> But there's no evidence of the first imaginary stages of the future
> Romance language either. I simply don't understand your logic.
Yes there is. We see it in graffiti, inscriptional errors, and even in the
development of the written language. Here's a longish example:
Plautus writes dico quod venit "I say that he comes."
Terence writes dico eum venire "I say him to come"
Classical Latin almost entirely moves over to the second construction
(though there are traces of the first).
The first construction re-appears in the slightly artificial written "vulgar
Latin", alongside the second, mid1st century AD.
In the fourth century AD, Jerome, writing a rather relaxed form of Latin for
his Vulgate translation of the Bible, says dico quod and dico quoniam.
(This is the first form, with confusion of quoniam = quod = because, and
quod = that.)
So we can guess that in the first 2 centuries BC, "dico quod" became marked
as an expression to avoid in posh Latin - but it went underground and
remained the normal expression in actual spoken Latin. And guess which form
emerges when the spoken Latin is finally written down? ego dico quod. (je
dis que, etc).
There is absolutely nothing strange going on. There is no break! Only a
divergence between formal written Latin and the spoken language.
Peter