[KS] Languages in Korea

michael Robinson robime at indiana.edu
Fri May 27 09:39:58 EDT 2005

Dear List: 

I fear Professor Bentley is being unfair to my colleague 
CHris Beckwith in his recent posting. 

Beckwith actually does not "claim that Japanese is a
Koguryo language" but that Japanese is genetically
related to Koguryo (more precisely, he says that the
Japanese-Ryukyuan and Puyo-Koguryoic language families
are genetically related), which is very far from being
the same thing. Bentley's remarks about Beckwith's
supposed "lack of knowledge about Japanese historical
phonology" are uncalled for. It would be sufficient to
say that Beckwith's views about Japanese historical
phonology are different from Bentley's, and to argue
specific points to show how Beckwith's forms are wrong
and Bentley's are right. 

Bentley says of Beckwith's etymology of Japanese _me_
'eye', "It is unclear why this etymology is brought
up, as he has no Koguryo word for 'eye' in his
database, unless this is simply an attempt to
ultimately relate Japanese back to Chinese." But
Beckwith's goal is not unclear, it is openly
proclaimed in the chapter title and the beginning of
the chapter. His aim is to show that the languages
discussed share some vocabulary and some other
developments, the aim being to try and find where the
Proto-Japanese-Koguryoic people might have come from.
He clarifies the purpose of the etymologies on the
second page of the chapter, where he says, "The
phonological characteristics of most such words
indicate they were borrowed into Japanese-Koguryoic
from Chinese and other languages..." The same applies
to Beckwith's etymology of Japanese _ti_ 'blood,
milk', which is in the same chapter. Whether or not
Beckwith's etymologies are correct may be debatable,
but Bentley claims he _knows_ what is correct: "The
true etymology of 'eye' would be *ma-i or perhaps
*ma-Ci." Bentley's discussion is restricted to two
words which are not evidence of the Japanese-Koguryoic
theory presented in the book (and are not presented as
such), so he is criticizing Beckwith for trying to do
something with them that he is not trying to do. How
can they then cause any 'suspicion' about the book's
goals and conclusions, which are entirely different? 

I know that Beckwith's book is a difficult read
(especially for a non-linguist), but it is an
important book. It, and Beckwith, deserve better

Mike Robinson

  ----- Original Message ----- 
  From: John Bentley 
  To: Koreanstudies at koreaweb.ws 
  Sent: Wednesday, May 25, 2005 9:08 AM
  Subject: Re: [KS] Languages in Korea

  Mr. Atkinson brings up an interesting though controversial question.

  Unfortunately, I don't think there is any one easy answer. Scholars
  are generally divided on the question, not because of the methodology
  so much as the lack of data on previous historical polities and
  the languages spoken by those people.

  This question posed by Mr. Atkinson perhaps needs to be defined
  a bit better, as the terminology is part of the issue. The difference
  concerns what scholars consider a language and what they consider
  a dialect. Leaving politics aside, most scholars at least pay lip service
  to the idea that a language is mutually unintelligible to another, while
  a dialect differs from another dialect in some vocabulary or morphology,
  but is still intelligible to speakers of a related dialect. A dialect is a
  handy word, because each dialect is thus a part of the language family.
  This is a helpful tool for national unity and what not. This basic
  definition of language and dialect, however, would turn many European 
  *languages* into mere dialects, while many dialects in China would 
  now become full-blown languages. And there's your problem.

  In Japan linguists call the Ryukyuan languages dialects, even though
  these are unintelligible to any mainland Japanese speaker. In fact,
  as a Ryukyuanist, I believe there are at least four 'languages' in the
  Ryukyus. But the term 'language' implies nationality, and that opens 
  up a can of worms, so people agree to call Ryukyuan a set of dialects.

  I am not sure where Mr. Atkinson's friend came up with the number
  six, but if we stick to a basic linguistic definition of language, we
  may be able to claim anywhere from five to twenty languages on the
  peninsula during the three kingdom period. The problem is that some
  who make claims about what language was what on the peninsula 
  are not being completely open and honest, as we have no written records 
  *in the languages* of Paekche, Koguryo, Kaya (Kara), Silla, or peninsular Wa. 
  So how do they know? I have written about the language of Paekche, and 
  have provided information and etymology for roughly 80 words from the 
  Paekche corpus (Bentley, New Look at Paekche Korean: data from Nihon
  shoki, Language Research (Ohak yon'gu), vol. 36.2, 417-443). But we have 
  very little surviving data on morphology or syntax or other data that would help
  us see Paekche as a language, and not merely a list of words. My own work
  has suggested that Paekche and Silla were related 'languages', but that
  is simply a scholarly hypothesis, not a fact. I don't go around claiming that
  this is proven.

  Lately Chris Beckwith has come out with a very provocative book titled,
  Koguryo: the language of Japan's continental relatives (Brill 2004). It is
  not my intention to review the book here, but suffice it to say this
  relationship (the claim that Japanese is a Koguryo language) is based
  on the tenuous comparison of about 140 Koguryo etyma. To me this is like
  looking at the skin of an elephant through a microscope and trying to
  guess what the animal is. It is not impossible, mind you, just highly
  difficult, and requires great skill.

  Just to give one reservation I have with Beckwith's work: it bothers me
  to see his lack of knowledge about Japanese historical phonology (which 
  forms half of this theory). Let me give just two examples. Beckwith 
  reconstructs Proto-Japanese-Ryukyuan 'eye' as *mika or *miak (2004:157-58).
  This is based *entirely* on the Hateruma island word miN (N is a velar nasal) 'eye'. 
  He sees the velar nasal going back to a velar -k. This is then compared with Old 
  Chinese *mek 'eye' and Old Tibetan myig 'eye'. Gary Oyler, in 1997, did an MA
  thesis on the problem of -N in Hateruma, and concluded that the -N is secondary.
  It only occurs word finally, and there is no pattern to which words have the N and
  which do not. It is completely random. My own work has found this same phenomenon 
  in Yonaguni, an island not far from Hateruma, but the N is attached to different 
  nouns than those in Hateruma (so the development of this nasal was independent
  on the two islands). This velar nasal is simply a relic of morphology that the speakers 
  have reanalyzed as part of the noun, kind of like American speakers spelling hafta
  (< have to), where they treat two original words as one. It *cannot* be reconstructed 
  as part of the proto-form. The true etymology of 'eye' would be *ma-i or perhaps *ma-Ci. 
  So Beckwith also has the vowels wrong. This destroys the comparison with Chinese
  and Tibetan. It is unclear why this etymology is brought up, as he has no Koguryo 
  word for 'eye' in his database, unless this is simply an attempt to ultimately relate
  Japanese back to Chinese.

  Another problem is his grouping of homonyms. He reconstructs a Proto-Japanese 
  form *tiu 'liquid' based on ti 'blood, milk' (2004:154). The accent for ti 'blood' and 
  ti 'milk' are different, so I seriously doubt these two words have a common etymology.
  Also internal reconstruction suggests that 'blood' may go back to *to-i, while 'milk' goes 
  back to an earlier *tu-i (cf. Sam Martin's Japanese Language Through Time, Yale Univ.,
  1987:545). Again, Beckwith provides this etymology to make a connection with
  Chinese, but as all his Chinese cognates have /i/ as the nuclear vowel, these etymologies 
  also must be rejected.

  Thus, if Beckwith's grasp of Japanese historical phonology has serious weaknesses,
  then I am suspicious of his other conclusions. This does not mean his theory is wrong, 
  mind you, but it has not solved the problem, in my opinion. I know others have claimed
  that Koguryo and Japanese may be related, and I do not intend to refute that claim. All 
  I want is for the work to be well-grounded in the historical linguistic methodology 
  (as well as phonology), and take account of what we already know about Japanese 
  historical phonology and its development. It is sad there are so many who really do not 
  understand the languages of Ryukyuan (but quote etyma from these languages as
  proof of X or Y), as these hold the key, I believe, to helping us better understand
  the relationship of Japonic (Japanese and Ryukyuan) to its neighbors.

  In the end, Mr. Atkinson should ask his friend who the source of this claim
  about a) the peninsula had six languages, and b) one migrated to Japan is.
  We should remember that while Chinese and peninsular records talk about
  the Wa, we do not even know what language the Wa spoke. It could have
  been Ryukyuan, or Western Japanese, Eastern Japanese, or Central Japanese.
  Or it could have been a language that eventually died out. There are so many yet-to-be-
  answered questions regarding this thorny debate. But that is probably what makes
  it so popular as a topic of dicussion.

  John R. Bentley
  Northern Illinois University
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://koreanstudies.com/pipermail/koreanstudies_koreanstudies.com/attachments/20050527/b3888c32/attachment-0002.html>

More information about the Koreanstudies mailing list