Subject: summary : buccalization

several weeks ago i posted a query asking for instances of " buccalization " , the development of a glottal stop into an oral stop . the reason for the query was that i ' m compiling a catalogue of lenition and fortition types for a textbook , and this was the only type in my catalogue i could n't find an example of . i received seven responses . none of them turned up an absolutely ironclad example of a glottal stop developing into an oral stop , though the things that did turn up were nonetheless striking . four languages were cited . maru , a tibeto - burman language , has turned syllable-final zero into [ t ] or [ k ] , depending on the preceding tone . it is possible , but not certain , that this change proceeded via glottal stop . ( robins burling , 1966 , ` the addition of final stops in the history of maru ' , _ language _ 47 : 581-586 ; anatole lyovin , 1968 , ` notes on the addition of final stops in maru ' , _ project on linguistic analysis _ 7 ( berkeley ) . ) mandarin chinese optionally allows syllable-initial zero to be realized as any of several segments , including a glottal stop , a velar nasal , or a velar or uvular voiced continuant . ( yuen - ren chao , _ a grammar of spoken chinese _ , p . 20 . ) there is reason to believe that some of these initial zeros derive from earlier glottal stop . winnebago has undergone the change [ - r ? ] - > [ - t ? - ] between vowels , and , if i understand the reply correctly , the rhotic itself may be epenthetic in origin . american english has its celebrated case of ` no ' > ` nope ' , possibly via glottal stop . the same is true of ` yep ' , if this derives directly from ` yeah ' and is not analogical . ( and i have noted that i myself sometimes have ` welp ' for ` well ' . ) that 's it . it really does look as if the glottal region is a vast sink from which no segment ever returns . it is not obvious why this should be so , since , as one respondent points out , the development of [ ? ] to [ p ] , [ t ] or [ k ] under the influence of neighboring [ u ] , [ i ] or [ a ] does not seem intrinsically implausible , and indeed it is reported that early european linguists working in southeast asia sometimes misheard and mistranscribed glottal stops in exactly this manner . my thanks to richard coates , lance eccles , james kirchner , bill mahota , john koontz , david solnit and scott delancey for their responses . perhaps i should also have inquired about cases of [ h ] > oral segment , but i did n't think of it . any further information in this vein will be gratefully received . larry trask cogs university of sussex brighton bn1 9qh uk larryt @ cogs . susx . ac . uk
