Is a relative pronoun commonly used as a third person pronoun? (Metamorphoses I.583-587)

Question

In this short passage by Ovid, the pronoun "quam" seems to be used as a third person pronoun.

Inachus unus abest imoque reconditus antro
fletibus auget aquas natamque miserrimus Io
luget ut amissam. Nescit, vitane fruatur,
an sit apud manes; sed quam non invenit usquam,
esse putat nusquam atque animo peiora veretur.

Here's my translation:

Inachus alone is absent, hidden deep within a cave, increasing the waters with his tears and, most miserable, mourning his daughter Io as lost. He does not know whether she enjoys life, or is among the shades; but [since] he has not found her anywhere, he thinks her to be nowhere, and dreads worse things in his mind. (Metamorphoses I.583-7)

You can see that I treated "quam" as a third person pronoun, translating it as "her". I have a few questions about this.

1. Am I correct in doing so? ("quam" = "her")
2. Is the relative pronoun commonly used as a third person pronoun, in classical Latin?
3. Do you think that Ovid may have preferred "quam" to "illam" because of its consonance with "usquam" and "nusquam"?

Answer 1

Quam here is being used as a normal relative, and could not be replaced with illam, since the relative clause quam non invenit umquam is the accusative subject of the verb esse in the accusative and infinitive construction introduced by putat in the main clause. A literal translation is: "[the one] whom (quam) he does not find anywhere, he believes to be nowhere".

Answer 2

In this case, it is a normal relative use of qui, but with the antecedent (like [eam] quam) included in the relative pronoun. You could translate it literally as "but he thinks [her] whom he does not find anywhere to be nowhere", relative clause in italics.

---

Although it doesn't apply here, using qui as a demonstrative/personal pronoun is not uncommon. The Dutch term is relatieve aansluiting (I have forgotten the Latin term), when a relative pronoun is used where we would expect a demonstrative or personal pronoun. We were taught to translate such forms of qui as if it said et/at ille/hic, as it fits the sentence.

The distinction between relative, demonstrative/personal, possessive, and interrogative 3rd-person pronouns is relatively recent, such that e.g. Homer still often uses what is to us the article as a relative pronoun, and the relative as a possessive.

In Latin, too, the relative pronoun developed relatively recently (probably around late prehistory), out of a demonstrative or interrogative qui. At first this qui was always used interrogatively, or demonstratively with some subordinating or similar conjunction or particle, but then later it came be used alone as a subordinating pronoun (relative pronoun).

The construction used by Ovid is probably a remnant of that.

In German, the phaenomenon is called relativ(isch)er Anschluß. You may find more information in Kühner–Stegmann, Ausführliche Grammanik der lateinischen Sprache (1914), second volume, part 2, paragraph 197 (p. 319). (When the Germans call something ausführlich, you know it is twice that.)

The English term is connective relative clause.