Subject: query : chinese grammatical markers

i ' m interested in any information regarding the diachronic development of the following grammatical markers in various chinese languages ( apologies for the email-mutilated transcriptions ) : ( 1 ) mandarin : [ de ] ( the " genitive " / " relative " marker ) ( 2 ) mandarin : [ ge ] ( the general numeral classifier ) ( 3 ) shanghainese : [ ge ] ( the " genitive " / " relative " marker ) ( 4 ) hokkien : [ e ] ( a ) ( the " genitive " / " relative " marker ) ( b ) ( the general numeral classifier ) ( 5 ) teochew : [ kai ] ( a ) ( the " genitive " / " relative " marker ) ( b ) ( the general numeral classifier ) ( 6 ) cantonese : [ ge ] ( the " genitive " / " relative " marker ) [ go ] ( the general numeral classifier ) [ dik ] ( the ligature ) more specifically : question 1 : which of the above markers are historically / etymologically related to which others , ie . deriving from a diachronically reconstructable common source ? ( eg . are mandarin [ de ] and cantonese [ dik ] diachronically related ? , etc . etc . ) question 2 : ( specific to hokkien and teochew ) most native speakers of hokkien and teochew , who are also speakers of mandarin , are of the opinion that there are two distinct markers exhibiting accidental homophony : a " genitive " / " relative " [ e ] / [ kai ] corresponding to mandarin [ de ] , and a " classifier " [ e ] / [ kai ] corresponding to mandarin [ ge ] . question 2a : is it demonstrably the case that [ e ] and [ kai ] are the respective products of diachronic coalescence of distinct forms at some earlier stage of the language ? question 2b : are there any valid synchronic arguments for positing two distinct lexical items , ie . two [ e ] 's in hokkien , and two [ kai ] 's in teochew ? ( i consider the choice of characters conventionally used to represent these markers as irrelevant . ) note : i am a general linguist , not a sinologist , and i do n't read or speak chinese . i would greatly appreciate either your own views on the above questions , or references to published works addressing these questions . however , if the references happen to be in chinese , i would be extremely grateful if you could provide a synopsis of their contents . thanks , david gil national university of singapore ellgild @ nusvm . bitnet
