Characters repeating across word boundaries

[MarcoP]

Thanks, Marco, those are some great graphs again. I shall add them to the graphs thread and link to your thread for ease of retrieval.

Might there be something special about the examples where initial [o] is allowed to follow final [o]? For example, may a specific word-initial combination be excluded from this set, causing the discrepancy in the data?

Thank you for your comment, Koen!

I have tried to investigate the subject some more, but I couldn't find anything clear.
These 4 tables compare the most common words:
ending -o, 1. most common words before NOT(o) and 2. most common words before o-
starting o-, 3. most common words after NOT(o) and 4. most common words after o-

I have manually highlighted words to show that there is a large overlap between the tables (the highlighting isn't 100% accurate). I cannot see nothing special in the o- starting words that appear after -o, nor in the -o ending words that appear before o-.
Another line of investigation could be examining word ending with -o in general. They are not terribly frequent, but not too rare either (a total of about 1000 occurrences). To avoid the appearance of -o.o-, Voynichese must either replace the ending -o before o-, or replace the starting o- after -o. While Emma has already done much work analysing o-, the -o suffix is much more obscure to me. A good field for further work....