R A Brown, On 05/05/2010 07:00: > And Rosta wrote: >> R A Brown, On 02/05/2010 15:23: > [snip] >>> Now IMNSHO any generative analysis (or any other analysis) of that >>> Latin sentence that fails to take into account that the first three >>> words (i.e. as far as the verb) is a complete, potentially meaningful >>> and *well-formed* sentence in itself - the rest amplifies and >>> resolves an uncomfortable semantic tension - is not truly reflecting >>> the Latin sentence. >> >> Can one have something like "warrior-NOM wept prince-NOM", "the >> warrior prince wept", > > bellator lacrimavit princeps - Yep, that's fine. > >> where "warrior" and "prince" are both substantives (assuming there is >> some way to tell if a word is a substantive)? > > There's the rub. If you look up _bellator_ in the Lewis & Short > dictionary, you'll find it given as a masculine substantive with many > examples of its use. But then you come to: > "II. Esp. (like amator, arator, ventator, etc.; v. Zumpt, Gr. § 102; in > close apposition with another _subst_., and taking the place of an > _adj_.): _warlike, ready to fight, martial, valorous_ ...." > > Clearly _bellator_ may be used as a substantive or an adjective, tho the > latter use in mainly, tho not entirely, found in poetry. > > _princeps_ is also used both an an adjective & as a substantive. Fascinating. On the assumption, then, that no sentence of the form "noun-X verb noun-X", where X is the same case, is categorically impossible, then I would tentatively propose an analysis in which: (1) nouns in the Latin clause are not complements but rather are predicative adjuncts, as in English "*A coward*, he ran away", "*Cowardly*, he ran away"; (2) these adjuncts are predicational phrases where the predicate is the noun and the subject position is lexically empty; (3) the actual complements of the verb are 'positions' that are lexically empty or perhaps occupied by inaudible or audible pronouns; (4) the complement positions are linked to the subject positions within the predicational adjuncts by 'movement' (i.e. they are interpreted as 'coreferential'), and this mechanism determines the case inflection of the noun. The resulting syntactic structure is kind of near-minimal in not imposing more structure than necessary, since the minimal logical form is "warrior(x), & wept(x), & prince(x)", a conjunctive list with no necessary hierarchical structure. --And.