Augustine, City of God, Book 1, cap. 8

[hlawson38]

Context: Why do misfortunes/good fortune happen to good and bad people? Augustine has already explained why the bad things are spread around. This is in the larger context of Augusine's argument that the sack of the city of Rome should not be blamed on Christianity.

Similiter in rebus secundis, si non eas Deus quibusdam petentibus euidentissima largitate concederet, non ad eum ista pertinere diceremus; itemque si omnibus eas petentibus daret, non nisi propter talia praemia seruiendum illi esse arbitraremur, nec pios nos faceret talis seruitus, sed potius cupidos et auaros.

Translation: It works the same with favorable things, if God out of plain generosity did not grant them to some of those asking, we might say He has nothing to do with them [good things]; likewise, if He gave them [the good things] to everybody asking for them, we might think them the owing to [superficial] devotion, and devotion of this kind would not make us godly but greedy and covetous.

Problem: I can't make out the grammar of serviendum in relation to the clause non nisi propter talia praemia seruiendum illi esse arbitraremur.

[Hylander]

seruiendum esse is an impersonal infinitive, in indirect discourse, dependent on arbitraremur.

Allen & Greenough 196 dubs this the "second periphrastic conjugation", which is formed from the gerundive plus forms of the verb esse.

http://www.perseus.tufts.edu/hopper/tex ... 99.04.0001

Since this is impersonal, the neuter singular form of the gerundive is used.

But to translate into acceptable English, we have make the dative singular indirect object illi the subject of a passive verb. The verb "serve" in English takes a direct object, not an indirect object like Latin seruiendum.

"we would not think that he should be served except [nisi] on account of such rewards".

Literally, perhaps, "that it should be served to him", but this isn't English.

In contrast to English "serve", seruio takes the dative/indirect object of the person to whom service is rendered, and Latin doesn't allow constructions like the English "pseudo-passive", where the indirect object of an active verb can be replaced by the subject of a passive verb, e.g., "John gave him a book" > "he was given a book by John."

This is impossible in Latin: passive constructions can only be formed using the direct object of an active transitive verb as the subject of a passive verb. Passive constructions with an intransitive verb, with or without an indirect object, have to be rendered by an impersonal construction, as here.

*ille seruiendus est (or in indirect discourse, *illum seruiendum esse) is impossible in Latin, so impersonal illi seruiendum est/esse must be used.

[hlawson38]

Many thanks Hylander; I need to review yourgenerous reply, and think it through.

[hlawson38]

After reflection, I think I see now why the grammar of this clause baffled me.

non nisi propter talia praemia seruiendum illi esse arbitraremur

1. I missed the indirect discourse aspect of the grammar, in which serviendum esse functions in the usual way.

2. That clarified, another stumbling block came into view. I was trying to grasp non nisi propter as if it were a single word controlling talia praemia. Hylander's trial translation made the scales fall from my eyes: "we would not think that he should be served except [nisi] on account of such rewards". I could see the non as controlling the verb arbitraremur, which made the logic of the clause clearer.

Thanks to Hylander for towing me out of the sand bed once more.

[Hylander]

non nisi, literally, "not unless" or less literally, "not ... except", is sometime written as one word, nonnisi, and usually is equivalent to "only".

Here, you could alternatively translate: "we would only think he should be served on account of such rewards" or "we would think he should be served only on account of such rewards". The sense would be the same as "we would not think he should be served except on account of such rewards", but maybe the "only" translations flow a little better in English.

[hlawson38]

Thanks again for the helpful comments, Hylander. These double negatives must have come natural to the Romans, but I have to slow down and explicitly trace the logic when I see them.