Michael Beddow wrote:

> 
> The classic example of this, long before TEI -- or SGML -- was thought of,
> was John Burrows' encoding, in a database-like format, of Jane Austen's
> novels. [see J.B., Computation into Criticism, Oxford 1987]. 

This is indeed the locus classicus for the kind of markup Martin has in 
mind. It doesn't actually predate the TEI that much, and doesn't predate 
SGML at all. More interestingly, perhaps, I can now reveal that I 
personally wasted several weeks of my life on the thankless task of 
converting that "database-like format" into kosher TEI SGML shortly 
afterwards. I say "thankless" advisedly, because OUP then decided they 
didn't see the point of publishing electronic texts anyway.

Not a problem for TEI markup of course: just use the "who" attribute on 
<sp>.

...

 > he got statistically stronger differentiations by considering
> the frequency and collocations/colligations of such "stopwords" in the
> speech assigned to specific characters than he did from more apparently
> "characteristic" stylistic or lexical features.

I'm surprised to see someone generally so careful with language as 
Michael implying that "collocation" and "colligation" are 
interchangeable technical terms. Most people working in corpus 
linguistics these days tend to distinguish them rather sharply. E.g.
"Colligation is a type of collocation, but where a lexical item is 
linked to a grammatical one. Surprising, amazing and astonishing are 
nearly synonymous. We can say it is astonishing/suprising/amazing, but 
we tend to say it is not surprising and not the others- surprising 
colligates with the negative." 
(http://www.usingenglish.com/glossary/colligation.html)