All questions
Question 1
Marcus dicit Iuliam in horto esse.
In the sentence above, what is the grammatical role of 'Iuliam' and why does it appear in that form?
- 'Iuliam' is the accusative subject of the infinitive 'esse,' functioning as the subject of the indirect statement introduced by 'dicit,' because indirect statement in Latin requires the subject of the infinitive to be in the accusative case rather than the nominative. (correct answer)
- 'Iuliam' is the nominative subject of the main verb 'dicit,' agreeing with the implied subject of the sentence, because Latin frequently shifts the grammatical subject into accusative form when two clauses are combined into a single statement, and 'dicit' is the finite verb that controls the entire construction.
- 'Iuliam' is a direct object of 'dicit' in the accusative case, indicating the person about whom Marcus is speaking, while 'in horto esse' serves as a separate predicate phrase that modifies the main clause independently without forming part of an indirect statement structure.
- 'Iuliam' is an accusative of respect modifying 'esse,' specifying in what regard something exists, because Latin uses the accusative of respect when a verb of saying introduces a location clause with 'esse,' making 'Iuliam' an adverbial rather than a subject element.
Explanation: Whenever you see a verb of speaking, thinking, or perceiving in Latin — like dicit, putat, or videt — your mind should immediately jump to indirect statement (oratio obliqua). Latin constructs indirect statements using an accusative subject + infinitive, which is fundamentally different from English, where we simply say "Marcus says that Julia is in the garden" with a subordinate clause.
In this sentence, dicit is the main verb ("says"), and everything that follows — Iuliam in horto esse — is the reported content. Because this is an indirect statement, the subject of that reported content (Iulia) must shift into the accusative case, giving us Iuliam. The infinitive esse ("to be") is the verb of the indirect statement, and Iuliam is its accusative subject. Answer A correctly identifies all of this: Iuliam is the accusative subject of the infinitive esse within the indirect statement triggered by dicit.
Answer B is wrong because Iuliam is accusative, not nominative, and Latin never shifts a main clause subject into the accusative — the subject of dicit is the implied "he/she/Marcus," not Iuliam. Answer C misidentifies Iuliam as a simple direct object, which would mean Marcus is saying Iuliam the way you'd "say a word" — but that ignores the infinitive esse entirely and breaks apart a unified indirect statement construction. Answer D invents a rule that doesn't exist; the accusative of respect modifies adjectives or other nouns, and is never used to introduce subjects of infinitives.
Your study tip: memorize the trigger — verb of saying/thinking/perceiving → accusative + infinitive for indirect statement. That pattern appears constantly in Latin prose.
Question 2
Femina putat filium suum Romam ire.
Which element of this sentence signals that an indirect statement construction is present, and what does that element require of the rest of the clause?
- The verb 'putat' (she thinks) is a verb of mental activity that signals indirect statement; it requires the subject of the embedded clause ('filium') to appear in the accusative and the embedded verb ('ire') to appear as an infinitive, producing the accusative + infinitive pattern. (correct answer)
- The reflexive pronoun 'suum' signals indirect statement, because reflexive pronouns in Latin can only appear inside embedded clauses; its presence requires 'filium' to be treated as a nominative and 'ire' to be read as a finite verb in an implied subordinate clause introduced by 'ut.'
- The place name 'Romam' in the accusative signals indirect statement, because Latin uses the accusative of destination exclusively inside indirect statement clauses; this accusative destination form therefore requires 'filium' to be nominative and 'putat' to govern the phrase as a locative expression.
- The infinitive 'ire' itself signals indirect statement and simultaneously requires 'putat' to be reinterpreted as a passive verb and 'filium' to shift into the dative case as the logical subject, because motion infinitives in embedded clauses regularly take dative rather than accusative subjects in classical Latin.
Explanation: When you encounter a Latin sentence and need to identify an indirect statement, your first move should always be to look for a head verb — a verb of speaking, thinking, knowing, or perceiving — because that verb is what triggers the entire construction.
In this sentence, putat ("she thinks") is precisely that kind of verb. Verbs of mental activity like puto, credo, and scio introduce indirect statement in Latin using the accusative + infinitive pattern: the subject of the embedded thought shifts into the accusative case, and its verb appears as an infinitive. Here, filium (accusative) serves as the logical subject of the embedded clause, and ire (the infinitive of eo, "to go") serves as its verb. Together they express the content of what the woman thinks: that her son is going to Rome. Answer A correctly identifies this chain of requirements.
Answer B misidentifies suum as the trigger. While reflexive pronouns do behave specially inside indirect statement (referring back to the main subject), suum is a consequence of the construction, not its cause — and indirect statement never uses ut with a finite verb.
Answer C confuses the accusative of motion toward a place (Romam) with indirect statement. This accusative is a standard Latin motion construction and appears freely outside indirect statements; it triggers nothing on its own.
Answer D inverts the logic entirely. The infinitive ire is a result of the head verb putat requiring it — infinitives don't trigger passive reinterpretation or dative subjects in classical Latin.
Study tip: Always scan for your head verb first. If it expresses thinking, saying, or perceiving, expect accusative + infinitive to follow.
Question 3
Senatus nuntiat hostes victos esse.
The infinitive in this sentence is a perfect passive infinitive. How does this affect the meaning of the indirect statement, and which translation is most accurate?
- The perfect passive infinitive 'victos esse' indicates that the enemies' defeat occurred before the time of the main verb 'nuntiat'; the sentence means 'The senate announces that the enemies have been defeated,' reflecting completed prior action. (correct answer)
- The perfect passive infinitive 'victos esse' indicates that the enemies' defeat will occur after the time of the main verb 'nuntiat'; the sentence means 'The senate announces that the enemies will have been defeated,' because perfect infinitives in indirect statement always express future time.
- The perfect passive infinitive 'victos esse' indicates action simultaneous with 'nuntiat'; the sentence means 'The senate announces that the enemies are being defeated right now,' because the passive form shifts the time reference to the present moment of the announcement.
- The perfect passive infinitive 'victos esse' is interchangeable with the present passive infinitive in indirect statement; the sentence means 'The senate announces that the enemies are defeated,' and the choice of perfect versus present infinitive carries no meaningful temporal distinction in this construction.
Explanation: When you encounter indirect statement in Latin, the key question to ask is: when does the infinitive's action occur relative to the main verb? Latin uses three infinitive tenses in indirect statement to signal time — present for simultaneous action, perfect for prior action, and future for subsequent action. This question tests whether you understand that the perfect infinitive signals completed prior action.
In Senatus nuntiat hostes victos esse, the main verb nuntiat is present tense ("announces"). The infinitive victos esse is a perfect passive infinitive, formed from the perfect passive participle victos plus esse. Because it is perfect, it tells you the defeat happened before the announcing — the enemies were already defeated at the time of the announcement. Answer A captures this exactly: "The senate announces that the enemies have been defeated." The perfect infinitive correctly places the defeat in the past relative to nuntiat.
Answer B is wrong because perfect infinitives never express future time — that role belongs to the future infinitive (e.g., victos iri). This is a clear reversal of how Latin tense in indirect statement works. Answer C is wrong because simultaneous action would require a present passive infinitive (vinci), not a perfect one; the passive voice does not shift time reference to the present. Answer D is wrong because the distinction between perfect and present infinitives is grammatically and semantically meaningful — swapping them changes when the action occurs relative to the main verb.
A useful memory anchor: Present = same time, Perfect = before, Future = after the main verb. Keep this tense chart in mind whenever you translate indirect statement.
Question 4
Scio te hoc facturum esse.
A student is unsure whether 'facturum esse' represents a present, perfect, or future infinitive and how it changes the translation. Which analysis is correct?
- 'Facturum esse' is a passive periphrastic infinitive indicating obligation, since 'esse' combined with any participial form creates an obligation construction inside indirect statement; the sentence therefore means 'I know that this must be done by you,' shifting the subject to an impersonal meaning.
- 'Facturum esse' is a perfect active infinitive because all infinitive forms ending in '-urum esse' are categorized as perfect in Latin; it therefore expresses completed action prior to 'scio,' and the sentence means 'I know that you have done this,' with the act of doing already finished.
- 'Facturum esse' is a present active infinitive in an emphatic or reinforced form; the addition of 'esse' to 'facturum' simply intensifies the present sense of the action, so the sentence means 'I know that you are indeed doing this right now,' with no shift in time relative to the main verb.
- 'Facturum esse' is a future active infinitive formed from the future active participle ('facturum') combined with 'esse'; it expresses action that will occur after the main verb 'scio,' so the sentence means 'I know that you will do this,' indicating future intent relative to the present moment of knowing. (correct answer)
Explanation: When you encounter an infinitive inside an indirect statement (oratio obliqua) in Latin, your first job is to identify its tense — and Latin has three infinitive tenses: present, perfect, and future. Each signals a different time relationship to the main verb.
The form facturum esse breaks down as the future active participle (facturum, from facio) combined with the infinitive esse. This is the standard recipe for a future active infinitive in Latin: future active participle + esse. Because the main verb scio is present tense, a future infinitive in the indirect statement means the action happens after the moment of knowing. So D is correct: "I know that you will do this" — future intent relative to scio.
Answer A misidentifies this as a passive periphrastic (the "must be done" construction). The passive periphrastic uses the gerundive + esse, not the future active participle. Facturum is masculine accusative singular, agreeing with te, not a gerundive in a passive construction.
Answer B falsely claims that -urum esse endings signal perfect infinitives. In reality, the perfect active infinitive ends in -isse (e.g., fecisse). The -urum ending is the hallmark of the future active participle, not the perfect.
Answer C invents a category — an "emphatic present" — that simply does not exist in Latin grammar. Adding esse to facturum does not intensify a present meaning; it creates a completely distinct future infinitive form.
Study tip: Memorize the three infinitive tenses as a trio — present (facere), perfect (fecisse), future (facturum esse) — and always check which tense your infinitive matches before translating indirect statements.
Question 5
Dux credit milites fortes esse.
In this sentence, 'fortes' is an adjective. A student must determine its grammatical function and case. Which explanation is correct?
- 'Fortes' is an accusative plural adjective, but it functions as a predicate of 'credit' rather than of 'esse,' because adjectives inside indirect statement always agree with the subject of the governing main verb rather than with the accusative noun of the embedded infinitive clause.
- 'Fortes' is a nominative plural adjective serving as the predicate of the main verb 'credit,' agreeing with an implied nominative subject of that verb; 'milites' is a separate accusative direct object of 'credit,' and 'esse' is a complementary infinitive completing 'credit' without forming an indirect statement.
- 'Fortes' is an ablative plural adjective of description modifying 'milites' in an ablative absolute construction with 'esse'; 'milites fortes esse' therefore means 'with the brave soldiers being present,' providing circumstantial background information for the main verb 'credit.'
- 'Fortes' is an accusative plural adjective agreeing with 'milites,' which is itself the accusative subject of the infinitive 'esse' in the indirect statement; together 'milites fortes' functions as the subject + predicate adjective of the infinitive, meaning 'the soldiers are brave.' (correct answer)
Explanation: When you see a verb like credit ("believes/thinks") followed by an accusative noun and an infinitive, that's your signal that an indirect statement is being constructed. In Latin, indirect statement after verbs of thinking and believing follows a specific pattern: the subject of the infinitive appears in the accusative case, and any predicate adjective describing that subject must also be accusative, agreeing with it in case, number, and gender.
That's exactly what's happening in D, the correct answer. Milites is accusative plural, serving as the subject of the infinitive esse. Fortes is accusative plural, functioning as the predicate adjective of that infinitive — it describes what the soldiers are. Together, milites fortes esse means "the soldiers to be brave," which the commander (dux) believes. D correctly identifies both the case and function of fortes.
Answer A contains a real-sounding but invented rule — adjectives in indirect statement do not shift agreement to the main verb's subject. They stay with the accusative subject of the infinitive. This is a dangerous distractor because it sounds grammatically sophisticated. Answer B misreads the whole structure by treating milites as a standalone accusative object of credit and esse as a complementary infinitive — but credit doesn't take a simple direct object here; it introduces indirect statement. Answer C invents an ablative absolute, which is impossible since esse cannot anchor an ablative absolute, and milites is accusative, not ablative.
Your study tip: memorize that in indirect statement, the infinitive's subject must be accusative, and all predicate adjectives match that accusative subject — not the main verb's subject.
Question 6
Mater videt puellam in via currere.
This sentence uses 'videt' (she sees) rather than a verb of saying or thinking. Which statement best explains whether this still qualifies as an indirect statement construction?
- 'Videt' introduces indirect statement only when it appears in the passive voice; in the active voice, 'videt' takes a direct object ('puellam') followed by a separate participial phrase, so 'currere' should be reanalyzed as a present active participle modifying 'puellam' in a participial rather than infinitive construction.
- 'Videt' cannot introduce indirect statement because only verba dicendi (verbs of saying) and verba putandi (verbs of thinking) are permitted to govern the accusative + infinitive construction; 'puellam currere' must therefore be analyzed as a simple accusative direct object paired with a complementary infinitive, with no indirect statement present.
- 'Videt' is a verb of perception, and Latin extends the accusative + infinitive indirect statement pattern to verbs of perceiving (seeing, hearing, feeling) as well as verbs of saying and thinking; therefore 'puellam... currere' is a legitimate indirect statement meaning 'she sees that the girl is running' or 'she sees the girl running.' (correct answer)
- 'Videt' can introduce indirect statement only when the governed infinitive is in the perfect or future tense; because 'currere' is a present infinitive, the construction here cannot qualify as indirect statement and must instead be read as an infinitive of purpose indicating the goal or reason for which the girl was observed.
Explanation: When you encounter a Latin sentence with an accusative noun followed by an infinitive, your first instinct might be to look for a verb of saying (dicendi) or thinking (putandi) — but that's only part of the picture. Latin also extends the accusative + infinitive indirect statement construction to verbs of perception: seeing (videre), hearing (audire), feeling (sentire), and similar sensory verbs.
In Mater videt puellam in via currere, the verb videt (a verb of perception) governs puellam in the accusative and currere as a present active infinitive. This is a textbook indirect statement: the mother perceives a situation — "that the girl is running" — rather than a direct, unmediated observation with no embedded proposition. Answer C is correct because it accurately identifies videt as belonging to the broader class of verbs that legitimately introduce the accusative + infinitive construction.
Answer A is wrong on two counts: there is no rule restricting videt to passive voice for indirect statement, and currere is clearly an infinitive, not a participle — the two forms look different and function differently. Answer B reflects the common misconception that only verba dicendi and putandi govern indirect statement; this is an oversimplification that Latin grammar explicitly contradicts with verba sentiendi (perception verbs). Answer D invents a nonexistent tense restriction — Latin never limits indirect statement to perfect or future infinitives, and an infinitive of purpose would require a different construction entirely (often ad + gerundive or a purpose clause).
Your study tip: memorize three verb categories for indirect statement — saying, thinking, and perceiving — so you aren't caught off guard when audio or video introduces an accusative + infinitive.
Question 7
Romani audiverunt regem magnum exercitum habuisse.
The main verb 'audiverunt' is in the perfect tense, and the infinitive 'habuisse' is also perfect. A student claims this means the king's having-of-the-army happened at the same time as the Romans' hearing. Is the student correct?
- Yes, the student is correct. When the main verb is in the perfect tense, the relative tense rule is suspended, and all infinitives in the governed indirect statement default to expressing simultaneous action; 'habuisse' therefore refers to the same time as 'audiverunt.'
- No, the student is incorrect. Infinitive tense in indirect statement is relative, not absolute. The perfect infinitive 'habuisse' indicates that the king's having a large army occurred before the time of 'audiverunt'; the sentence means 'The Romans heard that the king had had a large army,' not that both events were simultaneous. (correct answer)
- No, the student is incorrect, but for a different reason: the perfect infinitive 'habuisse' in indirect statement always refers to future time relative to the main verb, regardless of the main verb's tense; the sentence therefore means 'The Romans heard that the king would have had a large army.'
- Yes, the student is correct in this specific case because 'audio' is a verb of perception rather than a verb of saying; verbs of perception suspend the relative tense rule and require the infinitive to express time simultaneous with the main verb, making 'habuisse' refer to the same moment as 'audiverunt.'
Explanation: When you encounter indirect statement in Latin, the single most important principle to internalize is that infinitive tense is relative, not absolute. The infinitive doesn't tell you when something happened on a calendar — it tells you when it happened in relation to the main verb.
Here's the system: a present infinitive means the action happens at the same time as the main verb; a perfect infinitive means the action happened before the main verb; a future infinitive means the action will happen after the main verb. This rule holds regardless of what tense the main verb itself is in.
So in Romani audiverunt regem magnum exercitum habuisse, the main verb audiverunt is perfect ("heard"), and the infinitive habuisse is also perfect ("to have had"). Applying the relative tense rule: the king's having a large army occurred before the Romans heard about it. The correct translation is "The Romans heard that the king had had a large army" — answer B.
Answer A is wrong because no such "suspension" of the relative tense rule exists when the main verb is perfect. The rule applies universally across all main-verb tenses. Answer C invents a false claim — perfect infinitives never indicate future time; that role belongs to the future infinitive (habiturum esse). Answer D is a tempting distractor because audio can introduce indirect statement, but verbs of perception follow the same relative tense rules as verbs of saying; there is no special exception for them.
Your study tip: memorize the three-infinitive chart (present = same time, perfect = before, future = after) and apply it mechanically every time you see an indirect statement, no matter what tense the main verb is.
Question 8
Pueri credunt magistrum libros portare.
A student translates the passage as: 'The boys believe that the teacher carries the books.' Which analysis of the Latin syntax best supports this translation?
- 'Magistrum' is the accusative subject of the infinitive 'portare,' and 'libros' is the direct object of 'portare'; together they form the accusative + infinitive indirect statement governed by 'credunt,' which is a verb of believing that regularly introduces this construction. (correct answer)
- 'Magistrum' is the accusative direct object of 'credunt,' meaning the boys believe in the teacher, while 'libros portare' is a purpose infinitive phrase explaining why the teacher is trusted; the construction therefore reflects a complementary infinitive pattern rather than indirect statement.
- 'Magistrum' is a nominative predicate noun attracted into the accusative by proximity to 'credunt,' and 'portare' is a complementary infinitive that completes the meaning of 'credunt,' so the sentence does not technically employ indirect statement but rather a standard complementary infinitive pattern.
- 'Magistrum libros portare' forms a passive periphrastic construction in which 'portare' carries an obligatory sense and 'magistrum' is an accusative of agent; the sentence should therefore be translated as 'The boys believe the books must be carried by the teacher.'
Explanation: When you see a Latin sentence containing a verb of mental activity (thinking, believing, knowing, saying) followed by a noun in the accusative case and an infinitive, your first instinct should be to look for indirect statement using the accusative + infinitive construction. This is one of the most tested Latin syntax patterns at this level.
In Pueri credunt magistrum libros portare, the verb credunt ("they believe") is a classic head verb that introduces indirect statement. Magistrum appears in the accusative case, functioning as the subject of the infinitive portare, not as the direct object of credunt. Libros is then the accusative direct object of portare. The whole infinitive clause — magistrum libros portare — serves as the content of what the boys believe. Answer A correctly identifies all of these relationships, making it the right choice.
Answer B misreads magistrum as the direct object of credunt (as if the boys "believe in" the teacher) and miscategorizes libros portare as a purpose infinitive — but purpose infinitives express why an action is done, not the content of a belief, and they typically require ad + gerundive in Latin.
Answer C invents a non-existent grammatical process: "accusative by proximity" is not a real Latin rule. Magistrum is straightforwardly accusative because it is the subject of the infinitive in indirect statement, full stop.
Answer D confuses portare (a present active infinitive) with a gerundive, which would be portandum. The passive periphrastic requires a gerundive, not an infinitive, so D's analysis collapses immediately.
Study tip: Memorize the core verbs of saying, thinking, knowing, and perceiving — they almost always trigger accusative + infinitive indirect statement in Latin.
Question 9
Cicero scribit Caesarem in Gallia bellum gerere.
Suppose a second-year Latin student argues that 'Caesarem' must be dative because it follows a verb ('scribit') that sometimes takes a dative object in other constructions. How should this argument be evaluated?
- The argument is correct. When 'scribo' means 'to write to someone,' it takes a dative object, and 'Caesarem' should be read as dative singular of a third-declension noun; 'gerere' is then a complementary infinitive completing an implied verb of obligation introduced by the context of the subordinate clause.
- The argument is partially correct. 'Scribit' can take a dative object in some constructions, so 'Caesarem' is genuinely ambiguous between accusative and dative; however, because 'bellum' also appears in the clause, the dative reading is preferred, since having two accusatives in the same indirect statement would be ungrammatical in Latin.
- The argument is incorrect. 'Scribit' here introduces an indirect statement, not a dative-object construction; 'Caesarem' is accusative, functioning as the subject of the infinitive 'gerere.' The accusative ending (-em) distinguishes it from the dative singular, and the indirect statement pattern requires the accusative, not the dative. (correct answer)
- The argument is correct in principle but misapplied. While 'Caesarem' is dative here, 'scribo' introduces indirect statement by governing both a dative subject and an accusative infinitive, which is an alternative indirect statement pattern used specifically with verbs of writing and reporting in classical Latin.
Explanation: Whenever you see an accusative noun paired with an infinitive after a verb of saying, knowing, or thinking, your first instinct should be: this is indirect statement. Latin expresses indirect statement through a construction called the accusative + infinitive (ACI), and scribo is one of the verbs that can introduce it.
In this sentence, scribit introduces an indirect statement: Caesar is doing something (waging war in Gaul). The pattern demands that the subject of the infinitive appear in the accusative case — hence Caesarem, accusative singular of Caesar (a third-declension noun). The infinitive gerere is not complementary or subordinate in the way the student imagines; it is the main verb of the embedded indirect statement, and bellum is its direct object. Everything clicks together: scribit [main verb] + Caesarem [accusative subject of ACI] + gerere [infinitive] + bellum [accusative object of gerere]. Answer C is correct.
Answer A is wrong because it misidentifies Caesarem as dative and invents an "implied verb of obligation" — there is no such thing here, and gerere is not a complementary infinitive. Answer B is wrong on two counts: Caesarem is not genuinely ambiguous (the -em ending is only accusative for third-declension Caesar), and two accusatives in an ACI is perfectly grammatical when one is the subject and one is the object of the infinitive. Answer D fabricates a "dative subject + accusative infinitive" indirect statement pattern — this construction simply does not exist in classical Latin.
As a study tip: always check whether a verb of communication introduces an ACI before assuming it takes a dative. When it does, remember that two accusatives in the same clause is a feature, not an error.
Question 10
Philosophus dicit se multa scire.
In this sentence, 'se' appears as the subject of the infinitive instead of a noun or pronoun in a different form. What does the use of 'se' here indicate, and how does it affect the meaning?
- 'Se' is the ablative form of the reflexive pronoun indicating separation from the philosopher; it modifies 'multa' to convey 'many things apart from himself,' and 'scire' functions as an infinitive of result, so the sentence means 'The philosopher says that many things lead to knowledge independently of himself.'
- 'Se' is the reflexive pronoun used as the accusative subject of the infinitive 'scire' in indirect statement; it refers back to the subject of the main verb 'dicit' (the philosopher himself), so the sentence means 'The philosopher says that he himself knows many things,' indicating the philosopher is speaking about himself. (correct answer)
- 'Se' is the nominative form of the reflexive pronoun placed inside the indirect statement clause without case change, because reflexive pronouns are indeclinable in classical Latin; it refers not to the philosopher but to a third party mentioned earlier in a broader discourse context.
- 'Se' is the accusative reflexive pronoun, but in indirect statement it refers to the person being addressed rather than the speaker; Latin uses the reflexive in embedded clauses to signal the intended audience, so the sentence means 'The philosopher says that you (the listener) know many things.'
Explanation: When you encounter a sentence with an infinitive following a verb of speaking or thinking (dicit, putat, credit), you're looking at indirect statement — one of Latin's most important grammatical constructions. The key rule: the subject of the infinitive appears in the accusative case, not nominative.
Now apply that to Philosophus dicit se multa scire. The main verb is dicit ("says"), which introduces an indirect statement. The infinitive scire ("to know") needs an accusative subject — and that subject is se. Because se is the reflexive pronoun, it refers back to the subject of the main verb, the philosopher himself. So the sentence means "The philosopher says that he himself knows many things." Answer B correctly identifies all of this: se is accusative, it functions as the subject of scire, and it reflects back to the philosopher as speaker.
Answer A is wrong on multiple levels — se is not ablative here, it is not modifying multa, and scire is not an "infinitive of result." That interpretation invents a grammatical role that doesn't exist in this sentence. Answer C is false because reflexive pronouns in Latin are absolutely not indeclinable — they fully decline through accusative, genitive, dative, and ablative forms. And se in indirect statement refers to the main subject, not a vague third party. Answer D misunderstands reflexivity entirely — the reflexive pronoun points back to the grammatical subject of the main clause, never to the listener.
Your study tip: whenever you see se inside an indirect statement, immediately ask "Who is the main verb's subject?" — se refers to that person, making the embedded clause self-referential.