Okay, so here are some quantitative results for Q1 and Q2 (haven't time for other quires yet) for paragraphs starting with k.
In Q1, 10 folios out of 16 (hereinafter I count recto/verso as two folios) have paragraphs starting with k. In total, 12 paragraphs out of 35 in this quire start with k.
In Q2, only 1 folio out of 14 (12r/v are lost) has paragraphs starting with k, or, more precisely, it's only one paragraph - out of 25 total for this quire.
We are exploring sequences of line-initial characters.
For all those paragraphs, k always precedes any other character. There is only one occurrence where a line-initial k is not paragraph-initial, that is f2v, first paragraph, and here k is the second in sequence - that is, it is placed directly after the first k.
The general sequence order is as follows (I distinguish between s and sh, c and ch):
k
y
d
o
q
s
sh
(there is also one occurrence of ch - since it's only one, we can't assume any ordering for it)
If any of these is missing in the paragraph, the relative order of remaining characters is nonetheless preserved. In some places there are duplicates in a row, e.g. d-d.
There are of course exceptions to this rule, but not many:
f2r, first paragraph, line 5 begins with o after the fourth line began with sh. After that, the order is restored.
f2r, second paragraph, line 4 begins with y after the third line began with q. After that, the order is restored.
f3v, line 5 begins with y after the fourth line began with q. After that, the order is restored.
f5v, line 6 begins with o after the fifth line began with sh. After that, no further lines are present.
f6v, line 4 begins with y after the third line began with q. After that, the order is restored.
f13v, line 4 begins with o after the third line began with q. After that, the order is restored.
Note that the types of exceptions are fewer in number than the exceptions themselves: it is either o after sh, y after q, or o after q.
What's remarkable, is that all these exceptions occur no earlier than in the fourth line of a paragraph. What's even more remarkable, there never is more than one exception in a paragraph - that is, kinda auto-repair (of the broken order) is in place.
Two quires are perhaps too little for generalizations. I'm looking forward to check further quires.
In Q1, 10 folios out of 16 (hereinafter I count recto/verso as two folios) have paragraphs starting with k. In total, 12 paragraphs out of 35 in this quire start with k.
In Q2, only 1 folio out of 14 (12r/v are lost) has paragraphs starting with k, or, more precisely, it's only one paragraph - out of 25 total for this quire.
We are exploring sequences of line-initial characters.
For all those paragraphs, k always precedes any other character. There is only one occurrence where a line-initial k is not paragraph-initial, that is f2v, first paragraph, and here k is the second in sequence - that is, it is placed directly after the first k.
The general sequence order is as follows (I distinguish between s and sh, c and ch):
k
y
d
o
q
s
sh
(there is also one occurrence of ch - since it's only one, we can't assume any ordering for it)
If any of these is missing in the paragraph, the relative order of remaining characters is nonetheless preserved. In some places there are duplicates in a row, e.g. d-d.
There are of course exceptions to this rule, but not many:
f2r, first paragraph, line 5 begins with o after the fourth line began with sh. After that, the order is restored.
f2r, second paragraph, line 4 begins with y after the third line began with q. After that, the order is restored.
f3v, line 5 begins with y after the fourth line began with q. After that, the order is restored.
f5v, line 6 begins with o after the fifth line began with sh. After that, no further lines are present.
f6v, line 4 begins with y after the third line began with q. After that, the order is restored.
f13v, line 4 begins with o after the third line began with q. After that, the order is restored.
Note that the types of exceptions are fewer in number than the exceptions themselves: it is either o after sh, y after q, or o after q.
What's remarkable, is that all these exceptions occur no earlier than in the fourth line of a paragraph. What's even more remarkable, there never is more than one exception in a paragraph - that is, kinda auto-repair (of the broken order) is in place.
Two quires are perhaps too little for generalizations. I'm looking forward to check further quires.