Pronouns are linguistic pointers. They can point to someone or something that exists in the actual environment:
Or they can point to something in the "conversational environment," roughly, the set of mentioned things:
Some aspects of pronoun use are obvious and familiar:
Others are more obscure -- and naturally, more interesting to linguists. For example, it's obvious that "he" in (1) and (2) can refer either to John or to some other male.Pronouns must agree with their antecedents in person, number, and gender.
John is going to NY. He (*she/*it/*they/*we/*you) leaves at noon.
But notice that the first option, "he/him" refers to "John", doesn't work for (3).
The only real difference between (1) and (3) is that the pronoun is farther away from "John" in (1) than in (3). More specifically, they are in separate clauses. So, a first hypothesis concerning the distribution of names and pronouns might be
Naturally things aren't quite that simple. The pronouns in (5)/(6) are in the same clauses as their antecedents, but only (6) is impossible.
What's the difference between (5) and (6)?
Of course, there is a grammatical way of saying (3), but we have to use a special type of pronoun called a "reflexive," as in (7).
The sentences in (3)/(4) and (7)/(8) demonstrate an interesting fact about pronouns and reflexives in English: they are (virtually) in complementary distribution (meaning, where you get one, you can't get the other and vice versa).