Return of the minimal pairs (when is a morpheme not a morpheme?)

[petegray]

> Originally Latin had a vocalic [u] and a semi-vowel [w], both written with
> the <u/v> graph (the two shapes were simply variants of the same
> grapheme).  ...., they remained more or
> less interchangeable until the late middle ages.

The u shape was unknown to the Romans.  They had only capital V.  When small
letters were developed, about 600 AD or so, the small form of V was u, in
all contexts.  This is how the Oxford Classical texts are still printed:
   Vbi at the beginning of a sentence, ubi elsewhere.
   Venit at the beginning of a sentence, uenit elsewhere.

But this pattern is totally unroman.

The late development also explains why Greek and Latin developed different
small forms from the same capitals (eg from A B M N K etc).  If this had
happened earlier, the two scripts might have been more similar!

Peter