From Latin Cases to Romance Prepositions

Classical Latin can encode a relation in a noun ending where English normally needs a preposition: discipulō can mean “to/for the student,” gladiō “with a sword,” and Rōmae “at Rome.” The Romance languages generally make such relations more explicit with prepositions and fixed word order. That outcome, however, was not a sudden replacement of “cases” by “prepositions.” Prepositions were already central to Classical Latin, case endings remained vigorous in educated Late and Medieval Latin, and different spoken Romance varieties preserved reduced case contrasts for different lengths of time.

This page therefore follows functions, not a simplistic before-and-after story. It complements the synchronic account of case endings and functions and the broader survey of Late Latin structural change.

The Classical baseline: endings and prepositions cooperate

A Latin preposition does not replace case marking: it selects a case. Ad takes the accusative, ā/ab, dē, and ex take the ablative, while in distinguishes motion toward with the accusative from location with the ablative. Bare cases meanwhile cover many relations without a preposition.

The visible ending is therefore the first diagnostic. Before supplying an English preposition, identify the form and ask which case function the governing word licenses. The case functions without English prepositions page develops that reading habit.

Why analytic marking gained ground

Several changes reinforced one another in speech. Final syllables became less perceptually distinct, vowel quantity ceased to carry the old contrasts, and some endings were already identical: dominō may be dative or ablative, while many neuter nominatives and accusatives coincide. Prepositions offered an audible cue before the noun, and increasingly stable word order helped identify participants. None of these facts by itself mechanically “caused case loss.” Speakers could maintain syncretic paradigms through syntax and context, as Latin had long done.

The documentary record is uneven. Formal prose often keeps traditional endings because writers learned them in school. In less classicizing Late Latin, inscriptions, itineraries, technical texts, and translations, prepositional competitors become more frequent or extend to new environments. A rising frequency is evidence of restructuring; a single example is not proof that an entire case had vanished.

The last two sentences are a minimal contrast. English uses “with” for both. Latin distinguishes means from accompaniment more sharply, although later usage can broaden prepositional marking.

Expansion does not mean invention

Many constructions that later became important were already possible in Classical Latin. Dē could express source, separation, topic, material, or a part taken from a group; ad could mark a goal, purpose, approximation, or person addressed; cum could mark accompaniment and manner. Diachronic change often consists in an old construction becoming more frequent, less semantically restricted, and eventually obligatory after inherited endings merge.

The Vulgate phrase below is source-faithful and therefore has no pedagogical macrons:

Ūnus dē duodecim uses dē for a member selected from a group. Such partitive dē is not “bad Latin” and was not invented by Romance speakers; Classical Latin already supplies precedents. The historical issue is its distribution beside the partitive genitive, not its mere existence.

Both constructions can be Latin. In later texts the prepositional pattern may align more visibly with Romance forms such as Italian uno degli studenti or French un des étudiants, but a reader must date the Latin evidence before calling it Proto-Romance.

Recipient marking: a particularly tempting shortcut

Late and non-classicizing texts may use ad with a human goal in contexts where a Classical author might prefer a dative, especially around speech, sending, and movement. Yet dīcere ad aliquem, “speak to someone,” itself has Classical credentials because the addressee is conceptualized as the direction of speech. It does not prove that the writer can no longer form a dative.

To diagnose a genuine change, examine a corpus: Does ad + accusative occur with verbs that formerly selected a dative? Is the dative still used elsewhere by the same writer? Does the construction cluster by region, genre, or translation source? A Romance-looking phrase can be an extension of native Latin ad, a translation choice, or a stable late pattern; those explanations are not interchangeable.

From reduced contrasts to Romance systems

As unstressed final vowels and consonants changed, spoken paradigms increasingly merged. The accusative became the major source of many Western Romance noun forms, but “all Romance nouns come from the accusative” is only a useful first approximation. Romanian retains a merged nominative–accusative versus genitive–dative contrast in parts of its noun phrase; Old French and Old Occitan preserved a two-case opposition for centuries; pronouns throughout Romance preserve case-like distinctions more strongly than ordinary nouns.

The endpoint is consequently plural: French à, Spanish a, and Italian a continue Latin ad in many functions; de/di descendants extend possessive and partitive relations; other relations rely on new compound prepositions or word order. Written Latin, meanwhile, continued to use six-case morphology throughout the Middle Ages. Spoken Romance development and learned Latin composition overlap chronologically but are not the same system.

A parsing procedure for late evidence

When a construction looks Romance-like, use four checks:

1. Parse the actual case. After dē, for example, the noun is still morphologically ablative.
2. Check whether Classical Latin already allowed the construction. Innovation may lie in frequency or range.
3. Compare alternatives in the same text. Coexistence is normal during change.
4. Locate the witness. A fourth-century biblical translation, a sixth-century legal document, and a twelfth-century school text cannot be treated as one dialect.

This method prevents premature translation. Dē frātribus first means “dē governing an ablative plural”; only context decides “from the brothers,” “about the brothers,” or “one/some of the brothers.”

Common Mistakes

1. Replacing every Classical dative with ad because English uses “to.”

2. Assuming de plus ablative always equals a possessive genitive.

3. Calling any prepositional alternative corrupt Latin.

4. Projecting the Romance endpoint into every Late Latin text.

Key Takeaways

Romance prepositional systems grew from resources Latin already possessed. Phonological mergers weakened some ending contrasts, while ad, dē, cum, in, and other markers broadened or became obligatory in particular functions. Read each Latin form synchronically before placing it on a historical path: parse its governed case, compare same-text alternatives, and keep Classical, Late Latin, early Romance, and learned Medieval Latin distinct.