This dissertation finally establishes phonetics as part of linguistics. This conclusion might not surprise you, but it denies the foundational claims of some radically-minded, self-hobbling numbskulls, to wit, Chomsky and Halle in the Sound Pattern of English (SPE), and then after them most of the world of Academic, Let's Keep Ourselves Separate from the Rest of Reality, Linguistics, who blindly followed their quite stupid -- oh, let's just say arrogantly ignorant (or maybe to put it kindly they just don't have very good hearing) -- view that Language is (exclusively) Discrete.
Allow me to argue, again, here, this central conclusion.
The fact, established in detail through cross-dialect production studies in this dissertation, is that people control smoothly and precisely in production (and are evidently sensitive in perception) to some essentially continuous, phonetic dimensions of language production; they manage this control quite differently in dialects which are essentially the same phonologically. There may be no empirical basis to distinguish two dialects from each other in their category-based, discrete sound systems, while the two dialects implement corresponding or essentially the same categories with quite different phonetic realizations.
It is no longer the null hypothesis that linguistic knowledge is discrete. Just as whistling can swoop continuously across frequencies in patterns, which cannot be modelled accurately using the discrete diatonic scale, similarly languages and dialects select sets of sharply defined places on the continua of phonetic production to use as the targets for their pronunciations of vowels. Why? Because humans control these continuous dimensions precisely and that control is part of their linguistic knowledge.
Just as musicians in the 1500s were able to replace previous twelve-notes-per-octave scales with the innovation of equal temperament through an exact tuning of the frequencies of the various notes; similarly, dialects are able to exactly tune the physical sounds that implement in articulation and acoustics the discrete categories of phonology. These tuning choices are part of the linguistic knowledge of speakers of a particular dialect. They specify the detailed characteristics of the "accent" with which that dialect or language is spoken.
Thus linguistic phonetics is hereby established as a new and undeniable subcomponent of linguistic theory per se; phonetics is not the mere universal and mechanical implementation of phonology, but a specifically linguistic (language- and dialect-specific) domain of learned, arbitrary, conventional, patterned behavior that forms a layer of the language production (and perception) system.
In short, linguistic phonetics, or the study of accent, is part of linguistics. So let's stop pretending that it's not.