The Moon Reflected in a Thousand Rivers: Literary and Linguistic Problems in Wŏrinch'ŏn'gang chi kok

ABSTRACT

Despite its importance as one of the very first literary works written using the newly invented Korean script in mid-15th century Chosŏn, the Wŏrinch'ŏn'gang chi kok 月印千江之曲 (Songs of the moon reflected in a thousand rivers; henceforth, Wŏl kok) has been little studied or appreciated, especially in English. This paper surveys the scholarly literature to date on both literary and linguistic problems in the Wŏl kok and suggests that the relative paucity of literary research on this work as compared to studies of a linguistic nature is due to the general difficulty of understanding the text without a detailed knowledge of both Middle Korean and Buddhism (and especially the biography of the Buddha). After outlining some of the debates about the authorship, original language, and relative chronology of the Wŏl kok vis-à-vis the Yongbiŏch'ŏn ka and the Sŏkpo sangjŏl, the bulk of the paper focuses on one pesky grammatical issue in Middle Korean (defined here as the language of the 15th and 16th centuries) as exemplified in the Wŏl kok: the alternation of -·ke-/-·Ge- vs. -·e- in certain verb endings. I propose a new approach to transitivity in Middle Korean based on Hopper and Thompson's (1980) notion of "discourse transitivity" and show how a treatment of -·ke-/-·Ge- as "Low Transitive" and -·e- as "High Transitive" improves on earlier analyses of this alternation and also helps our understanding (and translations) of the Wŏl kok.

Introduction

The Wŏrinch'ŏn'gang chi kok 月印千江之曲 (Songs of the moon reflected in a thousand rivers; henceforth, Wŏl kok) is well known (if little studied) as one of the very earliest literary works to be written using the newly invented Korean script in mid-15th century Chosŏn. Compiled as they were by fervent Neo-Confucian scholars at the Korean court, the dynastic records are relatively quiet about the detailed origins of the text. We know simply that when Sejong's queen, Sohŏn wanghu 昭憲王后, died in the third month of 1446, the grieving Sejong asked Prince Suyang (later King Sejo) only two days later to compile a Buddhist text as a means to pray for the queen's happiness in the next life. The text that Suyang compiled was the Sŏkpo sangjŏl 釋譜詳節 (Abbreviated and particularized life of the Buddha, 1447), a biography of the Buddha, and the Sillok annals recount that when Suyang showed his father his finished draft in the seventh month of 1447, Sejong was so moved that he produced an epic Buddhist hymn in Korean based on it—that epic hymn is the Wŏl kok and it was printed using movable metal type some time in late 1448 or early 1449.¹

The text that has come down to us is incomplete, comprising only the first of what must have been three volumes.² The first volume contains 194 cantos, which is already considerably longer than the 125 cantos of the Yongbiŏch'ŏn ka 龍飛御天歌 (Songs of dragons flying to heaven; henceforth, Yongbi) printed in 1447 to commemorate the founding ancestors of the new Chosŏn dynasty. Because Sejo later combined and revised both his earlier Sŏkpo sangjŏl and the Wŏl kok and printed them in 1459 as the new Wŏrin sŏkpo 月印釋譜, it has been possible to more or less reconstruct all of the original songs. It was unclear for many years just how many cantos comprised the work, but with the discovery in 1998 of the 25th and final volume of the Wŏrin sŏkpo,³ it became clear that canto number 583 preserved in this text was the concluding verse. Given that volumes 3, 5, 6, 16 and 24 of Wŏrin sŏkpo remain unaccounted for, the total number of cantos attested is 377.5.⁴ Nonetheless, the Wŏl kok is more than four times longer than Yongbi and is a substantial work indeed.

As Ko Yŏnggŭn (1993, 93) reminds us, the Wŏl kok is a seminal text for students of the history of both the Korean language and of Korean literature. However, if my own recent attempt at surveying the academic literature on Wŏl kok is at all accurate, research to date in both fields has limped along somewhat one-sidedly. On the one hand, scholars of the history of Korean literature, and particularly of the history of Korean verse in general and of the court-sponsored eulogies of the akchang 樂章 genre in particular, have complained of a certain neglect, denigration, and pejoration of both the Yongbiŏch'ŏn ka 龍飛御天歌 (Songs of the dragons flying to heaven; henceforth, Yongbi) and the Wŏl kok, the two most representative works in this genre. It certainly seems that scholars of Korean historical linguistics have devoted more attention to the Wŏl kok than have scholars of literature, but here the skewedness is of a different sort: the preponderance of research on the Wŏl kok has focussed on questions of orthography and phonology at the expense of morpho-syntax.⁵

Thus, the bulk of this paper will focus on one pesky grammatical issue in Middle Korean (defined here as the language of the 15th and 16th centuries; henceforth MK) as exemplified in the Wŏl kok, before closing with some remarks on the intertwined issues of vernacularization and 'diglossia.' But before diving into the Wŏl kok grammatical data, it might be interesting for some readers to highlight a few salient issues that I have encountered in my literature review.

One problem I have noticed—no doubt related to the relative paucity of literary research on the Wŏl kok as compared to studies of a linguistic nature—is the general difficulty of understanding the text without a detailed knowledge of both MK and (even more crucially) Buddhism and especially the biography of the Buddha. That is, the Wŏl kok is not exactly a work of literature that one can just pick up, read, and enjoy as literature, unless one has already internalized all the details of the various episodes from the life of the Buddha. In other words, one has to have read and internalized the contents of the Sŏkpo sangjŏl for any of the cantos in the Wŏl kok to make sense—a task rendered no less easy by the fact that so few volumes of the Sŏkpo sangjŏl remain extant. Indeed, if we are to take at face value what King Sejo wrote in his preface to the Wŏrin sŏkpo 月印釋譜 of 1459, this is precisely what King Sejong did when he composed the Wŏl kok in the first place: he read the Sŏkpo sangjŏl first, took inspiration from it, and composed the songs. The prevailing assumption in the scholarly literature is that, although the two works are by and large contemporaneous, Sejo's Sŏkpo sangjŏl must have preceded Sejong's Wŏl kok.

There are two intertwined issues here: authorship and chronology. The latter is especially difficult, given the relative silence of the Sillok and other official records on the Wŏl kok. Pak Pyŏngch'ae (1962b, 107) opines that compilation and composition of the Wŏl kok must at least have been begun prior to the Sŏkpo sangjŏl, because "… only once the compilation of the [Literary Sinitic] Sŏkkabo 釋迦譜 [Life of Śākyamuni, upon which the new vernacular work, i.e., Sŏkpo sangjŏl, would be based-RK] had made some progress and the general outline of its contents emerged, would it have been possible to compile and compose Buddhist hymns [ch'anbulga 讚佛歌, i.e., the Wŏl kok-RK] on it." Sa Chae-dong (1982)⁶ and Peter Lee (2003) also seem to think the chronology was the other way around—that the Wŏl kok came first, and that the Sŏkpo sangjŏl functioned as a sort of gloss to it; thus, Lee refers to the Sŏkpo sangjŏl as a "companion volume" to the Wŏl kok.

There is also a persuasive minority view that Sejong was not in fact the primary author. For example, Pak Pyŏngch'ae (1962b, 1974/1991, 1986) has presented interesting arguments that the primary author was in fact Kim Suon 金守溫 (1410–1481), a rather iconoclastic court official who was as notorious for his pro-Buddhist leanings as he was highly regarded for his skills in poetry, and who was known to be closely involved with Sejong's alphabet-related projects. Sa Chaedong (1971) likewise casts doubt on the brief mention in the Sejo Sillok to the effect that Sejong authored the Wŏl kok (sejong soje Wŏrinch'ŏn'gang chi kok 世宗所製月印千江之曲, "The Wŏrinch'ŏn'gang chi kok composed by Sejong"), comparing it to the hyperbolic claim that Sejong single-handedly invented the Hunmin chŏng'ŭm 訓民正音 or vernacular script), and supposes (rather unconvincingly)⁷ that the real author was actually Chŏng Hyogang 鄭孝康. Gari Ledyard (1998, 333) also takes a somewhat dissenting view: "Sejong probably did not compose all of the verse himself, but it is generally believed that he took personal direction of a staff selected for the purpose, probably the same one that collaborated on the Sŏkpo sangjŏl." Cho Hŭng'uk (2003, 280) also casts doubt on the claim that Sejong was the primary author.

And with regard to the relative timing of the compilation and publication of both the Wŏl kok and the Sŏkpo sangjŏl, too, Ledyard (1998, 334) seems to think that the Wŏl kok came first (or at least was printed first): "we would naturally expect that the Prince would have deferred publication of his own book [the Sŏkpo sangjŏl—RK] until after his father's [the Wŏl kok] had appeared. One fact that tends to corroborate this is that the "Songs," unlike the Sangjŏl, does not follow the Sino-Korean orthography of the Tongguk chŏng'un in one important particular, the indication of the soundless terminal [h] in syllables ending in a vowel, following the principle stated in HCH (4.3.1)."⁸

The issue of the readability and/or literariness of the Wŏl kok is raised already in Hŏ Ung (1969, 262), who writes: "The one regrettable point is the problem as to whether the Wŏl kok was able to function independently as poetry (siga 詩歌). It is impossible on the basis of the Wŏl kok alone to know what it means, and one cannot help doubting to what extent it has any artistic value. In this regard the Yongbiŏch'ŏn ka is much the same, and the fact that we now today are unable to derive any poetic inspiration from them suggests that it is difficult to view them as works of literature."

Subsequent researchers would reject this view as both extreme and uninformed. Kim Chong'u (1965) gives an early and glowing appraisal of the Wŏl kok as a masterpiece of religious literature. Peter Lee in the same year (1965, 34–35) characterizes the Wŏl kok as standing "closer to the cult of devotional verse in Korea" and praises it as follows: "The language is sublime and elegant, commensurate with the theme, and brocaded with rich imagery. It is more smooth and natural than that of the Yongbiŏch'ŏn ka and is free from artificial parallelism and empty rhetoric. The section describing the birth of Siddhartha, in particular, has Longinian sublimity in tone and expression." Sa Chaedong (1982) is emphatic that the Wŏl kok is nothing short of an unprecedented Korean equivalent to Aśvaghoṣa's Buddhacharita—the epic poem in Sanskrit from the early second century CE that has served as the matrix for all subsequent biographical accounts of Gautama Buddha—in terms of its beauty and significance as a major piece of world Buddhist literature, and praises both its sustained sublime style and its masterful combination of narrative fiction and lyrical poetry. Pak Pyongch'ae (1986, 68) agrees, but Kim Chiyŏng (2015) takes the comparison with the Buddhacharita even further; whereas the Buddhacharita focuses on episodes in the Buddha's life after he attains Nirvana, the Wŏl kok is more balanced in its coverage of different phases of his lives. Moreover, it is a more supernaturally inclined narrative, emphasizing on the one hand the Buddha's virtues and salvific powers, while also portraying him as a paragon of morality and a divine source of refuge for all living beings. The Wŏl kok incorporates rich episodes about Buddha's previous lives from Mahāyāna texts like the Pŏphwa kyŏng 法華經 and Poŭn kyŏng 報恩經 to demonstrate his role in preaching the Dharma, while the episodes about his actual life emphasize his mercy and filial piety. Kim sees all of this as a kind of reaction to the early Chosŏn policy of ŏkpulsungyu 抑佛崇儒 or "oppress Buddhism, raise Confucianism."

Kim Taehaeng (1999) gives a useful discussion of the style and prosody of the Wŏl kok, while Hoyt (2000, 191) and Sa Chaedong (1982) find the poetic qualities of the Wŏl kok superior to those of the Yongbiŏch'ŏn ka. Ledyard (1998, 329–30), in his discussion of the Sŏkpo sangjŏl, Wŏl kok, and Wŏrin sŏkpo, laments the fragmentary nature of what remains of these works, all of which "… convince us that we are missing a great deal of literature that is at least equal, and perhaps superior, to the poems of the 'Song of the Dragons Flying Through Heaven,' many of which fall more readily into the category of ritual than that of literature. The Buddhist works, though all based on stories from the Buddhist canon, are not merely translations but often new, fresh writing. Their tone is devotional and, in that strange Buddhist way, elusive; their language is both rich and direct; their poems are inspired and not overly constricted in form; their narratives are interesting and detailed, with a beginning and an end." He characterizes the Wŏl kok as "a splendid collection of devotional verse" (333).⁹

Still, Korean scholars of Buddhist literature are unanimous in lamenting the relative neglect of court-sponsored akchang eulogies in Korean literary research. For example, in his article on the treatment of the akchang genre in North Korean research, Cho Kyuik (2006) shows how the authors of both Chosŏn munhaksa (1979) and Chosŏn munhaksa (1991) brand akchang as a kind of "reactionary and worthless" "literature of flattery and unconditional glorification," and goes on (37–38) to complain that other (South) Korean literary scholars have also tended to "either ignore or neglect" the genre—a view echoed by Cho Hŭnguk (2002) and Sinand Kim (2005).

A frequent theme in literary research on the Wŏl kok concerns the ways in which it parallels and serves as a kind of ideological counterbalance to the Neo-Confucian-inspired Yongbi. Peter Lee (1965, 34) writes that "The form of the Wŏrin ch'ŏn'gang chigok is similar to that of the Yongbi ŏch'ŏn ka. Each canto consists generally of two poems, and the number of syllables in a poem fluctuates commonly between twenty-one and twenty-four" and Hoyt (2000, 181) notes that "The form of the Songs of the Moon is generally the same as that found in the Songs of the Dragons. Sejong apparently intended to give this hymnbook the same prestige as the Confucian epic of the foundation of the dynasty" and concludes (192) that "the two epics represented two conflicting yet integrated ideologies, both of which were basic to the cultural perspective of the age."

The two different ideologies represented by the Yongbi and the Wŏl kok were certainly in conflict, but one wonders whether they were truly "integrated" anywhere other than in the person of Sejong himself and his immediate family members and associates at the Royal Court. It is interesting to compare Peter Lee's (1975, 10) characterization of the Songs of the Dragons as "a Confucian revenge against the Buddhist tradition, which, as heresy, the new kingdom attempted to displace," with analyses of the Wŏl kok by Kim Sŭngu (2005), Chŏng Soyŏn (2009), and Kim Chiyŏng (2015) that call attention to the ways in which the work highlights family relationships and a Buddhist approach to hyo 孝 or filial piety. For example, Kim Sŭngu (2005, 145) stresses that King Śuddhodana (Chŏngban wang 淨飯王) and Queen Māyā (Maya puin 摩耶夫人), who are referred to in the Sŏkpo sangjŏl by their official roles as 'the King' (wang 王) and 'his wife' (puin 夫人), are instead referred to in the Wŏl kok almost exclusively by means of the honorific kinship terms a·pa:nim 'honorable father' and e·ma:nim 'honorable mother,' respectively. Similarly, Sŏkchon 釋尊 and Yasudara (Yaśodharā)—referred to in the Sŏkpo sangjŏl by their royal titles of Crown Prince (t'aeja '太子') and Royal 妃子'), respectively—appear in the Wŏl kok as a·tol:nim 'honorable son' and mye·nol.·i¹⁰ 'honorable daughter-in-law,' respectively. Kim interprets this in the light of Sejong's grief for his departed queen, his cognizance of her own painful loss of her father and other relatives when earlier convicted of treason against the new state of Chosŏn, and his concern about the possibility of fratricidal strife among his sons after his own demise; thus, for Kim, the Wŏl kok was Sejong's attempt to portray Sŏkchon's family as a tighly-knit and harmonious unit, and simultaneously a warning to his sons. In similar fashion, Cho Hŭng'uk (2003, 285) finds it interesting that "the characterizations of Śākyamuni as a person depict him on the one hand as a merciful, Buddhist figure, and on the other hand as a Confucian figure who combines both military and literary achievements," citing both the depiction of his filial desire to ensure descendants, and his prowess in archery. Chŏng Soyŏn (2009, 202) likewise finds that the Wŏl kok emphasizes the relationship of filial piety between King Chŏngban and Sŏkka, thereby also reinforcing and upholding a Neo-Confucian world view.¹¹ Min Myŏngsuk (2016) calls attention to the text's portrayal of Sŏkchon as an exceptionally moral and ethical person, and claims that while the work is not exactly an overt rebuttal to his Neo-Confucian courtiers' constant critiques of his Buddhist proclivities, its contents nonetheless demonstrate that the teachings of the Buddha are not fundamentally inimical to those of Neo-Confucianism.

A final issue that surfaces repeatedly in the literature is that of the original language of composition of the Wŏl kok. Whereas it is widely accepted that the songs in the Yongbi were first composed in Literary Sinitic (hanmun) and then translated into Korean using the new script, the situation with both the Sŏkpo sangjŏl and the Wŏl kok is slightly more complicated. We know that Sejong had earlier directed Crown Prince Suyang (later King Sejo) to compile a biography of the Buddha based on the Shijia pu 釋迦譜 by Sengyou 僧佑 (445–518) and the 釋迦氏譜 by Daoxuan 道宣 (595–667), and all the evidence suggests that Sejo produced a composite life of the Buddha in Literary Sinitic based on these and (and other) Buddhist texts in Literary Sinitic before translating that into Korean in the new alphabet as the Sŏkpo sangjŏl. Indeed, recent research on this latter text has proceeded to the point where scholars are able to reconstruct the original hanmun text that was compiled from numerous other chŏgyŏng 底經 or "base texts" to form the basis on which the translation was based.¹² Whereas it has been widely claimed that the Korean language and style of the Sŏkpo sangjŏl is generally closer to spoken colloquial Korean of the day, and that Sejo's translation (if, indeed, that is what it was)¹³ was a looser 'meaning translation' (ŭiyŏk 意譯) rather than a stiffer 'direct translation' (chigyŏk 直譯), Kim Sŏngju and Cho Chunho (2017, 16) instead claim: "But if we reconstruct the original hanmun version of the Sŏkpo sangjŏl and compare it with the ŏnhae texts in the Sŏkpo sangjŏl, we can confirm that the hanmun original and ŏnhae texts are in an even closer 1-to-1 relationship than previous research has discussed."

The question is whether the similarly (and relatively) fluid and colloquial Korean language of the Wŏl kok is consistent with the notion that Sejong (or Kim Suon?) composed the Wŏl kok directly in the vernacular, on the basis of the vernacular Sŏkpo sangjŏl, rather than composing it first in hanmun based indirectly on Sejo's newly collated (and unfortunately non-extant) Literary Sinitic life of the Buddha in hanmun, and then translated that into Korean much as must have happened with the Sŏkpo sangjŏl. Traulsen (2016, 119) assumes without argument that both the Wŏl kok and the Sŏkpo sangjŏl were originally composed in Korean and that "they have no underlying original hanmun text," but this is patently not the case with the Sŏkpo sangjŏl and is difficult to determine in the case of the Wŏl kok. Chŏng Soyŏn dodges the issue of translation and (2009, 209) supposes simply that Sejo "converted" his Literary Sinitic life of the Buddha into Korean, and that the Wŏl kok is a versified version thereof. In any case, for our purposes here, it suffices to stress that the Korean language of both the Wŏl kok and the Sŏkpo sangjŏl is generally considered to be less compromised by "hanmun translation-ese" than other texts from the 15th century and thus more "natural" and closer than other such texts to colloquial spoken Korean, making it therefore of great value for investigating the history of the Korean language. Let us turn now to linguistic problems in Wŏl kok.

Orthography and Grammar in the Wŏrinch'ŏn'gang chi kok

The first point to emphasize in any discussion of the Wŏl kok and the history of linguistic research on this text to date is that the overwhelming preponderance of attention has been focussed on the special and often unique orthographic features of the text. These have been written about over and over again by numerous scholars in both English and Korean: in Korean, Hŏ Ung (1953 and 1969), Pak Pyŏngch'ae (1962a and 1974/1991), Nam Kwang'u and Sŏng Hwan'gap (1964), Kim Ch'agyun (1986), Yi Kimun [Lee, Ki-moon] (1992) [available in English as Lee (1997)], Ko Yŏnggŭn (1993 and 2003), Hŏ Ung & Yi Kangno (1999), and most recently, Chŏng Uyŏng (2014) are some of the most representative texts; in English, both Kim-Renaud (2000) and Ki-moon Lee (2009) devote considerable space to the orthography in the Wŏl kok.¹⁴ Thus, there is little point in rehearsing all the details here and instead I will simply mention some highlights before moving to a discussion of certain grammatical forms.

Orthography

One immediately striking orthographic feature of the Wŏl kok and one that is truly unique to this text is its "in-your-face" and "front-and-center" prioritization of the Chŏng'ŭm Korean vernacular script. Concomitant with this privileging of the vernacular script is a downplaying and sidelining of sinographs. That is, whereas all other texts of the period front-load sinographs and provide Tongguk chŏng'un-style Chŏng'ŭm Sino-Korean pronunciations immediately after them in a smaller typeface or font,¹⁵ the Wŏl kok in all such instances does exactly the opposite and instead front-loads the Chŏng'ŭm Sino-Korean pronunciation and appends the sinograph in a smaller size; Kim-Renaud (2000, 30) likens these pronunciation glosses to "back-stage prompts."¹⁶ Most Korean commentators read into this practice a proud patriotic or even proto-nationalist assertion of the Korean language, the Korean script and all things good and Korean on the part of King Sejong as opposed to adulation of things Chinese, and even no less an observer than Gari Ledyard opines: "The writers who make this judgment may be expressing more their own sentiments than Sejong's, but with some reservations that would take into consideration the differing nature of patriotic spirit in those days and these, I believe that they are right."¹⁷

Rather, and as Chŏng Soyŏn (2009) and others have shown, this particular feature was more likely tied to the target audience of the text, which Chŏng supposes to have been palace women and other palace employees (likely also certain monks and lowly palace functionaries) involved in performing the rituals at which the hymns would have been sung. This performance-related argument related to the orthography and sheer size of the text and its vernacular typeface has largely been overlooked. In the case of the Yongbi, which parallels the Wŏl kok in so many ways, we know that its verses were accompanied by ritualized music and dance for court ceremonies, and the tune (akkok 樂曲) and choreographic instructions (kamu 歌舞) for them are even recorded in the Akhak kwebŏm 樂學 軌範 (Guide to the study of music, 1493). Sa Chaedong (1982, 292) points to the line immediately preceding the passage cited above from Sejo Sillok, which reads: 授八妓諺文歌詞 令唱之 卽…, "handed to eight dancing girls the ŏnmun text of the song and bade them sing it—that is, [the Wŏl kok composed by Sejong]").¹⁸ Such dancing girls could hardly have been expected to have anything more than a basic acquaintance with sinographs; what they needed was an easy-to-read score. Another more sensible approach to the different orthographies in Yongbi and Wŏl kok is that of Kim Wanjin (1996) who likewise discerns a difference in orthographies according to readership, but also believes that Sejong's ultimate objective was experimentation with the harmonious textual blending and coexistence of cosmopolitan hanmun and sinographs with vernacular Korean.

The other orthographic feature that has attracted a great deal of attention from researchers is the way that the Wŏl kok—like the Yongbi but more consistently and with some differences—departs from the phonemic and 'surfacy' spelling laid out in the Hunmin chŏng'ŭm (haerye) and instead adopts in certain limited environments a deeper and more abstract morphophonemic spelling that anticipated already in many ways the Unified Han'gŭl Orthography pioneered first by Chu Sigyŏng and then by the Han'gŭl Hakhoe in the early twentieth century and that is still in use today.

For example, numerous researchers praise the compiler of the Wŏl kok for his advanced "grammatical consciousness" (i.e., his awareness of underlying forms and attempts to distinguish them orthographically) because in certain cases nouns are written separately from following vowel-initial nominal case particles, and verb stems are written separately from following vowel-initial endings: such cases are restricted to nouns ending in the sonorants -m, -n, -l, -ng, -z, and verb stems ending in -m and -n. But nouns and verb stems ending in voiceless consonants are not afforded this morphophonemic treatment, such examples are rare overall, and the Wŏl kok is not entirely consistent in its experimental spellings.

Here are some examples below. (All examples in the following exposition are presented in both han'gŭl and Yale romanization, followed by the English translation of Olof (2009) and with occasional reference to the German translation of Sasse & An (2002)).¹⁹

Spelling of consonants other than -p, -t, -k, -s, -m, -n, -ng, -l in coda position before pause or consonant

1) Nr. 7ta·sos ·kwoc :twu kwo·c i"Five flowers and two …"

2) Nr. 178·SYEYCWON ·ol mac·nazo·Wo·mye cum·key nam·k i tulGye·nulkwu·chye :pwoyzop·kwo cwocco·Wa·wo·ni"He met the Honoured One, and a big tree was raised, so he was clearly visible, and had to follow Him back again."²⁰

3) Nr. 99nam·k i nwoph·kwo ·two"Even if a tree is high …"

but (also no. 99):

·SSYWULQ·PEP ·i nwop·ta hon·tol"Obwohl man sagt, seine magischen Künste seien hoch entwickelt …""Even if they boast of their magic …"

4) Nr. 129·PPEM·CI :cuz ·ul ·pwo·sya"Er sieht Brahmacārin-Verhalten""He saw the conduct of the brahmanas, and …"

As can be seen from the examples from canto 99 above, the Wŏl kok was not entirely consistent in its orthography.

Archaic forms of the infinitive²¹ vowel -·e/·a-

There is some controversy in the scholarly literature as to how to interpret cases where the infinitive vowel is written with -·a even after a stem with vocalization that would require -·e according to the MK rule of vowel harmony. Martin (1992) prefers to interpret such examples as archaisms or holdovers from an earlier stage of the language when the shape of this morpheme was just -·a and before vowel harmony had arisen in the language. Opposing this historical phonological approach are scholars like Pak Pyŏ ngch'ae (1962a) and Ko Yŏnggŭn (1993), who prefer to attribute it to the same "grammatical consciousness of underlying forms" mentioned above, on the assumption that the compiler made a conscious orthographic decision to treat -·a as the underlying or basic form of this two-shape ending. As Ko Yŏ nggŭn (1993, 106) notes, however, the extant first volume of Wŏl kok is a kyojŏngbon or galley proof, where, for example, some of the shallow spellings seen just above have been corrected by hand to deeper, morphophonemic spellings;²² and yet this particular spelling of the infinitive vowel is not corrected anywhere in the text, and moreover survives unmolested into the later Wŏrin sŏkpo. It also occurs in other early MK texts like the Samgang haengsilto 三綱行實圖 (henceforth, Samgang).

5) Nr. 3na·la s :CHYEN ilpeza"[five hundred felons] robbed the national treasures"

6) Nr. 38sa·hwoy ·lol kol·hoy·ya coy·cwo ·lol :mwot mi·ta"He was looking for a son-in-law, but as he did not believe in His talents …"

7) Nr. 76·khun LYWONG ·ol ci ·za"[Der Flußgott] macht einen großen Drachen …""He was changed into a big dragon …"²³

8) Nr. 188ta·sos LA·CHALQ:NYE y kwol :epsun :cuz ·ul ci·za"The five female nach'al monsters showed their abhorrent faces …"

but cf. also Nr. 90 for another inconsistent example:

9) Nr. 90KKYEY ci·ze"verfasst eine Gatha …" "… composed a hymn of His lecture."

10) Nr. 123cwu·kwum sa·lwo·m ol te·la"I discarded life and death, and …"

11) Nr. 124ku ·mul ·ey ·MWOK·YWOKhol·ssoy SAM·TTWOK ·i :ep·sa"I bathe in its waters, the Three Evils are gone, and …"

12) Nr. 153SYA·LY·PWULQ ·uykungey mwu·la"He asked Saribul, and then …"

13) Nr. 155SYWU·TTALQ ·i … ·SYA·LI·PWULQ ·ul :mwot mi·ta"… hat er kein Zutrauen zu Śāriputra""[Sudal] … lost all confidence in Saribul …"

14) Nr. 157NGWOY:TTWOW SAM·QUK·MEN ·i NG WANG s al·ph oy ·tu·la"In great numbers the disbelievers gathered before the king."

15) Nr. 192·pol·h ol tulu·si·ni :pwo·poy ·yey s kwo·c i tu·la"He lifted His arm and jewel-flowers rained down …"

but

16) Nr. 101·YWOK:HWA ·lol ho·ma ·psku·sya·HHAY·ho·zoWo·l i :epsul·ssoy LYWONGTTANG ·oy ·tule·kasi·ni"He had already extinguished the fire of passions,so no one could harm Him, and He entered the dragon-hall."

17) Nr. 168·howo·za wu·zu·m ul wu·za"[Śāriputra] lacht vor sich hin und …""He was smiling, and …"

18) Nr. 86hungcyengpa·ci ·tolh ·i ·kilh ·ol :mwot ·nye·a"The merchants could not continue their way …"

In another interesting approach to these forms, Werner Sasse (p.c.), co-author of the volume whence the German translations here, wonders whether there might be a connection with the meter of the text. Each canto contains two lines, and each line is composed of three clauses (phrases).²⁴ Sasse notes that the position of infinitive vowel -·a instead of the -·e predicted by vowel harmony is invariably at the end of a clauses (phrase), whereas the expected forms in -·e occur in the middle of a clause.

Pleonastic/echo infinitive vowel -·a

One feature of the orthography in Wŏl kok that is less commented on but that links it with the spellings in the Samgang is the tendency with verb stems in final a to add an extra 'echo' version or pleonastic copy of the infinitive vowel -·a. Ko Yŏnggŭn (1991) has shown on the basis of certain orthographic, lexical, and grammatical features that even though the Samgang was not printed until the 1480s, the vernacular translations that appear in the upper margin of each page must have been completed while Sejong was still alive, contemporaneously with texts like the Yongbi, Sŏkpo sangjŏl, and Wŏl kok. There are only a few such examples in Wŏl kok, all with ˚na- 'arise; appear; come forth':²⁵

19) Nr. 3CYENG·SYA ·lol :ti·na·a ka·ni"[five hundred felons] … passed by His hermitage"

20) Nr. 41:LYEYCCYWUYEN ·i swo·sa·na·a"… and a sweet water well sprang up …"

21) Nr. 188nwun ·ey ·pul·na·a ·penkey ·kotho·ni"… and from their eyes fire jetted like flashes of lightning."

Morpho-syntax

Now let us turn our attention to morpho-syntax; more specifically, to the vexed question of the effective²⁷ morpheme -·ke-/-·Ge- and its allomorph in just -·e (there is also a rarely encountered first-person allomorph in -·ka-/-·Ga- and just -·a). There is an entire range of MK forms that incorporate the effective morpheme, with and without the initial k or G, and traditional Korean scholarship has struggled to explain their usage.²⁸

In an early study of the MK facts, Kono (1950) concluded that the function of the effective morpheme was to derive "intensive" or emphatic stems in opposition to "plain" stems in Zero or infinitive -·e. But Kono mistakenly tried to find the origin of the modern infinitive vowel -e of mek-e etc. in the lenited form of -·ke- (i.e., -·Ge-), and thus missed the MK alternation of -·ke-/-·Ge- vs. -·e-. Had he not missed the phonological facts, he might well have solved the problem, as nearly all his examples show the distinction to be discussed below.

Han Haengja (1964, 129, 133–35) identifies and discusses the alternation, claiming that -·ke-/-·e- "expresses all manner of functions of counterfactuals and conditions, and of tenses." Her discussion is confused, and she supposes that ·hoya·nol is 'past perfective,' whereas ·hoke·nul is simply a conditional, thus assimilating what are really facts about transitivity (see below) to tense-aspect. Along similar lines, Ch'oe T'aeyŏng (1965) noticed the alternation, and tried to claim that this marker has its origin in a "past perfect;" in 15th-century Korean, it supposedly marked a "hypothetical condition." Yu Ch'angdon (1963) also correctly identified the alternation, but could find no motivation for it. Reviewing the conclusions of previous Korean scholars, he says: "There are various hypotheses about 'suppositional,' 'conditional,' and 'tense/emphasis' functions, but it has no function other than emphasis." Hŏ Ung (1975) likewise could find no regularity, and the book-length annotated translations by both Pak Pyŏngch'ae (1974/1991) and Hŏ and Yi (1999) have nothing useful to say about any of these forms within the Wŏl kok itself.

The attempt to find a tense/aspect distinction in the use of -·ke-/-·e- has a long pedigree, especially in Japan. The first such suggestion can be traced back to Maema (1923), who called -·ke- the "demi-past" and -·a/-·e the "past." Similarly, Yi Sŭnguk (1967) saw -·ta-/-·te- and -·ka-/-·ke- as opposing members of a system of imperfective and perfective aspects, respectively. Shioda (1985), in a study restricted to the ending -·ke·nul/-·e·nul, also adopted an aspectual analysis: for her, -·ke·nul means "simply the state" in the case of adjectives and verbs of being, but "completion of change = perpetuation of result" in the case of intransitive change-of-state predicates, and "perpetuation/repetition of the action" in the case of transitives. As for -·e·nul, she supposes it signifies an action perceived as a single accomplished whole. Shioda admits though that her analysis encounters difficulties when extended to other endings that incorporate -·ke-/-·e-.

Ko Yŏnggŭn (1980) provides the most original and interesting attempt at sorting out the -·ke- facts, and overall I concur with his view in its broad outlines. He seems to have uncovered a transitivity distinction whereby verbal endings with -·ke-/-·Ge- usually appear with intransitive verbs, adjectives, and the copula, and endings in -·e- occur with transitives. But Ko Yŏnggŭn takes a conservative, all-or-nothing view of transitivity: simply, is the verb transitive or not? After reviewing some of his examples below, I will suggest that a more liberal and discourse-based view of transitivity saves many of his "exceptions."

It is impossible to summarize all of Ko's findings here, but he divides the various endings which show the alternation into four types, according to how well his hypothesis accords with the known examples.

A. YES-YES
means both -·ke- and -·e- conform:

22) ·i     kwo·ki nik-ke·ta (Nogŏltae 1: 22a)this meat ripen-intr.-decl."This meat is cooked/well-done."

23) ·CENGho-n ·kil·h ol il·h-e-·ta (Sŏk 23: 19b)correct     way -acc. lose-tr.-decl."They have lost the correct Way."

The only other pair that checks out consistently besides ·ho-ke-·ta ~ ·ho-·ya-·ta is ·ho·ke ci·la ~ ·ho.ya ci·la. Both are high-frequency forms.

B. YES-NO
means the -·ke- forms check out, but the -·e- forms do not.

24) ·yel:twu ·TAY·KEP ·i ·cho-ke-·za LYENHWA y ·phu-ke·tun (Wŏlsŏk 8: 75)twelve Kalpa nom. be full-intr.-if lotus-nom. bloom-"For the lotus flower blooms only after 12 Kalpas have passed."

but examples in -·e-·za are not always transitive in any obvious way (cited from Martin 1992, 542):

25) CCO-PI s :hoyngtye·k ul ·ho.ya ·za ho·l i 'l ·ss oySA-MI '·la ho·n i '·la (1447 Sŏk 6: 2b)"for they must perform deeds of charity therefore they are called śramaņera (religious novice)."

26)ne-huy ·tol·h i SOYNG :SO pe·swul :i·l ol ·him ·pseKKWUW·ho.ya ·za ho ·li'·la (Wŏlsŏk 10: 14b)"you people must endeavor to pursue the casting off of birth and death."

27) ·TWOY·TAP·ho.ya ·za     ho·l i '·la (Nŭngŏm 1: 44a)"will have to reply."

and there are examples of intransitive verbs taking -·e-·za like the following:

28) a·to·l oy ·ptu·t i     ho·ma ·kh-e-·za (← khu-·e-·za; not *khu-·ke·za) …(Pŏphwa 2: 224b)son gen. intention nom. already great-tr.-only if"Only once the son's intentions have become great …"

It is noteworthy that in many cases similar to this last one, the corresponding kugyŏl annotation of the original Literary Sinitic has the expected -·ke-·za form. If there really was an earlier transitivity distinction between -·e-·za and -·ke-·za, clearly the -·e-·za was already pushing out forms in -·ke-·za on its way to becoming the modern "only if" pattern in -e ya, and more recent research suggests that -·e-·za functioned independently of -·ke-·za, while the latter form alternated with -·kwo-·za in terms of transitivity.²⁹ The only other ending pair that belongs to this category is the adnominal ·ho-ke-n ~ ·ho.ya-n (also ·ho·ya-n). Like the two YES-YES endings above, ·ho-X-·za and ·ho-X-n are relatively frequent.

C. NO-YES
means the -·ke- forms show some exceptions, while the -·e- forms do not. Here there are ten ending pairs, vs. only two in YES-NO.

29) NGWANGs PWUIN hona·h i … :tele·Wun ·ptu·t ul ·ho-ke-·nul (Sŏk 24: 49)King's     wife     one-nom.     dirty     intention-acc. do-intr.-"One of the King's wives [loved the Crown Prince] and was harboring dirty intentions [toward him] …"

30)·CHYWULQKA ·lol :CHYENG ·ho-·zoW-a-·nol … (Wŏlsŏk 10: 17a)leave-home-acc. request do-hum.-tr.-"Requested to [leave home and] enter the priesthood …"

Other ending pairs in this category:³⁰

1. a). khe·n ywo (← *·ho-ke-·n ywo) ~ ·ho-ya-·n ywo(ma·lon)

2. b). ho-ke-la ~ ho-ya-la

3. c). ·ho-ke-·ni (stonye) ĥo-ya-ni (stonye)

4. d). ·ho-·ke-ni·Gwa ~ ho-ya-niwa

5. e). ·ho-·ke-nma·lon ~ ·ho-·ya-nma·lon

6. f). ho·li-Ge-·nul [Yale: ho·l i 'Ge-·nul] ĥolyenul

7. g). holi-Ge-ni (ston) ~ ho·lye·ni(ston)

   [Yale: ho·l 'ye·n i = ho·l [i] 'Ge ·n i; ho·l 'Ge. ·n i 's·ton]

   (cf. Ko Yŏnggŭn 1980, 73)

Many of these are already rare forms in MK and are poorly attested.

D. Finally, NO-NO
means both members of a pair show exceptions to the transitivity hypothesis. There are just three such pairs, and all are rare:

1. a). ·ho-ke-·ni ~ ·ho-ya-·ni

2. b). holi-Ge-ta ~ holye-ta

3. c). holi-Ge-niwa ~ holy-e-niwa

   [Yale: ho·l 'ye·n i ·Gwa = ho·l [i] 'Ge ·n i ·Gwa]

In the cases of (C:f-h) and (D:b-c), matters are clearly complicated by the widespread MK phonological rule whereby k lenited to G (and later to zero) after i, y, z, l, a state of affairs that must have rendered the alternation opaque in forms like these, on top of which we should recall that the MK future morpheme -li- was in the process of crystallization via grammaticalization of earlier adnominal copular structures like *ho-l[q] i i-Ge-niwa (analogous to modern ha-l kes ikeniwa). In other words, if Martin's analysis of these future forms as copular in origin is correct (putting to one side the issue of how synchronic or diachronic his analysis is), the question of a transitivity alternation in them is moot; in origin, they were all copular and thus intransitive and we should not expect any transitive versions of them. The considerable differences between Ko's and Martin's analyses also indicate the complexity of the forms, a situation complicated even more by the relative paucity of attested examples—in a number of other cases it is difficult to ascertain one side of a pair simply because there are no attested examples of one or the other of -·ke- or -·e- appearing with the ending in question, and we must simply wait for more texts to turn up. Here are some more examples that bear out Ko's hypothesis:

31) nil·Gwey ho·ma tatot-ke-·ta (Sāk 24: 15b)week     already arrive-intr.-decl."A week has already transpired."

32)wu·li     ·pap ·two me·k-e-·ta (Nogŏltae 1: 57a)we     rice too eat-trans.-decl."We have also finished eating our meal."

33) NGUYSIM-toWoyn ·kwo·t i         is-ke-·tun (Wŏlsŏk Intro: 20)doubt-like         place-nom. exist-intr.-cond."Whenever there was a doubtful point …"

34) nay ·emi     cwuk-ke-n ·ti     a·ni     wo·la·ni (Wŏlsŏk 21: 27)my mother die-intr.-adnom. since not     long"It has not been long since my mother died."

As alluded to above, some of the exceptional members of the NO-YES and YESNO pairs can be reconciled with a historical explanation: the "irregular" forms of the pairs were already pushing out the opposing member and were on their way to becoming the sole modern representatives of the old form. Thus, modern Korean ha-ketun, ha-keniwa, ha-keni, ha-kenmanun, ha-lyeni(wa), hay ya (← ·ho.ya-·za) and ha-lyenmanun were all victorious over their opposing members.

A Promising Avenue for Future Research: Discourse Transitivity

Transitivity, not merely in the sense of a verbal feature, but in the sense of a discourse phenomenon à la Hopper and Thompson (1980), provides an attractive solution to some of the recalcitrant examples of the -·ke-/-·e- alternation in MK. For Hopper and Thompson, transitivity is not an either-or concept but a more-or-less notion that can be thought of as a continuum ranging from Low-Transitive to High-Transitive. They outline the following parameters of "cardinal transitivity" in a clause:

Table 1.

Parameters for Cardinal Transitivity in a Clause

Note also that, while not on this table, first person ranks higher than non-first person on such a scale (cf. Silverstein 1976).

I propose to salvage Ko Yŏnggŭn's transitivity hypothesis for -·ke-/-·e- with a Revised Transitivity Hypothesis that views transitivity as a continuum in this way: while Ko views transitivity as a strictly all-or-nothing proposition in his article, many of his "exceptions" can be explained away with this discourse-based definition of transitivity. Thus, a number of "irregular" examples where allegedly intransitive -·ke- appears on an otherwise transitive verb, but with an accompanying Zero-marked object, look like noun incorporation or noun-stripping—a classic case of unaffected and/or non-individuated objects, and one where many languages mark the new incorporated structure as intransitive:

35) :TWO Ø ·TUK[ho-]-·ke-si-n ·ti (Pŏphwa 5: 120a)Way Ø get do -intr.-hon. since"Since acquiring the Way …"

36) CWUNGSAYNG ·i … ·wa         ·PEP Ø         tut-ke-·nul (Wŏlsŏk 13: 50b)the masses nom.     came-and Law     listen-intr."The masses came and listened to the Law (Law-listened)."

37) ·nay … HYANG Ø     phwuy·Gwu-·zop-ka-·ni (Wel 23: 88−9)I         incense Ø     sprinkle-hum.-1st person intr. sequential"I sprinkled incense (for the Buddha)."

As we might expect from a Discourse Transitivity approach to the -·ke-/-·e- alternation, negated verbs that are otherwise transitive frequently take Low-Transitive endings, and these constitute another interesting class of "exceptions" in Ko's original formulation. There are few such examples in the Wŏl kok, but evidence that MK :mwot, for example, can have a de-transitivizing effect on clauses is easy to find elsewhere.³¹ Note also the large number of examples in -·ti :mwot·ke 'yla (← -·ti :mwot ·ho-ke ·i·la), many of which involve the verb :a-l- "know:"

38) ·pwom s HUNG ·ey     :a-ti :mwot-ke-'y·la (Tusi 22: 16b)spring-gen. fluorish-dat. know-neg. mwot-intr.-cop."I did not know, for it was high Spring."

Here are a couple examples from the Samgang:

39)SAMNYEN :sal·Gwo ·wos a·ni paske·nul"[When Jiang Ge's mother died, he went to her grave and] lived there for three years without removing his mourning attire."

Here, the transitivity is muted both by the negation and by the non-punctual, extended nature of the action.

40):salo·m i ·cho·ma :mwot ·pwoke·nul"people could not bear to look on [whenever Yu Sŏkchin's father fell unconscious because of his affliction] …"

Here, transitive verb ˚pwo- 'see; watch' is both habitual and negated, and thus Low-Transitive.³²

A Brief Detour into the Prehistory of Korean Transitivity

Finally, note cases like the following:

41) SSYENG·PPWULQ·ho-ya-n ·ti (Wŏlsŏk 17: 22)attain-Buddha do-tr.- since"Since attaining Buddha-hood …"

Ko Yŏnggŭn lists this as an exception, because the sinographs have the intransitive meaning "become 成 + Buddha 佛," and because the same phrase usually appears with Low-Transitive ·ho-ke-n ti.³³ But an equally legitimate analysis would be to take SSYENGPPWULQ as a verbal noun object of transitive ˚ho- "do:" "since doing the Buddha-becoming." Modern K is famous for allowing "double object constructions" of the type:

42) Hankwukmal ul     kongpu lul     hantaK. language acc.     studying acc. does"studies Korean"

again with the verbal noun marked with an accusative. The invasion of Chinese loans via the {Verbal Noun+ [obj marker +] ˚ho-} conduit may well have helped to throw the older Korean transitivity alignment off kilter, and render the whole system opaque.

Garrett (1988), citing Haiman (1980), mentions that in Hua, ergative case-marking has been extended from its original domain into subjects of certain intransitive verbs, "… in the first instance, to verbs which represent old abstract NPs plus hu- 'do' …" Givon (1980) cites a similar case of incipient ergative-to-nominative re-analysis from Sherpa. Some verbs allow the subject to be marked with either the absolutive or the ergative:

43) (a) nga     laga kyaa-yin (absolutive subject)l-abs. work do-aux.-perf.

a·ni     ·cwue·nul"did not give it [to her]""I worked" (lit.: "I did work")

(b) nyee laga kyaa-yin (ergative subject)I-erg. work do-aux.-perf."I worked" (lit.: "I did work")

Givon notes:

"… anyone familiar with a similar process of creating complex verbs via object incorporation elsewhere in the Sino-Tibetan family … would know that these constructions are on their way toward lexicalization. And since this process is a massive one in the Sherpa lexicon… the potential for creating a beach-head of ergative subject marking in an essentially intransitive-stative domain of the verbal paradigm is unmistakable."

The analogous Korean process of {Verbal Noun + 'do'}-lexicalization has been equally 'massive,' and may well have been the beach-head for the Korean re-analysis, too.

Back to the Main Road: High- and Low-Transitive in MK

In any case, according to a more nuanced and discourse-based interpretation of transitivity, the affix -·e- would be a "High-Transitive" marker while -·ke- would be a "Low-Transitive" marker. I first explored these facts in a conference paper in King (1988), but have refrained from publishing anything on the topic because I wanted to muster a richer set of examples based on a more in-depth analysis of the earliest MK texts. The text with which I am most familiar is the Samgang (see King (forthcoming-b)), which does indeed present a rich variety of useful examples for the Revised Transitivity Hypothesis, and I will be publishing the results of my study in book form at a later date, but would like to take this occasion to analyze some of the relevant examples from the Wŏl kok.

For our purposes here, it is the alleged counter-examples that need the most explanation: cases where a) a transitive verb appears with the Low-Transitivity allomorphs in -·ke- ~ -·Ge-, and b) intransitive verbs that appear with the High-Transitivity allomorph -·e-, but since they are generally rarer in any case, let us begin with some examples of High-Transitive endings on transitive verbs.

Examples of High-Transitive -·a·nol/-·e·nul on Transitive Verbs

Examples of High-Transitive -·a·nol/-·e·nul are considerably less frequent than those with Low-Transitive -·ke·nul/-·Ge·nul, etc. While they do indeed tend to occur with transitive verbs, according to the Discourse Transitivity Hypothesis the more distinguishing features are telicity, punctuality, and total affectedness and high individuation of the object. Let us see how well the Wŏl kok examples accord with these features.

44) Nr. 7·wos ·kwa ma·li ·lol ·LWO·TYWUNG ·ey ·phye·asi·nol"In the middle of the road He spread out His clothes and hair …"

The verb here is ·phye- 'unfurl, unfold, spread out' followed by High-Transitive honorific -·asi·nol (which, as Shibu (1990, 242) notes, never appears as *-·^u/o·sya·nol). The High-Transitive ending here lends a nuance of finality, totality, and total affectedness of the object (Sumedha's deerskin garment and long hair), which he has spread out on a muddy road for Buddha Dipamkara to walk on (see Olof 2009).

Verba dicendi (verbs of speaking and quotation) tend more often than not to take the High-Transitive form of this ending. Wŏl kok provides several examples of :mwu-t- 'ask' and :solW- 'report to/inform a superior.' It is not always clear how to reconcile these verbs with the semantics of high transitivity. On the one hand, in reported/indirect speech, the complement functions like a direct complement. On the other hand, one can imagine that the 'asking' and 'informing' is performed once and unequivocally (punctuality).

45) Nr. 15·CCYENG·PPEN ·i mwu·lesi·nol CYEM:CYA y ·PHAN·ho·zoWwo·toy"Asked by Chŏngban, the soothsayers gave their judgement …"

46) Nr. 179ka·soy ·YANG :mwu·lusi·kwo ·nwun :men nap mwu·lesi·nol·SYEYCWON s :mal ·ol :wuz·Wi ne·ki·ni"The Honoured One asked him to describe his wife's looks, then asked him about a blind monkey, so he thought His words were ridiculous."

See also:

47) Nr. 82pwuthye s pwoncung ·ul TTANNGWANG ·i :mwut·coWa·nol"King T'an requested confirmation of His Buddhahood, so …"

Note also this causative of 'ask' in mwul-·Gi-:

48) Nr. 131TTYWUW·TTALQ ·oy QAN:PWUW ·lul ·SYEYCWON ·i mwul·Gyesi·nol"The Honoured One inquired after Chodal, who said …"

Unfortunately, none of German 'fragen' or English 'ask,' 'inquire,' 'interrogate,' etc.—with their usual implications of an iterative and unbounded action—capture the High-Transitive behaviour of these verbs in MK. Perhaps French 'demander' would be a closer approximation.

Here are the examples with :solW-:

49) Nr. 23CHYENGQUY kuy·pyel ·ul sol·Wa·nol a·pa:nim kis·ku·si·ni"Da die Diener in blauen Kleidern die Nachricht überbringen        ist der Vater erfreut und …""When he was informed by a servant, the father rejoiced        And …"(There are two more examples of this form in Nr. 27.)

Here again, Olof's use of passive "was informed" is unfortunate and inaccurate; a closer approximation would be "When the blue-clad servants informed him of/told him/reported to him the news." The action itself is punctual and total with no equivocation.

Another example:

50) Nr. 145·SYEYCWON :malsso·m ol ·HWAZIN ·i sol·Wa·nolkwo·tay :alo·sya     ·nwuns·mul ·lwo ye·huysi·ni"Die Worte des Von-aller-Welt-Verehrt sprach der Verwandlungsmensch,        und auf der Stelle verstehend entließ sie ihn unter Tränen.""The words of the Honoured One were conveyed by an incarnation,        and then she understood and woefully parted from her son."

Again, Olof's passive jars: the incarnation "conveyed/relayed/reported" the Honoured One's words.

Another interesting verbum dicendi is ˚˚ha-l- 'slander; commit slander or libel against; defame; calumniate,' etc.

51) Nr. 10:kyeci·p i ha·la·nol NILWUW y ·na·kasi·ni"Da die Frauen im Zank liegen,       verläßt Nṻpura den Palast.""Because the spouse was slanderous, Iru left."

Here again, Olof's "was slanderous" fails to capture the more active and totalizing implications of the Korean form: "The first queen committed a heinous act of libel against Iru, whereupon he left."

Another verbum dicendi that typically takes the High-Transitive ending is MK nilo-, nil·Ge 'say; tell' (→ modern ilu-, ille 'id.,' and showing the regular correspondence between MK -lG- and modern -ll-). Note that the 'big G' in nil·Ge·nul belongs to the verb stem, and not the ending. The implication is of a punctual, once-and-for-all telling and there are several examples in Wŏl kok:

52) Nr. 82pwuthye s kuy·pyel ·ul ·TTISSIN ·i nil·Ge·nul"The Earth Goddess gave news about His Buddhahood, and …"

53) Nr. 156nay ci·puy ·wa 'ys·non SAMWON ·i        ·LYWUKSO ·wa kyes·kwulq ·tol NGWANG s·kuy nil·Ge·nul""The monk who is my guest will compete with the six masters,"        he told the king …"

54) Nr. 110·mwom ·i twu·la·wo·sya SAM·SSI·HHYEN     nil·Gesi·nol"He appeared again, and spoke of the Three Manifestations, so …"

55) Nr. 116·SYWOWSSI·SSO nil·Gesi·nol QWUWTTAYA y tutco·Wo·mye"He spoke of the time of His childhood, and both Ut'aya and the Son listened to him."

56) Nr. 152CCYENGSSYENG ·u·lwo :pwoy·zoWol·ssoy SOTHYEY ·lol nil·Gesi·nol"He saw his sincerity, and He spoke of the Four Truths."

Another verbum dicendi from the Wŏl kok is KHWEN˚ho- "exhort; recommend to do; urge to do:"

57) Nr. 170kaya·m oy sa·li 'l :pwoy·Gwo ·mwom tas·ki 'l ·KHWEN·hoya·nol"Er deutet auf das Leben der Ameise        und rät, sich zu vervollkommnen …""Sudal also felt sad when informed of the ant's life,And this was an exhortation to improve himself."

A more literal rendering would be: "[Śāriputra] showed [Sudatta] the ant's life and urged him to engage in self-cultivation, whereupon …"

Finally, here are some verbs other than verba dicendi that occur with High-Transitive -·e·nul in the Wŏl kok:

58) Nr. 64CYWUK ·ul :cwasi·kwo pali ·lol te·tyesi·nolTHYEN·TYEY·SYEK ·i ·THAP ·ay ko·chwozo·Woni"He ate the gruel and threw the rice bowl away,    and the deva Chesŏk placed it in a pagoda …"

The verb is te·ti- 'throw (away); toss,' the predecessor of modern tenci-; the meaning here is more akin to "fling down; discard"—punctual and total.

59) Nr. 76·khun LYWONG ·ol ci·za     ·SYEYCWON s mwom·ay kam·a·nolCCOPISIM     ·o·lwo     :mal a·ni     ·hosi·niHWAMAN ol moyng·ko·la CWON:CYA s me·li ·yey yen·ca·nolSSINTHWONG·LUK ·u·lwo mwo·k ol kwu·ti moy·ni"Der Flußgott macht einen großen Drachen        und wickelt sich um den Körper des Von-aller-Welt-Verehrt;        der Von-aller-Welt-Verehrt schweigt barmherzigen Gemütes?? macht einen Blumenkranz        und kränzt das Haupt des Ehrwürdigen        mit Geisteskraft bindet er den Kranz fest um den Hals.""He was changed into a big dragon, winding itself around the        Honoured One,        and in His compassion He did not speak."He made a garland and raised it over the Revered One's head,        then with miraculous power he twisted the garland firmly around his neck."

The first example in kam·a·nol is with the verb ˚˚kam- "wrap around; envelope; enclose with." The spelling with -m in the coda of the first syllable at first blush suggests *kam·Ga·nol (or even *:kam·Ga·nol), but this is another example of the "grammatical consciousness" of the compiler of the Wŏl kok and his attempt with verb stems in final -m and -n to write the underlying shape consistently. Here, the dragon Mucilinda has wrapped himself (completely) around the Buddha to protect him during a violent thunderstorm (Olof 2009). The verb in the second example is yenc- "place on top of; put away on top of." The German translation here is superior to Olof's: "made a garland of flowers and 'crowned' Sŏkchon with it."

60) Nr. 179TWOW·LITHYEN ·ul :pwoysi·kwo ·TTI·WOK ·ol :pwoy·yesi·nol·SYEYCWON s :mal ·ol kis·pi     ne·ki·ni"Der Buddha zeigt ihm die Trāyastriṃśa-Himmel, und dann zeigt er ihm die Höllen        Nanda empfindet diese Reden beseeligend.""He showed him the Tori-heaven, then the hells, so he thought His words a source of joy."

The verb here is (causative, transitive) :pwoy- "show it; display it" and the implication is that the Buddha showed Nanda both the Trayastrimśās (Tori) Heaven and Hell in their entirety (or at least in a way conceived of as a once-andfor-all punctual and totalizing event).³⁵ Here is another example from the Samgang without the honorific:

61)       :SO·HHWALQ :twu ·CCO ·lol ·sse·a     :pwoyya·nol"wrote down the two sinographs for "live" and "die" and showed them to him"

Examples of High-Transitive Endings on Intransitive Verbs

Finally, the Wŏl kok evinces some examples of otherwise intransitive verbs taking a High-Transitive ending where the suddenness, categoricalness, and/or final once-and-for-all-ness trumps the intransitivity:

62) Nr. 151SSYWU·TTALQ ·i :LYEY ·lol :mwol·la hon pen ·two a·ni     two·la·nol"Sudatta kennt die rechte Höflichkeit nicht und umschreitet ihn kein einziges Mal …""In his ignorance Sudal did not even complete one full circle …"

The verb is ˚˚two-l- "turn, revolve; turn around; go around/go back" and the totality of the event (of his non-action) is reinforced by the hon pen ·two "[not] even once."

63) Nr. 137·KWUYKZIN s NGUYSIM ·i ho·ma :ep·seni·Gwa"… so the citizens were not suspicious any more."

The verb here is :eps- "not exist; die," but with the High-Transitive ending in -·eni·Gwa parallel to the much more frequent Low-Transitive -·keni·Gwa we can interpret the semantics as "[the people's doubts] had already disappeared entirely."

64) Nr. 178·SYEYCWON ·ol mac·nazo·Wo·mye cum·key nam·k i tulGye·nulkwu·chye :pwoyzop·kwo cwocco·Wa·wo·ni"Er trifft auf den Von-Aller-Welt-Verehrt,        der große Baum wird herausgerissen        und so sieht er ihn gezwungenermaßen und folgt ihm.""He met the Honoured One, and a big tree was raised,        so he was clearly visible, and had to follow Him back again."

The verb is tul-Gi- "be/get raised," a passive derived from tu-l- "hold up; lift; raise." So the basic semantics are intransitive, but here we must imagine a nuance along the lines of the German translation in Sasse and An: "the great tree [behind which Nanda was hiding] was suddenly raised up out of the ground, whereupon …" The nuance is of a sudden and calamitous or momentous and punctuated event.

Other High-Transitive Examples

65) Nr. 28…·mwom ·khe kuwu·nil LYONG ·olhyen·mas pelG·ey pi·nul ·ul spola·nywo"… countless leeches were sucking underneath the scales of the big squirming dragon."

Martin (1992, 426) analyzes this form as effective modifier -·an (with allomorphs -·en, -·kan/-·Gan, and -·ken/-·Gen) + interrogative postmodifier ·ywo and gives the following example:

66)·hyen ·ma s CANG-NGEM ·kwa ·hyen ·ma s KWONG-·YANG ·iSSYANG-·SSYWUY ·lol phyea·nywo (1459 Wŏl 17: 23b)"how much pomp and how much offering of food have unfolded favorable omens?"

Another example form the Wŏl kok:

67)·SSYWUY·SYANG :pwoysi·non ZYELOY s·kuy    ·hyen·ma s ·CYWUNGSOYNG ·i me·li :cwos·soWa·nywo"Wie viele Lebewesen beugen ihr Haupt vor dem gute    Vorzeichen tragenden Vollendeten …?""… countless people bowed deep under His auspicious five-coloured cloud."

Here the verb is MK ˚˚cwoz-, cwo·za meaning "kowtow" + deferential -:zoW-. Though the putative object me·li "head" is non-individuated, if we imagine a multitude each making their one-time kowtow, the punctual, telic semantics are confirmed. It is also interesting to note that all three examples here of -a·nywo collocate with ·hyen·ma s in a rhetorical question that functions as a kind of emphatic construction, likewise compatible with High-Transitive marking.

68) Nr. 92THAM·YWOKSIM :kyesi·kenma·lon hon :nath thelek :spwu·n ul    KWONG·YANGKWONG·TUK ·ey ·NYELQPPAN ·ol ·TUK·hoya·ni"Even before He prevailed over avarice, they obtained Nirvanaon the merit of worshipping just one single hair of His."

The first highlighted form in :kyesi·kenma·lon is uncontroversial, with Low-Transitive concessive -·kenma·lon on :kyesi- "be (honorific)," but the second form is more interesting. The verb is ·TUK˚ho- "obtain; acquire; secure; get" + High-Transitive -·a·ni (the corresponding Low-Transitive form is ·hoke·ni) and here once again we must imagine a sudden, once-and-for-all attaining of Nirvana.

69) Nr. 127ku·nal s SSYANG·SSYWUY ·lol :ta sol·Wa·lingis·ka"… die guten Omen, wer könnte sie alle beschreiben!""Words fail to describe the propitious omens of that day!"

The verb is :solW- "report to a superior" and Martin (1992, 417) analyzes the ending as -a·l i '-ngi s ·ka: effective -·a + prospective modifier -·^u/ol + postmodifier i ("fact")… + elided copula + polite -ngi + adnominal s + interrogative ka. In any case this is an example of High-Transitive -·a in a rhetorical-cum-exclamatory question: "how could one possibly tell all of the propitious signs?!" The totalizing nature of the semantics is reinforced by adverb :ta "all; in their entirety."

Examples of Low-Transitive -·ke-/-·Ge- with Transitive Verbs

Perhaps the most interesting "exceptions" are those of -·ke-/-·Ge- occurring with otherwise transitive verbs. The Discourse Transitivity Hypothesis predicts that in such cases, the semantics will skew towards atelicity, non-punctuality (habitual or repetitive or prolonged activity), negation, irrealis, and unaffected and/or nonindividuated objects.

70) Nr. 14THYEN·AK ·ol ·CWUWkhe·nul CYETHYEN ·i cwocco·Wo·ni"Heavenly music was performed, and all devas followed Him…"

The verb here is ·CWUW˚ho- "perform [music]" and the sense is clearly of an atelic, protracted and/or repeated activity. Thus, even though the clause contains an overt object marked with the object marker ·ol, the verb morphology is Low-Transitive. Again, Olof's passive is somewhat unfortunate: "[unspecified agent(s)] were playing heavenly music, whereupon all the Heavens/devas followed him …"

71) Nr. 16THYEN ·kwa :KWUY ·Gwa y tut:copke·nul"devas and spirits listened to Him [as he explained the Law …]"

Here the verb is tuT- "listen; hear; obey" + deferential -:zoW- and the sense is clearly durative: the devas and spirits were listening to him over a period of time.

72) Nr. 100QAN:PWUW ·lul :mwutcop·kwo ·PPEN :cwasywo·sye :CHENGkhe·nul"He greeted Him courteously, and invited Him in for a meal."

At first blush, this example of a verbum dicendi with Low-Transitive -ke·nul would appear to be a counter-example to what was said above about this class of verbs, but if one imagines Kāśyapa receiving Buddha and enjoining him several times or repeatedly to eat (as would be polite), there is not problem. In any case, Kāśyapa invited him to eat more than once.

73) Nr. 100PPANG ·ol a·ni patco·Wa ·PEP·u·lwo mak:sopke·nul"He did not offer a room and with tricks tried to stop Him …"

The verb is mak- "block; fend off" + deferential -:zoW- and the implication is that Kāśyapa tried (over a period of time and with various subterfuges) to deter Sŏkchon from staying (here Olof's translation works nicely). It is also irrealis.

74) Nr. 187CYETHYEN ·tol cwo:ccopke·nul KWANGMYENG ·ul ne·phi·sya"Whilst the devas followed Him, He emitted His radiance …"

The verb cwoch- "follow," while transitive, is usually non-punctual and atelic; literally, then: "While the devas were following Him…"

75) Nr. 189·SYEYCWON s kulu·mey ·yey KAM·LWO ·lol spu·liGe·nul"In the shadow of the Honoured One they sprinkled Sweet Dew…"

The verb spu·li- "sprinkle" has a durative, distributive, and non-individuating semantics. Thus: "they were sprinkling Sweet Dew (here and there)…"

76) Nr. 43·CCYENGKETHYEN :CWOWPPYENG ·i cwu·kun pelG·ey towoy·ya·nol        ·pwosi·kwo·za :anti·si     ·hosi·ni"Clean Vase, the Deva of the Pure Abode, became a dead insect,        and upon seeing this, it was as if He understood it all."

The MK verb towoy- "become" is as intransitive as they come, but if one imagines a sudden metamorphosis or once-and-for all transmogrification, this example can be translated as: "suddenly turned into a dead insect." Similar instances of High-Transitive morphology in examples involving metamorphosis are:

77) SSYENG·PPWULQ ·ho.yan ·t i … ·KEP ·i.n i"it has been kalpas (eons) since I became a Buddha" (Wŏlsŏk 17: 22a)

78) CHYE·KWEN towoy·yan ·t i SAM-NYEN ·i :mwot ·cha i·sye (Sŏk 6: 4ab)"it is less than three years since he suddenly became a man with wife and children, and …"

Conclusions

Curiously, since Ko Yŏnggŭn's seminal article of (1980), surprisingly little more has been published on the question of transitivity and the MK -·ke/-·Ge- ~ -·e-alternation. Kim Yŏng'uk (1996) makes a bold attempt to mobilize pre-15th-century kugyŏl data on the problem, and Yi Kŭmyŏng (1999) likewise dwells on pre-15th-century sources, but is more focussed on big-picture diachronic questions than on the synchronic question of how these morphemes actually functioned in the earliest Chŏng'ŭm texts. The most interesting post-Ko paper I have seen is Chŏng Hŭich'ang (2004) who adds a number of new and useful observations. For example, he notes that patterns expressing purpose or intention (ŭido) like -·u·l i '·la ˚ho-, -·kwo ·cye ˚ho- and -·wo·lye ˚ho- typically take (Low-Transitive) -ke·tun, and even cites Hopper and Thompson (1980) in his bibliography, but does not otherwise engage with their ideas. Thus, on the face of it, the fact that patterns expressing intentions take Low-Transitive endings could be a counter-argument to the Revised Transitivity Hypothesis, but the key point with these patterns is not the volitionality, but the irrealis nature of these patterns: a verb form does not qualify for High-Transitive endings unless and until the activity has actually taken place (which is one reason negated transitive verbs also tend to take Low-Transitive endings). Thus, 'trying to do' and 'intending to do' patterns will normally take Low-Transitive endings, whatever the basic transitivity of the verb stem in question. For example (from the Wŏl kok):

79) Nr. 152CCYENGSSYENG·u·lwo :CHYENG·hozop·kwo              CYENG·SYA ci·zwu·lye khe·nul"Devoutly he asked permission to build a seminary …"

Another valuable observation is that the causative pattern in -·key ˚ho- always takes (High-Transitive) ·ho.ya·ton rather than ·hoke·tun. Chŏng Hŭich'ang also mentions verba dicendi, but offers little in the way of detailed analysis, and overall tries to build a case around the vague notion of haengdongsŏng or "activity" as the triggering factor for (what I am calling) High-Transitive endings. He does, however, present some nice minimal pairs as follows with the verb mek- "eat" (2004, 423–24):

80) chon ·kuywun ·ulwo kisnon ki·chw·um un ·chwum i molkoni (Kugŭpkan 2: 9a–b)        te·wun mul me·ketun :cams·kan mecuk·hono·ni"In the case of a cold cough, the spittle is clear, so if you drink down/swallow hot water, it subsides for a short time."[RK: competely affected object and/or punctual action]

81) :salo·m o·lwo ·sye YANG ·ol me·ke·tun YANG ·i cwu·ke :sa·lom (Nŭngŏm 4: 30a)towoy·mye :salo·m i cwu·ke YANG towoy·ya"When, as a human, you consume/eat up a sheep, the sheep dies and turns into a human, while the person dies and becomes a sheep."[RK: punctual action, totally affected object]

82) (Kugŭpkan 2: 82a)·chwum i ·thiwa·thye ·pap ol :cye·ki mekke·tun"If spittle wells up [in the throat], and you eat [rice] sparingly …"[RK: object not totally affected]

83) ip i     pol·la ipsi[G]wul i :cye·ke ·pe·liti     :mwot·hoya        (Kugŭpkan 3: 5)        pap ·mwot mekke·tun"If you are unable to eat because your mouth is level and your lips are too small to open …"[RK: negation/irrealis]

84) (Wŏlsŏk 10: 24b)mot·aki ·non :pe·m i mu·le me·ke·nul"As for the oldest child, a tiger carried him off and ate him up."[RK: individuated object, punctual action]

85) (Wŏlsŏk l0: 25b):pem ·kwa ·ilhi ·tol·h i mwu·tem ye·le cwu·ke·m ul mekke·nul"The tiger(s) and wolves opened up the graves and were eating away at the corpses."[RK: non-individuated object and/or non-punctual or protracted action]

At any rate, I hope the discussion and examples above will serve as a stimulus to more detailed and nuanced investigations of the vexed semantics of the MK -·ke/-·Ge ~ -·e alternation.

In closing, let me mention the important questions of vernacularization and 'diglossia' in the history of Korean language and writing. In many ways the most important work to appear in recent years on questions of vernacularization in comparative perspective is that of Sheldon Pollock—especially Pollock (2006) with its discussion of vernacularization in the Sanskrit Cosmopolis as compared with vernacularization in European Latinitas. It would require more than another paper to explore this topic in detail from a Korean perspective, but I note here that one of Pollock's central arguments about the first wave of vernacularization in South Asia was that the crucial impetus invariably came from a royal court and royal patronage (and not from religious communities). The parallels between Pollock's South Asian examples and the 15th-century Chosŏn example with King Sejong and his closest family members and associates in the Korean royal court inventing a new vernacular script and penning highly self-conscious vernacular literary works like the eulogies in the Yongbi and the Wŏl kok, the prose in the Sŏkpo sangjŏl, and the harmonization of the two in the Wŏrin sŏkpo, seems to call for closer comparative scrutiny.

Finally, "diglossia." In King (2015) I have already outlined a number of the problems with characterizing Korea's complex pre-modern ecology of spoken and written registers as 'diglossic,' but I close by noting that research on the Yongbi and the Wŏl kok is frequently guilty of the same oversimplifications and caricatures that so often take the place of more nuanced investigation. A typical example along these lines is Chŏng Soyŏn, an otherwise fine scholar who nonetheless seems to have built an entire research profile around utterly uncritical use of the term "diglossia" (see Chŏng Soyŏn 2009 and 2015). Thus, in Chŏng (2009, 188) she writes: "Historically, our nation has used two languages" and then proceeds to posit the Yongbi and the Wŏl kok as two antipodal texts on a simplistic scale of Sinophilic (Yongbi with its Chinese-style poems, Sinitic philological apparatus, privileging of sinographs, and concomitant lack of vernacular pronunciation glosses, etc.) vs. Korea-philic (Wŏl kok with all the orthographic features enumerated at the beginning of this paper). The situation is far more complex and interesting than anything the term "diglossia" was originally designed to describe, and we do Korean language and writing a disservice by continuing to characterize Korea's rich and variegated system of spoken and written registers over the centuries as "diglossic."