As to IIII being earlier than IV, mallam, your asking for evidence is beyond my present library at home and, it seems, the net. 'IV' is used in the classical texts I have but that's not to say that a comparative view of ancient texts wouldn't show IV written earlier than IIII. However, we may look to why the Romans used I and V. If you put your hands flat on the table, there's an answer 'to hand'. Each finger forms an I. The findex finger and thumb form a V. From the left that reads I, II, III, IIII and V. Looking to the right hand we start with the V of thumb and forefinger, and read V, VI, VII ,VIII and VIIII. If we keep our hands flat and have the thumbs overlapping, the thumbs form X. Reading to the right we have X, XI, XII, XIII, XIIII. It may not escape notice that to the left of the X we have IIIIIX, IIIX, IIX, IX and then X.
Now, the question is, would an Ancient Roman think first of writing by copying his fingers up to IIII on his left hand alone and only later realise that he could have written IV , using his index finger and the adjacent V ? I think he would have thought of that IV only after he thought of the obvious way. Having hit upon that, someone would see that the same device could be used for VIIII by using the forefinger by the X, which was formed by the thumbs, and putting IX.
Scribes would not only save ink and a little time by writing IV but IV is readily recognised, whereas IIII might be misread for III. In fact, Roman scribes sometimes wrote the last I in a series as j (thus: iij for iii), to make the text clearer.
We know that IIII and IV did coexist, as said before. Whoever wrote the wikipedia entry 'Roman numerals' has a theory about IIII and IV. They say that the Romans put IIII because they put IV in IVPPITER for Jupiter, the chief god.That would explain inscriptions having IIII though IIII does occur a lot on inscriptions and not just when Jupiter figures LOL. It would make more sense to note that U was carved as