English, like other languages, resists the duplication of nouns in sentences, so it replaces duplicated nouns with what are called pronouns. (No one is sure why languages resist such duplication.) The nouns that get replaced are called antecedents.
Consider sentence 1:
1. Rahul liked Divya, so Rahul took Divya to a movie.
The duplication of the proper nouns Rahul and Divya just does not sound right to most people because English generally does not allow it.
The duplicated nouns are replaced, as in sentence 1a:
1a. Rahul liked Divya, so he took her to a movie.
Notice that sentence 1b also is acceptable:
1b. He liked her, so Rahul took Divya to a movie.
In this instance, however, sentence 1b is not quite as appropriate as 1a because the sentence lacks a context. Real sentences, as opposed to those that appear in books like this one, are part of a context that includes the complexities of human relationships; prior knowledge related to past, present, and future events; and, of course, prior conversations. The pronouns in sentence 1b suggest that Rahul and Divya already have been identified or are known. This suggestion is contrary to fact. In sentence 1a, on the other hand, Rahul and Divya appear in the first part of the sentence, so the pronouns are linked to these antecedents without any doubt or confusion about which nouns the pronouns have replaced. At work is an important principle for pronouns: They should appear as close to their antecedents as possible to avoid potential confusion.