An email from Malte Liesner alerts me to an imprecise formulation. On pg. 58 in discussion the pronunciation of Latin /t/ I write "t was probably a true dental stop and not an alveolar as in English, to judge from the evidence of the Romance languages." In fact, the way I use dental vs. alveolar, although traditional, is not quite correct. In Spanish, French, Italian and Romanian /t /and /d/ are generally produced as denti-alveolar laminals. The primary place of articulation is the alveolar ridge, but the active articulator is the blade (lamina) of the tongue not the tip. Because the blade is used to make the closure the tip of the tongue may be visible at the teeth. In English, in contrast, for most speakers the tongue tip is used for
/t /and /d/ and hence these are alveolar apicals. Since the Romance languages agree on this laminal articulation I assume that it was simply inherited from Latin.
Also on pg. 55 I give the date of Sturtevant's The Pronunciation of Greek and Latin as 1920, but that is the first edition. In the bibliography I give the 1940 date of the second edition.Image from http://sail.usc.edu/~lgoldste/General_Phonetics/Constriction_Location/index.html