Old Latin Forms in Early Inscriptions

Early Latin inscriptions preserve a changing language before the spelling and prose habits familiar from Cicero. They are not defective Classical Latin. Forms such as med, GNAIVOD, QVEI, and FOIDERATEI can preserve older endings or sound stages; other oddities reflect an epigraphic alphabet, local convention, formula, damage, or uncertain word division. The editorial problems connect directly with Capitalization and Word Division.

“Old Latin” here covers evidence from roughly the sixth through second centuries BCE, not one stable synchronic grammar. The sixth-century Duenos inscription, the Forum cippus conventionally associated with the Lapis Niger, third-century Scipionic epitaphs, and the senatorial decree on the Bacchanalia of 186 BCE are separated by centuries, genres, and material practices. Comparing them can reveal change, but combining every form into a single composition style creates an artificial language nobody wrote at one time.

Read the object before normalizing the language

An inscription is material evidence. Letter shapes, direction of writing, damage, abbreviations, and the absence of modern word division affect every linguistic claim. The Duenos text is written continuously around a vessel; the Forum cippus is fragmentary and written in a boustrophedon-like arrangement; official bronze and funerary stone follow different conventions.

Use this order:

1. keep the edition’s transcription and brackets;
2. propose word division without silently treating it as certain;
3. identify endings, stems, and formulae;
4. compare a later Classical outcome;
5. translate only to the degree the text permits.

A compact map from early evidence to Classical outcomes

The table aligns forms for reading; it does not claim a direct one-step spelling substitution in every case. Sound change, analogical reshaping, orthographic convention, and morphology interact.

The Duenos inscription: secure forms inside an uncertain whole

The Duenos inscription is commonly dated to the sixth century BCE. Its overall interpretation remains disputed, so a responsible guide should not print one smooth translation as though every word were secure. One often segmented source-faithful sequence is:

MED beside Classical mē and FECED beside Classical fēcit are comparatively transparent. DVENOS is historically related to Classical bonus and is often understood here as “a good man,” but it has also been taken as a personal name. The object speaks in the first person, a familiar inscriptional strategy.

The second line is not a replacement for the source. Its lower case, macrons, and spaces are pedagogical analysis. If DVENOS is read as the adjective “good,” a Classical prose normalization would be:

The contrast mēd : mē illustrates loss of final -d in the pronoun. Duenos : bonus belongs to historical lexical phonology, not to a Classical rule by which b can freely alternate with du.

The Forum cippus: fragments require restraint

The archaic cippus beneath the Lapis Niger pavement is often dated around the late sixth or early fifth century BCE, though the exact dating and interpretation remain debated. Much of its text is lost. A famous fragment is conventionally transcribed:

SAKROS stands beside later sacer; ESED is commonly compared with a later form of esse, often normalized in the formula as estō or interpreted through an archaic modal form. The fragment does not license certainty about the complete law.

Early alphabets did not distribute C, K, and Q exactly as standardized Republican and Classical orthography later did. Consequently K in SAKROS is inscriptional evidence, not a separate case ending or a permission to spell Classical sacer as saker.

The Scipionic epitaphs: morphology in a readable genre

The epitaph of Lucius Cornelius Scipio Barbatus, probably inscribed in the third century BCE, is much more extensive and intelligible. Its source-faithful opening includes:

GNAIVOD PATRE contains two ablatives. GNAIVOD preserves ablative -d, lost in the later Classical paradigm; the title formula’s word order and all-capital transcription belong to the inscriptional presentation.

Another source-faithful portion reads:

QVOIVS corresponds to cuius, VIRTVTEI to dative virtūtī, and PARISVMA to parissima. The single s in PARISVMA and the vowel spellings show that later spelling conventions must not be projected backward mechanically.

This epitaph is stylized praise. Its morphology is historical evidence, while its parallel virtues and career list are also features of funerary genre. Neither dimension should be mistaken for everyday conversation.

The Bacchanalian decree: change visible in official prose

The bronze copy of the senātūs cōnsultum dē Bacchānālibus dates to 186 BCE. It is much closer to Classical Latin than the Duenos text, yet it preserves systematic older spellings and official formulae:

A Classical teaching normalization is:

Compare QVEI : quī, FOIDERATEI : foederātī, ESENT : essent, EXDEICENDVM : ēdīcendum, and CENSVERE : cēnsuērunt. The final example shows that the perfect third-person plural ending -ēre remains available in Classical Latin; not every difference is an obsolete Old Latin ending.

Another source-faithful prohibition reads:

NEIQVIS corresponds to nē quis, HABVISE to habuisse, and VELET to vellet. The syntax is already recognizable: negative nē quis introduces the prohibited person, and imperfect subjunctive vellet belongs inside reported official instructions. Archaic spelling does not imply syntactic chaos.

What an early form does—and does not—prove

An inscription can securely prove that a sequence of letters was cut on a particular object, subject to editorial reading. Its linguistic interpretation can still be uncertain. A recurring form across independent inscriptions is stronger evidence for a productive feature than one damaged occurrence. Formulaic legal language may preserve older morphology after ordinary speech has changed, while an artisan’s spelling may reveal local convention rather than a language-wide norm.

English speakers sometimes approach an unfamiliar early form as a misspelling of the “real” Classical word. The direction is wrong: Classical mē, quī, and foederātī are later outcomes within a history. Conversely, it is equally wrong to call every later Romance-like feature “already Romance.” Each form must be dated and situated.

Common Mistakes

1. Silently replacing the inscription with Classical spelling.

2. Treating final -d as productive Classical morphology.

3. Giving a confident translation of a damaged or disputed text.

4. Combining centuries into one invented ‘Old Latin’ voice.

Key Takeaways

Early inscriptions preserve evidence, not failed Classicism. Quote their spelling faithfully; mark editorial segmentation and normalization; parse older endings before translating; and keep uncertain texts uncertain. Above all, treat Old Latin diachronically: a sequence from several centuries and genres can illuminate change, but it cannot be collapsed into one uniform grammar.