How The Roman Number System Actually Works
Most people learn Roman numerals once in elementary school and never think about them again. I spent about six months years ago trying to correctly read and transcribe a bunch of poorly carved milestone stones from the province of Dalmatia, and it turned out my understanding of Latin Words For Numbers was pretty surface level. The problem wasn't memorizing I through M. The problem was figuring out what to do when the inscriptions themselves didn't follow the rules you were taught. Here is the basic framework. Romans didn't use a positional system. There is no zero. There are no place values in the way we understand them. Each symbol carries its own value and you add or subtract based on placement. It is clunky but it gets the job done for everyday counting.
Latin Words For Numbers In Practice
The core symbols you need are I for one, V for five, X for ten, L for fifty, C for one hundred, D for five hundred, and M for one thousand. These are the building blocks. Everything else is built from these through addition and subtraction. Numbers one through ten work more or less linearly: I, II, III, IV, V, VI, VII, VIII, IX, X. Four uses subtraction. You put I before V to mean one less than five. Same logic applies to nine with I before X. This subtractive principle is what trips people up most of the time. Once you pass ten the pattern holds. Eleven is XI, twelve is XII, fourteen is XIV. Twenty is XX. Thirty is XXX. Forty is XL because you subtract ten from fifty. Fifty is L. Sixty is LX. One hundred is C. Two hundred is CC. Three hundred is CCC. Four hundred is CD because you subtract one hundred from five hundred. Five hundred is D. Six hundred is DC. Seven hundred is DCC. Eight hundred is DCCC. Nine hundred is CM because you subtract one hundred from one thousand. One thousand is M. Two thousand is MM. Three thousand is MMM. For larger numbers the system gets awkward quickly. Four thousand would be MMMM. There was actually a separate convention for numbers above three thousand where you put a bar over a symbol to multiply it by one thousand. So a barred V meant five thousand. A barred X meant ten thousand. This isn't standard classical Latin though. It shows up more in medieval and Renaissance manuscripts where scholars needed to write larger numbers on paper. If you are reading actual Roman inscriptions you will almost never see barred numerals. They are a later invention.
Here is the thing I learned the hard way from that inscription work. Roman numerals were not always written in strict subtractive form. IV is standard for four but you will see IIII carved on clock faces and ancient monuments all the time. IX is standard for nine but IC also appears, especially in military and astronomical contexts. CM for nine hundred is standard but IM shows up occasionally. The subtractive rule I learned in school is a later standardization. The ancient Romans were more flexible than textbooks suggest. Another edge case that cost me a few hours of confusion involves the number sixty. You would think it is LX. It is. But in some medieval manuscripts you see 60 written as LX and then also as XXXIII in abbreviated forms that look completely wrong until you realize the scribe was using a shorthand I wasn't familiar with. Roman numeral abbreviations became a mess during the Middle Ages. Different monasteries developed their own conventions. If you are working with a specific manuscript the standard rules might not apply. The subtraction rules have limits that nobody mentions. You cannot subtract a symbol from another symbol more than one place value away. That means you can write IV for four or IX for nine or XL for forty or XC for ninety or CD for four hundred or CM for nine hundred. But you cannot write IL for forty-nine or ID for four hundred ninety-nine. Forty-nine has to be written as XLIX. The subtractive piece has to be the next lower order. One place value only. This restriction exists because the system was never designed for heavy arithmetic. It was designed for labeling and record keeping. When you need to do math in Roman numerals you convert to Arabic numerals, do the calculation, and convert back. Nobody in ancient Rome was balancing ledgers in Roman numerals. They used an abacus or finger counting for that.
If you need to generate Roman numerals programmatically there are straightforward conversion algorithms available online. My own workflow for that milestone project was to write a small Python script that converted Arabic numerals to Roman and then cross reference the output against the carved stones. The script caught a handful of cases where my eyes had glossed over ambiguous carvings. One stone had what I initially read as MDCLXVI for 1666 but the carving was worn enough that it could have been MDCCLXVI for 1766. The context — the emperor listed on the stone — made the correct reading clear. That kind of ambiguity is exactly why blind reliance on conversion tables can fail you with actual historical sources. The system also has practical limits. There is no symbol for zero so you cannot represent fractions or negative numbers. Writing something like 1994 requires MMMMMM if you wanted to be strictly additive and reach near infinity before running out of room, which is absurd. That is why the barred notation existed at all, even if it was not classical. The largest purely additive number you would realistically carve is around three thousand something. Beyond that you start needing alternative notation or just switching to a word form. The Romans themselves often just wrote out the number in words on official documents rather than bothering with numerals for large quantities. For anyone doing serious work with Latin inscriptions the best resource I found was the Inscripitoncs Latinae Seleeta compendium volumes and the Epigraphic Database Heidelberg online. These give you the standard readings and the variants. When you see a form that doesn't match your textbook, it is usually either a regional variation, a scribe's abbreviation, or a later medieval insert. Checking the source catalogue usually sorts it out within a few minutes instead of guessing. Roman numerals look simple until you actually have to read them at scale and figure out why five different stones from the same region spell the same number five different ways.
Get the Full Details
