I’ve spent the past three weeks looking into occurrences of the ‘4’ shape as an alphabetic (and alphanumeric), attested before 1440. The research wasn’t difficult, though it was tedious and necessarily included scripts for which I found no sure identification, but overall it was not so difficult that it needs a whole blogpost here. The most time-consuming part is not collecting formal versions of scripts, but testing the homogenised ‘official’ version against historical examples. Omniglot is a convenient place to begin, if you are interested to follow that question.
Here are two illustrations showing unidentifed script, and both – if they are different scripts – have been mentioned by Voynich writers.

===============================
I wouldn’t say that my investigation of the ‘4’-shape as numeral is complete. There are examples from Armenian, Syrian and Byzantine mss which I haven’t addressed but it was gratifying to find, after back-checking and cross-checking what I had done that my findings accord with the best-qualified commentators’ opinions on the manuscript before imagination and speculation came finally to supplant informed opinion as preferred basis for ‘Voynich’ narratives.
Specifically: the results accord with the views of Georg Baresch who had the longest certain familiarity with the manuscript and who said its contents had not originated in the culture of western Europe. In his time, of course, and even to as late as the twentieth century – as witness works catalogued by British and French libraries – Jewish works were classed as ‘oriental’.
Again, the findings accord with the view of Erwin Panofsky, as it was given in 1932, that the manuscript was from ‘Spain or somewhere southern’ and displayed characteristics both Arabic and Jewish with (perhaps) something of Kabbalah in it.
And finally, they accord with the opinion relayed by H.P. Kraus’ assistant in the early 1960s, and which said that specialists had agreed on a date of manufacture “about 1400” and focused on Italy as most likely place of manufacture.
The early occurrence for that ‘4’ shape as numeral; the pattern of its subsequent dissemination, and the lines of diaspora from the south-western Mediterranean during the last decades of the fourteenth century, allow us to see how those those separate evaluations need not be supposed incompatible with the manuscript’s internal evidence, given the historical events, lines of regular travel and population movement over the period from c.1350-c.1430 AD. I have supposed, and may be proven mistaken in supposing, that whoever wrote the Voynich glyph had a hand accustomed to writing the numeral so.
The same events promise to shed light on the manuscript’s codicology, but I won’t elaborate on that point.
In the next post, I’ll resume the series ‘How to Voynich’ which was broken off to look more closely at the ‘4o’ after noting* Rainer Hanig’s passing comment that “it “seemed obvious” the Voynichese ‘4’ was meant for the letter ‘q’.
*passage was reproduced in earlier post.
I had intended to pursue the question of the ‘gallows’ glyphs, but as you’ll see from those two ‘unknown’ scripts illustrated above, the solution to that question may be better left to specialists in palaeography.
Since my survey considered only some of the areas in which we see Italian-and-Jewish interaction, and omitted other important centres where use of the ‘Arabic’ numerals occurs during the century from 1350-1450 AD, I can only offer a conditional conclusion about the ‘4’-shape as numeral and as Voynich glyph: I’d suggest those who are chiefly interested in Voynichese should be wary of assuming that the ‘4’ shape denotes the letter ‘q’ in any instance let alone in all; and also be very careful about supposing that the usual transcription into Roman letters is ‘as good as’ reproducing the original. As one example of why I reached that conclusion – the illustration below shows forms for ‘8’ and ‘9’ as they are found on a single folio of a manuscript cited by Hill. Just one example, I know, but enough to make the point. In EVA transcription, its use of ‘q’ might obscure distinction in the original between (say) the letter ‘q’, the numeral ‘4’ and even the numeral ‘8’ – just for a start. Some alphabets include two or more letters whose forms, to an untrained eye, appear similar to each other, to this ‘8’ and to the Voynich ‘4’ shape(s).

Afterthought.
Just by the way – here’s a cipher alphabet from the eastern Mediterranean. Early 14thC.
