The 'Chinese' Theory: For and Against

[Jorge_Stolfi]

But what I said in the post above was that I see the exact opposite:  final m m seems underrepresented at the ends of paragraph that we can see clearly (i.e. those with clear indents).  It does not seem common at paragraph end, let alone more likely to occur at paragraph end.  It underperforms.  Quite badly.   

What I meant was that m can be used as evidence of end-of-paragraph when conditions (1.) and (2.) are false: that is, when the next line has no p or f, and has full width.

But I am afraid that you are right.  This criterion is wrong.

Sigh.  I have to check again all breaks in the SPS.  Fortunately most of them (I think) are determined by criteria (1.) and/or (2.).  For the others, I will have to rely on (4.) (wider interline spacing) and/or (5.) (near a star) and/or (6.) (it is the last line on the page).