CC BY
103
Yasunori Takeuchi
Otani University, Kyoto
DIRECTION TERMS IN KHITAN1 1. Introduction
Khitan is a language which was once used by the Khitan people, who established the Liao Dynasty (907-1125) in Eastern Eurasia. The Khitan language is thought to be related to the Mongolic languages [Janhunen 2003] and its sources have been preserved in Chinese transcriptions and Khitan scripts, comprised of two distinct writing systems: Large Script and Small Script. With large numbers of Khitan epitaphs having been unearthed, significant progress has been made in deciphering them. However, the work is still far from complete.
In this study, we examine direction terms in Khitan2. The previous studies presumed the readings of these word forms by referring to their equivalents in Mongolian3. However, Khitan has characteristics different from those of Mongolic languages and it is impractical to make direct assumptions about Khitan based on Mongolian. Therefore, it is important to reconstruct the word forms through a philological analysis, without a speculative reconstruction based on simple comparison. In addition, Khitan has several written forms capable of expressing the same directional meanings, but it is not fully understood whether they were differentiated and, if so, how. Accordingly, in this study, we begin by reconstructing the word forms. Then, on the basis of these word forms, we demonstrate that Khitan has a derivational suffix -d and discuss its possible cognate in Mongolian.
1 This study was supported by a Grant-in-Aid for JSPS Fellows (23-9506).
2 For the previous studies, see Toyoda [1992], Aisin Gioro [2004: 65-67], and Wu Yingzhe [2004].
3 Toyoda [1992] suggested the same word forms with Mongolian for Khitan as dorona 'east,' hörene 'west,'emüne 'south,' and ümere 'north'.
2. Direction terms in Khitan
A summary of the Khitan direction terms, as identified in previous studies, is shown in (1).
(1) a. ft 'east'
b. XM ~AM 'east'
c. + 'west'
d. 'west'
e. h ' south'
f. 'south, lower'
g. — 'north'
h. 'north'
i. AM 'upper'
j. 'middle'
k. 'left'
l. 'right'
To reconstruct the word forms, it is first necessary to uncover the phonetic values of each character. Recent studies of the phonetic values of Khitan characters include Kane [2009], Chinggeltei [2010], and Yoshimoto [2012]. We examine each form below.
2.1. ft 'east'
Phonetic values in the previous studies: Yoshimoto [2012] <dor> Chinggeltei [2010] <dor> Kane [2009] <deu>
Kane [2009: 43] regarded ft as an allograph of % <deu> and applied the phonetic value <deu> to ft , while Yoshimoto [2012: 138] pointed out that ft and % were different graphemes. Yoshimoto [2012] and Chinggeltei [2010] likewise posited the phonetic value <dor> for ft , based on the assumption that ft and ^"ifc (both of which mean 'east') have the same word form. We agree with Yoshimoto [2012] that ft and % were different graphemes. However, as we indicate in Section 3, ft and ^"ifc do not share the same word form, so it is incorrect to apply the phonetic value <dor> to ft .
2.2. XM (AM ) 'east'
Phonetic values in the previous studies: Yoshimoto [2012] <umu-ur> (<do-ur>) Chinggeltei [2010] <ju-ru> (<t-ru>) Kane [2009] <yu-ur> (<ud-ur>)
X and A were previously considered to be different graphemes. However, it is difficult to distinguish clearly between the two characters as Wu Yingzhe and Janhunen [2010: 63-64, 80] assert and it is more rational to consider X as an allograph of A . In the following, we examine the phonetic values of X (A ) as well as M .
In the previous studies, AM was considered to have a similar form to WM doruna 'east', with the phonetic value <do> being applied to A and <ur ~ ru> to M .
As for the phonetic value of X , according to Aisin Gioro and Yoshimoto [2011: 128], the person's name written as X in the Khitan epitaph was transcribed as Wumei (EM4 u mu) in Liaoshi and X was given the phonetic value <umu>. In addition, Shen Zhongwei [2009] conducted an analysis of rhymes in Khitan verses and found that X rhymed with # . Regarding the phonetic value of # , according to Ji Shi [1996: 249], the person's name written as
, seen in the Khitan epitaph, correspond to the Chinese transcription Tumeilitongwa H^M^^ (EM thu mu li thu\7 ua) and # was given the phonetic value <thum>. Thus, X can presumably be pronounced as <um>.
Some previous studies observed the phonetic value of M by comparing the Khitan ordinal numbers with numerals in Mongolian5.
(2) a. ^M^ <ed3-[?]-er> *dj[?]er 'second (masculine)' (cf. WMjirin 'two (mainly used for females)')
b. >IM<& <aq-[?]-er> *q[?]er 'third (masculine)'
(cf. WM yurban 'three')
c. <ed-[?]-er> *d[?]er 'fourth (masculine)' (cf. WM dorben 'four')
4 Early Mandarin forms are from Ning Jifu [1985].
5 Studies on the numerals of Khitan include Ji Shi [1986], Chinggeltei [1997], Janhunen [2003: 399-400], Aisin Gioro [2004: 70-71, 179-187], and Wu Yingzhe [2007: 126-175].
In the previous studies, ^M was considered to correspond with jir- in Written Mongolian, while >IM corresponded with yur-, and ^M with dor-. Naturally, the phonetic value containing <r> was applied to M .
However, M was used in the transcription of the name of the ethnic group ❖ЙМЛ <ed-aq-[?]-in> *da^[?]-/-n Tangut-E-GEN 'of Tangut' (cf. MM6 tangyut 'Tangut') in the epitaph of Xiao Zhonggong
l.14 [Ji Shi 1996: 127]. Thus, it was more rational to consider its phonetic value as <ud> instead of <ur>. If we assume the phonetic value <ud> for M , we can say that the forms of the ordinal numbers in Khitan could be considered close to the forms of Written Mongolian and Middle Mongolian, as shown below, supporting this hypothesis.
(3) a. <ed3-ud-er> *djuder ~ djude.r 'second (masculine)'
(cf. WMjituger 'second')
b. >IM<& <aq-ud-er> *quder ~ qude.r 'third (masculine)'
(cf. WM yutuyar, MM quta'ar ~ qutu'ar 'third')
c. <ed-ud-er> *duder ~ dude.r 'fourth (masculine)' (cf. WM dotuger 'fourth')
Based on the above discussion, the written form X M was assumed to be *umud.
2.3. + 'west'
Phonetic values in the previous studies: Yoshimoto [2012] <[?]> Chinggeltei [2010] <or(?)> Kane [2009] <WEST>
It is clear that the meaning of + is 'west', but there are no direct clues as to its pronunciation.
2.4. 'west'
Phonetic values in the previous studies: Yoshimoto [2012] <s, sa, as-i> Chinggeltei [2010] <s-i> Kane [2009] <s-i>
Each character used to transcribe 'west' was figured out based on the correspondences with Chinese transcriptions and there is fairly
6 Middle Mongolian forms are from Kuribayashi [2009].
general agreement that <is> and <i> were the phonetic values for and ^ . Thus, this word form can be reconstructed as *si, suggesting that it was borrowed from the Chinese word xi H (EM si) 'west.'
2.5. * 'south'
Phonetic values in the previous studies: Yoshimoto [2012] <d^r> Chinggeltei [2010] <dele> Kane [2009] <sio>
According to Bao Yuzhu [2006] and Aisin Gioro [2010], *>l corresponds to the tribal name Diela (EM tie la). As Wu Yingzhe [2010] pointed out, *>l was also written as , and we can infer
the phonetic value of 'south' based on these three characters.
£ <id>, # <sr> and # <sq> obtained their reliable phonetic values from the correspondences with Chinese transcriptions.
Assuming that *>l <[?]-aq> and £# # <id-sr-sq> are transcriptions of the same word, we can reconstruct its form as *dereq, giving the phonetic value <dsr> to * .
Thus, it is appropriate to reconstruct * <dsr>'south' as *der.
2.6. ~ ~ 'south, lower'
Phonetic values in the previous studies:
Yoshimoto [2012] <ad-^r>, <9d-j^-^r>, <9d-j^-^r-t,d>
Chinggeltei [2010] <t, th-le>, <t, th-e, ie-le>, <t, th-e, ie-li, la-th, t>
Kane [2009] <t-ge>, <t-ie-ge>, <t-ie-ra-d>
Basically these transcriptions were used to mean 'south' in Khitan texts, but the difference with * <dsr> *der (it also means 'south') leaves room for a variety of interpretations. Ji Shi [1996: 11, 29] pointed out that the original meaning of was 'lower', suggesting a different meaning of 'south' in * <dsr> *der. In contrast, Bao Yuzhu [2006: 11] pointed out that and * had the same word form.
Among these characters, ^ <ed>, <s>, # <sr>, and £ <id> obtained their reliable phonetic values from correspondences with Chinese transcriptions. Also, if # was thought to correspond with #£ <sr-id> or <s-sr-id>, it is reasonable to consider that # was expressed
as <srd>. From these assumptions, it can be said that the word form of ^fi <ed-srd>, <ed-s-srd>, and <ed-s-sr-id> is *dsrd.
2.7. — 'north'
Phonetic values in the previous studies: Yoshimoto [2012] <[?]> Chinggeltei [2010] <xoi> Kane [2009] <ABOVE, NORTH>
It is known that — meant 'north', but there are no direct clues as to its phonetic value.
2.8. — £ 'north'
Phonetic values in the previous studies: Yoshimoto [2012] <[?]-t, d> Chinggeltei [2010] <xoi-t\ t> Kane [2009] <ABOVE, NORTH-d>
— £ <[?]-id> was sometimes used to mean 'north'. Its difference in usage with — is discussed in Section 3.
2.9. AM 'upper'
Phonetic values in the previous studies: Yoshimoto [2012] <u-ur> Chinggeltei [2010] <u-ru> Kane [2009] <ü-ur>
For A <u>, a reliable phonetic value is obtained from its correspondence with Chinese transcriptions. As for M <ud>, refer to 2.2. Thus, the word form of AM <u-ud> can be inferred as *u:d.
2.10. ЯЪ 'middle'
Phonetic values in the previous studies: Yoshimoto [2012] <dau-du, ud> Chinggeltei [2010] <ta-th, t> Kane [2009] <dau-dû>
For Я <dau>, a reliable phonetic value is obtained from its correspondence with Chinese words. Ъ at times alternates with ^ <ord>, just as and £ ^Ъ* 'Chongxi МШ
(reign title)'. Thus, it can be considered that the phonetic value of ^ is close to ^ <ord> which contains the phonetic value <rd>. Wu Yingzhe and Janhunen [2010: 87] noted the possibility that the phonetic value of ^ was <rdu>. Since ^ follows K <dau>, we can make a reasonable prediction that the phonetic value of ^ is <urd> which starts with the vowel <u>. Accordingly, we can conclude that the word form of K^ <dau-urd> is *daurd.
2.11. 'left'
Phonetic values in the previous studies:
Yoshimoto [2012] <tji-gi, ga-n>
Chinggeltei [2010] <tf, tjh-k-an>
Kane [2009] <c-g-en>
There is a consensus on the phonetic values used to transcribe this word. We reconstruct this word as *djigm, which certainly is a cognate of WMjegun 'left'.
2.12. 'right'
Phonetic values in the previous studies:
Yoshimoto [2012] <ab, b-ar, sr-a-an>
Chinggeltei [2010] <p-li, la-a-an>
Kane [2009] <b-ra-a-an>
As can be seen above, there is little disagreement on the phonetic values of other characters. We reconstruct this word as *bsran ~ bsra.n, which certainly is a cognate of WM barayun 'right'.
From the description above, the forms of the Khitan direction words can be summarized as (4).
(4) a. ^ *[?] 'east'
b. *umud 'east'
c. + *[?] 'north'
d. *si 'north'
e. * *der 'south'
f. *derd 'south, lower'
g. - *[?] 'north'
h. *[?]d 'north'
i. *u:d 'upper'
j. ** *daurd 'middle'
k. *d3igm 'left'
l. *bsran ~ bsra.n 'right'
As can be seen above, only the word forms of ^A — *djigin 'left' and *bsran ~ bsra.n 'right' display correspondences with those
of Mongolian. Others do not. Furthermore, it is noteworthy that ' <der> *der 'south' and ^fi <ed-erd> *derd 'south, lower' are not used to mean the same word, although they have the same root. It is possible that the form ^fi <ed-erd> *derd 'south, lower' consists of ' <der> *der 'south' plus a suffix -d. Similarly, there are sets of different words that are differentiated by the presence or absence of -d, such as — *[?] 'north' and — £ *[?]d 'north'. It is possible that the differences in the presence or absence of -d were used systematically in Khitan. In Section 3, we will discuss the differences between the usages with and without -d.
3. Derivational suffix -d
In this section, we examine the differences between word forms with and without the suffix -d. Examples in (5) are the word forms without -d.
(5) a.
<ord-u-und-i
*ordu-ndi court-ABL
h ft
der [?]> der [?] south east
'from the court to the southeast'
[the Epitaph of Yelu Renxian MWt
SH'fr l.33]
b. * h +
<en-ad-ab-ad-i der [?]> *nadaba-di der [?]
palace (^#)-ABL south west 'from the palace to southwest'
[the Eulogy for Emperor Xingzong l.2]
c. —
<[?]
*[?] north
ft
[?]
[?] east
ip-o'-or-in>
flo'or-l-n route-E-GEN
'of north east route
[the Epitaph of Yelü Zongjiao
sn^i
l. 13]
d. *
<der u-ud> *der u:d south division 'Southern Division (^1^)'
[the Epitaph of Xiao Dilu l.14]
Word forms with the suffix -d are shown in (6).
(6) a. Affi —
<ed-erd ig-ig-in> *der-d giy-i-n south-d capital-E-GEN 'of Southern Capital
[the Epitaph of Han Gaoshi l.20]
b.
<[?]-id ed3-ab-oq-on> *[?]-d djaboq-o-n north-d Zubu (®h)-E-GEN 'of Northern Zubu'
[the Epitaph of Yelu Renxian l.46]
c.
ЛЯФ
<[?]-Id ed-erd d3ur Ig-ur-is>
*[?]-d der-d djur gur-s
north-d south-d two country-PL
'two countries in the north (Liao) and south (Song)'
[the Eulogy for Emperor Xingzong l.10]
Examples in which the context can be comprehended sufficiently are rare, yet there seems to be a tendency in which forms with -d are used when they are followed by nouns, such as 'capital' or 'country.' This indicates characteristics similar to the suffix WM -dU, which forms an adjective that contains the meaning ofthe root [Poppe 1954: 41, 50; Ozawa 1997: 252], as shown in (7).
(7) a.WM dumda-du 'central' < dumda 'center' b.WM doruna-du 'eastern' < doruna 'east'
The suffix -d in Khitan should be considered a cognate of -dU in Written Mongolian. The same phenomenon can be observed in ft *[?] 'east' and %M (AM ) *umud 'east,' as shown in (8).
(8) a. ^ + ft
<iu-ul pu-us-u'-er [?]-in [?]>
*iul pusu'er [?]-i-n [?]
cloud blessing moutain-E-GEN east
'east of Qingyun Mountain (Mff^)'
[the Eulogy for Emperor Daozong l.6]
b. AM <um-ud
*umu-d east-d
Al.fi
ig-iq-in>
giq-i-n
capital-E-GEN
'of Eastern Capital (^M)'
[the Epitaph of Han Gaoshi
.25]
As in the examples of 'north' and 'south', if we assume that % M consists of ft and -d, then it may be possible to apply the phonetic value <umu> to ft .
In (9), we examine direction terms other than 'north', 'south', 'east' and 'west'.
(9) a. AM Affl-
<u-ud ig-iq-in>
*u:-d giy-i-n
upper-d capital-E-GEN 'of Upper Capital
[the Epitaph of Yelu Dilie l.25]
b.
<dau-urd
*daur-d middle-d
ib-eq>
beq son
'second son )' [the Epitaph ofHan Dilie l.9]
As can be seen in (9), A M *u:d 'upper' and K ^ *daurd 'middle' are followed by nouns and it can be inferred that -d found in them is likely to be a suffix.
In conclusion, we presume that Khitan is a fairly different language from Mongolian in that they do not share the same word forms of direction terms. In the meantime, it is suggested that the shared morphology of Khitan and Mongolian exemplified in the analyses of -d supports the genetic relationship between the two languages.
Abbreviations
abl — ablative; e — epenthetic sound; EM — Early Mandarin; gen — genitive; MM — Middle Mongolian; pl — plural; WM — Written Mongolian;
--morpheme boundary, character boundary; < > — phonetic value; < —
derived from (the left side is derived from the right side); [?] — unknown meaning, unknown phonetic value.
Bibliography
Aisin Gioro 2004 — Aisin Gioro Ulhicun. Qidan yuyan wenzi yanjiu [Research on the Khitan Language and Script]. Kyoto: Töa rekishi bunka kenkyukai, 2004.
Aisin Gioro 2010 — Aisin Gioro Ulhicun. Yaonian shi di nian xianzhi kehan yu taoweisi diela bu — yi qidan wen «gu zuo long hujun shang jiangjun zhengliang gongchen jianxiao taishi zhiyanyu changwen muzhi» wei zhongxin [Xianzhi Khan of the Yaonian clan and the Taoweisi Diela tribe: Mainly based on the Gu Zuo long hujun Shang jiangjun Zhengliang gongchen Jianxiao taishi, Zhiyanyu Changwen Muzhi in the Khitan Script] // Ritsumeikan bungaku 616, 2010. P. 1013-1024. Aisin Gioro, Yoshimoto 2011 — Aisin Gioro Ulhicun, M. Yoshimoto. Kan hantö kara nagameta kittan joshin [Khitan and Jurchen as viewed from the Korean Peninsula]. Kyoto: Kyoto University Press, 2011. Bao Yuzhu 2006 — Bao Yuzhu. Qidan xiaozi A ji qi tihuan zi yanjiu [The study on the Khitan Small Script A and its substitute] // Neimenggu daxue xuebao [Journal of Inner Mongolia University] 1, 2006. P. 8-12. Chinggeltei 1997 — Chinggeltei. Qidan yu shuci ji qidan xiaozi pin dufa [Khitan numerals and graph combinations in the Khitan Small Script] // Neimenggu daxue xuebao [Journal of Inner Mongolia University] 4, 1997. P. 1-9. Chinggeltei 2010 — Chinggeltei. Qingge'ertai wenji di 5 juan qidan wenzi yanjiu [Collected Works of Chinggeltei. Vol. 5. Studies on Khitan Literature]. Chifeng: Inner Mongolia Science & Technology Press, 2010. Ji Shi 1986 — Ji Shi. Guanyu qidan shuci duyin wenti [The question of the pronunciation of Khitan numerals] // Neimenggu daxue xuebao [Journal of Inner Mongolia University] 4, 1986. P. 89-100.
Ji Shi 1996 — Ji Shi. Milin wenjing — qidan xiaozi jiedu xincheng [Ask for the Road in the Mystic Forest: A New Process of Interpretation of the Khitan Small Script]. Shenyang: Liaoning minzhu chubanshe, 1996.
Janhunen 2003 — J. Janhunen. Para-Mongolic // J. Janhunen (ed.). The Mongolic Languages [Routledge Language Family Series 5]. London — New York: Routledge, 2003. P. 391-402.
Kane 2009 — D. Kane. The Kitan Language and Script [Handbook of Oriental Studies 19]. Leiden — Boston: Brill, 2009.
Kuribayashi 2009 — H. Kuribayashi. 'Gencho hishi' Mongoru-go kanji on'yaku boyaku kango taisho goi / Word-Index to the Secret History of the Mongols with Chinese Transcriptions and Glosses [CNEAS Monograph Series 33]. Sendai: S. n., 2009.
Ning Jifu 1985 — Ning Jifu. Zhongyuan yinyun biaogao [The Tables of the Zhongyuanyinyun]. Changchun: Jilin wenshi chubanshe, 1985.
Ozawa 1997 —S. Ozawa. Mongoru-go bungobunpo kogi [A Lecture on the Grammar of Written Mongolian]. Tokyo: Daigakusyorin, 1997.
Poppe 1954 — N. Poppe. Grammar of Written Mongolian [Porta Linguarum Orientalium. Neue Serie 1]. Wiesbaden: Harrassowitz, 1954.
Shen Zhongwei 2009 — Shen Zhongwei. Qidan xiaozi yunwen chutan [Preliminary investigation into the epitaphic rhymes written in the Khitan Small Script] // Minzu yuwen 3, 2009. P. 37-50.
Toyoda 1992 — G. Toyoda. Guanyu qidan xiaozi de fangwei he yixie shuci [On the directions and some numerals in the Khitan Small Script] // Zhongguo minzushi yanjiu [Chinese Ethnics Research] 4, 1992. P. 71-81.
Wu Yingzhe 2004 — Wu Yingzhe. Guanyu qidan xiaozi zhong de fangwei mingcheng «dong» [On the direction word for «east» in the Khitan Small Script] // Neimenggu daxue xuebao (Renwen shehui kexue ban) 1, 2004. P. 51-53.
Wu Yingzhe 2007 — Wu Yingzhe. Qidanyu jingci yufa fanchou yanj iu [A Study on the Grammatical Category of Khitan Nominals]. Hohhot: Inner Mongolia University Press, 2007.
Wu Yingzhe 2010 — Wu Yingzhe. Qidan xiaozi «diela bu» kaoshi [Research on the «Diela tribe » in the Khitan Small Script] // Minzu yuwen 5, 2010. P. 72-77.
Wu Yingzhe, Janhunen 2010 — Wu Yingzhe, J. Janhunen. New Materials on the Khitan Small Script: A Critical Edition of Xiao Dilu and Yelu Xiangwen [Corpus Scriptorum Chitanorum 1 / Languages of Asia 9]. Folkestone: Global Oriental, 2010.
Yoshimoto 2012 — C. Yoshimoto. Kittan shoji no onka suitei oyobi sokan mondai [Reconstruction of phonetic values of the Khitan Small Script and related problems] // Ritsumeikan bungaku 627, 2012. P. 129-157.
