There's no contradiction here. Orberg says essentially the same thing as Wheelock and Allen & Greenough, though pmda
misread him originally as saying the opposite. When a participle has verbal force, i.e. when it describes the time, circumstances, condition, reason, etc. during, under, or for which the force of the main verb applies, the ablative singular must always be -e. So likewise when the participle is actually a noun, e.g. amans
"lover". However, when the participle acts as a mere attribute to a noun which describes it only in general terms, it will usually have its ablative in -i, just like most other adjectives.
furrykef wrote:Perhaps a better term would be 'adjectival', but yes: the ablative is always -ī for third-declension adjectives and -e for third-declension nouns (including adjectives behaving as nouns), and this applies to the present participle as well. (This applies only to the present participle, though. Future and past participle are first/second declension rather than third, and follow the normal rules.)
With the ablative absolute, -e is used: "Caesare dūcente, vīcimus." -- "With Caesar leading, we conquered."
As an annoying exception, comparative adjectives always take -e too (longus = long, longior = longer, abl. longiōre).
You're speaking a bit too broadly, I think. First of all, there are a few 3rd declension adjectives, such as vetus, veteris
and (as you mentioned) all the comparative adjective forms, which are not i-stem and thus have their ablative in -e instead of -i. But more importantly, almost all neuter nouns of the 3rd declension in fact have ablatives in -i, with only a handful of exceptions. The masculine and feminine nouns generally have -e, it's true, but even here certain i-stem nouns, especially those which have identical nominative and genitive forms (e.g. ignis
), very often exhibit ablatives in -i, though the Romans themselves weren't always so consistent on this point as we might like them to have been.
furrykef wrote:The "ā patre amantī" seems to be equivalent to Orberg's "ā puerō dormiente". I'm not clear why there's the discrepancy... either one of the two is mistaken, or maybe the usage varied depending on time period or region.
They're not quite equivalent: a patre amanti
means just "by the loving father", and not "by the father as/while he loves". On the contrary, a puero dormiente
means precisely "by the child as he sleeps", which can be rendered more simply as "by the sleeping child".