home edit page issue tracker

This page pertains to UD version 2.

NumForm: form of a numeral

Values: Digit Reference Roman Word

This feature is assigned to numeric expressions with an exact value and describes how these are formally represented in the text. It usually applies to numerals labeled as cardinals.

Numbers expressed in a symbolic form (digits, Roman numerals…) act as “placeholders” and do not show their morphology, if any; as such, they are not marked for Case, Number, Gender, and similar.

In the case of non-literal numeric expressions, the lemma corresponds to the form: no “generic lemma” is used.

Digit: numeral expressed by digits

Digits (a form of positional numerical notation) appear either in very late Latin texts, after their widespread adoption in Europe, or in later edited versions.


Reference: complex reference key

A reference e.g. to a passage in a text or similar, written by means of numbers (usually digits or Roman numerals), punctuation marks and/or symbols. A reference is considered a complex token which may be marked as a “compound”. In future, such expressions might be split and treated as multi-word tokens.


Roman: numeral expressed by a Roman numeral

Additive numerical notation using letter-like symbols. Often used as cardinals, but also with other functions. Very frequent in Classical texts. Every number is treated as a single (possibly complex) token.


Word: numeral written in letters

A numeral written as the corresponding word; morphology possibly applies to the token. Complex numerical expressions are usually tied together by the flat deprel.


NumForm in other languages: [cs] [es] [hy] [hyw] [ka] [la] [lt] [mdf] [orv] [sl] [u]