Minimal Old Chinese and Later Han Chinese: A Companion to Grammata Serica Recensa

PREFACE

PREFACE Bernhard Karlgren's Archaic Chinese (= Old Chinese, OC) as presented in his standard work Grammata Serica Recensa (GSR) of 1957 has long been outdated. This present manual is an attempt to update GSR with a relatively simple "Minimal Old Chinese" (OCM, for OCMinimal ) which incorporates those features on which there is broad agreement among investigators today. Though this manual can be thought of as a Grammata Serica Recensa update, I will refer to it as Grammata Serica Companion (GSC) for short. The perception of the mysterious nature and confusing state of OC has been shaped by several factors. First, GSR does not present the data in a transparently organized fashion, so that only patient scrutiny reveals what the OC system is, behind Karlgren's bewildering diacritics and phonetic symbols. The ordinary user has to take Karlgren's, or anyone's, authority at face value. Secondly, the experts' arguments are so specialized and arcane that only the initiated are in a position to follow them. The Introduction to this manual attempts to provide an overview over some of the terms and issues, demystify OC if you will, so that outsiders may have some notion of the data, sources and theories on which expert arguments are based. Thirdly, superficially scholars do not seem to agree on much, because they debate unclear issues and not the many features of OC on which there is a tacit consensus. Furthermore , old settled issues in Middle Chinese (MC) and OC phonology are periodically raised again so that the non-expert must conclude that almost all about OC is still up in the air. Since the publication of GSR, historical linguists have tried to simplify and systematize Karlgren's reconstructions, have suggested emendations or their own OC systems which sometimes look as different from each other as if they were different languages. Compare, for example, (P: = Pulleyblank): Karlgren GSR 'today' 4 kL~m 'remember' ~ ni~m 'offspring' r tsi~g 'plum' $ 1i~g 'offense' ml dz'w~d 'think, be' 'liE dLw~r 'little bird' 1E ~Lw~r Baxter 1992 k(r)j~m nims tsj~? rj~? dzuj? wjij Sagart 1999 -im [am~_]anem-s "tsi? bt(~)-wij btU[j] Pan 2000 OCM krwm bm mqlUIms nfms splw· ts~? b-rw· r~? sblul· dzui? wi P: kwj~l tui After Karlgren, the field seems to have fallen apart. The occasional user of OC material probably finds it difficult or impossible to tell which OC proposals are just an author's latest theories, ideas and speculations, and what is actually widely accepted.) For the purposes of OCM, we will attempt to separate generally agreed OC features from more speculative and probing hypotheses - however valuable and insightful they may be - which are often presented in such definitive language that an unsuspecting reader may think he now has the OC language in front of him to work with. I Witness comments like this by the Indo-Europeanist Douglas Adams: "There are a number of 'competing' systems of [OC) reconstruction (Karlgren, Pulleyblank, Li) whose inherent likelihood and mutual interrelationships can baffle the outsider" (JIES 23, 3e4, 1995: 401). ix PREFACE The Introduction outlines basic issues in OC phonology since GSR, and the rationale for OCM, a relatively simple form of ~C, a minimum on which most investigators may agree, and which shows that OC is not quite as enigmatic and complex as it often appears. The OCM forms are "minimal" in several respects: they incorporate only the more widely accepted insights into OC gained since GSR was published, but leave out more speculative proposals with their often complex OC reconstructions; OCM is based on simple and less complex hypotheses and assumptions than some other proposals (see Intro. sections 6, 8 and 9); OCM is written in a simple form, similar to recordings of modern dialects. Unfortunately, these objectives need to be compromised on occasion because it is necessary to decide a detail on which there is no consensus, in order to be able to put something on paper. But these doubtful cases will be clearly pointed out. The user will find, it is hoped, a simple, transparent form of OC that may be useful. This is the essence of this endeavor. As pointed out in section 9.2.5 ofthe Introduction: The user of this manual can add phonemic elements to OCM forms as he may deem appropriate, but it is difficult, even impossible, for an unsuspecting user to visualize an initial cluster presented to him with a questionable...