[split] Measure uncertain spaces?

[Koen G]

I split this off the conversation here: You are not allowed to view links. Register or Login to view.

Basically, Patrick reported that when [r] is followed by [a], there are about as many instances where they have a space between them, as instances where they don't. In addition to that, he also reports that [r] followed by [a] is the pair with the relative highest amount of uncertain spaces.

Put simply, whether there is a space between EVA-a and r or not is not as easy to predict as in other Voynichese glyph pairs, and this coincides with a high amount of uncertain spaces as well. We can wonder whether this uncertainty lies with the scribe, or with the transcriber.

I thought this pair would be good to focus on, and maybe someone could find a machine solution to measure the pixels between each of them. However, when I actually looked at instances of [ra] without space, I realized it would be more complex:
* This is very often in four-letter words like "aral".
* A lot of those [ar] in our transliteration files are actually [or] in the MS...