Volō, Nōlō, Mālō, Numbers,
Nouns of Limited Form and Variable Meaning
Learning numbers is important:
Caesar XL barbaros videt does not mean “Caesar sees the very large
The normal Greek equivalent
to unus, una, unum is the rather different εἷς,
(heis, mia, hen), but in the dialect spoken in
Ionia (Western Turkey) the term for the face of a die with a single
dot is οἴνη
(oine = una) or οἶνος (oinos = unus).
Laboring under such handicaps
as the lack of a specific term for zero or a million, the Latin numbering
system will perhaps strike the modern reader as rather cumbersome. It
may be worth observing, therefore, that few surviving Latin texts require
a detailed knowledge of Roman mathematics. (Vitruvius’ De Architectura
is a salient exception.) Fractions are rarely found outside specialist
texts, and will not be discussed in this course. Slaves who were specially
trained to calculate compound interest deserve particular admiration.
The reckoning of the year in
relation to the birth of Jesus Christ (i.e. Anno Domini “In
the Year of our Lord”) was first introduced in AD 525 by a monk named
Dionysius Exiguus (“Tiny Denis”); in devising this system, he was
the first person in the Roman world known to use the number zero, which
was introduced via Arab scholarship from India. The symbol 0 for zero
is commonly supposed to be derived from ὀ (omicron), the first letter
of the Greek word for “nothing” (οὐδέν [ouden]). The French
astronomer and priest Denis Pétau devised the use of BC dating in his
De doctrina temporum of 1627. It is a pity that he did not use the
other Denis’ zero; interposing a year between BC and AD would have
avoided the problem of calculating the actual date of the millennium.
Latin has no word for “million”,
which is in fact derived from mille + the Romance augmentative
suffix -one, i.e. “a large thousand”. Latin uses the periphrasis
decies centena milia, i.e. “ten times one hundred thousand”.
“One and one, two; two and
two, four [unum et unum duo, duo et duo quattuor] was a loathsome
chant to me” (St. Augustine, Confessions 1.13.22, in one of
his many outbursts of disgust at his schooldays). [The symbols + and
-, abbreviations respectively for et and perhaps m(inus),
were introduced in the late 15th or early 16th century, whereas the
symbol =, with its parallel lines, was devised in 1557 by Robert Recorde,
a Welsh mathematician, “bicause noe 2 thynges can be moare equalle”.]
The genitive and dative singular
forms of unus, -a, -um, in -ius and -i,
may initially seem rather surprising, but they occur also in many pronouns
and some other adjectives (see Chapters 13, 17, and 18).
Despite the seeming illogicality,
unus is occasionally found in the plural, declining like the plural
of carus. It is used in the plural to refer to something as a
set (e.g. unae manus “one pair of hands”) and with nouns
which have a plural form but singular sense (e.g. una castra
“one camp”). In English, “ones” can be used as a noun or pronoun:
e.g. “Bay horses are more handsome than gray ones”. Spanish and
Portuguese use the plural of uno to mean “several”; cf. also
Italian alcuni and French quelques-uns.
Latin once had three grammatical
numbers, the singular, the dual, which referred to two people or things,
and the plural, which referred to three or more people or things. There
were dual inflections for all parts of speech which retained singular
and plural inflections; “the two little pigs were happy” will have
had dual endings for “little”, “pigs”, “were” and “happy”,
and they will all have been distinct from those used for the plural.
“Two” might have been omitted altogether, since the dual forms of
the other words will have given a clear indication of the number of
pigs. The ending -o in duo and ambo “both”
(which declines like duo) is almost the only vestige in Latin
of this IndoEuropean dual number. (The neuter plural nominative, vocative
and accusative forms of nouns, pronouns and adjectives almost universally
end in -a. Since duo and ambo are really duals,
they are not true exceptions to this rule.)
The abandonment of the dual
can be seen in other languages: in Old Irish, it is only found in company
with the word dau (“two”) to mark it, and, although it survived
quite strongly in classical Greek, it did so mostly in fossilized expressions.
Even in Sanskrit, morphologically the richest IndoEuropean language,
dual forms are not well marked, the same forms being used for several
cases. It is found now only in Icelandic, the Slavic languages Slovenian
and Wendish (Sorbian), and some dialects of Lithuanian. Arabic, and
Hebrew to a lesser extent, use the dual. The English word “both”
and its German equivalent “beide” come from the same root as
The Romans were convinced that
Latin was more closely akin to Greek than it actually is. They were
therefore puzzled that Latin showed no trace of the dual, even though
remnants of it were so prominent in Greek. Quintilian felt the need
to argue against the quite implausible view that such third person plural
perfect indicative active verb-forms as scripsere and legere
(for scripserunt and legerunt; see the Fetutinae
on Alternative Word-Forms) are actually dual forms (Education of
the Orator 1.4.42).
It may seem strange to us that
there should be a dual as well as a singular and a plural, since we
think of two as so obviously a plural. To a Greek mathematician, however,
it might seem otherwise, if one thinks of the dual as “having properties
of both singular and plural simultaneously” (Ps.-Iamblichus The
Theology of Arithmetic 10).
secundus means second
of many (e.g. “February is the second month of the year”), whereas
alter means second of only two (e.g. “If day is first, night is
second”). alter has a slightly irregular declension; see Chapter
The stem of viginti,
triginta etc. indicates a now lost word for “ten”, just as,
more obviously, the stem of ducenti, trecenti etc. is
centum. St. Isidore (Etymologies
3.3) suggests that “twenty (viginti) is ten twice born, with
the letter v used instead of b” (viginti dicti quod
sint decem bis geniti, v pro b littera posita).
To refer to 18 and 19 as
duodeviginti “two from twenty” and undeviginti “one
from twenty” respectively may reflect an Etruscan counting-pattern.
Compound numbers from twenty-one
to ninety-nine may be written not only as viginti unus, nonaginta
novem etc., but also as unus et viginti, novem et nonaginta
etc.; compare the old-fashioned “one and twenty”, “nine and ninety”
and the German einundzwanzig, neunundneunzig. It has been
calculated that Livy puts the single digit before the multiple of ten
283 times (always with et vel sim.; e.g. septem et viginti),
after it 147 times (usually without et vel sim.; e.g. viginti
septem). It is perhaps slightly preferable to give the lower number
first in a sentence such as unum et viginti porcos video “I
see twenty-one pigs”, thus distracting emphasis from the only partial
agreement of unum with porcos. There was, however, a certain
fluidity in the formulation of compound numbers: viginti et unum
porcum video, odd as it may seem, would also be permissible.
It is not possible to use
centum, an indeclinable adjective, as a noun to say “I see many
hundreds of pigs”. Depending on the context, however, both centum
and mille can be used as adjectives to denote an indefinite large
number; hence centum porcos video (with no word for “many”)
can mean “I see hundreds of pigs”.
Specific high numbers, other
than exact hundreds and thousands, were not generally used as frequently
as we might expect. (It seems remarkable that Livy, in his surviving
corpus of just over half a million words, should have a mere 430 instances
of compound numbers written out in words; see above.) It is perfectly
possible to say “The wolves eat four hundred and fifty seven thousand
eight hundred and ninety three pigs”, but a Roman would more probably
write multos/innumerabiles porcos devorant lupi.
“I give food to 2,200 pigs”
is duobus milibus [sc. porcorum] et ducentis porcis
cibum do, a somewhat unsatisfactory construction, since the substantival
milibus requires the partitive genitive porcorum, whereas
the adjective ducentis defines the dative noun porcis.
For further discussion of the partitive use of the genitive, see Chapter
“Cardinal” is derived from
cardo, cardinis masc. 3 “hinge”, i.e. the principal numbers.
“Ordinal” is derived, less interestingly, from ordo, ordinis
masc. 3 “order”, “sequence”.
Most indeclinable adjectives
are cardinal numbers. The opposing pair frugi “decent” and
nequam “worthless” are the only other common adjectives which
do not decline. (nequam does, however, have fully declining comparative
and superlative forms, nequior and nequissimus; see Chapter
Since most cardinal numbers
do not decline, it is therefore difficult to indicate their precise
function in the syntax of a sentence: decem et octo sex triplo anteit
would be indecipherable, and hence “Eighteen is three times six”
is expressed as decem et octo numerus sex numerum triplo anteit
“The number eighteen is triply in advance of the number six”.
Most distributive numbers higher
than deni are rarely found.
Adverbs denoting “x times”
are rarely found higher than decies.
When a French audience calls
for an “encore” (Fr. “again”), they sometimes shout bis,
with an irregularly old-fashioned sounding of the final s (contrast
brébis, dis, fis, lis, pis, ris,
Both the additive and the subtractive
system were in use for letters signifying numbers; hence both IIII and
IV meant 4.
88 is particularly cumbersome,
requiring eight letters: LXXXVIII.
There is very little evidence
for M used as an abbreviation for mille or milia
before the 15th century.
nolle is constructed
from ne + volo. For the negative prefix ne-, with
a short e, a form which never stands as an independent word,
cf. also e.g. nemo “no one” (ne + homo; see
Chapter 13), neque “neither”, “nor”, nescio “I
do not know” (ne + scio “know”; see Chapter 21),
neuter “neither” (ne + uter; see Chapter 13),
nullus “no”, numquam “never”. (non “not”
is thought to have originated as ne + unum [see the
Fetutinae on Alternative Word-Forms] = “not one thing”.) It
is not to be confused with the interrogative suffix -ne, for
which see Chapter 4, nor with the negative adverb/conjunction nē,
for which see Chapter 22 etc.
Latin has no parallel to the
modern English use of “swine” (cognate with “sow”
and Germ. Schwein) which refers literally to pigs only as a zero
plural (compare “one sheep, two sheep”, but the same form is used
as a singular as a term of abuse.
masc. 2 “children” is related to liber, libera,
liberum “free”, but it is not known precisely how. Donatus’
explanation, liberalitate regendi sunt, propter quod liberi dicuntur
“Children are to be controlled in a liberal manner, and hence they
are called liberi” (in his commentary on Terence’s Adelphi
57) looks distinctly half-hearted.
The plural of gratia
is used most notably in the idiom gratias agere + dat. “to
give (literally “drive”) thanks to someone”.
Unlike English, Latin always
requires a plural verb when the subject is plural, even if the sense
is singular; note the distinction between aedes magnae
sunt (meaning “the house is big”) and “politics
is a strange affair”.
When a plural form is used
to refer to a single object (e.g. castra, castrorum neut.
2 means “camp”), it is not always possible to distinguish whether
a singular or a plural sense is intended. Hence, castra Romana magna
sunt means both “The Roman camp is big” and “Roman camps are
big”; indeed, since castrum is also found in the singular,
with a different sense, the intended meaning might be “The Roman forts
vis is an irregular
noun. In the singular, it declines thus: vis, vis,
vi, vim, vi (but the genitive and dative are very
rare); the plural is regular, but is formed on a different stem from
the singular: vires, virium, viribus, vires,
However improbable Nigidius
Figulus’ discussion of the contrasting origins of first and second
person pronouns may seem, it would seem that he is drawing on similar
reasoning propounded by the Stoic philosopher Chrysippus in his discussion
of the origins of the corresponding Greek pronouns; see Galen, On
the Doctrines of Hippocrates and Plato 2.2.8ff. Since ancient etymologies
often give insights into the way people thought about their world, they
should always be given a degree of respect, even if their conclusions
about particular words can be rather odd.
An example of poetic play with
etymologising: Horace describes the Ides of April, the birthday of his
patron, Maecenas, as Idus .../qui dies mensem Veneris marinae/findit
Aprilem “the day which splits April, the month of marine Venus”.
It is a compliment to Maecenas, who was descended from the Etruscan
kings, that Horace should allude in the verb findit to the supposed
derivation of Idus from an Etruscan word iduare, meaning
“divide”, and, in describing Venus as marina, he is also
hinting at the supposed derivation of Aprilis from Aphrodite,
the goddess having been born from the foam (in Greek ἀφρός [aphros]) of the sea.
Roma may actually be
an Etruscan tribal name (i.e. non-IndoEuropean), but it was a useful
coincidence for Roman propaganda that the Greek form of “Rome” (῾Ρώμη)
was identical with a word meaning “physical power”. This point,
noticed by the Greeks themselves already in the 2nd century BC, is made
several times in Roman literature; for example, the Augustan elegist
Tibullus says “Rome, your name is ordained by destiny for world-rule”
The etymology of nox from nocere is possibly encouraged by the Greek euphemism for “night” εὐφρόνη (euphrone “the kindly minded [time]”), or indeed by the standard word for day, ἠμέρα (hemera), which could be construed as the feminine form of the adjective ἤμερος (hemeros “gentle”).