accusative infinitive construction with a relative pronoun

[Carpe Piscem]

Does anyone know if it was ever acceptable to split the subordinate clause when using a relative pronoun? i.e., are these both ways of saying, "They loved him whom they saw loved the dog."

Ipsum amverunt, quem viderant canem amare.
Ipsum amaverunt, quem canem amare viderant.

The second is more grammatically correct I think, but I believe I've run into examples of the first construction too. (Usually by medieval writers who were using Latin as a second language.)

Piscem

 

[Bitmap]

The second is more grammatically correct I think, but I believe I've run into examples of the first construction too. (Usually by medieval writers who were using Latin as a second language.)

Latin word order is free, both relative clauses are fine. I would probably add an eum to the ipsum in the main clause, though.

 

[Pacifica]

Whether you add eum to it or not, ipsum is rather emphatic, making it more like "they loved him himself/they loved that very person who..." There's no such emphasis in your English version at first sight, so I'd suggest the more neutral eum instead of (rather than in addition to) ipsum. That said, maybe you were influenced by medieval authors, some of whom routinely use forms of ipse without much apparent emphasis at all, basically where is would be used in classical Latin.

Note that quem canem amare viderant (and every word order variation of it) is ambiguous: which of quem and canem is the subject and which is the object? They're both accusative. Latin authors often avoided this sort of thing by turning the construction into the passive: a quo canem amari viderant.

The aspect of the verb amaverunt could also be debated: perfect or imperfect? The imperfect seems more likely at first sight, but it would ultimately depend on context. Now, if you only made up that sentence to ask about the word order in the relative clause, I guess it doesn't hugely matter.

 

[Serenus]

Whether you add eum to it or not, ipsum is rather emphatic, making it more like "they loved him himself/they loved that very person who..." There's no such emphasis in your English version at first sight, so I'd suggest the more neutral eum instead of (rather than in addition to) ipsum. That said, maybe you were influenced by medieval authors, some of whom routinely use forms of ipse without much apparent emphasis at all, basically where is would be used in classical Latin.

Yes, and this phenomenon later shows up in Italian as the personal pronouns esso/essa/essi/esse 'he/she/they(men)/they(women)'. (These pronouns have now in the 21st century been pretty much abandoned in favour of lui/lei/loro, but you know, they used to be common.)

 

[Carpe Piscem]

Thank you all for your responses. Yes, the ipsum is emphatic. I was sort of making up a simplified version of a more complex sentence I ran into. The ipusm there was emphatic because of the context: "this very man I've been talking about." It is true though that a lot of my reading is medieval Latin, and they did love to use ipse!

As for the ambiguity with whether quem or canem is the subject, that was a little intentional. I'd always thought word order did matter a fair bit in accusative-infinitive constructions, with the first accusative being the subject of the clause and the second being the object -- or at least almost always. So while my main question was whether it was ever an acceptable practice to insert the main verb between elements of the subordinate accusative-infinitive clause, I'd also wondered how, if it was ok, it affected the rules of word order. With relative pronouns in general, it seems tricky because you would usually always have the pronoun at the beginning of the clause anyway. (This is perhaps why AICs with relative pronouns don't seem very common.)

Piscem

 

[Godmy]

I think the word order matters a bit more* in an AcI, since the language behaves there completely analytically (zero distinction of the syntactic roles via an inflection). That is, the SOV should be naturally preferred.


^ you have a really entertaining nickname

*I'm talking a preference or a tendency of the native speakers, not an actual rule

 

[Pacifica]

I'd always thought word order did matter a fair bit in accusative-infinitive constructions, with the first accusative being the subject of the clause and the second being the object -- or at least almost always.

I wouldn't trust that rule overmuch.

I guess there may be a tendency to that effect in instances where ambiguous AcIs do occur (I haven't researched the question), but the truth is that those ambiguous AcIs are usually avoided in the first place — and when they do occur, I'd say it's primarily the context, rather than the word order, that clarifies things; but even when you'd think the context would make things clear, the best authors usually choose to avoid the theoretically ambiguous acc.-acc.-inf. construction. Note, however, that it's perfectly OK to use it when the very nature of the object and verb makes the former quite unlikely to be taken as the subject of the latter (e.g. scio te libros legere).

 

[Pacifica]

(This is perhaps why AICs with relative pronouns don't seem very common.)

They seem common enough to me (but, like with all other AcIs, ambiguity tends to be avoided there, too, by turning things into the passive when needed).

 

[Dantius]

with the first accusative being the subject of the clause and the second being the object -- or at least almost always.

"almost always" is definitely an overstatement in this case