It is by now a well-established fact in Korean kugyŏl glossing studies that the Korean practice of glossing canonical hanmun 漢文 texts underwent a major change between the late Koryŏ and early Chosŏn periods. In essence, an earlier system of ‘interpretive kugyŏl’ (sŏktok kugyŏl 釋讀口訣) that used both sides of the hanmun line in conjunction with ‘back marks’ (yŏktokchŏm 逆讀点) in such a way as to convert (in effect, translate) the hanmun text into a Korean text with Korean nominal particles, Korean verbal endings and Korean word order while also rendering~reading some of the sinographs in vernacular Korean (see Chung 1999 and Paek 2005), gave way by the end of Koryŏ to a system of ‘sequential kugyŏl ’ (sundok kugyŏl 順讀口訣 ; also referred to as ŭmdok kugyŏl 音讀口訣) that used only the right-hand side of the hanmun line, read the sinographs exclusively in their Sino-Korean pronunciations, and placed vernacular Korean morphological markers at natural clause breaks in such a way as to punctuate the text while leaving the Chinese word order intact (see Nam Kyŏngnan 2005/2006 and 2009).

However, despite the relative abundance of early Chosŏn texts with both kugyŏl and han’gŭl2 glosses, the ongoing hunt for and focus on Koryŏ-period (and earlier) materials has led scholars by and large to neglect texts from the mid-fifteenth and sixteenth centuries. In my survey of pre-Imjin Korean texts with kugyŏl glosses held in North American library collections (King 2010), I called attention to an edition of the Ch’ŏllo kŭmgang kyŏng 川老金剛經 held in the Asami Collection at the University of California-Berkeley3 that falls into this quasi-‘no-man’s land’ in kugyŏl studies, and this paper is a more detailed examination of that text.

In his excellent description of the Asami collection, Fang (1969) describes this edition as follows: “28.40 Ch’ŏnno kŭmgang kyŏng 川老金剛經 [천노금강경] by Tao-ch’uan 道川 (twelfth century). 1387 and earlier. 1 vol. Woodblock.

What is most interesting about this work (and Fang’s description of it) for our purposes here is the fact that it contains numerous kugyŏl glosses; in fact, the entire main text of the Kumgang kyŏng and Yefu Daochuan’s poetic commentary is glossed with rather elegant kugyŏl. Oddly, Fang in this rare instance makes no mention of the presence of these markings in his description of the edition, despite his otherwise careful attention to this detail in his descriptions of other works in the Asami collection. Equally odd is the fact that the Haeoe sojang Han’guk chŏnjŏk munhwajae website (http://koreanbooks.nricp.go.kr/) appears to have overlooked this important work. O Yongsŏp (2008:90-97) also gives a description of this work, but-true to form with his Korean sŏjihak colleagues-neglects to mention the kugyŏl annotations4 .

My initial excitement about this text owed to two reasons: its status as a late-Koryŏ edition (of which there are precious few in North American collections), and the verbal morphology reflected in the annotations. At first blush, the latter appear to have been added in the late fifteenth (or perhaps early sixteenth?) century; if so, this would make them the oldest kugyŏl markings available in any North American library collection. The kugyŏl markings are also interesting for their hybrid nature; the Middle Korean politeness marker -ngi5 is rendered here with the hunmin chŏng’ŭm symbol , while most of the other kugyŏl glosses are marked with traditional kugyŏl graphs (yakcha kugyŏl 略字口訣 ―kugyŏl glosses created by abbreviating full-form Chinese characters). Traditional kugyŏl annotational conventions had no straightforward means of representing this syllable (-ngi in Yale romanization) with Chinese characters or their abbreviated forms. Thus, Kim Namgyun (2000:95) notes that whereas the edition he examined (which he dates to 1387, and which he assumes to have been glossed shortly thereafter) renders this Middle Korean politeness marker with the clumsy combination + , and other earlier kugyŏl-annotated editions of this text used just or (i.e. i, ignoring the velar nasal).

In general, Kim Namgyun (2000) provides a useful point of comparison for a more detailed examination of the Asami edition’s kugyŏl. But first please recall that the Kŭmgang kyŏng 金剛經 was one of the four main textbooks in the early Chosŏn-era Buddhist clerical training programs. Nam Kwŏnhŭi (2000:163) notes that the Ch’ŏllo or Yabu (Yefu 冶不) version of the Kŭmgang kyŏng was the most popular of various different editions, as witnessed by the fact that numerous copies of the 1387 edition survive today (many of which have kugyŏl annotations). Nam Kwŏnhŭi (ibid.:173) distinguishes two main types of edition of the Ch’ŏllo Kŭmgang kyŏng― (A) and (B). The Asami edition is type (A).

In addition to three officially designated ‘treasures’ also noted by Chŏng Chaeyŏng (2007)―Pomul 974, 1127 and 919 (the latter bound together with the Pŏmmang kyŏng 梵網經)―Nam writes that he has examined six other editions. Kim Namgyun (2007:92) describes four editions in personal collections (the same as those described by Nam) as well as one edition in the Tōyō Bunko, but the Asami edition appears to have gone unnoticed thus far in Korea (except, that is, for O Yongsŏp 2008, who neglects to note the presence of kugyŏl glosses). For more on the Kŭmgang kyŏng in general, and on the Ch’ŏllo Kŭmgang kyŏng, in particular, see also Chŏng Chaeyŏng (2007), Ko Ikchin (1974), Nam Kwŏnhŭi (1997) and Nam Kwŏnhŭi (1998:559-562).6

1This work was supported by the Academy of Korean Studies Grant funded by the Korean Government (MEST) (AKS-2011-AAA-2103). My thanks also to Professors Nam Punghyon, Jung Jaeyoung, Gernot Wieland, and John Whitman for helpful comments on an earlier version of this paper presented at the “Workshop on Korean Historical Linguistics, with a Special Focus on Kugyŏl Materials,” hosted at the University of British Columbia, July 21-23, 2010. I am also grateful to two anonymous peer reviewers for their comments and corrections.  2Because the term ‘han’gŭl’ for the vernacular Korean script was not coined until the 19teens and did not gain wide acceptance until the 1930s, it is—technically speaking—anachronistic to use this term to refer to the Hunmin chŏng’ŭm 訓民正音 (or Ŏnmun 諺文) before the modern period, but in this paper I follow standard South Korean (anachronistic) practice for the sake of convenience.  3As of April 9, 2013, the entire text is available online for viewing and PDF download at the Internet Archive at this URL: http://archive.org/details/chonnokumganggyo008800rich.  4See also Chŏn (2006) for an annotated translation into Korean of this work and the more recent Yŏngnam Taehakkyo Minjok Munhwa Yŏn’guso (eds.)(2012) for a reproduction and transcription of the mid-13th century edition held at Yŏngnam University along with an overview of relevant research. (This Yŏngnam University edition does not have any kugyŏl glosses.)  5Henceforth, all pre-modern Korean linguistic forms are rendered in Yale romanization as per Martin (1992). Other Korean forms (proper names and technical terms) are rendered in McCune-Reischauer romanization. Transcription conventions for the glossed text are as follows: Chinese text from the Diamond Sūtra is in bold, while text from the Ch’ŏllo commentary is in plain. Righthand glosses are in subscript, while lefthand glosses are in superscript.  6The officially designated treasures are: Pomul no. 974 owned by the Seoul Museum of History (Sŏul Yŏksa Pangmulgwan); Pomul no. 919 (bound together with the Pŏmmang kyŏng 梵網經) owned by the Adan Mungo; and Pomul no. 1127 owned by the National Library of Korea. However, this latter copy appears to be unglossed. We can also mention: an early nineteenthth-century edition owned by Tongguk University; an edition from 1869 owned by Koryŏ University and the National Library of Korea; the Toyo Bunko copy; the Tenri Library copy from the Imanishi collection (glossed, but most of the glosses have been obliterated by a later reader); a nineteenth-century edition with glosses owned by Waseda University Library; etc. Chŏng Chaeyŏng and Kim Sŏngju (2010) report a late Koryŏ~early Chosŏn glossed fragment found recently at Pulgapsa Temple.

Please refer to the chart below for a full listing of the kugyŏl graphs deployed in the Asami edition and for comparisons with the edition studied by Kim Namgyun (2000) and the Song Sŏngmun edition of the Nŭngŏm kyŏng 楞嚴經 (1401). The latter two editions are thought to contain markings that can be considered typical or representative of early Chosŏn sundok (順讀 ‘sequential reading’) kugyŏl7 .

Including the hunming chŏng’ŭm graph , then, our Asami edition deploys a total of 53 different kugyŏl graphs:

This is already a significant decrease in the total number of such graphs deployed in Koryŏ-era kugyŏl texts, but not quite so low as the 40 or so typically found in mid- to late Chosŏn texts.

The following kugyŏl graphs are new (new, that is, compared to the edition examined by Kim Namgyun):

7As opposed to the Koryŏ-era sŏktok (釋讀 ‘interpretive reading’) kugyŏl that essentially translated the hanmun text into vernacular Korean with Korean word order.  8The chart is adapted from that in Kim Namgyun (2000:93-95), but also draws on Nam Kyŏngnan (2009).  9Note that is often rendered in such a way that it is difficult to distinguish from (tu).  10Henceforth, unless otherwise noted, all forms in italicized Yale romanization placed beneath kugyŏl markers represent the corresponding Middle Korean han’gŭl forms in the Samgahae 三家解.  11But note that there are also two examples of in this same vocative usage: (金剛 4b:09) and (金剛 5a:09). In both cases, the 1482 Samgahae ŏnhae records ya.

The Asami edition of the Ch’ŏllo Kŭmgang kyŏng contains a total of approximately 210 different t’o (grammatical particles and endings), as opposed to the 302 found by Kim Namgyun in the edition with the earlier glosses that he examined. The sections below examine locative endings (2.1), Ho-deletion vs. o-deletion-plus-aspiration (2.2), optional deletion of –s- (2.3), and (za) and (iza) (2.4). 

The locative markers deployed in the kugyŏl glosses reveal the transitional nature of the overall kugyŏl glossing in the text. That is, we find a mixture of earlier (Koryŏ) locatives in just (or with the the topic~contrastive marker ) and locatives corresponding grapheme-for-grapheme with the MK locatives ·ey, ·ay, and ·yey as they occur in the Kŭmgang kyŏng samgahae. MK also had locative markers in ·oy and ·uy, but the Ch’ŏllo Kŭmgang kyŏng does not have anything corresponding to these. Thus, the locative markers found in our text are as follows:

The kugyŏl locative markers correspond exactly to the expected MK shapes in the majority of cases, but in a number of other examples (clustered mostly toward the beginning of the text), the marker appears, corresponding to all three of MK ·ey, ·ay, ·yey in the Samgahae. This is an orthographic archaism reflecting an earlier stage of kugyŏl marking which, in turn, was probably reflected in one or more earlier kugyŏl-glossed version(s) of the texts consulted by the glossator(s) of the Asami edition.

2.1.1. Examples of corresponding to Samgahae ey, ay, yey

1) 에 ey

2) 애 ay

3) 예 yey

2.1.2. Examples of corresponding to Samgahae ayn

2.1.3. Examples of (ay) corresponding to Samgahae yey

Examples of corresponding to Samgahae yey are again transitional or hybrid, in the sense that a) they preserve the older uniform locative and b) they nonetheless incorporate the final y of MK yey (and ay and ey). Note that all MK examples with locative yey occur after forms ending in i or y.

In contrast with these five examples of corresponding to Samgahae yey, there are twenty-seven examples of = Samgahae yey; again, these are always after forms ending in i or y (including 西 and 水 , which had the MK shapes syey and sywuy, respectively). For example:

Note that there is also one example of corresponding to yeyn:

2.1.4. (ey) corresponding to Samgahae ey, yey

There are more than one hundred examples of = ey, and just two examples of = yey:

Needless to say, these two examples should be treated as mistakes for on the part of the glossator.

It is not uncommon in glossed texts of all periods for the stems i- and ho- of the copula and the dummy verb ‘do; be’, respectively, to delete, leaving just the verb endings in the gloss. In our text, too, forms of the MK verb ho- ‘do; be’ from the Samgahae beginning with effective –ke- are frequently treated in the kugyŏl glosses as though the preceding stem in ho- were deleted, whereas the ŏnhae glosses show aspirated khe- from deletion of the first vowel in h[o]-ke-13 :

It is tempting to interpret examples like the following as somehow reflecting an archaic holdover of an interpretive glossing practice whereby 有 was read in vernacular Korean ([is-]ketun), but there is no other evidence for such a suggestion, and here we must assume ho-deletion in light of the ŏnhae gloss in ketun:

A similar deletion phenomenon can be observed with = -kwo or khwo (from h[o]-kwo):

But one also finds examples of just :

We also find cases of optional deletion of ho- before ya (i.e., deletion of before ):

Our text offers nine examples of what Martin (1992:736) calls the modulated processive emotive indicative assertive in -nwosta (< *-(·)n[o]-·wo-s-ta), rendered as - in the kugyŏl glosses.

However, there are also two examples where the same ŏnhae gloss honwosta is rendered rather than :

It is tempting to interpret these examples as cases of optional deletion of -s- in this ending, but the following lone example of kugyŏl (normally representing the extremely frequent copular form in ·ilwo·ta) corresponding to the ŏnhae gloss in honwosta suggests that we are dealing instead with a scribal error on the part of the glossator, who was mechanically transferring glosses from the Samgahae to his text:

The kugyŏl character appears in just two contexts in our text: as part of the verb form corresponding to MK ·ho.ya ·za and to hoyaza in the ŏnhae glosses (ten examples), and as the nominal particle ~ corresponding to MK ·za ~ ·i·za (usually just nominative ·i + ·za, but see Martin 1992:592 for other possibilities). This particle sequence had an alternate form in y·za, usually only after Chinese characters ending in a vowel; han’gŭl words ending in a vowel usually took just ·za15. Our text has just two examples of the particle. Because the kugyŏl glosses had no way to indicate a semi-vowel, the yza of the Samgahae gloss in the second example was carried over as kugyŏl:

12There is just one example of   13One of the anonymous reviewers of this article made the interesting observation that ho-deletion is in fact rather infrequent in kugyŏl texts compared to copular i-deletion, and therefore questioned whether the examples in this section should be taken as instances of ho-deletion. To be sure, in some cases like these it is entirely possible that the deleted verb stem is the copula rather than ho-. However, the presence of parallel kugyŏl forms with ho- in the text along with the fact that the glossator was often copying from Samgahae han’gŭl glosses that clearly included ho- make this a rather vexed question. Note also that examples like 金剛 3b:08 一枝無孔笛 with just (and where the Samgahae shows hoya in any case) cannot be taken as instances of copular deletion, since there was no such combination as * (copula + -a). A comparison with other glossed versions of the Ch’ŏllo Kŭmgang kyŏng will likely clarify this issue somewhat, but note also that Nam Kyŏngnan (2009:89) records frequent examples of deletion of ho- before -ke-, and indeed, claims that deletion was the default practice. She also writes in footnote 29 on the same page that the materials she studied frequently delete ho- before .  14NB: The ŏnhae has 惜醉人 instead of 識醉人.  15However, the ŏnhae translation in the Samgahae renders our second example here as y·za.

By now it should be clear that the kugyŏl glosses in the Asami Collection of the Ch’ŏllo Kŭmgang kyŏng bear a close relationship to those in the 1482 Kŭmgang kyŏng samgahae 金剛經三家解 .

One obvious piece of evidence suggesting that the glossator relied heavily on the Samgahae for his glosses is the occasional appearance of kugyŏl glosses with shapes atypical for traditional glossing practice. For example, the traditional kugyŏl shape for the topic~contrast particle was just , but our text has five examples with the shape  corresponding to the MK shape ·non and to the ŏnhae gloss non:

And in yet another case the same gloss is written in han’gŭl, suggesting yet again that the glossator was copying mechanically from the Samgahae:

In the following example, the Ch’ŏllo Kŭmgang kyŏng and the Samgahae ŏnhae gloss agree in recording and hwomiye:

But the (or at least a) glossator of the Ch’ŏllo Kŭmgang kyŏng has added han’gŭl 호미녀 hwomistonye in the righthand margin, while the Samgahae ŏnhae translation (not the ŏnhae kugyŏl gloss) reads:

We seem to be dealing with an older vernacular glossing pattern for 何況 ... in N ·i·ye (or nominalized V-·w^o/u ·m i·ye) that was replaced by MK N ·i·sto·n ye (V-·w^o/u ·m i·sto·n ye).

One of the more interesting features of the glosses in our text is the use of han’gŭl alongside abbreviated character kugyŏl glosses to create hybrid glosses for verb ending strings that incorporate the MK bound politeness marker –ngi17 . As noted already in the introduction above, earlier (pre-han’gul) glosses from late Koryŏ and early Chosŏn had to resort to combinations like + (reported by Kim Namgyun for the glossed copy of the Ch’ŏllo Kŭmgang kyŏng he examined) or just or (as in, for example, the Adan Mungo copy of our text). Here is a list of all the different hybrid glosses using in our text along with their corresponding ŏnhae glosses (note the occasional appearance of ᅌᅵᆺ):

While our text is relatively consistent in deploying exactly where the Samgahae does, there are nonetheless a few lapses:

It is tempting to interpret these cases of -less glosses as holdovers (archaisms) from earlier glossed versions of our text. This is especially true in the cases of the glosses in on 不 , as these are yes-no questions (‘... or not?’) where other glossed versions of our text often leave the preceding main verb (生 in our examples here) unglossed and gloss only 不 . This suggests that the glossator of our text relied on an earlier glossed version for his glosses on 不 , but transferred the gloss with from the Samgahae because his earlier glossed version had to gloss to copy.

Besides the hybrid glosses seen above, there are a few other examples of glosses either partially or wholly in han’gŭl demonstrating yet again the glossator’s reliance on the Samgahae.

The following example mixes han’gŭl and an abbreviated character gloss for MK ·i·ye (see also section 3.2 above):

In two other examples, the han’gŭl kugyŏl is transferred intact from the Samgahae into the position reserved for traditional abbreviated character glosses at the bottom lower righthand corner of the Chinese character:

16Unless preceded by “[lit.],” translations from the Diamond Sūtra are taken from Price and Wong (1969), and translations of the Ch’ŏllo commentary are taken from Qian Su (2002).  17See Martin (1992:714) and 1996 for this marker.

Thus far we have focused on the glosses deployed in the traditional position for ‘sequential’ (sundok 順讀) kugyŏl glossing: the righthand margin of the hanmun line, commencing from the bottom righthand corner of the last character in the glossed string. Historical surveys of the development of kugyŏl glossing practices in Korea point out that Koryŏ-period glossing was similar in nature to Japanese kanbun kundoku 漢文訓讀 . The glossing conventions used both sides of the hanmun line in one integrated process that made extensive use of vernacular glosses as opposed to just Sino-Korean chunks read in Sino-Korean pronunciation: the reader proceeded first down the righthand side until a ‘go back’ sign (yŏktokchŏm 逆讀點) was reached, at which point he went back up the lefthand side of the line to the first gloss there, and continued up the lefthand side until he ran out of ‘back marks’, at which point he resumed reading from the point with the last gloss on the righthand side. Both sides of the line were used in a process that rendered the hanmun text into a Korean text in Korean word order with substantial use of vernacular glosses for many of the Chinese characters and with Korean nominal and verbal morphology throughout.

Diachronic studies of Korean glossing practice emphasize that this sort of interpretive glossing practice that reached its peak during Koryŏ was waning by the end of that dynasty and was more or less moribund by early Chosŏn. Instead, it was replaced by a sequential glossing (also known as ŭmdok 音讀 kugyŏl or ‘sound’ glossing because the Chinese characters were read according to their Sino-Korean pronunciations rather than their vernacular glosses) practice that read the hanmun straight down the line in chunks of what William Skillend (in the slightly different context of late Chosŏn vernacular fiction) called ‘Sino-Korean holophrase’―now without the benefit of ‘back marks’ or accommodations for Korean word order, and therefore without the need for glosses on the lefthand side of the line. The exact chronology and details of the shift from interpretive to sequential (or ‘sound’) glossing, though, are not entirely clear, and sequentially glossed texts from early and even mid-Chosŏn can contain traces of techniques for indicating both vernacular Korean word order and vernacular Korean glosses (hundok 訓讀) of Chinese characters18 .

It is therefore yet another interesting feature of our text that it occasionally deploys glosses―often in han’gŭl, but also in abbreviated character kugyŏl and mixtures of the two―along the lefthand margin of the hanmun line of a more ‘interpretive’ (sŏktok 釋讀) type.

The lefthand glosses in the Ch’ŏllo Kŭmgang kyŏng have two sources: in the case of the main text from the Kŭmgang kyŏng (which is not provided with ŏnhae translations in the 1482 Samgahae), the glosses are copied over from the Kŭmgang kyŏng ŏnhae translations (1464), and in the case of the Ch’ŏllo 川老 commentary, the glosses are copied over from the ŏnhae translations in the Kŭmgang kyŏng samgahae ŏnhae. However, both types of lefthand gloss stop abruptly at the point where volume one of the Kŭmgang kyŏng ŏnhae ends, suggesting that the glossator had access to only the first of the two volumes, and simply stopped glossing at that point.

Below are some of the more interesting examples of lefthand glosses, along with the source passages and English translations. Lefthand glosses are marked in bold, as are the matching strings in the source text.

4.1.1. Glosses copied from the K？mgang ky？ng ？nhae (1464)19

The very first example is of especial interest because it is also the only example in our text of true ‘interpretive’ glossing: three of the characters have dots to their left indicating the Korean word order (the dots are vertical in the original)20 :

Note also that the glossator’s analysis of the ‘should; ought to’ pattern glossing Chinese 應 suggests a contemporary analysis as -(^u/o)lq·ti + copula, whereas Martin’s analysis (reflected in the transcription) is more diachronic.

The next example and many others following it suggest that a kind of vernacular glossing (hundok 訓讀) regime lurked vestigially just below the surface, even while actual glossing practice had already become largely sequential. The gloss in 샨 syan on 說 ‘speak; say; tell’ can be interpreted as a relic of the earlier practice of ‘final sound suffixation’ (marŭm ch’ŏmgi 末音添記) whereby the vernacular gloss to a Chinese character was hinted at by adding a gloss denoting the final segment or syllable of the vernacular equivalent. Thus, the 샨 here indicates that 說 is to be read as the Korean verb stem nilo- + honorific modulated modifier -·syan(< *-si-·wo-n):

In this particular case, the point is reinforced further by the marginal gloss in the bottom left gutter: ... (nilosya[n] with the final ㄴ cut off. Finally, note also the on 我 “I”, reinforcing evidence (as if it were needed) that (di)graphs like ㅐ (ㅏ + ㅣ) really were diphthongs at the time of glossing in the late fifteenth or early sixteenth century.

More examples of marŭm ch’ŏmgi 末音添記 ‘final sound suffixation’ include 當^기 = pan·toki, = :nyey, = il·hwum i, = :nayywo·m i and 無^슨 = :ep·sun in the examples below:

In the following example, the glossator has glossed 無 with epsulsoy where the 1464 source text has :epsul·ssoy, but this is consistent with the abandonment of geminate spellings in cases like this starting with the Wŏn’gak kyŏng ŏnhae 圓覺經 諺解 in 1465:

In the next example, there is a difference in the interrogative endings between the ŏnhae and the kugyŏl glosses (ka and Ga in the kugyŏl glosses vs. ·ye in the Kŭmgang kyŏng ŏnhae), suggesting that the glossator was not always entirely mechanical in his work:

And the same (very occasional) independence of spirit can be seen in the next example, where the ㄹ gloss on 他人 is not supported by the ŏnhae, and is thus the work of the glossator21 (The initial ㄹ in 리를 is a mistake, and the final ㄹ is hard to make out):

Likewise, in the next example the glossator has subsituted for what should be :

The following righthand gloss also happens to be a nice example of the nominal use of MK processive modifier –non:

4.1.1.1. Other glosses from the K？mgang ky？ng ？nhae

In just a couple of cases the Kŭmgang kyŏng ŏnhae glosses are entered not in the lefthand margin but directly at the righthand side of the relevant Chinese character or just beneath it, showing again that the glossator was using this text as a crib. For example, in the following two cases, the translation gloss is placed immediately to the right:

4.1 .2. Glosses copied from the Samgahae ？nhae translation

The most interesting glosses copied over from the Samgahae are those that are rendered in ‘final sound suffixation’ (marŭm ch’ŏmgi 末音添記) style―examples like the following, rendered in more detail below:

This example is interesting because the Sino-Korean pronunciation for 慧 is glossed as colloquial vernacular hyey in contrast with the artificial Tongguk chŏng’un 東國正韻 reading HHYWUYEY in the Samgahae. Note also the mixture in mid-word of han’gŭl and abbreviated kugyŏl gloss in 어리 = e·lini ·lol.

4.1.2.1. Other glosses from the Samgahae ？nhae

As with glosses copied over to the lefthand margin from the Kŭmgang kyŏng ŏnhae, in a few cases the Samgahae ŏnhae glosses are entered not in the lefthand margin but directly at the righthand side of the relevant Chinese character:

18In this regard, see especially Fujimoto (1992; 1993).  19Kŭmgang kyŏng ŏnhae examples are cited from the version available online from the Hangeul Museum at: www.hangeulmuseum.org/sub/information/bookData/total_List.jsp?d_code=00423&g_class=01.  20For discussion of similar word order glosses in an earlier edition of the Kŭmgang kyŏng from Koryŏ, see Chŏng (2007).  21The only other example in the text of a gloss not supported by either the Kŭmgang kyŏng ŏnhae or the Samgahae is the gloss here on 中 corresponding to MK ·eys: (金剛22b:03)

The Asami edition of the Ch’ŏllo Kŭmgang kyŏng contains a substantial number of marginalia (more than thirty), written by hand in the upper and lower gutters, in the remaining space in lines of the main text with few Chinese characters, and occasionally in the righthand margin of the page. These marginalia are invariably translation glosses copied over verbatim (albeit without the pitch-accent dots) from the explanatory notes to the Samgahae ŏnhae passages. Most of the translation glosses like these copied into the upper and lower gutters were cut off on either the top or bottom, respectively, when our text was remounted and rebound at some later date, but the translation glosses copied into either the righthand margin of the page or the blank lines within the text proper are mostly intact. For example:

a) In the right-hand margin of 金剛 4a we find the handwritten gloss:

This glosses the characters 堂堂 in 金剛 4a:02 ... and is copied from Samgahae 1:28B column 2, line 2:

堂TTANG堂TTANG ·on 盛·SSYENGhol ·s i·la.

Note, however, that the han’gŭl gloss here has been converted into an abbreviated character kugyŏl gloss.

b) In 金剛 5a:04 the line ends halfway down the page with 離空王殿 after which the following is entered by hand:

Here again the glossator has converted into an abbreviated character kugyŏl gloss the following note from 二 14 column 8:

c) In 金剛 5b:08 the line finishes halfway down the page with:

This is followed by the following handwritten translation gloss:

Cf. Samgahae 二 30, column 8, where the note reads:

That is, the glossator has copied verbatim from the Samgahae, including the replacement of 毫 for 頭 , but the note is hybrid (abbreviated character kugyŏl plus han’gŭl) in its execution.

d) The short line in 金剛 6a:04 ends with 舊百花香 and is followed by the handwritten gloss:

This glosses 五湖 in the preceding line:

Once again, the gloss is hybrid, incorporating both abbreviated character kugyŏl and han’gŭl.

e) 金剛 12b:03-04 contains the passage:

The bottom half of line carries the gloss:

Here again the glossator has copied verbatim from the Samgahae, including its use of 律調 instead of 格調 (note on 3:12, column 3):

This study of the glosses in the Asami Collection copy of the Ch’ŏllo Kŭmgang kyŏng was undertaken in the first instance to provide, for the first time ever in English, a detailed analysis of a Buddhist work in hanmun from traditional Korea with kugyŏl glosses and thereby to provide a tentative model for similar studies in future. It has also aimed to introduce the oldest such glossed hanmun text held in a North American library collection.

Though the text itself was printed in 1387 at the very end of the Koryŏ dynasty, our analysis has shown that the glossator of our text relied heavily on both the 1464 Kŭmgang kyŏng ŏnhae and the 1482 Kŭmgang kyŏng samgahae for his glosses. It is impossible to know when exactly the text was glossed, but it was likely within twenty to forty years of the printing of the Samgahae in 1482―perhaps as early as the 1480s. That the glossator likely completed his work before the end of the 15 th century is suggested by the presence of certain archaisms in the kugyŏl, such as the numerous locatives marked with older and the one example of word order dots.

But perhaps the more enduring lesson to be learned from this particular text concerns the impact that the invention of the hunmin chŏng'ŭm had on inscriptional practices in general and glossing practices in particular, and we have suggested tentatively that the availability of this new inscriptional technology may have led to the existence of a new kind of ‘two-dimensional’ glossing practice that used both sides of the hanmun line again, but in ways different from the earlier integrated ‘interpretive’ glossing of the high Koryŏ. In other words, the claim is not that the invention of the new vernacular script led to the demise of the older ‘interpretive glossing’ system (it was already dead by the mid-1400s), but that once the Hunmin chŏng’ŭm was available, glossators seized on it to add vernacular glosses to the (now unused) lefthand side of the hanmun line in ways that partially reprised the role of interpretive glossing on a limited scale and supplemented the new ‘sequential glossing’ system (which was rather unforgiving for less advanced readers of hanmun).

A related and parallel lesson would be the enduring impact of Sejong’s and especially Sejo’s projects to provide canonical Buddhist texts with authoritative kugyŏl glossings and ŏnhae translations22. Indeed, the temptation is to imagine our glossator sitting down sometime in the early 1480s with our (clean) text on one side, an earlier glossed version beside it, and a copy of the 1464 Kŭmgang kyŏng ŏnhae; he starts to copy over the older glosses, only to come into possession of the newly printed Samgahae, after which he abandons the earlier glossed version and copies most of his glosses from the new ŏnhae version and its (han’gŭl) kugyŏl glosses. In other words, our text is an eloquent demonstration of the powerful standardizing effect of Sejo’s kugyŏl projects in the Kan’gyŏng togam 刊經都監 , of which the Samgahae was one (albeit much delayed in publication, appearing only in Sŏngjong’s time).

Glossed texts from early Chosŏn―particularly messy ones full of both abbreviated character kugyŏl glosses and han’gŭl glosses―have not received the same attention and care as have texts from Koryŏ and Silla, but deserve to be studied in more detail because of the need to understand better the transition from interpretive to sequential glossing, and to appreciate the role that the newly invented vernacular script played in glossed texts once it was invented. In this regard, then, it will be useful in future to compare the Asami edition glosses with those in other known editions of the Ch’ŏllo Kŭmgang kyŏng with kugyŏl glosses.

22See Plassen (forthcoming) for a useful discussion.