k/t gallows reduplication

[pfeaster]

The plot is only qualitatively comparable with Patrick's, since benched gallows cth ckh are excluded from his plot but included in mine.

For what it's worth, my corresponding plot in that paper likewise includes the benched gallows -- it uses the same selection criteria as the preceding Table 3, which I should probably have clarified.

I suppose one line of speculation might be that the "preferences" of [k] and [t] for different parts of the line increase the probability of continuous blocks of text occurring in which only one or the other of these glyphs appears -- but of course it's frustratingly difficult to tell what's the cause and what's the effect.