Peter Collier wrote:
[snip
> 
> The language is VL based.  

Good.

> Ablatve forms do merge with the accusative, 
> and so logic (and OTL fact) moves all the prepositions over to governing 
> the accusative. 

Which is precisely what happened in the real world *here*.

> BUT, and this is where the grain of doubt enters my mind:
> 
> ~ The Romance languages lose case disitnction generally, and quite early 
> on -  my Romconlang does not.

Case distinction in nouns remained in France, at least in _langue 
d'oïl_, until the early 12th century (not sure how long it survived in 
the _langue d'oc_ of the south, but it was certainly still extant in the 
early Medieval period). And case distinction in nouns remains until the 
present day in Romanian.

It is true that it has disappeared before the earliest attested writing 
in Italian or in Iberian Romance. But Spanish survivals of old 
nominatives, e.g. Dios, Carlos, do point to an earlier period where a 
two case system similar to that of _langue d'oïl_ and _langue d'oc_.

> ~ In (very) Old French at least there is only a Nom/Oblique dsitinction 
> (I'm not familiar enough with the history of the other natangs to cite 
> here), so the prepositions could be said to govern the oblique case, 
> rather than the acccusative. 

It could be, except that the oblique case of old French and old 
Provençal is clearly derived from the Latin accusative, as the plural 
forms make abundantly clear.

We find in actual written Latin a great confusion between the ablative 
and the accusative after prepositions, e.g.
(a) ab hortu[m], con quen, cum libertos, ex donationem, pro salutem, pro 
hoc ipsud, sine lesionem, de hoc ipsud, de carnem;
(b) contra ipso loco, venit in civitate sua.

(a) are all examples of prepositions which classical purists use only 
with the ablative; in all these attested examples they are followed by 
the accusative.
(b) are examples of prepositions which would be followed by the 
accusative in classical Latin, but are followed by the ablative here. 
This may be due either to hypercorrection, or simply reflect the 
colloquial merging of accusative and ablative.

> Either way, the preopositions do not govern 
> *nominatives* - but IMC the accusative has merged with the nominative 
> and there is only a Nom/Gen/Dat distinction.

It is true that prepositions do not govern the nominative in Latin of 
any sort (nor AFAIK in any IE language). But in Romanian the nominative 
and accusative have merged; Romanian differs from your conlang in that 
the genitive and dative have also merged, giving just a two case system 
in the modern language: nom/acc ~ gen/dat (strictly speaking, a three 
case system as separate vocative forms exist in the singular). The 
Romanians have no problem with prepositions governing the nom/acc. case 
(but see below).

> ~ The pre-Roman substrate in the region uses the Dative also for 
> indirect objects, and in situations where (Classical) Latin uses/used 
> the Ablative.

NO! Classical Latin used the _dative_ for indirect objects.


> So I have poor old Octavio who is confused, because his teacher is 
> telling him his writing is inaccurate because such and such a 
> preposition governs some 'ablative' case he has no practical concept of 
> (other than it sometimes seems a bit like the dative, which he uses a 
> lot), whereas in the same circumstance that Gaulish guy on the market 
> stall is using what sounds to Octavio like the nominative, but his 
> Chattian grandfather, 

The Gaulish guy would surely be using the accusative, if Old French or 
Old Provençal is anything to go by!!

The ablative doesn't sound a bit like the dative in the singulars of 
1st, 3rd, 4th (I guess the pedantic teacher would insist on the dative 
ending -ui in the fifth declension). Poor old Octavio, of course, would 
have to learn which nouns belonged to the 4th & 5th declensions, as 
these declensions had long since disappeared from the spoken language. 
If his teacher is insisting on him writing correct Latin according to 
the Classical, Octavio, unless he is particularly obtuse, should long 
have been aware that the language he is being made to write is different 
in very many respects from the language he speaks, and very different 
from the language of the Gauls he may meet.

He would surely just learn the prepositions and their cases by rote - 
as, indeed, he would have to learn much else. This must have been a very 
familiar situation during the so-called 'dark ages', as the western 
Empire crumbled.
----------------------------------

Benct Philip Jonsson wrote:
 > What can I say other than that I agree with your thinking:
 > the grammatical distinction between prepositions governing
 > the acc. and the dative would be carried over as a feature
 > borrowed from the Germanic substrate. Granted it would no
 > longer matter which prep governed which case in Latin, but
 > rather whether they indicated motion or location, plus

So, Latin _ad_ which always governs the accusative in Classical (and 
Vulgar) Latin, whether it means _at, near_ or _towards_ would be 
reconstructed?

No preposition governs the dative in Classical Latin. If Peter's Romance 
conlang has such a strong Germanic substrate that the Germanic 
accusative (motion towards) and dative (no motion) feature gets carried 
over into it, so that, for example _a(d)_ may be followed by the 
accusative or the dative, then this is an *innovation* - it has nothing 
to do with the Latin ablative whatsoever.

As I have observed, the Latin acc~abl after preposition distinction was 
moribund even in Classical Latin (surviving with only four prepositions, 
one of which was fairly uncommon) and was clearly as dead as the 
proverbial dodo in the spoken language.

> In Old High germanican
> you should also get a number of preps governing the
> genitive, of course. 

Yes, indeed.

It is interesting that in Romanian, as well as prepositions governing 
the nom/acc case, we also find prepositions governing the gen/dat case. 
*But these have nothing to do with the ancient Latin ablative.* This is 
an innovation. (One source I checked distinguishes between prepositions 
that govern the dative and those that govern the genitive; I do not know 
by what criteria this distinction was made.) I suspect - tho I not 
nearly as familiar with the development of Romanian as I am with the 
other Romance langs - that this may be in part influenced by the Slav 
languages which neighbor it.

My view is that Peter should forget the Classical ablative. Either the 
prepositions will simply govern the nom/acc. or, like Romanian, may also 
govern the genitive & dative. But the most plausible reason for that IMO 
would be the influence of Germanic.

If there is a strong Germanic influence, I expect expect it to be 
reflected in other ways, particularly in the restructuring of the verbal 
system.


-- 
Ray
==================================
http://www.carolandray.plus.com
==================================
Frustra fit per plura quod potest
fieri per pauciora.
[William of Ockham]