of sensible if arbitrary restrictions on "cutting short" roman numerals. However, I disagree with its claim that this is "the modern way" to do it or that it has more claim to ubiquity than the other variants. (talk) 21:21, 12 April 2009 (UTC)

Neither of them is right. The only correct way to express 99 in Roman numerals is LXXXXIX. --Kar98 (talk) 04:36, 28 October 2011 (UTC)
Piffle. XCIX is by far the most commonly used and is therefore correct. If you accept including subtractive numerals (e.g. IX) in the system, as you apparently do if you advocate "LXXXXIX", then there is no reason not to accept XC and therefore also XCIX as well. (talk) 20:52, 10 February 2012 (UTC)
Agreed, if you wanted to write 99 without using any subtractives it would be LXXXXVIIII. That said, since Roman numerals are written in pretty much whatever version you want, any of the above is technically "correct." (talk) 15:37, 11 June 2013 (UTC)
There's a germ of truth in that somewhere - but of course if Roman numerals really were "written in pretty much whatever version you want" they'd very often be ambiguous, or inconsistent and hard (or impossible) to read. A bit like the way English was spelled in (say) the time of Shakespeare - is the first line of Hamlet's famous soliloquy "Oh that this too, too solid flesh would melt", or "Oh that this too too sullied flesh would melt"? You'll actually see both in different editions of Hamlet - because Shakespeare's own spelling is so haphazard it simply isn't clear which he meant, and both words fit the context quite well. Assuming we want to use Roman numerals to unambiguously designate actual numbers, and not just for decoration, and that we want people to be able to read them, as well as thinking how pretty they are, then it is a good idea to stick to the rules as given in the article, which are clear, unambiguous, and "correct". Common variations are in fact few and far between. "IIII" is quite a common way of writing "IV" (especially on clocks and coins) - and "LXXXX" is occasionally seen for "XC". The only other variation I've ever seen more than one instance of is "IIXX" for "XVIII", which is derived from Latin idiom - they used to say "twenty less two" for "eighteen". But these forms, not to mention other, more fanciful variations, are honestly best avoided if you want to write a Roman numeral that has an unambiguous and plainly read "Arabic" equivalent, and will be read in the correct way by anyone familiar with the system. That is all the word "correct" implies in this context. --Soundofmusicals (talk) 20:18, 11 June 2013 (UTC)

Template:romannumeral up for deletion[edit]

{{romannumeral}} has been sent to WP:TFD (talk) 07:49, 27 April 2009 (UTC)

Double subtraction?[edit]

Somebody I know is "XXIIX" according to his Facebook status today, using "IIX" to mean 8. It's always been my understanding that a given instance of a symbol can only be subtracted from once. Is he just plain wrong, or was such double subtraction once deemed valid? On a quick look through the article, there doesn't seem to be any comment on it. -- Smjg (talk) 12:51, 18 June 2009 (UTC)

You only subtract with one numeral (Always keeping it as simple as possible), and it's always the second to last numeral (the numeral that's smaller than the last one in the sequence), so in this case it's VIII for 8, not IIX) Nizzemancer (talk) 06:30, 14 July 2011 (UTC)

I HAVE seen "IIXX" for "XVIII". The Romans may have used this form for all I know (look up the Latin for "eighteen"). It's not "correct" according to the usual rules we use nowadays, anyway.--Soundofmusicals (talk) 20:43, 11 June 2013 (UTC)

Dating Roman Numerals[edit]

It seems to me (and I would be grateful for comment) that this article suffers from assuming that the Roman system of numerals were invariant throughout the life of the Roman Empire and also that the Roman Catholic church maintained these same methods after the demise of Rome.

As I understand it, research at Housesteads suggests that such assumptions are invalid for the language, never mind counting, and that the average Legionaire spoke a languege wildly different (not a mere dialect of Latin) from that subsequently used by the Church.

My point is that asertions as to methods of counting must be associated with a date to which they apply, or a period over which they are believed to have been in use. With over a thousand years of Empire, changes are to be expected.

Clearly in a system of written numerals which did not employ a zero as a device for maintaining the position of other digits, some mechanical system which avoids this grave shortcoming is essential in support of written records, but what were they uing in 750BC?


Drg40 (talk) 17:40, 22 August 2009 (UTC)

Actual "mechanical" calculations used to be done on an abacus, which of course works on a place system - but the idea of writing numbers down according to the number of beads in the column of an abacus (what our modern number boils down to) is, believe it or not, NOT a self-evident idea at all. The Indians seem to have been the first to think of it, and from them it passed to the Persians and then the Arabs (and finally to us backward Europeans, several centuries late!). But writing numerals has the same relation to counting (or calculation) as speech has to writing. Pre-literate people can count (and calculate for that matter) every bit as well as we can - they just can't write the result down). Your point about Roman Numerals changing over the centuries is of course perfectly true - perhaps we need more detail in the "history" section? Know any good sources? --Soundofmusicals (talk) 20:58, 11 June 2013 (UTC)

The Roman numerals: IIII vs IV section[edit]

It troubles me that all the references in this section are from the late 14th century and beyond. Don't we have any authentic evidence from the time of the Roman empire? I'm sure you'd find lots of errors and incosistancies if you studied 10th Century Old English by looking at its use by the Portuguese in the 20th! ~~ Dr Dec (Talk) ~~ 17:13, 19 September 2009 (UTC)


The article says they used nulla for zero. Unary numeral system says they used nullae and sloppily translates this as nothing. Wouldn't it be nulli, nullae or nulla according to the gender of what was counted?-- (talk) 02:45, 28 November 2009 (UTC)

I've only seen nulla or nullae used as a number for the first entry in a table of epacts, with all other entries being Roman numerals. Cyclus Decemnovennalis Dionysii is an example of this usage. Early medieval authors used nulla in this situation, while later medieval authors used nullae. I've never seen nulli used in this situation. Translations into English of works by medieval authors containing such a table replace nulla or nullae by "zero" or "0" and replace the Roman numerals by the appropriate Arabic numeral. — Joe Kress (talk) 08:35, 28 November 2009 (UTC)

The article says: "About 725, Bede or one of his colleagues used the letter N, the initial of nulla, in a table of epacts, all written in Roman numerals." Do you have a reference? User: —Preceding unsigned comment added by Bussinchen (talkcontribs) 15:55, 20 July 2010 (UTC)

It is in Corpus Christianorum, Series Latina, Vol. 123C: Opera Didascalica. Ed. C.W. Jones. — Joe Kress (talk) 16:26, 20 July 2010 (UTC)

New usage needed under "Modern usage?"[edit]

The Modern usage section mentions using Roman numerals to denote monarchs and Popes, but it can also be used in regular family lines. For example, Harrison Greeley -> Harrison Greeley, Jr. -> Harrison Greeley III -> Harrison Greeley IV, and so on. —Preceding unsigned comment added by (talk) 02:11, 12 December 2009 (UTC)

I've added a section to Modern Usage about how the BBC uses Roman numerals for copyright dates. I wrote the BBC guidelines on copyright dates in 1990 and researched the whole field of Roman numerals in order to come up with a system that would be clear and last for a few hundred years. It all arose from a BBC design manual being published that gave the example year as "©BBC MCMLXXXX" when I believed it should be "©BBC MCMXC". The BBC did use normal numerals in the 1970s but found that people complained too much about the number of repeats or the age of the programmes. People could just read the copyright date too easily! I didn't put this in the article as I have no references to prove it - it's just the BBC legend. GrahamPadruig (talk) 00:00, 3 April 2010 (UTC)

It's still just a BBC legend, in spite of the claimed sources added. I've moved what we know up into the bulleted list. — Arthur Rubin (talk) 09:32, 26 April 2010 (UTC)
It seems the BBC would like to abolish many things, and when they make an effort, all there is to show for it, are all the complaints.-- (talk) 11:44, 14 September 2010 (UTC)

I really don't know whether this is the place to mention this or not - apologies if it isn't - but the reference to Jupiter's moon states "currently" 63. Surely this should be qualified by some date or other as more moons might well be discovered in due course. (talk) 18:01, 4 July 2010 (UTC)

Need a new clock picture[edit]

The photo used to illustrate the calendars and clocks section is a poor choice, being an idiosyncratic 24-hour analog face (with a zero, even!). The numerals are not positioned the way the text describes. Spark240 (talk) 00:12, 21 January 2010 (UTC)

I added a more typical clockface (and left the other one too) Modest Genius talk 17:29, 24 January 2010 (UTC)


explain why the roman system is not a place value system? and what are the diadvantages of the Roamn system —Preceding unsigned comment added by (talk) 21:25, 24 January 2010 (UTC)

Just try to do arithmetic using it? (Hint - that wasn't what it was for - they used an abacus instead.--Soundofmusicals (talk) 21:01, 11 June 2013 (UTC)

Archaic Numerals[edit]

I've seen the numeral "XIIX". What does this mean? AmericanLeMans (talk) 17:47, 12 March 2010 (UTC)

18 —Preceding unsigned comment added by (talk) 02:37, 17 January 2011 (UTC)

More often seen as "IIXX", I think. But either way it's an "abnormal" form. We usually write "18" as "XVIII". --Soundofmusicals (talk) 21:03, 11 June 2013 (UTC)

Added footnote references for use of letter "j" to replace terminal letter "i"[edit]

I was reading this article and noticed the (unreferenced) mention of the use of a letter j to replace a terminal letter i, especially for medical prescriptions, a practice I'd never heard of. Since there were no citations for this information, I decided to research it on-line for myself. In my opinion, most of the more recent sites that mention this practice appear to be simple "echoes" of this Wikipedia article and, therefore, would be unsuitable as citable references. When I limited my searches to much older texts (e.g., Google Books), I found multiple instances describing this practice. From those, I've selected two, which I hope are sufficient. Bgpaulus (talk) 18:55, 15 March 2010 (UTC)

I have seen this in 15th-16th century works, but I need to dig up citations.Dogface (talk) 05:29, 10 May 2011 (UTC)

Symbols not visible on Firefox[edit]

In the section "Alternate Forms," non-ascii-character symbols come through as weird tiny numbers in boxes (21/80 in a box), rather than as what they're supposed to be (e.g. a lemniscate). This may be a problem with my version of Firefox, but others probably have the problem too.

"Sometimes CIƆ was reduced to a lemniscate symbol (ↀ) for denoting 1,000. John Wallis is often credited for introducing this symbol to represent infinity (∞), and one conjecture is that he based it on this usage, since 1,000 was hyperbolically used to represent very large numbers. Similarly, 5,000 (IƆƆ) was reduced to ↁ; and 10,000 (CCIƆƆ) was reduced to ↂ." (talk) 20:03, 15 May 2010 (UTC)


Seems that this page could use some reordering. Perhaps origins first (though it would equally be ok to leave symbols first, for quick perusal/explanation). The subrtractive principle needs to be in with the Symbols section, since the principle is mentioned well before the principle itself is explained. Perhaps Zero should be before symbols or after fractions? After the above, then IIII and IV, then Modern Usage and Clock Faces (or clock faces then modern usage, whichever fits well). —Preceding unsigned comment added by (talk) 18:18, 2 June 2010 (UTC)

I concur with the anon comment. I am a casual but not-infrequent visitor to this page and find it clumsy to navigate.--Cruickshanks (talk) 01:40, 10 July 2010 (UTC)

IC is possible[edit]

In Ancient Rome, writing numbers in the form IC for 99 instead of XCIX was allowed. Similarly, 999 could be written as IM. This is not mentioned in the article. (talk) 23:19, 15 October 2010 (UTC)

If you would provide a reliable source for this, it might be included in the article. — Arthur Rubin (talk) 23:29, 15 October 2010 (UTC)
Hmm... My Latin teacher said that, and there are probably classical documents somewhere that have these numerals. I realize that she does not count as a source, but I might be able to find something. (talk) 01:32, 16 October 2010 (UTC)
Here is a link to an article from Wolfram MathWorld. Specifically, it says:
"It should also be noted that the Romans themselves never wrote M for 1000, but instead wrote (I) for , (I)(I) for , etc., and also occasionally wrote IM, IIM, etc. (Menninger 1992, p. 281; Cajori 1993, p. 32). " (talk) 03:13, 16 October 2010 (UTC)
I don't think that meant 999, I think it was a way of writing 1000, 2000 by multiplying what was before by 1000. The IV and IX maps obviously to the abacus but IL for instance doesn't. Dmcq (talk) 09:25, 28 October 2011 (UTC)

(I) versus M for thousand[edit]

I'm new to Wiki Talk, so please forgive me any formatting errors.

There is an inconsistency in the "Modern Roman numerals" table.

I ran all the numbers of that table through a "roman number to string" converter using the roman "digits" further on.

The table mixes M and (I) for thousand. Note: I'm using the digits in parenthesis to avoid Unicode conversion issues.

In essence, it comes down to these values in the table being inconsistent:

    6666, '(V)MDCLXVI'
    1666000, '(M)(D)(C)(L)(X)(V)(I)'
    3888000, '(M)(M)(M)(D)(C)(C)(C)(L)(X)(X)(X)(V)(I)(I)(I)'

With 6666, M is used for thousand. With 1666000 and 3888000, (I) are used for thousand.

That is inconsistent, so it should either be mentioned that the table is inconsistent, or the table should be made consistent by making a single choice the digit representing thousand: M or (I).

Roman "digits" used for conversion and testing:

   1000000, '(M)'
   900000, '(C)(M)'
   500000, '(D)'
   400000, '(C)(D)'
   100000, '(C)'
   90000, '(X)(C)'
   50000, '(L)'
   40000, '(X)(L)'
   10000, '(X)'
   9000, '(I)(X)'
   5000, '(V)'
   4000, '(I)(V)'
   1000, 'M'
   900, 'CM'
   500, 'D'
   400, 'CD'
   100, 'C'
   90, 'XC'
   50, 'L'
   40, 'XL'
   10, 'X'
   9, 'IX'
   5, 'V'
   4, 'IV'
   1, 'I'

Jpluimers (talk) 11:39, 22 October 2010 (UTC)

Use of 'Quotes' to make pluralizing 'I's, 'V's and 'X's more legible?[edit]

This might make the article more legible on first read. --ButterSoda (talk) 14:40, 31 December 2010 (UTC)

You are correct that legibility is a bit of a problem (example "twenty Is, four Vs, and four Xs"), but an apostrophe is not the solution. Many people here (I'm one of them) hate incorrect usage of apostrophes, and we automatically change "in the 1990's" to the correct "in the 1900s". Johnuniq (talk) 03:32, 1 January 2011 (UTC)
Buttersoda did not propose using apostrophes to form the plural, he proposed using quotes around each Roman numeral. — Joe Kress (talk) 09:38, 3 January 2011 (UTC)
Ooops, I must have had apostrophes on my mind when I saw the OP. Sorry about my mistaken comment. Here is how some of the text might appear (original and proposal):
With IIII, the number of symbols on the clock totals twenty Is, four Vs, and four Xs
With IIII, the number of symbols on the clock totals twenty 'I's, four 'V's, and four 'X's
The single quotes do not look quite right to me, particularly when thinking about the consistency problems (one I and two 'I's). I don't think there is a good solution. I am tempted to investigate a template to set the font to something serifed for every instance on the page (e.g. wikitext might appear as {{rn|I}}s to produce Is), but that seems overly complex (and we would need help from WP:VPT to see if browser or other problems would be likely). Johnuniq (talk) 05:59, 4 January 2011 (UTC)

Multiples by 100,000 | Nested multiplication[edit]

According to Mathworld there was an additional way of writing large numbers, that is by putting a box with no bottom around the number. This should be included in the article, as it shows how to write extremely large numbers with Roman numerals. image

In addition, the mentioned article tells of the Roman practice of using nested multiplications of 10. Thus, quoted, “The Romans sometimes used multiple parentheses to denote nested multiplications by 10, so (I) for 1,000, ((I)) for 10,000, (((I))) for 100,000, etc. (Cajori 1993, p. 33).”

CannedMan (talk) 17:55, 1 May 2011 (UTC) (fixed web link formatting —Coroboy (talk) 03:24, 2 May 2011 (UTC))

File:Westerkerk MDCXXX.jpg Nominated for Deletion[edit]

Icon Now Commons orange.svg An image used in this article, File:Westerkerk MDCXXX.jpg, has been nominated for deletion at Wikimedia Commons for the following reason: Deletion requests May 2011
What should I do?
A discussion will now take place over on Commons about whether to remove the file. If you feel the deletion can be contested then please do so (commons:COM:SPEEDY has further information). Otherwise consider finding a replacement image before deletion occurs.

This notification is provided by a Bot, currently under trial --CommonsNotificationBot (talk) 06:53, 30 May 2011 (UTC)

Remove programming section?[edit]

I have three objections to the programming section:

  1. No obvious target audience. Who is it exactly that benefits from this?
  2. Requires specialist knowledge to understand. Out of all people who come to the "Roman Numerals" article for information, what percent will understand C++?
  3. There has been no standardization of how to convert decimal numbers to Roman Numerals. The algorithms present *a* way to convert numbers, not *the* way.

--RSLxii 16:38, 19 October 2011 (UTC)

I agree that the programming section is not helpful for this article (despite the accurate assertion that it is a common programming exercise). Further, if such code were warranted, a language like Python should be used with no tricky constructions so that it reads like pseudocode and would be comprehensible to an interested passer by.
However I am not happy with the deletion of the Unicode numerals section since the "see also" to Unicode numerals#Roman numerals does not have what appears to be the same useful information (i.e. I am hoping someone will review the deleted material and copy what is useful to the target, using "moved from [[Roman numerals]]" in the edit summary to satisfy WP:CWW).
One trivial point about both changes: syle would be "Programming conversion" (lowercase c) and "Roman numerals" (lowercase n). Johnuniq (talk) 21:59, 19 October 2011 (UTC)
I was going to move the material, but then I saw that the info here contradicted the info there. I would if I knew more about unicode....--RSLxii 17:01, 20 October 2011 (UTC)
Well, I gave it a shot. The unicode numerals article now has the info from this article. If anyone wants to double-check my work.... --RSLxii 17:36, 20 October 2011 (UTC)
It is an improvement to this article, however I do not know enough to comment on the other article.Beefcake6412 (talk) 17:41, 20 October 2011 (UTC)

MDCCCVIL = 1844[edit]

Strange Roman numeral

The lighthouse at Røsnæs near Kalundborg, Denmark, was built in 1844. On the side of the lighthouse the year is indicated as MDCCCVIL (1800 + 50-6). Is the enclosed picture interesting enough to put on the article page, or is it simply so weird that it holds no relevant information? What do you think? --Oz1cz (talk) 13:43, 23 October 2011 (UTC)

By the way, the monogram is that of king Christian VIII. --Oz1cz (talk) 13:46, 23 October 2011 (UTC)
If there were fewer strange uses already listed, I would be happy to have it in the article. But I'd hate for the number of "regluar" roman numerals to be outnumbered by the strange cases in the article. That's my opinion, anyways.--RSLxii 15:00, 24 October 2011 (UTC)
That sounds reasonable. --Oz1cz (talk) 08:10, 25 October 2011 (UTC)
Just a comment - the proper Roman numeral for 1844 would of course be MDCCCXLIV - which wouldn't have fitted on the sign! A case of sign-writers licence? I don't think one need look further for an explanation of this thoroughly weird numeral. --Soundofmusicals (talk) 14:57, 19 February 2013 (UTC)

Historical and modern usage examples[edit]

"Because there has been no standardization, there may be multiple ways of representing the same number in Roman numerals.[2] Historical and modern usage include the following examples:

  • 9 as IX, 4 as IIII or IV (Forme of Cury, a manuscript from 1390)
  • 90 as LXL, 83 as XXCIII, 78 as LXXIIX, 6 as IIIIX.[3]
  • 5 as IIIII, eight as IIX, VV as 10[citation needed]
  • 29 as XXVIIII, 44 as XLIIII (Gate numbers on the Colosseum)[2]
  • 99 as XCIX (1524 German arithmetic textbook)[2]
  • 1606 as MCCCCCCVI (inscription in Sant' Agnese fuori le Mura church in Rome) [2]
  • 1910 as MDCCCCX (Admiralty Arch in London)
  • 1954 as MCMLIV (Trailer for the movie The Last Time I Saw Paris)[4]"

This section is simply both confusing and combines some common variations of the system with odd instances which simply look like errors. First of all:

  • 4 is commonly written as either IV or IIII,
  • so by the same principle 44 may be either XLIV or XLIIII.


  • also 9 (5+4) is written as either VIIII or IX
  • so by the same principle also 19 may be written as XIX or XVIIII
  • and by the same principle also 29 may be written as XXIX or XXVIIII, and so on.


  • by similar principles 40 may be either XL or XXXX, 90 may be XC or LXXXX, 400 may be either CD or CCCC, 900 may be either CM or DCCCC
  • therefore there is nothing strange in XCIX for 99, nor in MDCCCCX for 1910.

The rest of the examples:

  • 90 as LXL, 83 as XXCIII, 78 as LXXIIX, 6 as IIIIX.[3]
  • 5 as IIIII, eight as IIX, VV as 10[citation needed]
  • 1606 as MCCCCCCVI (inscription in Sant' Agnese fuori le Mura church in Rome) [2]

are most probably rather rare errors caused either by ignorance of the system, bad taste (willingness to be "original" can be a disaster when the willing person is completely untalented), or even are pure fantasies. — Preceding unsigned comment added by (talk) 00:15, 29 October 2011 (UTC)

The list of examples is not meant distinguish between 'normal' and 'strange'; it is only meant to show that numbers can be written in a variety of ways. In general the examples you complain about as being "probably rare errors" were deliberately picked by the sources cited to show how much variation was possible.
In the research I did on the topic, I found plenty of sources citing all kinds of supposed rules for the "right" way to make Roman Numerals. But despite my trying (I tried pretty hard), I could not find an authoritative source that said "here are the most prevalent set of rules being used for Roman Numerals." In fact, I found just the opposite--sources describing a vast Roman Empire where Roman Numeral use varied region to region. Sources describing the lack of a modern standard for writing Roman Numerals.
If you find some kind of proof that there is or has been a set of commonly-applied rules, by all means cite and correct what is in the article. If such a thing exists, I would welcome it--it was hard for me to write that section in a way that correctly describes the state of ambiguity in the rules. --RSLxii 02:26, 29 October 2011 (UTC)
I am sorry, but that's not the way it works. The burden of proof that "LXL" and such are correct, is at your side. From my point of view, it is enough to see that such examples are abnormal, so please provide reliable sources to support them. BTW. for IIIII, IIX and VV even the article says that there are no such sources. So, mixing such stuff with examples of regular, consistent, frequently seen variations of the system (such as IIII vs. IV) is, as I wrote above, confusing. And, if there are Roman inscriptions which confirm regular usage of "90 as LXL, 83 as XXCIII, 78 as LXXIIX, 6 as IIIIX", I'd like to see them, because at present I don't believe they're real (or, if you prefer, transcribed correctly). My sources ("Inscriptionum Latinarum Sylloge", pp. 13-14) only mention subtractions up to two in numerals above 10 (i.e. IIL = 48, which agrees with the Roman habit of subtracting up to two from every ten except the first one, for 48 is duo-de-quinquaginta, "two less than 50"); and subtraction from 10, and subtraction of more ones than II is not mentioned, so I would really want to see the otherwise unsourced IIX as 8 and IIIIX as 6. PS. BTW. seeing IIIIX I would interpret it as "14" (as digits written in the order of saying the number, as 14 is quattuor-decim (4+10). — Preceding unsigned comment added by (talk) 15:52, 31 October 2011 (UTC)
The burden of proof has been satisfied for all but the one uncited example. Feel free to remove that one if you like. It wasn't my addition anyways--I marked it as "uncited", hoping the original editor would provide a source. --RSLxii 17:07, 31 October 2011 (UTC)
I did a quick look for a source--maybe this is the one that was used for the uncited example: (look starting from page 31)

You mentioned that IIIIX you would interpret as 14. The source has similar examples (IIIX as thirteen, VIX as sixteen), which would probably be good to put into the examples section as well. --RSLxii 17:25, 31 October 2011 (UTC)
On further thought, I'd hate for an example like IIIX = 13 to confuse people even more than they need to would be better in the "history" section. --RSLxii 20:05, 31 October 2011 (UTC)
One more note: I remember now that, as I was putting this section together, I was uneasy putting in so many historical examples in a section that was really meant to be a practical introduction in how to read and write modern Roman Numerals. The average person isn't going to want to know all the historical quirks. But on the other hand, I could not find any source, let alone a modern source, that could be called an authoritative source of the rules. I left the examples just to show that such craziness is possible, given that no rules have been standardized. I would be much happier taking them out and saying "commonly, the subtractive principle is used in this way". But I need a source that says something to that effect. All the sources I've found suggest that no such source exists. --RSLxii 17:31, 31 October 2011 (UTC)
Hmmm thinking about it, I've got a better idea of how to structure this whole thing....give me a second--hopefully the final result will address your concerns as well. --RSLxii 20:26, 31 October 2011 (UTC)

OK, I ended up finding a not-so-great source (a computer programming book) that states that there is a common set of rules that get applied. I would feel much more comfortable if a history book made such a claim, but on the other hand, I doubt it's a controversial claim. The only problem with the source is that it doesn't actually state all the rules! So the way it's written now, I feel I've engaged in WP:SYNTH. If anyone can find better sources, I'd appreciate it. --RSLxii 21:04, 31 October 2011 (UTC)

Also, I'd like to see more modern examples of Roman numeral use. Please add if you have time. Thanks. --RSLxii 21:10, 31 October 2011 (UTC)

Use in TV and film[edit]

I'm not convinced that the use of roman numerals in TV and film is "an attempt to disguise the age of films or television programmes", given that things which have zero chance of ever being repeated in full - the news, sports coverage, Question Time etc - also have it in roman numerals. Having said that I don't have any alternative explanation. MFlet1 (talk) 16:16, 3 February 2012 (UTC)

Agree it sounds unlikely to me too, but it has a citation and doesn't say it too forcefully and you haven't anything implying otherwise or even a very good story for anythig different, so I'm afraid what you say comes under WP:Original research and can't go in. Dmcq (talk) 17:38, 3 February 2012 (UTC)
Fair enough. MFlet1 (talk) XII:XXXVI, VII February MMXII (UTC)


The article states that:

"The symbols "I", "X", "C", and "M" can be repeated three times in succession, but no more. (They may appear four times if the third and fourth are separated by a smaller value, such as XXXIX.) "D", "L", and "V" can never be repeated."

If this is true, then how does one write the number 3889, or anything higher than it? You would need the ability to use M more than three times or an additional symbol. Kag427 (talk) 01:05, 4 February 2012 (UTC)

MMMDCCCLXXXIX --RSLxii 18:07, 6 February 2012 (UTC)
Well, 4000 then? — Preceding unsigned comment added by Kag427 (talkcontribs) 01:58, 7 February 2012 (UTC)
See the "Large Values" section. --RSLxii 15:58, 7 February 2012 (UTC)

Actually in the article itself, just a few lines up, it says:

1910 as MDCCCCX (Admiralty Arch in London)

C repeats 4 times here — Preceding unsigned comment added by Mdivk (talkcontribs) 01:50, 25 September 2012 (UTC)

This is rather an old thread - but just butting in anyway - "M", as the largest symbol in R.N. is an obvious exception to rules about repeating more than three times. Modern conventional Roman Numerals are not really adapted to large numbers, we generally only use them for numbers less than 4000 anyway, but since there is no Roman numeral for 5000 and 10000 (at least in our modern system) big numbers WOULD need a long string of Ms. As for the Admiralty Arch date - this is a rather obvious error - it's much too recent to pass as an example of historic usage. --Soundofmusicals (talk) 15:11, 19 February 2013 (UTC)

Logical contradiction in Roman numeral rules[edit]


After the three rules for constructing Roman numerals, this statement is offered: "Following this additional set of rules, there is only one possible Roman numeral for any given number."

However, this statement does not follow from the rules. I can write the number 18 at least two different ways: XVIII or IXIX. As far as I can see, I have not broken any of the rules. This statement should be removed or clarified. Roland Deschain (talk) 05:04, 19 February 2012 (UTC)

I see four rules, not three. You have violated the fourth rule which requires that 18 be broken into 10 + 8. — Joe Kress (talk) 02:54, 20 February 2012 (UTC)
I just thought of two ways to write 8 using the rules listed: VIII and IVIV. Does anyone know of a rule we can cite that prevents the latter from being created? --RSLxii 17:09, 24 February 2012 (UTC)

these are roman numerels... 1 I 5 V 10 X 50 L 100 C 500 D 1,000 M 1,000,000 M + (-) on top — Preceding unsigned comment added by (talk) 17:50, 16 June 2012 (UTC)

Should this not go under Wikipedia: Wikiproject Numbers?[edit]

We are told at the top of the talk page of this article that this article goes within the interests of the Wikipedia: Wikiproject group called "Wikipedia: Wikiproject Mathematics". However, should it not also go what seems a more obvious wikiproject group for this article - that called Wikipedia:WikiProject_Numbers? I may drop a note at the talk page of this wikiproject group telling them that they should take an interest in this article. Thank you in advance to any one who can help here. ACEOREVIVED (talk) 20:52, 31 January 2013 (UTC)

V and 'false etymology'?[edit]

Soundofmusicals, I understand your point about references in the edit summary. The trouble is that the second sentence is a non-sequitur. Etruscans doing tallies doesn't mean that V did not originally represent a hand, etc.

I imagine that no one knows or ever can know, which is why I was surprised to see it described as 'false'. The reference at the end of a para is to a hard copy book with no page number so it cannot readily be verified. So without proper support or a clear line of thought, I don't think this page should be baldly saying that this theory is wrong. I've moderated my edit. asnac (talk) 08:03, 19 February 2013 (UTC)

Please clean up "false etymology" language

It looks like the text has reverted to using the phrase "false etymology" - can someone please fix this? It's tacked on to the end of the paragraph in an unsupported, and very distracting, way. I agree with Asnacclasnac, its hard to believe that this could ever be definitively disproven, and its annoying. At the very least, there needs to be an explanation given of how/when this theory of etymology was disproven, (with a citation of course) and the section should be labelled as a "disproven theory" early on, not labelled at the end as "false" which isn't at all neutral or appropriate. Thanks! Kduckworth (talk) 17:42, 2 January 2014 (UTC) Kate

Simplification of Reading Roman numerals section[edit]

The Reading Roman numerals section as it stood was liable to confuse those seeking to access what is (in essence) a simple system.

I have therefore simplified it without losing any of the essential information. At the start, I've retained the list of symbols. Then the next paragraph gives the system in a nutshell, as it's a cumulative, non place value system. And the next paragraph explains what so often throws people, the IV and the CM etc.

At the same time I've tidied up the inconsistent use of "rn".

I've retained virtually all the citations. asnac (talk) 12:24, 19 February 2013 (UTC)

On the rules as you've (ably) described them the Admiralty Arch example is highly eccentric (to be kind) - although ancient and medieval use of R.N. may have been less consistent, by modern standards it's a plain mistake (by the stone mason rather than the architect?). In any case it just won't do as one of the examples - we probably need one or two new ones at this point. A bit rough going into exceptions before we've established the unexceptional convention. --Soundofmusicals (talk) 13:39, 19 February 2013 (UTC)
I've fiddled with the section a bit (excuse multiple edits, I usually try to avoid these by editing offline). On reflection I reckon the Admiralty Arch mason was an old fellow - and started to engrave MDCCC (1800) - the century he was most at home in, realised his mistake, and added an extra C rather than start again with MCM (especially as he probably would have needed a new piece of stone - assuming the stone on which the inscription was engraved wasn't structural!). It was a dreadful example, anyway. --Soundofmusicals (talk) 14:49, 19 February 2013 (UTC)
Actually the Admiralty Arch is not hugely eccentric. British inscriptions often dispensed with the CM / IV convention. See for example this coin of William IV, though his predecessor, George IV, used IV. Or Pope Paul's tombstone which shows he died on iiii Sept 1965; shows the same relaxation about the convention in modern Rome. The Admiralty Arch will not have been a mistake, the layout would have been carefully planned: anyway at that time every schoolboy knew his Latin and knew his Roman numerals. But I agree that it's not a great example when there are so many standard numbers and dates around. How would one search for a suitable uncopyrighted inscription on Wikipedia? A tombstone would be the obvious thing. asnac (talk) 14:52, 19 February 2013 (UTC)
Must admit I always believed that there WAS a fixed set of rules, and exceptions to it were rare and mostly produced by ignorance. This is, I now recognise, NOT strictly the case at all. Goes to show we are all learning. On the other hand see my earlier comments a bit further down the page. Our first obligation is to be helpful. --Soundofmusicals (talk) 23:05, 8 March 2013 (UTC)
You may have kept most of the references, but you've omitted discussion of the major point of the Michael Stroh article, namely, that there is no universally accepted set of rules when it comes to Roman Numerals. Moreover, you've added all sorts of uncited facts/opinions:
"This reflects typical modern usage..."
"A common exception..."
"A peculiar example..."
"MCMX would be more conventional."
If you can find a good source that clearly states that:
* There is a agreed-upon "right way" of how to express Roman Numerals, and
* What that set of rules are,
I would welcome the clear-cut explanation of the rules that would accompany it. I searched high and low for such a source, and finding nothing better than a offhand comment in a computer programming handbook, was forced to agree with the Michael Stroh article that there really is no one set of rules, either historically or today. --RSLxii 19:50, 8 March 2013 (UTC)
I can certainly see your point. On the other hand to avoid giving clear, coherent rules altogether would be most unhelpful in the context of a popular encyclopedia - whether because we can't find an undisputed source that sets them out, or on the grounds that Roman numerals that break the rules as we have listed them can be found inscribed all over the place (as they can, I have learned). In a reference work like this, we do need a clear unambiguous set of rules that will enable the typical (fairly unsophisticated) reader to write a clear, recognisable Roman numeral of a given number, and to read (most) Roman numbers he is likely to come across, even at the risk of being a bit simplistic. Having given these rules, it is of course important that we point out that "rules" for the construction of Roman numerals have been honoured as much in the breech as in the observance. To be fair - we already did - although I have clarified this a bit. --Soundofmusicals (talk) 22:56, 8 March 2013 (UTC)
"Because we can't find an undisputed source" is a very good reason to remove text from a wikipedia article. But I'm confident we can come to a happy consensus. We both agree that a clear set of rules, aimed at people who know nothing about how to write Roman numerals, is important to the article. These are the changes I would like to see:
  • A statement at the top saying there has never been an authoritative set of rules for use of modern numerals, citing Michael Stroh
  • A statement that the rules presented in this article come from Source X, a source that most people will agree is reasonably trustworthy, if not exactly authoritative
  • A description of the rules from Source X, without opinionated language like "This reflects typical modern usage."
The trick is to find Source X. I tried hard to find it and failed. Texts aimed at adults assumed that people knew what the "modern" rules were, while texts for elementary school students invariably omitted the more complicated parts of the ruleset. In the end, I ended up taking rules piecemeal from many sources, and I think that was a mistake, because I can't explain in the article why I picked rules I did from so many sources. It may be better to cite a single computer programming book, one that includes code for creating a Roman numerals, even if it doesn't concisely state the rules. We could make the rules clear, just as long as they agree with the algorithm included in the source. Of course, ideally, we can find a source that includes the rules explicitly stated! If you know of a source, I would be happy to let you use it as a source for your description of the rules. --RSLxii 03:15, 9 March 2013 (UTC)
I don't think the current set of rules is a "mistake". However important "verifiability" might be (and it is indeed very important) there are occasions when we have to fall back on the observed facts of the matter, at least as an interim measure. There is, in this instance, no doubt or dispute that the rules (as we now have them) are a good concise summing up of "typical modern practice" whether or not they have been lifted whole from another text or not (if they haven't, then that is actually an advantage when all is said and done!). If you can improve our referencing here - or if the (rather) chaotic state of Roman Numeral usage can be given better coverage - then by all means, throw something at the wall and let's see if the spaghetti sticks! Although I would rather we gave the rules FIRST - and then talked about exceptions, and brought up any general statements about lack of "real" rules LAST rather than leading off with the bare statement that there are really no rules; if that were true we would have no business giving any, would we? And Roman Numerals would be totally unusable (and unused for that matter) and the need for this article would not arise. --Soundofmusicals (talk) 08:45, 9 March 2013 (UTC)

JAJVII In Scottish legal documents[edit]

This is an oddity I've run across, in reading documents from the 1700s. It turns out that this is a corruption, but an interesting one. An explanation can be found below. Should this be included?

--Paul — Preceding unsigned comment added by (talk) 22:17, 25 February 2013 (UTC)

Needs a page number, as it doesn't highlight assuming it's there at all. — Arthur Rubin (talk) 03:33, 9 March 2013 (UTC)

Usage on coins[edit]

Does anyone want to address unusual Roman Numerals on coinage? I'm not qualified and it seems this would indicate a "royal mandate" much more than the usage on clocks would suggest (Which is discussed in some detail with numerous theories given.) Mensch (talk) 08:30, 16 March 2013 (UTC)

Spanish Real using "IIII" instead of IV
Really i think rather too much is made of the "IIII on clocks" business - in practice "IIII" is a not uncommon variant of "IV" - and anyway clocks with "IV" are far from unknown. Carlos IV may well have preferred the longer form of the Roman numeral as being more imposing - or perhaps this was a whim of the designer of his coinage? In any case no big deal, especially as we do make the point that RM usage (especially historical usage, before our 20th/21st desire for "scientific" exactitude was brought to bear) was always a bit inconsistent, especially when it came to "subtractive" notation. --Soundofmusicals (talk) 00:26, 5 May 2013 (UTC)


Utilisation (or utilization, if you like) is a perfectly good word, with a very specific meaning. From Wikidictionary :

  1. The act of using something.
  2. The manner in which something is used.
  3. The state of being used.

None of these meaning apply where "usage" is used in the article. Essentially, here we are rather talking about the rules governing the way the "roman numeral" symbols are combined to form numbers.

Again - from Wikidictionary:

  1. The manner or the amount of using; use
  2. Habit or accepted practice
  3. The ways and contexts in which spoken and written words are used, determined by a lexicographer's intuition or from corpus analysis.
    1. Correct or proper use of language, proclaimed by some authority.
    2. Geographic, social, or temporal restrictions on the use of words.

All of the above meanings (except for 1.) clearly apply, in context, to the use of "usage" in the article.

My apologies if this is taking a sledgehammer to crack a nut - but I just wanted to make this abundantly clear to a well-meaning new editor.

--Soundofmusicals (talk) 00:17, 5 May 2013 (UTC)

Ordinal form[edit]

Is it proper or acceptable to use Roman numerals with modern vernacular ordinal suffixes, i.e. "the XXIIIrd Olympiad"? (talk) —Preceding undated comment added 23:30, 8 May 2013 (UTC)

Not sure about "proper" or "acceptable". It certainly wouldn't be usual. The Romans had spoken ordinal numbers of course (primus, secundus, tertius etc.) but they didn't, so far as I know, add notation to their numerals to distinguish cardinal and ordinal numbers. It wouldn't have worked anyway, since in Latin ordinal numbers are adjectives that must agree with their nouns, so the the endings would have to differ if the number, case or gender of the noun were different. Even though "our" Roman Numerals certainly don't follow Roman practice in every case (read the article), people usually write "The Games of the XXIII Olympiad" and read it as "the games of the twenty third Olympiad". But what is "usual" and what might be "acceptable" are not quite always be exactly the same thing. --Soundofmusicals (talk) 05:57, 9 May 2013 (UTC)

Australian Elections[edit]

In the guide for tellers in Australian elections (where voters have to indicate an order of preference for canditates on the ballot paper) there is the following paragraph-

Acceptable forms of numbering
For voting in the House of Representatives, voters may use a consecutive series in various styles – such as: numerals (1 2 3), words (one two three), roman numerals (I II III IV), or ordinal numerals (1st 2nd 3rd). In certain cases, a mixture of numbering sequences can be used, provided that the voter’s intention is clear.

But is this a notable "use" of Roman numerals? I don't feel I can legitimately argue this one - having called the editor who brought this up "mischievous", and deleted his edit without checking the reference properly. What do others think? --Soundofmusicals (talk) 22:40, 11 September 2013 (UTC)

Though I dug up the reference, being sufficiently interested by the claim to want to check the fact, now you mention it, it's trivial and localised when compared to the other items in the list. It doesn't really tell us much about Roman numerals, more about the flexibility of the Australian vote-counting system. I think it would be reasonable to delete it. If you agree do you want me to do the deed? asnac (talk) 16:10, 12 September 2013 (UTC)
That was exactly what I thought (I wonder if anyone ever DID vote using Roman numerals - or if it's just a commonsense provision about what to do it they did?) - I was worried about being seen to be a grumpy old man who was "right" even when he was wrong, or I would have already deleted it as "non-notable". By all means "do the deed", in fact I would be grateful if you would. --Soundofmusicals (talk) 23:47, 12 September 2013 (UTC)
Done. asnac (talk) 16:15, 13 September 2013 (UTC)

The "three hypotheses"[edit]

The little numbered indented headings on these seem to have attracted some dissatisfaction. I don't think that they warrant full-flown sub-section status, or even bolding, in fact I rather like them just as they are. Is there something in the MOS I am missing? --Soundofmusicals (talk) 21:28, 22 September 2013 (UTC)


Anyone else think that in view of the large number of edits needing to be reverted recently, that this page should be semi-protected? asnac (talk) 17:57, 16 October 2013 (UTC)

IV vs IIII revisited[edit]

The section on the tendency of clock faces to use the "IIII" alternative form of "IV" (4) has recently acquired a few new tags and a new illustration. I fear that most of the "explanations" (o.k., all of them after the first one) of how this anomaly came about ignore certain facts:

  • The Romans simply didn't have "clocks" (in our sense) to have faces, in fact the "modern" clock face dates from a time when modern numerals had already started to replace Roman ones in other contexts! And yet even the very first clock face used IIII (long before the time of Louis the XIV / XIIII).
  • The use of "IIII" for four has many precedents on coins (and other monumental inscriptions). It's use on clock faces really isn't that big a deal.

These two considerations negate (or render almost humorously unlikely) most of the "explanations". So should we say something like the following

these mostly very unlikely "explanations" have been offered, in some cases perhaps humorously, to account for this.

That might take some of the heat off the push to "verify or revert" every silly statement. --Soundofmusicals (talk) 18:25, 3 January 2014 (UTC)

It's clear that no one has actually any idea about exactly why, and most of the explanations, as you suggest, don't hold water. As you say, IIII was used anyway so is to be expected, so actually I think IV vs IIII is a red herring, the question is why is it IX and not VIIII. My own assumption is that VIIII would, as the longest number on the clockface, create problems of symmetry. Not a very exciting reason but it's probably something as prosaic as that. Still, we don't know and no one knows, and I don't think we should even acknowledge these batty speculations about Jupiter, clockmakers' moulds and children who can't count to four. Just delete the lot. If they can't be deleted then the reasons they're wrong should be explicit - however that would be WP:OR. asnac (talk) 09:30, 4 January 2014 (UTC)
I agree that it's a red herring. I fear the question should not be "why was IIII used instead of IV" but "why do some people think there was a strict rule and that any deviations from it need an explanation?" I just now thought to look in Kennedy's Revised Latin Primer and was amused to see no rules given at all, just a long list of arabic numerals and the Roman numerals, and Latin cardinals, ordinals etc including - with no preference or explanation -
4: IIII or IV
8: VIII or IIX
9: VIIII or IX
14: XIIII or XIV
80: LXXX or XXC
The OCD is explicit that "both [the additive and the subtractive] methods were used, sometimes even in the same document. Inscriptions seem to show a preference for the additive method, especially in officil contexts, and the preference is occasionally carried to the extent of ignoring the signs V and L (so IIIIIIviri often for VIviri and such forms as XXXXXX for LX)." NebY (talk) 16:31, 6 January 2014 (UTC)
OK, I'm going for it. I know what I've put is arguably OR but it gets rid of a swathe of OR and patent nonsense in the process. Anyone feel free to revert if you think I haven't improved it! asnac (talk) 19:12, 6 January 2014 (UTC)
Ah, that's better! And it saves me from pursuing my own bit of OR, looking for numerals on ancient sundials. :) NebY (talk) 20:15, 6 January 2014 (UTC)
If there was no preference at first, but clocks eventually settled on IIII and IX, that raises the question as to why one should be consistently additive and the other consistently subtractive. The mold hypothesis provides a practical reason for the convention; if we're able to substantiate it or cite it to credible sources, I think it should remain. It's not a matter of IIII vs IV, but of why a clocks were persistently internally inconsistent. — kwami (talk) 20:39, 6 January 2014 (UTC)

I believe the recent changes (removing sourced explanations, replacing them with a sentence that ridicules them, and adding a different and unsourced explanation "there was no established preference for the additive or subtractive principle", constitute a violation of WP:OR and WP:NPOV. It is clear that there is no consensus in the published literature for one explanation or another. We should say so in our article rather than presenting one of the explanations as the only accepted one when it isn't. To be clear, I think this explanation makes more sense than many of the removed ones. But "what I think makes sense" should not be our guideline for what to include here. —David Eppstein (talk) 03:46, 10 January 2014 (UTC)

What is "subtractive vs additive"?[edit]

(starting a new section as the above was getting a bit long and chaotic, also indented too far)

Eschewing the use of "V" and/or "L" clearly has utterly nothing whatsoever to do with any clash between "subtractive" and "additive" notation, what ever any source might say. I am deleting the relevant addition to the article from its present position, where it adds only confusion about what "subtractive" notation actually is. It may warrant re-inclusion in a more appropriate spot, but really, without "V" or "L" (and presumably "D") we are left pretty much in a realm of pure "tally" notation. It is really stretching things to call numbers recorded in this way "Roman" numerals at all. This "tally" notation may have well been used at various times and places, but needs its own paragraph, if not its own section, if it is decided to be notable enough for inclusion in this article at all.

I also think we are getting a bit hung up on the idea of clock faces using IIII AND IX together as being "inconsistent". It may simply indicate that "IX" became the almost invariable way of writing 9 at a time when "IIII" was still a fairly common way of writing 4. Because we have decided to call both of these usages "subtractive" does not mean that they necessarily happened at the same time and became "common" at the same rate! "IX" is a more useful shortening of "VIIII" (2 characters instead of 5) than is the case with "IV" vs "IIII" (2 characters instead of 4). It is at least perfectly possible that IX was introduced earlier, or caught on quicker, than IV. We're talking about common usage, at a time (the Middle Ages) when literacy was not a common acquisition, printing wasn't invented, and "standard usage" even among the literate was really something that was going to have to wait a few centuries! The whole idea of having a clock face in Roman Numerals at all is of course imitative - we are copying past usage, and if most "Roman numeral clocks" have used IIII in the past there is quite a strong likelihood we will follow suit. The exceptions (which include Big Ben!) tend to prove that there is no "real" reason at all. Simple as that? --Soundofmusicals (talk) 00:22, 7 January 2014 (UTC)

The fact that we imitate old clocks doesn't mean there was no reason for those clocks at the time.
But yes, of course not using V, L, D has nothing to do with being subtractive. As for the result being a tally system, that's what Roman numerals are: abbreviated tallies. — kwami (talk) 01:05, 7 January 2014 (UTC)
By the principle of Occam's Razor - we shouldn't go inventing fanciful "explanations" that raise more questions than they settle, especially where the simplest explanation is perfectly feasible. I am actually fine with the text as it stands.
In a way ALL systems of writing down numbers are tallies, of course. Just some are closer to being a "straight" tally than others. The Romans used V, L, and D (see the illustration at the top of the article). Those symbols are a pretty inescapable part of what are called "Roman" numerals. In the middle Ages they did all kinds of weird things - some of which we cover in the article - but then for a long time they had degenerated from a "superior" Roman civilisation in many ways. "Subtractive" and "additive" notation is another matter altogether of course. No one has been really consistent with that until quite recently. --Soundofmusicals (talk) 01:29, 7 January 2014 (UTC)
This is the wrong way to approach writing an encyclopedia. We should not be trying to persuade each other what the "right" reason is. We should be looking for a consensus of recent published authorities or, if there is none (and I think there is none) reporting on their lack of consensus. —David Eppstein (talk) 03:47, 10 January 2014 (UTC)
If you're talking about clocks, and why they use "IIII" more often than "IV" then that's what we already do, really. No need, necessarily, to list all (or any) of the explanations, especially when they are contradictory, and none of them are, for various reasons, very likely. --Soundofmusicals (talk) 04:38, 10 January 2014 (UTC)
Well, that's not what the article says right now. It says that all but one of the explanations are unlikely and that the remaining one ("when the first clock faces appeared in the Middle Ages, there was no established preference for the additive or subtractive principle" ... from which we infer that it's a historical accident) is the correct explanation. What is the source for taking this explanation over all the others? —David Eppstein (talk) 04:52, 10 January 2014 (UTC)
It's EXACTLY what the article is trying to say - something like "hey, there are lots of ways of explaining the use of four strokes rather than a stroke and a vee to mean 4 on most (r.n.) clock faces (and regnal numbers on many coins) - but given the chaotic state of spelling (of words as well as numerals) in the Middle Ages, when clocks were invented, these explanations seem to be modern attempts to "explain" something that's not exactly a problem crying out for an explanation. It just happens to be a very old convention (albeit one that's quite often broken). This (the fact that Roman numbers aren't that consistent, and have been a lot less consistent than they are now) IS exactly what Adams points out (albeit he doesn't specifically mention clock faces - his remarks are quite general enough to apply to rn usage there, as well as anywhere else). And I wouldn't call it an "explanation" anyway - more an avoidance of an explanation.
If the text as it stands isn't clear, in your view, what would you have in its place? The list of contradictory and highly speculative explanations might be rather fun elsewhere, but add only confusion here. We include links to them, if someone is determined to be mislead.
Do we even need this section at all, now it has (rightly, I think) been gutted? Why not "be bold" anyway, and suggest an alternative that doesn't require us to "synthesise", or require a reference we are most unlikely ever to find? --Soundofmusicals (talk) 21:11, 10 January 2014 (UTC)
While I was agonising over this, someone hopped in and very sensibly moved the whole thing. Well done! --Soundofmusicals (talk) 21:23, 10 January 2014 (UTC)
(ec) I agree that we haven't located a consensus of recently published authorities. We could list every explanation we've come across but that would mean including Asimov's suggestion and I'm not convinced he's an RS on this. Indeed, we're rather short of actual authorities so far and that's not too surprising; they're in much the same position we are. There are many speculative answers to the question but if you look critically at the question itself you ask whether, as a historian or as an encylopaedist, you should even try to provide a direct explanation. It might be better to simply recount what happened and implicitly - but only implicitly - challenge the questioner's notion that it "should" have been done this way or that.
So I've tried moving right away from a structure that looked a little like the residue of a breach of WP:NOTFAQ and folded the clock story into an alternative-forms section (renamed from "exceptional" because they weren't, then). I've added a reference to the OCD (yes, I know encyclopedias are deprecated but the quality's high and the authorship and sourcing fairly clear) which might benefit from reformatting - I couldn't fit two authors of the article and two editors into a handy template. NebY (talk) 21:39, 10 January 2014 (UTC)
I think this is pretty spot on - although I have tweaked it a little. Sometimes someone just needs to "cut the knot". --Soundofmusicals (talk) 21:54, 10 January 2014 (UTC)
Thanks - not least for the tweaks! NebY (talk) 22:08, 10 January 2014 (UTC)
I have read, but can't cite, an explanation for IIII on clocks which seems quite reasonable, without any appeal to what is and isn't permissible under "rules" that may or may not have existed in common understanding: symmetry. At 8 o'clock you have VIII. The mirror position, 4 o'clock, could be either IV or IIII, but IV would leave an unbalanced appearance: two letters heavily outweighed by four. IIII creates a pleasing visual symmetry.
I'm not proposing to add this without a citation; that would be WP:OR. But if anyone can find a good source, it should go in. --Thnidu (talk) 06:35, 20 August 2016 (UTC)

Just for fun[edit]

Looking for examples I found some records produced by a slightly querulous Mr Benchley. NebY (talk) 22:00, 10 January 2014 (UTC)

Fun indeed! There is a really wonderful Wiki Article on the man - of whom I had not previously heard (great, in what are statistically surely my last days on the planet, to learn new things about matters with which I have long been familiar, and (even better) to learn of completely "new" people, places, animals etc. etc.) - it is evident that he was a humourist/satirist of great talents. His objections to the use of Roman numerals (that few people can read them properly - and that they are a totally unnecessary affectation) is pretty much spot on really. --Soundofmusicals (talk) 03:07, 11 January 2014 (UTC)

Usage of Q instead of D[edit]

In the table of the usage of Roman numerals in the Middle Ages and the Renaissance it is stated that Q was used for 500 and became reduntant with D. I have found no sources for this and have also never seen this usage. Where does this come from? — Preceding unsigned comment added by (talk) 11:39, 16 January 2014 (UTC)

I've just now found this much in the Oxford English Dictionary (1933): "Q (in mediaeval notation) = 500". No examples, alas. NebY (talk) 13:12, 16 January 2014 (UTC)
There are a number of good sources which list Q as representing 500,000 in Roman times: Adkins' Handbook to Life in Ancient Rome, as well as Encyclopaedia Americana. I'd like to add this to the article, but it appears to be locked. (talk) 06:38, 26 March 2016 (UTC)
out of the 2 sources you listed, only the first one refers to Q as 500,000. The second one clearly says Q=500 and Q with a dash over it = 500,000. the dash rule would be the same 1000 multiplier as with the other characters. Dhrm77 (talk) 13:21, 27 March 2016 (UTC)

"Continental" use of Roman numerals[edit]

I have rewritten this without the extensive "lists of countries that do this" that seem to have crept into the text. This kind of usage is patchy, often local rather than national, and seems to to go in and out of fashion. In any case the list of countries are best summarised in a context like this. The graphic of Yeltsin's dated signature is probably as good a "verification" as we are likely to get for the habit of putting the month into Roman numbers (obviously, although we don't say it here, to avoid ambiguity), one of the other points is "grass is green" obvious, and yet another now has a reference, so I have taken the liberty of removing the tag. --Soundofmusicals (talk) 23:39, 29 January 2014 (UTC)

Roman numerals/Large numbers section[edit]

If CIƆ = 1,000 and IƆ = 500 then CIƆIƆ = 1,500. This is easy to confirm by looking at the photo to the right. There are more instances like this from books and other photos that confirm my correction. What is the explanation to revert my edit. If I am missing something here I would like to know. Either way this portion of the article has no references and I believe contains additional errors. Any ideas? i0rion 18:28, 9 February 2014 (UTC) — Preceding unsigned comment added by I0rion (talkcontribs)

I agree this section needs sourcing. Kennedy's Revised Latin Primer, which isn't quite the sort of source we'd like anyway, supports it in many ways including the examples for 5,000 and 50,000 but doesn't show 1,500 etc. I agree the picture on the right does not match the text and would be glad to know more about it; the object doesn't look 17th-century at first glance. The examples in the page pictured on the right don't entirely fit either; on the other hand, the description "mysticae numerorum" doesn't fill me with confidence either. As it stands, the table is internally consistent and is consistent with the text immediately above, "An extra Ɔ denoted 500, and multiple extra Ɔs are used to denote 5,000, 50,000, etc", a quite different system of notation from the one you've deduced. If you're correct, then at the very least we must completely rewrite that text and change the examples in the table for 10,500 and 100,500. So yes, please do bring your instances from books and other photos! NebY (talk) 19:07, 9 February 2014 (UTC)

K.I.S.S. ("Reading Roman Numerals" section)[edit]

"Rules" should be kept as simple as possible - no need to detail, for instance, usages that break them, such as IC instead of XCIX. Otherwise we could easily be asked to list every "wrong" Roman number ever propounded, which would run to an article many times the length of this one! Our own age, being inherently more "scientific" and much given to efforts to standardise and avoid ambiguity - actually uses Roman numerals in an unprecedentedly "consistent" way, many of our "rules" used to be broken fairly often (see the following section!!) In any case, the purpose of setting firm rules is really just to give the "new player" a guide that will enable him/her to translate a "standard" Roman numeral into a (so-called) "arabic" number and vice versa - not to list every possible pitfall for the unwary. -Soundofmusicals (talk) 01:05, 22 February 2014 (UTC)

XC vs LXXXX[edit]

Most people use XC for ninety. There is, however, an exception: some may use LXXXX depending on the situation under them. For example, MCMXCIV and MCMLXXXXIV are the Roman numerals that represent 1994, although the earlier is used far more often. Now, how often is XC used vs LXXXX? Angela Maureen (talk) 11:58, 17 March 2014 (UTC)

As explained in the article, "XC" is the "normal" way of writing 90 in Roman numerals (by the same principle as IV for 4, IX for 9, XL for 40, and so on). Many people (like teachers, for instance) would say, in fact, that IIII for 4, VIIII for 9, XXXX for 40, and LXXXX for 90 etc. are just "wrong", because they break the rule that you don't put four "same" characters in a row. We hesitate to be quite so positive about it in an article like this - because over the centuries people have not always "spelled" Roman numbers "correctly". They used to spell ordinary words lots of funny ways too. Books, printing, and now the internet, have made us much more fussy about correct spelling. Some "incorrectly spelled" RNs are more (historically) common than others - many, many years ago IIII was quite a common variation for IV (you still see it on clocks), and "CM" (900) used to be (at least occasionally) written as "DCCCC". I have never actually seen LXXXX used for 90 - although that doesn't mean it has never been used, and of course it is obvious what number it stands for. The reason is that at a quick glance it is so easily confused with LXXX (80). "MCMLXXXXIV" seems a very strange way to write 1994. Can you see why?. --Soundofmusicals (talk) 09:17, 18 March 2014 (UTC)

Large numbers section[edit]

There is something wrong with the examples

  • I for 1000
  • VI for 6000
  • |VIII| for 800,000
  • |XX| for 2,200,000

1000 is 1 * 1000 : ok

6000 is 6 * 1000 : ok

800,000 is 8 * 1000 * 100 : Is the 100 from the pair of lines to the left and right or is it *10 from the left one and *10 from the right one? (In otherwords can you just have a leftline for *10. Frome what I casually read, I was expecting the overline and the sidelines to both be * 1000, not *1000 and *100. SO is this example right?

2,200,000 is 20 * 1000 * 100 : Same questions as above, but where did the 22 come from?

-- SGBailey (talk) 22:05, 9 May 2014 (UTC)

It was added recently (11:32, 17 April 2014) by Soundofmusicals, with the one-hundred-thousand/million ambiguity being added at 11:38. My guess is that "2,000,000" was intended. I do not know if the current wording (Adding further vertical lines ... one hundred thousand, or a million) is correct, but would be inclined to think not. If it really was that ambiguous, a source would be needed—where/when was it one and not the other? Johnuniq (talk) 23:41, 9 May 2014 (UTC)
Oh dear, I think you're right too! Like to correct it? Either way will do - 22 as XXII, or XX as 20!!. But I think I got the decimal place right... I'm afraid that they WERE a bit loose about what multiple was represented by which "side-line". There's an old story (sorry, but I don't have a source off the top of my head) that the Emperor Tiberius substituted one value for another in a will or account or something (to his own benefit. of course). --Soundofmusicals (talk) 11:05, 10 May 2014 (UTC)

The symbol to be repeated for extra-large numbers doesn't load on all computers. All I see are blank spaces where obviously something was intended. →Twentydragon 00:28, 24 February 2015 (UTC)

"j" for "i"[edit]

This keeps being added at various points in the article, in spite of already being well covered. It is not really a variant "form", like "U" the letter "J" was not part of the original Latin alphabet, (the Romans themselves always used "I" and "V" where we would use "J" and "U"). While in English and French "J" is a consonant and a genuine separate letter - in most languages it is a vowel, and more or less just a variant of "I" anyway. I have given the Roman numeral "j" its own paragraph, anyway, which will hopefully make it more conspicuous. If editors intent on improving the article can miss the fact that we mention it - then presemably people consulting the encyclopedia also do so. --Soundofmusicals (talk) 23:03, 10 July 2014 (UTC)

Usage of U instead of V[edit]

On church of my town, there is text above door in latin that has roman numerals capitalised, but uses both V and U to represent value 5. — Preceding unsigned comment added by (talk) 05:57, 4 October 2014 (UTC)

Could you take a photo and upload it? It would be a nice addition. SamuelRiv (talk) 06:24, 4 October 2014 (UTC)

"tally marks" section needs paragraph split[edit]

second paragraph, halfway through, before the line "When the tallies were transferred to writing, ...". (talk) 00:36, 2 November 2014 (UTC)

Semi-protected edit request on 6 December 2014[edit]

Hi, I'm new to Wikipedia as far as contributing goes. I'd basically like to know if it's possible to add a link to my Roman Numeral converter page to the Roman Numerals article under the "External Links" section at the bottom of the page:

My converter has been written is Java as an applet, and as such I thought it might be a useful educational tool for beginner math/computer science students. I can answer any questions on the source code too. Here is the link to my Java Roman Numeral calculator:

Please feel free to review it fully and let me know if a link to it can be added. Thanks.

Oskar03 (talk) 12:18, 6 December 2014 (UTC)

@Oskar03: X mark.svg Not done: I don't believe this website would meet our criteria for external links. Also, there appear to be many other websites offering the same service, so I don't see how your webpage is more relevant than theirs. I'm sorry. G S Palmer (talkcontribs) 15:03, 6 December 2014 (UTC)
Quite. A number of calculator-creators have added their sites to the External Links, but we have no good reason to favour those sites above others. On the contrary, given how many there are and how easily they can be found, Wikipedia is not a repository of links applies. I've removed the existing ones from the External Links section. NebY (talk) 16:04, 6 December 2014 (UTC)

Clarification on Tally Mark Hypothesis[edit]

The article states: ″The tenth V or X along the stick received an extra stroke. Thus 50 was written variously as N, И, K, Ψ, ⋔, etc., but perhaps most often as a chicken-track shape like a superimposed V and I..." However, in the tally-mark system presented (IIIIVIIIIXIIIIVIIIIX, counting twenty marks), the tenth V would correspond to 105, not 50. 50 would correspond to the fifth X. Rangdor (talk) 19:07, 16 January 2015 (UTC)

Hence the phrasing "the tenth V or X" (i.e. the tenth instance of a symbol that isn't I), which means that both V's and X's add to the count to 10. Then the tenth V or X does indeed occur at 50. Double sharp (talk) 13:22, 13 May 2015 (UTC)

Semi-protected edit request on 24 May 2015[edit]

The roman numeral for 0 is missing. Can you please add it?

JacobDanS (talk) 19:53, 24 May 2015 (UTC)

Red information icon with gradient background.svg Not done: Well, that's the point. The Romans didn't have a numeral for zero, hence instead of being to write a large number like 1986 in decimal notation, they end up with monstrosities like MCMLXXXVI. Altamel (talk) 20:11, 24 May 2015 (UTC)
The Roman numeral for "nothing" is "N" (read the article). Still it is perfectly true that given the fact the Roman numerals aren't a "place" system - there is no need for the zeroes of an (Arabic) number like "2006" - we just write 'MMVI". --Soundofmusicals (talk) 05:23, 27 May 2015 (UTC)

"Reads like a manual"???[edit]

Someone put a really weird tag on the "how RNs work section" on these lines. I have removed it - in fact I would very strongly question the whole idea of a tag like this, at the very least it needs to be applied with great discretion - articles are never be "improved" by the addition of subjective opinion, based on a very highly personal reaction, to text that has in fact been "hammered" out by a number of editors over a considerable period of time. In this case one might very well remark "what the (naughty word expunged) do you expect it to read like?" In effect - this section IS a manual on how to construct a Roman numeral. This is altogether appropriate, as this is very probably exactly what the user is looking for. Ample space is given to the history of Roman numerals, including the fact that they are very often "mis-spelled", and in fact that there is really no immutable "standard" way of writing them (one of their endearing quirks, this, one might well argue). On the other hand there is a "usual" way of writing them, which happens to be fairly logical and consistent, and one really needs to give this (in some form) before one can talk coherently about the exceptions!

But all of this is really bye-the-bye! The point is that we don't (or shouldn't) put subjective criticisms into article - except for "citation required" tags of various descriptions, and they are not, or should not, by used as implied criticism of the substance of an article, but rather as a very specific request for a citation that would improve the articles credibility, or perhaps nuance a controversial statement.

In fact this is precisely what talk pages are for. If the original "tagger" had come here with a statement that he was unhappy with the general tone of the section in question (preferably with some well thought-out suggestion for specific things that might be changed, then this would have very much more appropriate. --Soundofmusicals (talk) 02:35, 31 May 2015 (UTC)

Yes, I dislike those in-article tags. I've seen discussion about showing them on the talk page instead, which I remember as having rather petered out inconclusively as all just too awkward to change now (but I might be misremembering and there may have been Very Good Reasons). Anyway, I have tried editing the text to make it a little less instructional and a little more descriptive - see what you think. NebY (talk) 12:59, 31 May 2015 (UTC)

Large numbers, again[edit]

I fear the table in § Large numbers may have errors. It shows 15,000 as CC|ƆƆƆƆ but the illustration beside it has CC|ƆƆ|ƆƆ. Similarly, the table shows 1500 as C|ƆƆ but the illustration below has C|Ɔ|Ɔ . It seems the content was first added, in list form rather than as a table, by in April 2005.[5] It wasn't referenced and the whole section's been marked as needing references since April 2013. I'm loathe to correct the table to match the illustrations as that would be WP:OR on my part. NebY (talk) 12:15, 31 May 2015 (UTC)

From Gildersleeve's Latin Grammar, 3rd ed., 1903:
§96 2. Numeral Signs.
D is short for IↃ, M for CIↃ. Adding Ↄ on the right of IↃ multiplies by 10; IↃↃ = 5000; IↃↃↃ = 50,000. Putting C before as often as Ↄ stands after multiplies the right-hand number by 2; CIↃ = 1000; CCIↃↃ = 10,000; CCCIↃↃↃ = 100,000. A line above multiplies by 1000; V̅ = 5000. A line above and at each side multiplies by 100,000: |X̅I̅I̅I̅I̅| = 1,400,000. These signs may be combined: thus, |X̅I̅I̅I̅| X̅X̅X̅V̅I̅I̅D or |X̅I̅I̅I̅| XXXVII MD = 1,337,500. Plin.,N. H. IV., 12, 24. Other signs are… [rest of the paragraph not copied because it contains multiple (probably) undisplayable characters].
The overbars are done by letting a roman digit (e.g. X) follow by unicode character U+0305 (combining overline).
Leen (talk) 17:55, 16 June 2015 (UTC)

Reference for line above numeral meaning 1,000 times[edit]

Here's a reference I found: Lindquist, Theodore. c1920. Junior high school mathematics, Volume 1. Scribner's. (talk) 20:19, 20 July 2015 (UTC)

Modern Usage.[edit]

One of the modern usages not mentioned is the use of roman numeral "M" in the United States Engineering along with US Customary Units as a multiplied Unit Prefix. For instance, SCF = Standard Cubic Feet, MSCF = 1,000 SCF, MMSCF = 1,000,000,000 SCF — Preceding unsigned comment added by (talk) 19:01, 21 August 2015 (UTC)

Edit request[edit]

"It was later identified as the letter D; an alternative symbol for "thousand" was (I) (or CIƆor CꟾƆ)"

Space missing before the second or. (talk) 16:26, 26 August 2015 (UTC)

Done (two of them), thanks. Johnuniq (talk) 02:07, 27 August 2015 (UTC)

Can you...[edit]

Give the actual numbers they stand for? — Preceding unsigned comment added by (talk) 22:57, 14 September 2015 (UTC)

Error under larger numbers[edit]

I'm new to this, but I noticed that there is an error under the larger numbers.

You've got VIII (with a line over it) to express 800,000 and XX (with a line over it) to express 2,000,000. This is incorrect. Putting a line over the numeral is akin to multiplying that numeral by 1000. So VIII (with a line over it) would actually be 8000 and to express 2,000,000, you'd have MM (with a line over it). The XX (with a line over it) is actually 20,000. — Preceding unsigned comment added by Dicoreijers (talkcontribs) 00:57, 28 September 2015

The text at Roman numerals#Large numbers seems ok, although it is a bit hard to follow and needs proper references. I don't recall reading any sources about large numbers, but according to the text it is the combination of the over-bar and the surrounding vertical bars. A forum such as this is not regarded as a reliable source, but I'm sure that one comment there is correct, namely that there was rarely a need to express large numbers, so there was no standard procedure, just ad hoc tricks. Johnuniq (talk) 01:45, 28 September 2015 (UTC)
This was actually a right dog's dinner - I have hopefully got it at least into an understandable condition, although this has necessitated a bit of a re-write. The main obvious changes, apart from clarification tweaks and elimination of repetition in the text, are as follows:
I have deleted the templates at the top of the section - which is no londer totally "uncited" (whether it is adequately cited is another question, it may or may not need additional "cn" tags at particular points). The confusion about which system is which has been cleared up by separating them, and putting each under their own sub-headings. The two more obscure systems, which were not meaningfully described, are simply omitted. They may be added back in, but only, please, with a citation, and a (perhaps very brief) description of how they worked.

Too large roman numerals[edit]

The roman numerals in this article are overemphasized. Please make them smaller. -- (talk) 20:18, 4 November 2015 (UTC)

? -Soundofmusicals (talk) 20:52, 4 November 2015 (UTC)
In my opinion, the roman numerals should have the same height than uppercase letters, e.g. 'XXI Olympic Winter Games' instead of 'XXI Olympic Winter Games.' -- (talk) 21:01, 4 November 2015 (UTC)
this version shows the numerals, tally marks and characters from other scripts reduced to the same height as the uppercase body text. I think the simpler text of the Roman numeric system section may work reasonably well in both versions but the larger font helps a lot in the History section. NebY (talk) 22:01, 4 November 2015 (UTC)

"Specific" European countries? (Roman numerals in dates)[edit]

We started with a list of them - this was getting added to willy-nilly, as such lists will. Sure it's vague, but what would you suggest? Fact seems to be that while the use of RN in dates like this is unknown or rare in the English-speaking world it quite common "on the continent". Even then we were pressed to find a source that actually said as much - our evidence in based rather on a few examples (one is given) and "common knowledge". Not ideal, but one of those cases where being too strict means we can't actually say anything at all. -Soundofmusicals (talk) 20:51, 4 November 2015 (UTC)

A list may not be the best solution indeed. But perhaps the region where roman numerals in dates is used can be specified somehow (e.g., eastern Europe, Romance countries, etc.)? -- (talk) 21:08, 4 November 2015 (UTC)
Don't really think we have that information with sufficient precision, in fact I would be surprised if it were a fixed rule anywhere. Our example is Russian - but is it common even there? Do you know, and if so have you a reliable source? --Soundofmusicals (talk) 07:21, 5 November 2015 (UTC)

Changes to lead etc by Soundofmusicals[edit]

The top of the article had been "attacked" by vandalism and other ill-advised edits - gallant and generally well meaning attempts to resolve the problems this had caused introduced other infelicities. I have tried to restore some order, in particular a lead section that is a genuine "lead" - and an early description that avoids the pitfall of reading more like a manual than an encyclopedia article.

  • I have tried to resolve the grammatical awkwardness of the combination of singular (numeric system) and the plural (numerals). In a sentence confusing the two, singular and plural verbs BOTH feel wrong.
  • In the fourteenth century the Middle Ages were just about over - not "dawning".
  • Matter that is either relatively minor (like the relationship with Etruscan numerals) or is important, but too complicated for a one sentence summary (the no negative numbers or "zero") has been cut from this lead. Both questions are (or should be) adequately treated later.
  • A table of "subtractive numerals" duplicated a clear verbal description and is not really necessary, in spite of being very obviously well-meant. Objections have (rightly) been raised in the past about this being an encyclopedia article rather than a manual.

Hope this leaves the article in as good a condition as I left it after (rather carelessly) removing it from my watchlist! --Soundofmusicals (talk) 11:24, 13 February 2016 (UTC)

I was hoping we'd see you here again. Thanks for cutting the Gordian knot of the first sentence; I kept trying to begin it with the article title and of course WP:LEADSENTENCE isn't that foolish. Thanks also for the reminder of how late Arabic numerals replaced Roman ones and for pruning the duplications. Hope you'll keep this on your watchlist, NebY (talk) 14:08, 13 February 2016 (UTC)

Roman numerals were used in rome and many other places too. CX means 90 and M means 1,000. btw this is karolina — Preceding unsigned comment added by (talk) 07:44, 18 February 2016 (UTC)

Semi-protected edit request on 3 June 2016[edit]

I realize that when describing fractions, the two images of reversed S's are inconsistent. For the sake of neatness, I would like it altered. Kylefreed13 (talk) 06:28, 3 June 2016 (UTC)

Yes check.svg Done, thanks. I went with Ƨ rather than 𐆓 as the latter doesn't appear correctly in my browser. Adrian J. Hunter(talkcontribs) 14:24, 3 June 2016 (UTC)
Fwiw, the latter character (U+10193 "Roman Sextula Sign" from 'Ancient symbols' block) does not appear for me either (Chrome 51, Windows 10, standard fonts). By symbol name alone U+01A7 "Tone Two" refers to the symbol used in an obsolete (as of 1982) script for the Zhuang languages 1960s -- as it's in 'Latin Extended B', it will available be in pretty much every Unicode font.
What's the dilemma in standardization where one choice is properly classified but not widely supported while the other fills the same requirements and is supported by most platforms but is technically classified as something completely different, thus if a cog in the interpretation system decided to change it could screw the whole semantics up? SamuelRiv (talk) 04:06, 25 June 2016 (UTC)