Filipino to Telugu: keeping track of the speaker inside a report
First-person agreement need not identify the current writer
Messick’s Telugu study gives a reported-speech example in which Raju says that he ran. The embedded pronoun tanu is glossed third person, but the running verb carries first-person agreement. This attested pattern means that a first-person-looking ending inside a Telugu report should not automatically be read as introducing the person writing the current message.
For a Filipino source reporting someone else’s statement, distinguish the current writer from the person whose speech is being reported. Those are two possible centers of reference. The Telugu construction can reflect the reported speaker’s perspective without making that speaker identical to the present writer. Replacing the referent simply to force a familiar person-agreement match could change who supposedly ran.
The study locates this phenomenon in particular embedded environments and investigates its restrictions. It is not permission to use any agreement ending with any pronoun. The useful translation lesson is narrower: a surface mismatch in a reported clause can be grammatical, so participant identity must be interpreted through the whole construction.
Filipino siya leaves another question open
NIU lists siya for he or she. A Filipino reported statement may therefore identify a singular person without marking that person as male or female in the pronoun. Telugu’s third-person system, as described by Messick, distinguishes gender and levels of formality in ways that are not supplied by the Filipino pronoun alone.
This is separate from identifying the reported speaker. Knowing that siya refers to the person who spoke does not necessarily establish the formality level or gender-sensitive Telugu reference. Conversely, selecting a plausible Telugu third-person form does not prove that the embedded action belongs to that person. The identity and the appropriate form of reference are related but distinct decisions.
A new clause is not always a new reported viewpoint
The Hawaiʻi Filipino conjoining lesson joins a sentence about Ben holding a pencil with a sentence saying that he wrote. It explains that the two clauses retain equal standing under the conjunction. That is different from an activity embedded under a saying predicate. A Telugu translation should not impose reported-speech perspective merely because the source contains more than one clause.
This helps distinguish a sequence of events from a report about what someone said happened. Ben performing two actions and another person reporting Ben’s account are different structures. If the source gives only the actions, adding a reporting frame would change the information source; if it explicitly reports speech, flattening it into direct narration would erase that frame.
An intended event is another kind of embedding
The Hawaiʻi gusto lesson places an eating activity inside a wanting expression. That is not the same relationship as reported speech, and it does not assert that the desired activity happened. Consequently, the Telugu study’s reported-speech example should not be generalized to every Filipino sentence containing two predicates.
Read what links the predicates: equal coordination, a report, a desire or another relation. Then preserve who speaks, who acts and whether the action is asserted or merely intended. This keeps an unusual but documented Telugu agreement pattern from becoming either a false error diagnosis or an excuse to alter the source participants.
Language references