Tangentially, there are studies that state that the information rate (information per time) is about the same in all languages, and languages with lower information density per word are spoken slower.
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.433...
(a lot of articles provide summaries, eg http://www.scientificamerican.com/article/fast-talkers/)
They cite three others:
http://www.researchgate.net/profile/John_Locke5/publication/...
http://www1.icsi.berkeley.edu/~steveng/2000/PDF/UninvitedGue...
I can't find access to the third one without digging since it's published in conference proceedings from 1961, but there's a talk with access (wow!): http://www.helsinki.fi/speechsciences/history/snds/Karlgren....
Further citation-following yields more books I cannot get a hold of (see the citations in the papers if you want to digg).