Hacker News new | past | comments | ask | show | jobs | submit login

Just remembered that Roman numerals were and still are common in legal documents and I have a vague memory that they were the standard for dates in legal documents at one time in the US and a copyright date is a legal disclaimer. Quick search brings up this

https://www.unrv.com/roman-numerals/use-roman-numerals-legal...

Perhaps we have a lawyer on the site who can offer more, search engines really are nearly useless these days.




I wonder if it made it harder to change dates back when contracts were written by hand. Similar to the now obsolete but preserved practice of writing a number in a contract both in digits and in words.

Nowadays the only contract people write by hand is a check, and it preserves the “write the number twice” practice.


> Similar to the now obsolete but preserved practice of writing a number in a contract both in digits and in words.

I have always seen that as some check sum? I.e. both need to be wrong and the same wrong to mess things up.


Yes, it's mostly to prevent someone from doodling and changing a 3 to an 8 somewhere in the amount.


Roman numerals are pretty easy to change, even compared to Arabic.


It’s interesting that Japan and China have separate more complex sets of numbers specifically to prevent this. https://en.wikipedia.org/wiki/Japanese_numerals#Formal_numbe...


TIL! Thanks. Fun fact, I just looked at my Aikido 2nd Dan certificate, and it uses a mix of formal and in-formal numerals. Informal for the certificate number and the date, and the formal character for '2' for my rank number.


Like many security policies, it merely has to be harder, not impossible.

There is an interesting decoding aspect where a "M" can never follow a "V" in a roman numeral representation. So its technically very easy to turn a written capital V into a written capital M but its not actually useful in practice.

Also there are word/letter spacing issues where technically one could remove "III" from the end of a numeral but it would probably result in typographically or graphic arts weird looking spacing.


What are some examples?


There's a German idiom "turning an X into a U" (U being a V) with the origin about changing the number on a chalkboard recording your purchases in a pub. In a hand-written contract you can do the opposite, adding two lines to turn a V into an X.


More generally, a fun fact many miss about Roman numerals is that you can do this with all digits (except I because that would require fractions):

X (10) -> V (5)

C (100) -> L (50)

M (1000) -> D (500)

It's less obvious in prose like this but it helps if you consider that these numbers would normally be drawn with straight lines (e.g. M would often end up looking more like |X| instead of |V|, which explains why cutting it in half would result in |> or D). This is also a handy way of remembering the less frequently used digits like L or D which are harder to remember because they don't really stand for anything (unlike C = centum and M = mille).


If something ends with “I“ you can add two more I’s and have it go from 1 to 3.

And so on.

Using that appending technique works to sneak the year forward in most cases - perhaps that was the intended benefit? We can update the copyright from (say) 1950 to 1951 without any obvious tampering.


This is why you draw a line before, and after amounts on cheques.


V -> VI -> VII

I imagine that the ones with preceeding values would be harder though, perhaps XIX -> XX for example


In handwritten Arabic numerals, you could maybe get away with:

3 -> 8

(non-German) 7 -> 9

(German) 1 -> 4

5 -> 6


Presumably changing I to V or removing prefix subtraction characters (e.g., XIX becomes XX).


I think it is the reason given by the second response in OP link which I find dubious in the sense of film aging but makes sense for written documents.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: