Serious efforts by modern scholars to interpret texts written prior to the fifteenth century in the Korean language using Chinese characters really begin with the work of Ogura (1929), and it is still true to say that they owe their chief debt to the work of Yang Chudong (1942). Although in many cases Yang’s interpretations were quite speculative, later scholars have rarely managed to improve substantially upon his work. Essentially, this is because the texts themselves are few in number, and they span just over a full millennium in time, from the earliest surviving evidence of native Korean literacy, the Kwanggaet’o Stele (A.D. 414), to the promulgation of the Hunmin chŏngŭm in 1446.
At current levels of knowledge, attempts to draw a linguistic map of the Korean peninsula for any period during this millennium would be a highly speculative enterprise, and so it is not surprising to find little consensus forthcoming on even quite basic theories. One of the few well-established hypotheses is, of course, that one cannot speak meaningfully of an entity called “the Korean language” during the Three Kingdoms Period (1st century B.C.-A.D. 668) and the Unified Silla Period (668-935). Yi Kimun (1972) is probably the most influential proponent of this view when he suggests that it was chiefly due to the centralizing tendencies of the Koryŏ dynasty with its heartland in the central part of the peninsula that people who spoke the separate but relatively similar languages of Paekche, Silla and Koguryŏ came to speak a common language.[358]
Thus, not only does the paucity of material make the exercise of cross-checking a rare luxury, but it is well-nigh impossible for modern scholars to make due allowance for the many and varied linguistic changes that took place during this period. In the area of linguistics, it is only by means of reconstruction from the more plentiful resources of Late Middle Korean (especially fifteenth-century Korean) that some very general idea of the phonetic reality of the forms transcribed by the Chinese characters can be gained, while in the area of orthography, it has only been in comparatively recent times that scholars have been able to view the evolution of pre-fifteenth-century writing systematically.
The decade of the 1970s was a particularly significant one for orthographical studies. A good deal of new research was stimulated by the rediscovery of several important new texts, and a new generation of scholars began to critically examine, rather than simply accept, the work of pioneers in the field. Not only has light been shed on specific problems of interpretation, but also it has been possible to advance reasonably comprehensive conceptual frameworks into which to place and view the heterogeneous texts. It has become clear that the writing system used in pre-fifteenth-century Korea (hereafter abbreviated to EKWS: the Early Korean Writing System) is indeed a “system”, in that while practices were not invented or given any definitive form at any one place or time, it is possible to discern the workings of a coherent tradition. This tradition appears to devolve primarily from the work of the Silla scholar Sŏl Ch’ong 薛聰 (active late seventh to early eighth century), who is believed to be the first to establish reasonably consistent principles in order to guard against the possibility of confusion in applying Korean readings to Chinese characters.[359]
At the same time, of course, the constant evolution of Korean culture over the immense span of time involved, and the variety of stimuli that caused people to have recourse to Korean-language writing ensures that this system, though reasonably consistent in its basic principles, has many facets. At the current level of knowledge it is in terms of basic, underlying principles rather than in terms of specific, definitive prescriptions enabling the interpretation of individual texts that EKWS is best approached.[360]
Nam (1981) has discerned four different ways in which Chinese characters were used in order to transcribe fully the Korean language in EKWS.[361] These are:
1) ŭmdok 音讀 : the sound and the meaning of the Chinese character are adopted directly into the Korean language transcription. The best example of this is the large number of Sino-Korean items in the Korean lexis. In Kyunyŏ’s songs, most of the Buddhist terminology, for example, is rendered via ŭmdok readings. As might be presumed from this example, the scope for using this mode was considerable in idu texts but limited in hyangch’al texts since both sound and meaning were bound to the Chinese language.
2) hundok 訓讀 : only the meaning of the Chinese character is adopted, with the phonetic value being assigned from the Korean equivalent. In the songs this is usually indicated by transcribing the final phoneme of the Korean word separately. Here the limitation was the requirement of clear lexical correspondences between the Chinese and Korean languages.
3) ŭmga 音假 : only the sound of the Chinese character is adopted, the meaning is ignored. This is straightforward transcription, such as is used for the Sanskrit namo 南無 (“Hail!”) in song and poem No. 2. The scope for using phonetic transcription characters in this way was clearly considerable, but the syllabic expression of the Korean sound system would have required many hundreds of them. Thus they were unable to provide a foundation for EKWS by themselves. Particular features of the ŭmga mode include the basic principle of one character-one syllable, though their vowels could be ignored to render final consonants.
4) hun’ga 訓假 : the meaning of the Chinese character is initially adopted in order to obtain the phonetic value of the Korean lexical equivalent, which is then used phonetically, not semantically.[362]
Although it was thus possible for one Chinese character to be used in up to four different ways, and thus in theory possess up to four different phonetic values, Nam finds that in practice this did not occur, and that a constant value seems to have been ascribed to each character prior to its actual usage.[363] Strong circumstantial evidence suggests that this came about initially as a result of the activities of Sŏl Ch’ong and that the systematic development of the hyangch'al script dates from this time. About one hundred years after Sŏl’s time, the Samdaemok 三代目, the only known major collection of Silla songs, or hyangga, was compiled under royal auspices (A.D. 888),[364] and approximately seventy years later, Kyunyŏ composed his songs.
Apart from the conferring of a consistent value on each character, consistent principles were employed in actual usage as well. In hyangch’al the most important of these principles established a basic, though not uniform, pattern of -dok readings (whether ŭmdok or hundok) being followed by -ga readings (whether ŭmga or hun’ga) or, in other words, semantic readings being followed by one or more phonetic readings.[365] Again, this pattern emerges as quite standardized in the years following the generation of Sŏl Ch’ong, and contrasts with the less ordered patterns of combination visible in texts before Sŏl’s time.
When the combined resources of EKWS methodology, Ch’oe Haenggwi’s poems and the original HYS passages on the BhCP are utilized, it becomes possible to interpret Kyunyŏ’s songs with a fair degree of clarity. Oddly, in the sixty-seven years since the rediscovery of the Kyunyŏ-jŏn, this approach has not yet been attempted. Although Yang (1942) gave literal Late Middle Korean interpretations for all twenty-five surviving hyangga, and very literal translations into Modern Korean of the fourteen SGYS hyangga, he did not attempt translations of Kyunyŏ’s songs into Modern Korean; Kim Chigyŏn (1977) arranged the Kyunyŏ, Yang, Ch’oe and HYS texts sequentially for each song, but again did not translate them into Modern Korean. And although Lee (1957-8) did, in fact, offer modern renderings of the songs in English, he did not translate the poems in the course of his translation of the Kyunyŏ-jŏn, and gives no sign of having consulted the HYS corresponding texts.[366]
This state of affairs in turn seems to reflect the modern fate of the Kyunyŏ-jŏn as a document often plundered for its linguistic treasures but practically ignored for its historical and literary content, despite the perils that such an approach entails. It also serves to underline the fact that any serious attempt to redress this imbalance of attention, and to evaluate the literary aspects of the Kyunyŏ-jŏn, needs to begin with a consideration of what can be gleaned about Kyunyŏ’s purpose in composing the songs, and then proceed to a consideration of the actual content of the songs in the light of that purpose.
At several places in the text of the Kyunyŏ-jŏn, Kyunyŏ’s songs are identified as ga 歌, and in one instance as hyangga 鄕歌 (KYJ 10A.5). The term sanoe 詞腦 is also attached to them (KYJ 10A.5), thus raising another basic issue of terminology, but also more importantly in the case of the latter term, defining Kyunyŏ’s purpose.
Hyangga
In both its etymology and in the contexts in which it is used in ancient sources, hyangga is a general term for songs composed in the Korean language. “Hyang” itself is a term of broad meaning “the country, locality, neighborhood” and occurs frequently in Silla and Koryŏ texts with the meaning of the Korean state itself. Thus hyangga may be translated as “the songs of our country”, and used to distinguish such songs from Chinese language songs and poems.[367]
Sanoe-ga
For early modern scholars, the view of Ogura (1929) and Yang Chudong (1942) that the terms hyangga and sanoe-ga (hina uta for Ogura, saenaennorae for Yang)[368] were synonymous held sway, and even when later scholars such as Cho Yunjae (1963) and Chŏng Pyŏng’uk (1952, 1972) began to distinguish between the two terms, they still tacitly accepted their fundamental similarity.[369] With the work of Hwang (1978), it became possible to view sanoe-ga as a distinct term, denoting a special form of hyangga.
In essence, Hwang defined the sanoe-ga not in terms of its formal structure but in terms of the actual use to which it was put. He sees the etymology of the term very much as it is in fact described in the text of the Kyunyŏ-jŏn, namely, that it is a “Koreanized” Chinese compound meaning literally “verse-brain” or “verse-thoughts” but with a meaning more akin to “versifying (spiritual) thoughts”. In examining the contexts in which the term sanoe appears in the Samguk yusa, he sees sanoe-ga as essentially a light-hearted, secular song form with accompanying music and dance, but possessing a flexible aesthetic range, depending upon the aim of the composer. Thus it could be used for the expression of sublime or profound sentiments, including spiritual thoughts, but it was nevertheless a secular and not a religious song form, and although the two extant sanoe-ga in the SGYS, Anmin-ga 安民歌 and Kip’arang-ga 耆婆郎歌, are refined compositions with serious subject matter, they are not as overtly spiritual or didactic as Kyunyŏ’s songs.[370]
In fact, what is known of sanoe from the Samguk yusa tends to support a fairly literal interpretation of Kyunyŏ’s own description of his purpose in composing the songs. This is stated in the Preface to the songs as follows:
Sanoe is a means of worldly amusement, while the Sovereign Vows are the very pivot of spiritual practice. Therefore, by means of it, one may cross the shallows to regain the depths; may follow the near-at-hand to reach the far-away.
For if one does not make use of secular ways, one cannot attract those of inferior capacity; and if one does not employ vernacular speech, one cannot reveal the path of the All-Encompassing Cause. Here, I rely on familiar things that people may easily understand in order to make known unfamiliar doctrines that are difficult to grasp.
Thus Kyunyŏ had no very high opinion of the intrinsic merits of sanoe, viewing it simply as a means to a rather didactic end. This attitude should, of course, be set against Hyŏngnyŏn Chŏng’s claim that “the Master was especially competent in sanoe" (KYJ 5B.1-2), a statement that seems to imply that Kyunyŏ composed such songs on a regular basis. However, not only does Kyunyŏ’s own statement of intent contradict this notion, and not only is there otherwise no hint of secular song composing by Kyunyŏ, but also a careful reading of Ch’oe Haenggwi’s Preface reveals no hint of Kyunyŏ enjoying a general literary reputation. Ch’oe’s praise is directed specifically at the songs and not geherally at Kyunyŏ as a composer of such songs.[371]
External evidence that points to Kyunyŏ as being much more concerned with couching specific material—the Sovereign Vows—in sanoe-ga form for didactic purposes, rather than exercising his talent as an accomplished composer, garners some support from the evidence of the songs themselves. In the exegesis that follows, it will be seen fairly clearly that while the first three songs bear all the signs of having been carefully constructed and well-executed, they rather ironically provide a standard by which to measure the ensuing eight songs. Despite the inherent limitations that attend any attempt to judge Kyunyŏ’s merits as a composer from a perspective so far removed in time and place, it seems fair to say that from almost any standpoint that one cares to adopt—use of imagery, formal elements of construction, development of theme, interaction of these three, etc.— none of these eight can be said to measure up to the first three. The impression that remains is of a composer unable to sustain inspiration; whether due to lack of talent, intent, or pressures of time, or all three, one cannot say.
Exegesis
The following exegesis employs the numbering system of Yang (1942) for ease of reference to what is still the standard work in the field. Yang treats the fourteen SGYS hyangga first, and the eleven Kyunyŏ songs are thus Nos. 15 to 25 (inclusive). Thus “15.1.1” refers to the first character in the first line of the first song. It should also be pointed out that so comprehensive was Yang that for the most part, subsequent scholars have been obliged to follow his readings. This is the case here unless otherwise indicated.
Each character is interpreted in terms of Nam’s four categories listed above, using the following abbreviations: hundok reading = HD, ŭmdok reading = UD, ŭmga reading = UG, and hun'ga reading = HG. Other abbreviations used are: topic particle = TP, subject particle = SP, object particle = OP, instrumental particle = IP, locative particle = LP, possessive particle = PP, verb stem = VS, verb ending = VE, final consonant = FC, final syllable = FS, final vowel = FV.
[The text of Kyunyŏ’s songs reprinted below has essentially followed the version in Yang (1942) which has standardized most variant forms of the Chinese characters—Editors.]