Latin Compounding

The idea that the Latin language itself is (or was) ill-suited to prolific formation of compounds of the kind found in Greek is hooey. Ancient Roman linguistic preferences do not exhaust the possibilities of the language itself.

Quintilian's value judgment:

Sed res tota magis Graecos decet, nobis minus succedit, nec id fieri natura puto, sed alienis favemus; ideoque cum κυρταύχενα mirati simus, incurvicervicum vix a risu defendimus.

(But all this suits the Greeks better. It has not caught on so much with us — not I think because of any innate property, but rather we favor loanwords and so admire κυρταύχενα whereas we can hardly keep from ridiculing "incurvicervicum")

 
Should not be interpreted to be an inexorable result of something intrinsic to the Latin language, as when one scholar suggested that "nominal compounding is not as developed in Latin as it is in other IE languages, since the prototypical length of Latin words is rather short".

The fact that, at least during the empire, Roman-era Latin was not especially given to creative compounds reflects less a linguistic property of Latin than the aesthetic and cultural judgments of those Latin writers whose works survived transmission from antiquity.

 I strongly suspect that resistance to innovative compounding in Latin was BECAUSE it was so associated with Greek, and thus was especially liable to strike the tender nerves of Roman linguistic insecurity. Some words from early Latin tragedy coined on Greek models (sonipes, flexanimo) got entrenched early enough to be grandfathered in.

Some Late Republican writers did coin freely when they wanted to. I think it's no accident that they tend to have been either (a) rather experimentally uninhibited in their Hellenizing or (b) in operating genres held in low esteem by the Roman literary elite. A single poem by Catullus contains such interesting items as hederiger (ivory-bearing), nemorivagus (grove-rover), silvicultrix (forest-tending), erifuga (master-fleeing). And this by no means exhausts the list of Catullan compounds. In the fragments attributed to Decimus Laberius, who worked in a low-prestige genre, there are 46 hapax legomena, and many of these too are obvious compound words. Three nonce-compounds to describe a bull (testitrahus, reciprocicornis, lanicutis) all occur in a single fragment of his. Both the Catullan examples, and those of Laberius, (and even more so comparable coinages found in Lucretius and the fragments of Accius) seem to be inspired by Greek precedent.

Glosses of Greek terms often contain as explanations perfectly serviceable compounds which simply seem not to have caught on in recorded latin. For example πλατυλόγος in one gloss is explained as "latiloquens", which — though a hapax word — is not just perfectly intelligible, but probably near-idiomatic given that that the adverb "late" qualifying verbs of speaking (including loqui itself) was no stranger to the language of Cicero or Quintilian, and that "lati-" as a compounding theme itself is well-attested in words like "laticlavium" and "latifolius."

That "latiloquens" never caught on in accepted literary usage probably has little to do with inherent unsuitability. There is no intrinsic reason why literary Latin could not have taken such a Greek lexical technique over wholesale, much as it did Greek metrical and generic conventions. That it did not do so is an accident of cultural history, and has little to do with the kinds of lexemes the Latin Language itself can or cannot withstand.