Subordinate Clauses in Indirect Discourse
Once Latin tips into indirect discourse, the subjunctive bug spreads. Every subordinate clause sitting INSIDE an ōrātiō oblīqua — relative, cum-clause, causal quod, conditional sī — flips to the subjunctive, even when it would have been a plain indicative in direct speech.
Grammarians call this the attraction of mood.
Caesar stacks them: Quī... fuissent, quod intellegerent quantam calamitātem intulissent, in Britanniam profūgisse* — one infinitive drags three subjunctives along. The trap is the exception: when the speaker breaks frame to vouch for their OWN fact (quae vidēmus* — "things WE see"), the indicative survives.
Telling speaker-aside from reported thought is the AP-level judgment call.
Subordinate clauses inside indirect discourse switch indicative ⟶ subjunctive — "attraction of mood" — translate them back to plain indicative in English.
Exception: a clause the speaker treats as their OWN factual aside (not part of the reported thought) keeps the indicative.
See It In Action
— B. G. ii. 14. 4
One infinitive (profūgisse) drags three subordinate verbs into the subjunctive — the framing dīxit sits up the page. Translate every subjunctive here as plain indicative.
— B. G. i. 36. 1
Four nested subjunctives, all secondary sequence behind respondit. Notice the symmetry: quī vīcissent / quōs vīcissent — both relatives surrender their indicative because the whole thought is reported.
— B. G. i. 14. 1
Even the deepest nest — a relative inside a causal inside the OO — flips to subjunctive. Commemorāssent is two layers down and still attracted.
— Cic. Cat. iii. 21
Here is the exception. Vidēmus stays indicative because Cicero is making his OWN factual aside — the seeing belongs to the speaker ("WE"), not to the reported thought. Speaker-aside survives; reported-thought attracts.
Both sit inside an ōrātiō oblīqua paragraph. The mood tells you whose voice you're hearing.
part of the reported thought — flip back to indicative in English
quod intellegerent
(he said) because they realized…
the SPEAKER's own factual gloss — translate as a plain narrator comment
quae vidēmus
things we (the speaker) see
Tip: Ask: whose perception is this? If the seeing/knowing belongs to the reported subject, expect subjunctive. If it belongs to the author breaking frame to vouch for a fact, indicative survives.
In Caesar's Quī eius cōnsiliī principēs fuissent, quod intellegerent quantam calamitātem cīvitātī intulissent, in Britanniam profūgisse*, why are fuissent, intellegerent, and intulissent* all in the subjunctive?
Study Tips
- •Default rule first: any verb inside a reported clause goes subjunctive. Don't try to think about it — translate the indicative back from the subjunctive automatically.
- •Tense follows the sequence-of-tenses rule from the governing main verb: primary head verb → present/perfect subj.; secondary head verb → imperfect/pluperfect subj.
- •Spot the indicative-aside by asking "is this the speaker's voice or the reported person's thought?" The author's own factual gloss ("things WE see," "the city THAT EVERYONE KNOWS") stays indicative.
- •Sentences with stacked quī... quod... cum... subjunctives almost always belong to a paragraph of ōrātiō oblīqua — read backward to find the framing dīxit / respondit / negāvit.