Expansion of the Roman Republic during the 2nd century BC. Very little Latin is likely to have been spoken beyond the green area, and other languages were spoken even within it.
The use of "old", "early" and "archaic" has been standard in publications of Old Latin writings since at least the 18th century. The definition is not arbitrary, but the terms refer to spelling conventions and word forms not generally found in works written under the Roman Empire. This article presents some of the major differences.
The concept of Old Latin (Prisca Latinitas) is as old as the concept of Classical Latin – both labels date to at least as early as the late Roman Republic. In that period Cicero, along with others, noted that the language he used every day, presumably upper-class city Latin, included lexical items and phrases that were heirlooms from a previous time, which he called verborum vetustas prisca,[4] translated as "the old age/time of language".
In the classical period, Prisca Latinitas, Prisca Latina and other idioms using the adjective always meant these remnants of a previous language, which, in Roman philology, was taken to be much older in fact than it really was. Viri prisci, "old-time men", meant the population of Latium before the founding of Rome.
The four Latins of Isidore
In the Late Latin period, when Classical Latin was behind them, Latin- and Greek-speaking grammarians were faced with multiple phases, or styles, within the language. Isidore of Seville (c. 560 – 636) reports a classification scheme that had come into existence in or before his time: "the four Latins" ("Moreover, some people have said that there are four Latin languages"; "Latinas autem linguas quattuor esse quidam dixerunt").[5] They were:
Prisca, spoken before the founding of Rome, when Janus and Saturn ruled Latium, to which period Isidore dated the Carmen Saliare
Latina, dated from the time of king Latinus, in which period he placed the laws of the Twelve Tables
This scheme persisted with little change for some thousand years after Isidore.
Old Latin
In 1874, John Wordsworth used this definition: "By Early Latin I understand Latin of the whole period of the Republic, which is separated very strikingly, both in tone and in outward form, from that of the Empire."[6]
Although the differences are striking and can be easily identified by Latin readers, they are not such as to cause a language barrier. Latin speakers of the empire had no reported trouble understanding Old Latin, except for the few texts that must date from the time of the kings, mainly songs. Thus, the laws of the Twelve Tables (5th century BC) from the early Republic were comprehensible, but the Carmen Saliare, probably written under Numa Pompilius (who according to tradition reigned from 715 to 673 BC), was not entirely clear (and remains so). On the other hand, Polybius, a Greek historian of Rome who flourished in the late second century BC,[7] commented on "the first treaty between Rome and Carthage", (which he dated to 28 years before Xerxes I crossed into Greece; that is, in 508 BC) that "the ancient Roman language differs so much from the modern that it can only be partially made out, and that after much application by the most intelligent men".
There is no sharp distinction between Old Latin, as it was spoken for most of the Republic, and Classical Latin, but the earlier grades into the latter. The end of the republic was too late a termination for compilers after Wordsworth; Charles Edwin Bennett said, "'Early Latin' is necessarily a somewhat vague term ... Bell, De locativi in prisca Latinitate vi et usu, Breslau, 1889,[8] sets the later limit at 75 BC. A definite date is really impossible, since archaic Latin does not terminate abruptly, but continues even down to imperial times."[9] Bennett's own date of 100 BC did not prevail; rather Bell's 75 BC became the standard as expressed in the four-volume Loeb Library[citation needed] and other major compendia. Over the 377 years from 452 to 75 BC, Old Latin evolved from texts partially comprehensible by classicists with study to being easily read by scholars.
Corpus
Old Latin authored works began in the 3rd century BC. These are complete or nearly complete works under their own name surviving as manuscripts copied from other manuscripts in whatever script was current at the time. There are also fragments of works quoted in other authors.
Many texts placed by various methods (painting, engraving, embossing) on their original media survive just as they were except for the ravages of time. Some of these were copied from other inscriptions. No inscription can be older than the introduction of the Greek alphabet into Italy but none survive from that early date. The imprecision of archaeological dating makes it impossible to assign a year to any one inscription, but the earliest survivals are probably from the 6th century BC. Some texts, however, that survive as fragments in the works of classical authors, had to have been composed earlier than the republic, in the time of the monarchy. These are listed below. Some authors, especially in recent texts, refer to the oldest Latin documents (7th–5th c. BCE) as Very Old Latin (VOL).[10]
Fragments and inscriptions
Notable Old Latin fragments with estimated dates include:
Old Latin surviving in inscriptions is written in various forms of the Etruscan alphabet as it evolved into the Latin alphabet. The writing conventions varied by time and place until classical conventions prevailed.
A part of old inscriptions, texts in the original writing system have been lost or transcribed by later copyists.[citation needed]
Old Latin is thought to have had a strong stress on the first syllable of a word until about 250 BC. All syllables other than the first were unstressed and were subjected to greater amounts of phonological weakening. Starting around that year, the Classical Latin stress system began to develop. It passed through at least one intermediate stage, found in Plautus, in which the stress occurred on the fourth last syllable in four-syllable words with all short syllables.
Most original PIE (Proto-Indo-European) diphthongs were preserved in stressed syllables, including /ai/ (later ae); /ei/ (later ī); /oi/ (later ū, or sometimes oe); /ou/ (from PIE /eu/ and /ou/; later ū).
The Old Latin diphthong ei evolves in stages: ei > ẹ̄ > ī. The intermediate sound ẹ̄ was simply written e but must have been distinct from the normal long vowel ē because ẹ̄ subsequently merged with ī while ē did not. It is generally thought that ẹ̄ was a higher sound than e (e.g. perhaps [eː] vs. [ɛː] during the time when both sounds existed). Even after the original vowel /ei/ had merged with ī, the old spelling ei continued to be used for a while, with the result that ei came to stand for ī and began to be used in the spelling of original occurrences of ī that did not evolve from ei (e.g. in the genitive singular -ī, which is always spelled -i in the oldest inscriptions but later on can be spelled either -i or -ei).
In unstressed syllables, *oi and *ai had already merged into ei by historic times (except for one possible occurrence of poploe for populī "people" in a late manuscript of one of the early songs). This eventually also evolved to ī.
Old Latin often had different short vowels from Classical Latin, reflecting sound changes that had not yet taken place. For example, the very early Duenos inscription has the form duenos "good", later found as duonos and still later bonus. A countervailing change wo > we occurred around 150 BC in certain contexts, and many earlier forms are found (e.g. earlier votō, voster, vorsus vs. later vetō, vester, versus).
Old Latin frequently preserves original PIE thematic case endings -os and -om (later -us and -um).
1 Intervocalic /s/ (pronounced [z]) was preserved up to 350 BC or so, at which point it changed into /r/ (rhotacism). This rhotacism had implications for declension: early classical Latin, honos, honoris; later Classical (by analogy) honor, honoris ("honor"). Some Old Latin texts preserve /s/ in this position, such as the Carmen Arvale's lases for lares. Later instances of single /s/ between vowels are mostly due either to reduction of early /ss/ after long vowels or diphthongs; borrowings; or late reconstructions.
2 Might have been an /f/ sound instead, perhaps in free variation.[15] Lloyd, Sturtevant, and Kent make this argument based on misspellings in early inscriptions, the fact that many instances of Latin /f/ descend from Proto-Indo-European */bʰ/, and the outcomes of the sound in Romance (particularly in Spain).[16]
There are many unreduced clusters, e.g. iouxmentom (later iūmentum, "beast of burden"); losna (later lūna, "moon") < *lousna < */leuksnā/; cosmis (> cōmis, "courteous"); stlocum, acc. (> locum, "place").
Early du/dw/ becomes b: duenos > duonos > bonus "good"; duis > bis "twice"; duellom > bellum "war".
Final /d/ occurred in ablatives, such as puellād "from the girl" or campōd "from the field", later puellā and campō. In verb conjugation, the third-person ending -d later became -t, e.g. Old Latin faced > Classical facit.
Morphology
Nouns
Latin nouns have grammatical case, with an ending, or suffix, showing its use in the sentence: subject, predicate, etc. A case for a given word is formed by suffixing a case ending to a part of the word common to all its cases called a stem. Stems are classified by their last letters as vowel or consonant. Vowel stems are formed by adding a suffix to a shorter and more ancient segment called a root. Consonant stems are the root (roots end in consonants). The combination of the last letter of the stem and the case ending often results in an ending also called a case ending or termination. For example, the stem puella- receives a case ending -m to form the accusative case puellam in which the termination -am is evident.[17]
In Classical Latin textbooks the declensions are named from the letter ending the stem or First, Second, etc. to Fifth. A declension may be illustrated by a paradigm, or listing of all the cases of a typical word. This method is less often applied to Old Latin, and with less validity. In contrast to Classical Latin, Old Latin reflects the evolution of the language from an ancestor spoken in Latium. The endings are multiple. Their use depends on time and place. Any paradigm selected would be subject to these constraints and if applied to the language universally would give false constructs, hypothetical words not attested in the Old Latin corpus. Nevertheless, the endings are shown below by quasi-classical paradigms. Alternate endings from different stages of development are given, but they may not be attested for the word of the paradigm. For example, in the second declension, *campoe "fields" is unattested, but poploe "peoples" is attested.
The locative was a separate case in Old Latin but gradually became reduced in function, and the locative singular form eventually merged with the genitive singular by regular sound change. In the plural, the locative was captured by the ablative case in all Italic languages before Old Latin.[18]
The stems of nouns of this declension usually end in -ā and are typically feminine.[19] A nominative case ending of -s in a few masculines indicates the nominative singular case ending may have been originally -s: paricidas for later parricida, but the -s tended to get lost.[20] In the nominative plural, -ī replaced original -s as in the genitive singular.[21]
In the genitive singular, the -s was replaced with -ī from the second declension, the resulting diphthong shortening to -ai subsequently becoming -ae.[22] The original form is maintained in some formulas, e.g. pater familiās. The genitive plural ending -āsōm (classical -ārum following rhotacism), borrowed from the pronouns, began to overtake original -om.[21]
In the dative singular the final i is either long[23] or short.[24] The ending becomes -ae, -a (Feronia) or -e (Fortune).[23]
In the accusative singular, Latin regularly shortens a vowel before final m.[24]
In the ablative singular, -d was regularly lost after a long vowel.[24] In the dative and ablative plural, the -abos descending from Indo-European *-ābhos[25] is used for feminines only (deabus). *-ais > -eis > -īs is adapted from -ois of the o-declension.[26]
The vocative singular had inherited short -a. This later merged with the nominative singular when -ā was shortened to -ă.[24]
The locative case would not apply to such a meaning as puella, so Roma, which is singular, and Syracusae, which is plural, have been substituted. The locative plural has already merged with the -eis form of the ablative.
Second declension (o)
campos, –ī field, plain m.
saxom, –ī rock, stone n.
Singular
Plural
Singular
Plural
Nominative
campos
campei < campoi
saxom
saxā, saxă
Vocative
campe
saxă
Accusative
campom
campōs
saxom
saxā, saxă
Genitive
campī
campōm
saxī
saxōm
Dative
campō
campeis < campois
saxō
saxeis < saxois
Ablative
campōd
saxōd
Locative
campei
saxei
The stems of the nouns of the o-declension end in ŏ deriving from the o-grade of Indo-European ablaut.[27] Classical Latin evidences the development ŏ > ŭ. Nouns of this declension are either masculine or neuter.
Nominative singulars ending in -ros or -ris syncopate the ending:[28] *agros > *agrs > *agers > *agerr > ager. (The form terr "three times" for later ter < *tris appears in Plautus.)
Many alternative spellings occur:
As mentioned above, the sound change -ei > -ẹ̄ > -ī leads to many variations, including the reverse spelling ei for ī. This spelling eventually appears in the genitive singular as well, though -ī is earliest and the true ending; cf. populi Romanei, "of the Roman people."[29], with both spellings in the same inscription.
Likewise, the sound changes -os > -us and -ōm > -om > -um affect the nominative and accusative singular, and the genitive plural.
One very early text, the Lapis Satricanus, has genitive -osio (an ending found in several other archaic languages descended from Proto-Indo-European [PIE], languages such as Vedic Sanskrit, Old Persian, and Homeric Greek) rather than -ī[30] (an ending appearing only in Italo-Celtic).[citation needed]. This form also appears in the closely related Faliscan language.
In the genitive plural, -um (from PIE *-ōm) survived in classical Latin "words for coins and measures";[31] otherwise it was eventually replaced by -ōrum by analogy with 1st declension -ārum.
The nominative/vocative plural masculine -ei comes from the PIE pronominal ending *-oi. The original ending -oi appears in a late spelling in the word poploe (i.e. "poploi" = populī "people") in Sextus Pompeius Festus.[32]
The dative/ablative/locative plural -eis comes from earlier -ois, a merger of PIE instrumental plural *-ōis and locative plural *-oisu. The form -ois appears in Sextus Pompeius Festus and a few early inscriptions.
The Praeneste Fibula has dative singular Numasiōi, representing PIE *-ōi.
A number of "provincial texts" have nominative plural -eis (later -īs from 190 BC on[33]), with an added s, by some sort of analogy with other declensions. Sihler (1995)[32] notes that this form appears in literature only in pronouns and suggests that inscriptional examples added to nouns may be artificial (i.e. not reflecting actual pronunciation).
In the vocative singular, some nouns lose the -e (i.e. have a zero ending) but not necessarily the same as in classical Latin.[34] The -e alternates regularly with -us.[35]
Third declension (consonant/i)
rēx, rēges king m.
ignis -is fire m.
Singular
Plural
Singular
Plural
Nominative
rēx
rēgeīs, rēgīs, rēgēs, rēgĕs
ignis, ignes
igneīs, ignēs, ignīs, ignĕs
Vocative
Accusative
rēgem
rēgeīs, rēgīs, rēgēs
ignim
igneīs, ignēs, ignīs
Genitive
rēges, rēgis, rēgos, rēgus
rēgom, rēgum, rēgerum
ignis
igniom, ignium
Dative
rēgei, rēgī, rēgē, rēgĕ
rēgebus, rēgebūs, rēgibos, rēgibus
igni, igneī, ignē
ignebus, ignebūs, ignibos, ignibus
Ablative
rēgīd, rēgĭd, rēgī, rēgē, rēgĕ
ignīd, ignĭd, ignī, ignē, ignĕ
Locative
rēgī
rēgebos
ignī
ignibos
Instrumental
rēge
igne
This declension contains nouns that are masculine, feminine, and neuter. The stem ends in the root consonant, except in the special case where it ends in -i (i-stem declension). The i-stem, which is a vowel-stem, partly fused with the consonant-stem in the pre-Latin period and went further in Old Latin.[36] I/y and u/w can be treated as either consonants or vowels; hence they are semi-vowels. Mixed-stem declensions are partly like consonant-stem and partly like i-stem. Consonant-stem declensions vary slightly depending on which consonant is root-final: stop-, r-, n-, s-, etc.[37] The paradigms below include a stop-stem (reg-) and an i-stem (igni-).
For a consonant declension, in the nominative singular, the -s was affixed directly to the stem consonant, but the combination of the two consonants produced modified nominatives over the Old Latin period. The case appears in different stages of modification in different words diachronically.[38] The Latin neuter form (not shown) is the Indo-European nominative without stem ending; for example, cor < *cord "heart".[39]
The genitive singular endings include -is < -es and -us < *-os.[40] In the genitive plural, some forms appear to affix the case ending to the genitive singular rather than the stem: regerum < *reg-is-um.[41]
In the dative singular, -ī succeeded -eī and -ē after 200 BC.
In the accusative singular, -em < *-ṃ after a consonant.[40]
In the ablative singular, the -d was lost after 200 BC.[42] In the dative and ablative plural, the early poets sometimes used -būs.[40]
In the locative singular, the earliest form is like the dative but over the period assimilated to the ablative.[43]
In the instrumental singular, the earliest form is an -e during its early days.
Fourth declension (u)
The stems of the nouns of the u-declension end in ŭ and are masculine, feminine and neuter. In addition there is a ū-stem declension, which contains only a few "isolated" words, such as sūs, "pig", and is not presented here.[44]
Personal pronouns are among the most common thing found in Old Latin inscriptions. In all three persons, the ablative singular ending is identical to the accusative singular. In Classical Latin, "ego" can be pronounced with both long and short final vowel, thus pointing at two versions of the pronoun possibly from Old Latin; the original form of the pronoun is reconstructed in Proto-Italic as *egō, with a long vowel.
ego, I
tu, you
suī, himself, herself (etc.)
Singular
Nominative
egō/ego
tu
—
Accusative
mēd
tēd
sēd
Genitive
mis
tis
sei
Dative
mihei, mehei
tibei
sibei
Ablative
mēd
tēd
sēd
Plural
Nominative
nōs
vōs
—
Accusative
sēd
Genitive
nostrōm, -ōrum, -i
vostrōm, -ōrum, -i
sei
Dative
nōbeis, nis
vōbeis
sibei
Ablative
sēd
Relative pronoun
In Old Latin, the relative pronoun is also another common concept, especially in inscriptions. The forms are quite inconsistent and leave much to be reconstructed by scholars.
queī, quaī, quod who, what
Singular
Masculine
Feminine
Neuter
Nominative
queī
quaī
quod
Accusative
quem
quam
Genitive
quoius, quoios, -a, -um/om (according to gender of whatever is owned)
Dative
quoī, queī, quoieī, queī
Ablative
quī, quōd
quād
quōd
Plural
Masculine
Feminine
Neuter
Nominative
ques, queis
quaī
qua
Accusative
quōs
quās
Genitive
quōm, quōrom
quōm, quārom
quōm, quōrom
Dative
queis, quīs
Ablative
Verbs
Old present and perfects
There is little evidence of the inflection of Old Latin verb forms and the few surviving inscriptions hold many inconsistencies between forms.[dubious – discuss] Therefore, the forms below are ones that are both proved by scholars through Old Latin inscriptions, and recreated by scholars based on other early Indo-European languages such as Greek and Italic dialects such as Oscan and Umbrian, and which also may be compared to modern Spanish.
Indicative present: sum
Indicative present: facio
Old
Classical
Old
Classical
Singular
Plural
Singular
Plural
Singular
Plural
Singular
Plural
First person
(e)som
somos, sumos
sum
sumus
fac(e/ī)ō
fac(e)imos
faciō
facimus
Second person
es
esteīs
es
estis
fac(e/ī)s
fac(e/ī)teis
facis
facitis
Third person
est
sont
est
sunt
fac(e/ī)d/-(e/i)t
fac(e/ī)ont
facit
faciunt
Indicative perfect: sum
Indicative perfect: facio
Old
Classical
Old
Classical
Singular
Plural
Singular
Plural
Singular
Plural
Singular
Plural
First person
fuei
fuemos
fuī
fuimus
(fe)fecei
(fe)fecemos
fēcī
fēcimus
Second person
fuistei
fuisteīs
fuistī
fuistis
(fe)fecistei
(fe)fecisteis
fēcistī
fēcistis
Third person
fued/fuit
fueront/-erom
fuit
fuērunt/-ēre
(fe)feced/-et
(fe)feceront/-erom
fēcit
fēcērunt/-ēre
In popular fiction
The Italian director Matteo Rovere has shot the 2019 film The First King: Birth of an Empire and the 2020–2022 TV series Romulus with dialog in a reconstructed version of Old Latin.
The linguists have had to make concessions for ease of filming and not going too much against the expectations of viewers.
For example, the character of the Lady of the Wolves is Lukwòsom Pòtnia (an allusion to Homer's Potnia theron) since Latin domina did not have the desired nuances.
Before rhotacism, Old Latin had lots of sibilants, so some had to be substituted to ease the actors' work.[45]
^Bell, Andreas (1889). De Locativi in prisca latinitate vi et usu, dissertatio inauguralis philologica. Breslau: typis Grassi, Barthi et soc (W. Friedrich).
^Buck, Carl Darling (2005) [1904]. A Grammar Of Oscan And Umbrian: With A Collection Of Inscriptions And A Glossary. Languages of classical antiquity, vol. 5. Bristol, Pa.: Evolution Publishing. p. 204.
Allen, William Sidney (1978) [1965]. Vox Latina—a Guide to the Pronunciation of Classical Latin (2nd ed.). Cambridge University Press. ISBN0-521-37936-9.
Goldberg, Sander M. 2007. "Antiquity's antiquity." In Latinitas Perennis. Vol. 1, The continuity of Latin literature. Edited by Wim Verbaal, Yanick Maes, and Jan Papy, 17–29. Brill's Studies in Intellectual History 144. Leiden, The Netherlands: Brill.
Lembke, Janet. 1973. Bronze and Iron: Old Latin Poetry From Its Beginnings to 100 B.C. Berkeley: University of California Press.
Mercado, Angelo. 2012. Italic Verse: A Study of the Poetic Remains of Old Latin, Faliscan, and Sabellic. Innsbruck: Institut für Sprachen und Literaturen der Universität Innsbruck.
Vine, Brent. 1993. Studies in Archaic Latin inscriptions. Innsbrucker Beiträge zur Sprachwissenschaft 75. Innsbruck, Austria: Institut für Sprachwissenschaft der Univ. Innsbruck.
Warmington, E. H. 1979. Remains of Old Latin. Rev. ed. 4 vols. Loeb Classical Library 294, 314, 329, 359. Cambridge, MA: Harvard Univ. Press.
Warner, R. 1980. "Word Order in Old Latin: Copulative Clauses." Orbis 29, no.1: 251–63.