Catalan to Zulu: distinguish the object clitic from agreement with the actor
One target word can preserve two different people
Keet and Khumalo analyse the Zulu example Ngimbona kusasa as I will see him tomorrow. Inside ngimbona, ngi- refers to the first-person subject and -m- to the object. The fact that both occur within one written verb does not make them the same participant.
Catalan also permits a short verb-and-clitic expression to recover someone previously mentioned. But its verb agreement and object clitic must be resolved separately before comparing the target. An original Catalan message saying that I will see the colleague must not become a message that the colleague will see me simply because both people remain somewhere in the translation.
The translated noun’s class matters for object agreement
The Zulu study’s concord tables distinguish person reference from noun-class agreement and list different object concords for different classes. Catalan el and la instead carry gender and number information associated with their source antecedents. Those source categories cannot directly supply a Zulu noun class.
If a Catalan message alternates between the book and the package, the target must first establish the Zulu nouns being used. Recovering either item through an object concord depends on that target antecedent. Choosing one concord for every source el would wrongly treat all Catalan masculine singular objects as members of a single Zulu class. Conversely, a change of target concord can be appropriate when the source clitic remains visually unchanged.
A recipient and a direct object can share a Catalan surface form
CPNL’s weak-pronoun account distinguishes determined direct objects from indirect objects. Catalan els can occur as a masculine plural direct-object form or as a plural indirect-object form, so its spelling alone does not identify the participant’s role.
Consider an original account about seeing colleagues and then making coffee for them. In the first action, the colleagues are the people seen; in the second, they benefit from something prepared. A translation that identifies the same people but assigns the same object relationship indiscriminately can misrepresent what was done. The Zulu study separately discusses applicative verbal extensions, reinforcing why a beneficiary relationship should not be reduced to a pronoun spelling match.
A repeated-reference word can instead concern a place
CPNL illustrates hi with going somewhere on foot, whereas en can recover an origin in a returning statement. Neither expression should be interpreted as a direct object merely because it avoids repeating a noun. Their antecedents belong to different relationships in the clause.
An exchange might say that the writer has the package and will go to the office tomorrow. If the next Catalan reply uses hi, carrying over the package’s Zulu object concord could confuse an item with a destination. Resolve whether the source recovers a person, thing, beneficiary, or location before interpreting a target word’s internal agreement. The meaningful test is the preserved relationship, rather than whether every Catalan clitic corresponds to one visible standalone target word.
Language references