When I was twelve and he started with me on indirect statement, Dr Gabb gave me three sentences. The first was the standard English: I think that the boy is tired. The second was how this was said in spoken Latin: credo quod puer fessus est. The third was how it had to be said in the written language: credo puerum fessum esse – the leading verb still in the indicative, but the event following from that shown by an accusative and infinitive (I know the boy to be tired). He explained that, while the spoken form creeps through into the high classics of the language, and is often seen before and after the period of the high classics, the accusative and infinitive construction is the only form for indirect statement accepted by examiners.

This done, he dropped me straight into Book I of Livy:

iam prīmum omnium satis cōnstat Troiā captā in cēterōs saevītum esse Troiānōs, duōbus, Aenēae Antēnōrīque, et vetustī iūre hospitiī et quia pācis reddendaeque Helenae semper auctōrēs fuerant, omne iūs bellī Achīvōs abstinuisse; Cāsibus deinde variīs Antēnōrem cum multitūdine Ēnētum, quī sēditiōne ex Paphlagoniā pulsī et sēdēs et ducem rēge Pylaemene ad Troiam āmissō quaerēbant, vēnisse in intimum maris Hadriatīcī sinum, Euganeīsque quī inter mare Alpēsque incolēbant pulsīs Ēnētōs Troiānōsque eās tenuisse terrās.

Will my readers sneer if I say that this took two days of struggle? Every time I managed to identify the accusative and infinitives and their dependent objects, I forgot how they all depended from the main verb.

But I am glad that Dr Gabb took the trouble to distinguish between the written and the spoken language. Most course books do not. Or those that mention the spoken form do so in a way that hides the obvious truth. They insist that Latin grammar reached its peak in the golden age of Cicero and Virgil, with the accusative and infinitive (AcI) reigning supreme as the preferred method of reporting speech and thought. They then dismiss the later emergence of subordinate clauses introduced by quod as a sign of decline—a grammatical softening, a Greek import, a concession to imprecision. But this narrative does not survive contact with the evidence. Dr Gabb is right. The truth is more subversive: quod-clauses were never new. They were always there, in the mouths of ordinary speakers. It is not that they replaced the AcI in Latin. Rather, they eventually broke through the literary wall that had excluded them for centuries.

I return to the standard course books. The argue that the AcI dominated Latin for reporting speech: dico te dormire (“I say that you are sleeping”). Because of its compactness and formal clarity, it became the standard in legal, philosophical, and oratorical prose. But in the late Republic and early Empire, an alternative began to appear in written texts: dico quod dormis (“I say that you are sleeping”). According to the standard view, this was a novel structure that gradually displaced the AcI under Greek influence. Yet this account falters under scrutiny.

Already in Plautus, writing in the early second century BC, we find:

Equidem scio iam filius quod amet meus istanc meretricem e proxumo Philaenium.
“I already know that my son is in love with that prostitute from next door, Philaenium.”

This is not an aberration. Plautus regularly uses quod (and quia) with finite verbs where classical prose would insist on an infinitive. His Latin is theatrical and conversational, aiming to reflect the language of the Roman street. And if quod is found here, it is unlikely to be an innovation. Far more plausibly, it reflects existing spoken usage. Another line from Plautus, Cistellaria 202:

Nunc scio ego quod te amavi et miser sum factus.
“Now I know that I loved you and have been made wretched.”

Here quod again serves to introduce a clause expressing a known fact, echoing the pattern of scire with quod that resurfaces in Petronius and Christian Latin.

Turning to the Satyricon of Petronius, a first century work saturated in colloquial speech, we see further support. The freedman Trimalchio declares:

Ego illi iam tres cardeles occidi, et dixi quia mustella comedit.

This use of quia with dixi is deeply revealing: it mimics everyday speech rather than learned style. In Petronius 45.10:

Sed subolfacio quia nobis epulum daturus est Mammaea.
“I sense that Mammaea is about to give us a feast.”

And in 71.9:

Scis enim, quod epulum dedi binos denarios.
“You remember that I gave a dinner at two denarii per head.”

These forms reflect a linguistic reality that pre-dated their literary reappearance. They do not mark the decay of Latin but the endurance of its vernacular foundation.

The AcI is not typical of natural language globally. It is a marked construction, requiring the speaker to suppress a finite verb and embed a subordinate clause without an explicit conjunction. It is cognitively denser and semantically more assertive. When someone says scio te venire, they are not merely reporting what someone said or thought. They are presenting the subordinate content as fact.

By contrast, scio quod venis allows for nuance. It can indicate speaker commitment, but it can also convey detachment or ambiguity. See, for example, Bellum Hispaniense (36.1):

Legati Carteienses renuntiaverunt quod Pompeium in potestatem haberent.

The quod-clause permits the narrator to report the envoys’ claim without endorsing its truth. In context, the city was divided; not everyone had accepted Pompey’s captivity. Had the author used an AcI, it would have implied that the statement was factually accurate. With quod, it becomes simply what was said.

Cicero, despite his preference for the AcI, also occasionally uses quod for emphasis or clarity. In Philippicae 2.91:

Optimum, quod sustulisti.
“It is a good thing that you have removed [him].”

Even here, we see the complement clause introduced by quod following a predicate adjective.

These subtleties would have mattered in everyday speech. Roman speakers needed tools for hedging, distancing, or attributing beliefs to others. And these tools were not confined to high rhetoric. They were the bread and butter of transactional speech—in gossip, testimony, complaint, and negotiation. The survival of quod and quia in the Romance languages (je dis que, dico che, digo que) confirms their deep embedding in the spoken Latin substrate.

Why then does the classical canon give pride of place to the AcI? The answer is sociolinguistic. From Cicero to Tacitus, Latin literature is the product of elite education and rhetorical training. Writers adopted and cultivated a style that mirrored the values of clarity and conciseness. The AcI, with its air of formal rigour, aligned with these goals. Constructions like dixit quod or scio quia were regarded as inelegant or even vulgar.

And so they were excluded. But they did not disappear. They remained alive in legal formulae, private letters, graffiti, inscriptions, and the unrecorded flow of conversation. Only in the later Empire, as Christian authors and non-elite voices gained ground, did these suppressed forms begin to re-enter the written record.

This is not a unique phenomenon. Standard languages often conceal the diversity of real usage. Just as Shakespeare’s English is not how London apprentices spoke in 1600, Cicero’s Latin is not how market traders or centurions spoke in 50 BC. The reappearance of quod in late Latin texts is not evidence of decay but of democratisation—a literary system finally acknowledging the spoken idiom.

St Jerome’s Latin translation of the Bible (the Vulgate) offers perhaps the clearest example of this process. Greek indirect statements often use either a conjunction like ὅτι or a participial clause: the AcI is also there, but, unlike in Latin, is not the only approved form. Jerome frequently renders all types from Greek using quod or quia.

For instance, in Romans 8:28, the Greek has:

οἴδαμεν δὲ ὅτι τοῖς ἀγαπῶσιν τὸν θεόν πάντα συνεργεῖ εἰς ἀγαθόν.
“We know that for those who love God all things work together for good.”

Jerome translates:

scimus autem quod diligentibus Deum omnia cooperantur in bonum.

Or see Acts 3:17:

οἶδα ὅτι κατὰ ἄγνοιαν ἐπράξατε
“I know that you did it on account of ignorance.”

Jerome translates:

scio quia per ignorantiam fecistis

These constructions mirror the Greek ὅτι, but also use a structure far more idiomatic to vernacular Latin than the AcI alternative (scimus omnia cooperari diligentibus Deum—scio vos per ignorantiam fecisse). His rendering reflects not only Greek syntax—which was part of the brief given him by the Pope—but also Jerome’s desire to produce a version of Scripture accessible to ordinary Latin speakers. In doing so, he confirms the vitality of quod/quia-clauses in the spoken idiom. His choice is not merely theological or literary—it is sociolinguistic.

What follows from this is that we must invert the old story. Quod/quia-clauses were not late interlopers. They were early and widespread. It was the AcI that functioned as a prestige construction, confined largely to formal prose and elite composition. The apparent “rise” of quod in the imperial and post-imperial periods was simply its re-emergence from concealment.

This re-evaluation is not merely technical. It reshapes our understanding of Latin as a living language. It encourages us to consider how social class, literary taste, and ideology affect what gets preserved as “correct”. And it challenges us to read surviving texts not just for what they say, but for what they suppress.

In a sense, quod has the last laugh. It lives on in Italian, French, Spanish, Portuguese, and Romanian—the languages people actually speak. The AcI, meanwhile, lies embalmed in school textbooks and church Latin. That, too, tells us something. When Latin stopped being a living language, it kept the grammar of its rulers. But while it lived, it spoke quod.

Selected Reading

• Cuzzolin, Pierluigi. Sull’origine della costruzione dicere quod. Firenze: La Nuova Italia, 1994.
• Adams, James N. “The accusative + infinitive and dependent quod/quia-clauses.” In Latin et langues romanes, 2005.
• Rosén, Hannah. Latine loqui: Trends and Directions in the Study of Latin Syntax. München: Fink, 1999.
• Kiparsky, Paul and Carol. “Fact.” In Progress in Linguistics, Mouton, 1970.
• Coleman, Robert. “Greek Influence on Latin Syntax.” Transactions of the Philological Society, 1975.
• Herman, József. Vulgar Latin. Penn State University Press, 2000.
• St. Jerome. Biblia Sacra Vulgata. Multiple citations from the Latin Vulgate translation.