Tense and Person Shifts in Reported Speech

Reported speech changes the center from which person and time are interpreted. A speaker’s “I,” “you,” “now,” and “tomorrow” must be understood from inside the original speech even though the narrator presents them in a new sentence. Classical Latin manages this recentering through accusative-and-infinitive syntax, reflexive pronouns, subjunctive sequence in subordinate clauses, and context-sensitive time expressions.

English often “backshifts” tense—“I am tired” becomes “she said that she was tired.” Latin infinitives do something more precise: present, perfect, and future infinitives express time relative to the reporting verb, not absolute present, past, and future.

Infinitive tense is relative

Consider direct aeger sum “I am ill.” After present dīcit and past dīxit, the reported illness is simultaneous, so both versions use the present infinitive esse.

The English tense changes; the Latin infinitive does not. Feminine aegram agrees with accusative sē, the subject of esse. The three basic relations are:

Perfect vēnisse is earlier than present dīcit. Future ventūram esse is later than past dīxit; ventūram agrees with Lucia. English “would” here reflects subsequent time in indirect statement. It is not a general translation of a subjunctive.

First person becomes an accusative subject

In direct speech, the speaker can say ego veniō. In indirect statement, that first-person subject normally becomes accusative sē when it refers back to the subject of the reporting verb.

Latin normally omits emphatic ego in the direct version, but it is shown here to make the person shift visible. Sē has no nominative and does not mark singular versus plural; the reporting subject and context supply its reference.

If the reported subject is someone other than the reporting subject, use an ordinary accusative noun or pronoun, not reflexive sē.

The English sentence is ambiguous because both Latin sē and eum can become “he.” Latin distinguishes them. This is one reason not to translate before resolving reference.

Possessives and indirect reflexives

The speaker’s direct meus “my” normally becomes reflexive suus when it refers to the subject whose viewpoint is being reported. A nonreflexive genitive such as eius points to someone else.

Reflexive suum agrees with librum, the thing possessed, not with Claudia. Its reflexive interpretation comes from reference, not gender. In subordinate clauses inside indirect discourse, sē, sibi, and suus can refer “long-distance” to the person whose words or thoughts are represented. That logophoric use is treated in Indirect Reflexives.

This sentence is not the ordinary local reflexive “the soldiers trusted themselves,” which would be a possible reading in another context. Reported viewpoint licenses reference back to rēx. Context and the semantics of the report must decide; Latin reflexives are systematic, but not every isolated sentence is unambiguous.

Second person depends on the new addressee

A direct “you” cannot be converted mechanically. The reporter must identify the original addressee. A second-person subject may remain tē or vōs if the report is still addressed to that person, become an accusative noun or eum/eam/eōs/eās when told to someone else, or become mē/nōs if the current narrator was addressed.

The direct words were something like tū nimis festīnās. Because the present narrator was that tū, indirect statement uses accusative mē and present infinitive festīnāre. There is no special “second-person indirect pronoun.”

Commands and questions keep finite subjunctives, so their personal endings may also shift:

Direct lēntius ambulā has second-person reference; indirect ambulārem is first-person singular because the narrator reports the command as addressed to “me.” The imperfect follows historic monuit. The subjunctive marks an indirect command, not uncertainty.

Subjunctive tense in the reported frame

Main assertions in indirect discourse use infinitives, but indirect questions, commands, and speaker-owned subordinate clauses use subjunctives. These normally follow sequence of tenses.

Perfect discesserim is anterior after primary quaerit; pluperfect discessissem is anterior after historic quaesīvit. The visible ego is legitimate because the narrator, not Marcus, is the subject of the embedded question and contrast may make the pronoun useful. Present and imperfect subjunctives would instead locate an action as simultaneous or prospective.

Authors can preserve primary tenses under a historic report to revive the speaker’s original perspective. That marked (formal) and (literary) repraesentātiō is a discourse choice, not an arbitrary exception. In normal prose composition, apply regular sequence unless a sustained viewpoint effect is intended.

Words for “now,” “here,” and relative days

Deictic expressions may also be recentered. Direct nunc “now,” hīc “here,” hodiē “today,” herī “yesterday,” and crās “tomorrow” can remain when the narrator deliberately preserves the original speaker’s standpoint. More distanced narration often uses tum “then,” ibi “there,” eō diē “on that day,” prīdiē “on the preceding day,” and posterō diē “on the following day.”

Neither conversion is automatic. If narrator and original speaker share the same place or day, hīc or hodiē may remain literally appropriate. If the report is remote, relative forms prevent confusion. Register and narrative distance matter: sustained deictic retention is especially noticeable in (formal) and (literary) indirect discourse.

An authentic network of shifted reference

Caesar reports Ariovistus’s reply with a dense alternation of reflexive and demonstrative pronouns. The source is reproduced without added macrons:

Ipsī and sēsē refer to Ariovistus, the reported speaker. Eum and ille/illum refer to Caesar, the other participant. In the second condition, sē again means Ariovistus as the person whose viewpoint organizes the report. The alternation keeps two third-person men distinct. Present velit is a vivid primary-sequence form within the past report, while esset follows regular historic sequence; Caesar’s mixed framing brings part of Ariovistus’s challenge closer.

The future active participle with perfect infinitive fuisse, ventūrum fuisse, represents the apodosis of a past contrary-to-fact condition here: Ariovistus says that he “would have gone.” This is more specialized than the ordinary future infinitive ventūrum esse and should not be substituted merely because the governing verb is past.

A conversion procedure

1. Identify the original speaker and addressee.
2. Decide whether the reported unit is a statement, question, or command.
3. For a statement, put its subject in the accusative and choose the infinitive by relative time.
4. Replace first- and second-person references according to their actual participants; use sē/suus for the represented speaker where syntax licenses the indirect reflexive.
5. Apply sequence to finite subjunctives in embedded questions, commands, and speaker-owned subordinate clauses.
6. Reevaluate “here/now/today” expressions from the narrator’s spatial and temporal position.

Common Mistakes

Backshifting the infinitive because the report is past

Using eum for the reporting subject

Making suus agree with the possessor

Translating an imperfect subjunctive as conditional “would”

Key Takeaways

Latin infinitive tense is relative: present is simultaneous, perfect anterior, future subsequent. The original speaker’s “I” commonly becomes reflexive sē as the accusative subject, while eum/eam points elsewhere; suus similarly tracks the represented subject and agrees with the possessed noun. Finite embedded clauses follow subjunctive sequence. Deictic words may shift or remain according to the narrator’s actual center and desired viewpoint. Structure—not an automatic English backshift—governs every conversion.