Inscriptional Repertoires and the Problem of Intra- versus Interlingual Translation in Traditional Korea

Abstract

This article brings together a series of examples demonstrating the wide range of inscriptional practices in premodern Korea and the ways in which they force us to reconsider modern and Eurocentric notions of translation. The premodern inscriptional spectrum in Chosŏn Korea was not a simple binary of cosmopolitan orthodox Literary Sinitic versus vernacular Korean in the form of ŏnhae exegeses but was a range of inscriptional styles that included idu and kugyŏl. The ways in which texts were inscribed, reinscribed, and transliterated between these different inscriptional styles, as well as the ways in which Chosŏn literati themselves understood the notion of yŏk (譯, “translation”) challenge modern-day notions of translation, on the one hand, but also invite an understanding of them as rather more intralingual than interlingual. They also force us to ask whether LS was conceived as a “foreign” language for literate Koreans in Chosŏn. The premodern Korean cases forces us to add script and inscriptional repertoire (including notions of orthography, notational system, munch'e 文體, etc.) to the list of the main factors that influence intralingual translation.

Introduction

In modern-day Korean parlance about translation in contemporary literature one frequently encounters references to “translation into han'gŭl,” “translation into Korean letters,” or (worse) “translation into Chinese letters.” This is a practice that has irked for me many years now, and I have always dismissed it as yet another instance of clumsy command of English on the part of the Koreans producing such phrases for the consumption of foreigners, combined with the widespread tendency on the part of linguistically untrained laypeople to conflate speech and writing in East Asia generally, and specifically to conflate “han'gŭl” and “Korean language” in Korea.

But the practice is not confined to statements produced in English; for example, it is quite normal for modern-day Koreans to say han'gŭl lo pŏnyŏkhaetta “translated it into Korean [script]”). Nor is this practice confined to academically untrained laypeople, for it pops up frequently in scholarly articles and books. With the term hanmun 漢文 the situation is similarly complicated, insofar as here too we have a widespread modern-day conflation of hancha 漢字, meaning “Chinese characters; sinographs” (as a writing system), with hanmun, meaning “classical Chinese; Literary Sinitic” (as a linguistic and literary system/code), and it is also reasonably common for Koreans to speak of “translation into hancha,” though normally hanmun does double duty for both the script (characters) and the written language.¹

My point here is that what I had once dismissed as a purely modern-day phenomenon owing to a lack of sufficient scripto-linguistic sophistication on the part of Koreans may in fact have much deeper historical roots. The same sorts of usage can be found in abundance in traditional Korean contexts and therefore raise important questions about premodern Korean ideologies of language, writing, and translation. These in turn raise theoretical questions about the status of language, writing, and translation in premodern Korean literary culture, about the relationship between Chinese language and writing and Korean vernacular language and inscription in traditional Korea, and about the vocabulary that we use today as we struggle to understand and better contextualize and historicize these issues.

In the two sections below, then, I address in turn the following problems: (1) the general question of intra- versus interlingual translation and the usefulness of Roman Jakobson's distinction between the two, both in general and with respect to premodern Korea; and (2) the conceptualization of translation in traditional Korea, with a focus on inscriptional ecologies and terms for and types of translation in traditional Korea. A consideration of translation practices in premodern Korea forces us to question whether modern, commonsensical ideas about translation and languages are valid for or helpful in understanding the complex relationship between different inscriptional repertoires in pre-twentieth-century Korea.

The Problem of Intra- versus Interlingual Translation in Premodern Cosmopolitan Literary Formations

Most theoretical discussions of translation appeal to Roman Jakobson's now classic (1959) three-way typology of interlingual, intralingual, and intersemiotic translation, where interlingual translation is typically seen as the commonsense default notion of “translation proper” between two different languages, intralingual translation is understood as essentially paraphrase using the same language, and intersemiotic translation is defined as translation of verbal signs by other types of signs (e.g., music or image).

Numerous scholars have already pointed out problems with Jakobson's formulation. Gideon Toury (1986: 1113, as cited in Berk Albachten 2014: 575) characterizes Jakobson's typology as “crude,” and David Bellos (2011: 311) opines that “Jakobson's attempt at clarification actually introduced a great muddle.” And in his stimulating discussion of Jakobson, John Sturrock (1991: 309) concludes that “there was no need for Jakobson to have categorized translation as he does, into the intralingual and the interlingual kinds, when both of these are forms of ‘rewording.’”

The fundamental problem with Jakobson's structuralist approach is that it presumes we can say with confidence what a language is, and different researchers from different theoretical approaches have seized on this weakness. For example, Jacques Derrida (1985: 173, as cited in Davis 2014: 587) notes that Jakobson's scheme “obviously presupposes that one can know in the final analysis how to determine rigorously the unity and identity of a language, the decidable form of its limits.” Naoki Sakai (2010: 26) makes another sustained assault on this notion of languages and limits and introduces the concept of “bordering” in an attempt to “reverse the conventional comprehension of translation that always presumes the unity of a language.” For Sakai, too, the problem with Jakobson's typology is that “the unity of a language has to be unproblematically presupposed” (26); for if “translation is not only a border crossing but also and preliminarily an act of drawing a border, of bordering” (32), there can be no discussion of such bordering unless the languages involved operate in a new “economy of the foreign” (29) and are “given as countable, like apples” (33), a state of affairs that requires the “idea of the unity of language” as a “regulative idea” (28)—a recent and indeed modern invention.

Another important scholar who has rendered a sustained and in many ways quite devastating attack against modern structuralist linguistics and its assumptions about discrete languages is Roy Harris, who, in a series of brilliant, iconoclastic works has proposed a new kind of “integrational linguistics” in opposition to Saussurian “segregational” linguistics and has demolished the modern “language myth in Western culture” along with other related myths (the dialect myth, the standard language myth, the idiolect myth, the myth of the native speaker, etc.).²Harris (2004: 195) shows how most current theories of translation are code-based and all rest on one “basic theoretical mistake: this consists in reifying words as invariants that have been given a collective authorization in order to make linguistic communication possible. Instead, integrational semiology proposes to see communication in terms of the continual creation of new signs, rather than the use of old signs already supplied from some hypothetical community store.”

More recently, Harris (2011) has dedicated a short essay to a new myth, the “translation myth.” He begins with the following exchange between Alice and the Red Queen in Lewis Carroll's Through the Looking Glass. The Red Queen asks:

As Harris (2011: 86) explains, “Alice's reply in effect sums up a whole orthodox tradition of thinking about translation. This tradition assumes that we cannot translate until we have identified two languages, one to translate from and one to translate into.” Harris goes on to note that Jakobson's classic formulation “makes little sense unless languages are construed as systems of determinate forms and meanings. In other words, languages are assumed to be understood as fixed codes,” and he reminds his readers that Jakobson even explicitly appealed to the notion of code when he defined “translation proper” as involving “two equivalent messages in two different codes” (Jakobson 1959: 233, as cited in Harris 2011: 86).

An additional problem with the limited research to date on intralingual translation is its presentist or modern bias; the question of translation in medieval cosmopolitan cultural formations is a relatively neglected topic.³ Insofar as discussions of intralingual translation deal with anything premodern it all, they tend to focus on the cases of “modernizing” translations of premodern texts into modern iterations of the “same” language. Thus, in her introduction to the Routledge Encyclopedia of Translation Studies, Mona Baker (2001: xvii) writes that “intralingual translation is not such a minor issue as the existing literature on translation might suggest. Intralingual translation figures far more prominently in the Greek tradition than interlingual translation: the major preoccupation in Greece has been with translating ancient Greek texts into the modern idiom. I know of no research that looks specifically at the phenomena of intralingual or intersemiotic translation.”

But a statement such as Baker's begs a number of questions. What are the criteria for intralingual translation? Do the two codes in question have to carry the same name? Do they have to be mutually intelligible? Ancient/classical Greek and modern Greek are surely not mutually intelligible, so does translating from one to the other count as intralingual translation simply because modern-day Greeks deem it so? To put it another way, might ideologies of language, writing and/or translation not also play a role in determining what counts as intra- versus interlingual translation? And are there perhaps differences between medieval/premodern and modern notions of translation in this regard?

The Sinographic Cosmopolis in general,⁴ and premodern Korea in particular, offer rich sites within which to explore these questions; a good place to start for our purposes here is the recent polemic between Kim Chŏngu and Kim Uktong over the question of intralingual translation in traditional Korea. Their disagreement can be summarized as follows. In his influential book from 2010, Translation and Korean Modernity (Pŏnyŏk kwa Han'guk ŭi kŭndae), the literary scholar Kim Uktong makes a number of controversial claims. Chief among them is that the Vernacular Exegesis of the Correct Sounds to Instruct the People (Hunmin chŏngŭm ŏnhae 訓民正音諺解, 1459)—and by implication, all premodern Korean texts in the ŏnhae 諺解 or “vernacular exegesis” genre—is an example of intralingual translation.

Kim Chŏngu, a linguist trained in the traditional “national language studies” mode who has turned his attention in recent years to translation studies, raises a number of objections to Kim Uktong's claim but is himself held captive by a number of modern myths. For example, when he writes (Kim Chŏngu 2013: 3), “Generally speaking, ŏnhae refers to the result of translating a document written in hanmun [Literary Sinitic, 漢文; henceforth, LS] into Korean,” he gives voice to at least two common modern-day assumptions held by Korean scholars: first, that the ŏnhae process is translation, plain and simple; second, that LS was a foreign language for Koreans before the twentieth century. Interestingly, he goes on to note that it was only vernacular explications of texts in LS that earned the designation ŏnhae; vernacular Korean explications of texts from Japanese, Manchu, or Mongolian were never called ŏnhae. One implication here that Kim does not pursue is that LS had a very different ontological status from these other languages. For Kim Chŏngu's part, the only bona fide examples of intralingual translation in Korean would be modern Korean versions of the vernacular renditions found in ŏnhae texts or South Korean renditions of North Korean texts that deploy different vocabulary and forms from South Korean standard.⁵

Kim Chŏngu also treats his readers to the famous first line from the Correct Sounds to Instruct the People (Hunmin chŏngŭm 訓民正音, 1446), the document (in LS) that promulgated the new vernacular alphabet, and makes the usual claim that the very wording implies an understanding of LS and vernacular Korean as “different languages.” But does it really? Figure 1 shows a photo of the original text followed by a transcription in Yale romanization (following the conventions in Martin 1992 minus pitch-accent marks) and three different renditions into English.

The translations in (A) and (B) are from Ledyard (1997) and (1998), respectively, while those in (C) are mine. The point is that modern-day Korean scholars have uncritically understood the phrase 國之語音 (kuk chi ŏŭm, “the speech sounds of [our] country”) as equivalent to modern Korean kugŏ 國語 meaning “national language,” and have blithely projected back onto earlier centuries an assumption of a consciousness of clear and well-defined boundaries around vernacular Korean (the “national language”) and (“foreign”) LS. Pace Kim Chŏngu, I see no hard evidence here that fifteenth-century Koreans perceived of LS as a foreign language.⁶ And when he counters (2013: 14–15) that, for Kim Uktong's argument to hold, “we would have to view the Chosŏn-era literati class with a command of LS as a kind of bilingual one that knew how to use both written Chinese and spoken Korean,” he at least begins to approximate the truth, even if missing the bigger point. He continues: “It is a logical contradiction, similar to claiming that Arabic is one of the Persian languages.” Here it is clear that Kim Chŏngu fails to understand the dynamics of cosmopolitan and vernacular in premodern literary cultures, and while he mentions medieval Europe and Latin, the point surely is that vernacularization in Europe happened many centuries before it did anywhere in the Sinographic Cosmopolis. In other words, what matters is not whether LS and vernacular Korean are “the same language” but whether Koreans in fifteenth-century Chosŏn (and things do not appear to have changed much on this score until the end of the nineteenth century) thought of LS as a “foreign language.” There is little evidence that they did, in which case ŏnhae texts surely cannot be considered interlingual translations in anything like the modern understanding of the term. Of course, this does not necessarily make them intra-lingual translations by default, but at the very least it calls into question the very distinction. The following section explores other problems with the supposed distinction between intra- and interlingual translation in premodern Korea.

Conceptualizing Translation in Traditional Korea

Language and Writing/Linguistic Codes and Inscriptional Ecologies

The first indications that the Sinographic Cosmopolis and its various inscriptional ecologies might pose serious challenges to traditional modern and Eurocentric notions of translation have come from scholars familiar with the history of writing in Japan. Japanese kundoku (訓読, “reading by gloss” or vernacular reading⁷) methods of reading kanbun, in particular, have attracted much attention in this regard. For example, Masaomi Kondo and Judy Wakabayashi (2001: 485) write:

Baker (2001: xvii) comments on Japanese kanbun kundoku as follows: “But was it translation? It seems to be something in between intralingual and interlingual translation, and I do not believe we have any theories that can account for this type of practice either.”

Though still little known outside Korea, analogous Korean practices of reading LS in and through vernacular Korean via practices called kugyŏl 口訣 pose exactly the same problem and have recently attracted the attention of Korean scholars of translation studies.⁸ For example, concerning the Koryŏ-era (918–1392) practice of “interpretive kugyŏl” (sŏktok kugyŏl 釋讀口訣) that was directly analogous to Japanese kanbun kundoku, Kim Chŏngu (2006: 1615), displaying again his rather modernist bias, is reluctant to count the practice as translation at all: “Because it did not result in fully-formed Korean sentences it is difficult to include it within the realm of translated materials. . . . Our position is to understand kugyŏl graphs and word order inversion marks as a type of writing system.” But Yi Yŏnghun (2011: 134) disagrees and cautions against the limitations inherent in what he sees as “uncritical” acceptance of Western-derived definitions of translation.

That Korean glossing techniques and Korean hundok/Japanese kundoku-type “reading-by-gloss” traditions challenge Western/Eurocentric and modern definitions of translation can also be seen in the recent work of Sŏ Minjŏng (2012), who laments that discussions of translation practice in premodern Korea have focused almost entirely on the ŏnhae 諺解 genre (exegeses of canonical Buddhist, Confucian, and other texts using the Korean vernacular script after its invention in the fifteenth century) while neglecting the diverse array of inscriptional modes that existed well before the invention of the vernacular script and persisted until the twentieth century alongside the vernacular script in what, in effect, became an even more complex inscriptional ecology. Thus, Sŏ Minjŏng (2012: 327–28) presents the following eight types of sentence, where English is the hypothetical source language instead of LS:

• (1) You go to school. (source text)

• (2) You_ga go to school_da. (ŭmdok kugyŏl translation)

• (3) You₁ go₃ to school₂ (sŏktok kugyŏl translation)

• (4) You_ga school_e go_da (idu 吏讀 translation)

• (5) You_가 school_에 go_다 (ŏnmun kugyŏl 諺文口訣)

• (6) You_( 유 )가 school_{( 스쿨 )}에 go_{( 고우 )}다 (ŏnhae text 諺解文)

• (7) 네가 school(스쿨)에 간다. (ŏnhae text 諺解文)

• (8) 너는 학교에 간다. (modern Korean translation)

As Sŏ notes, numbers 2 through 4, where no vernacular script is in sight but which nonetheless contain vernacular elements, have not been treated in modern Korean scholarship as examples of translation. The problem with such an approach, Sŏ (2012: 330–31) continues, is that contemporary Chosŏn sources refer to translations using the vernacular script not as ŏnyŏk 諺譯 (vernacular translations) but as ŏnhae 諺解 (vernacular unravelings/explications). In other words, if we concede that in Chosŏn there was not just one but several techniques for “explicating” source texts in LS, of which ŏnhae is simply the most representative for modern Koreans because of its reliance on the vernacular script, then we must also concede that idu translations and kugyŏl translations belong in the same category of “translation.”

This leads Sŏ (2012: 331–32) also to raise the question of inter- versus intralingual translation, and she similarly concludes that Chosŏn-era translated texts are difficult to explain using modern standards of translation. Sticking with Sŏ’s helpful exposition for now, let us examine a more concrete example of the process of idu translation, cited by Sŏ (2012: 333) from Pak Sŏngjong (2011):⁹

“Relexification” in (d) refers to replacing and/or expanding (usually monosyllabic) LS words with (usually bisyllabic) Sino-Korean words that were presumably current in spoken Korean of the time.¹⁰

Another Korean scholar active recently in the field of translation studies and the history of translation in Korea is Yu Myŏngu. Yu (2004: 70–71) goes so far as to suggest that one of the key motivations behind the invention in the fifteenth century of the Correct Sounds to Instruct the People was dissatisfaction with the shortcomings in idu. Needless to say, Yu takes the view that idu was a form of translation.¹¹

However, it should be noted that the problem in Korea is not confined to the early days of “interpretive kugyŏl” before the invention of the vernacular script in early Chosŏn, or to idu documents before and after the invention of the script, but persists throughout Chosŏn into the twentieth century. An excellent illustration of the overall problem from a more strictly literary perspective can be found in Chang Yusŭng's study (2005) of the Annals of the Flower King (Hwawang pon'gi 花王本記), a late Chosŏn fictional narrative that exists in parallel hanmun and vernacular Korean versions.

Chang's motivation for his study was the question of how to determine, in cases where parallel LS and vernacular versions of the “same” work of fiction survive (and there are many such narratives in late Chosŏn), which came first. The problem here is that previous studies have frequently argued, on the basis of the presence of kugyŏl-type constructions and turns of phrase in vernacular Korean fictional narratives, that such traces were evidence of translation from a LS original. But Chang (2005: 209–10) points out that, contra the prevalent understanding, kugyŏl-annotated texts are not simply either an intermediary stage in the process of translation into an ŏnhae text, or mere restructurings of an LS text into something resembling Korean. Rather, he argues, they are a form of modified or transformed LS inscriptional style (pyŏnhyŏngdoen hanmun munch'e ŭi hana), which in and of itself was perfectly viable as a means of communication and which was a systematic entity with its own structure and rules. Thus “it is impossible to take the presence of kugyŏl structures in a vernacular text as evidence that the vernacular text was translated from a hanmun text. It was customary inscriptional practice to record source texts in tandem” (Chang 2005: 209–10).

Chang (2005: 215) then goes on to point out that educated males in Chosŏn were typically unaccustomed to using the vernacular for expository prose in the first place: “The writing of prose in Korean and composition of fiction in Korean passed through a hanmun textual filter inside the consciousness of the writer.” This is precisely what Kichung Kim (1996: 3–4) has described as the process of “self-translation,” where “to express themselves in writing they had to first translate their thoughts and speech in classical Chinese.”

In effect, then, the very act of a Chosŏn author composing a work of vernacular fiction in Korean was an act of translation—“the act of translating a text fixed in one script into another script” (Chang 2005: 220). So the point here is that translation between LS and vernacular in premodern Korea appears to have been perceived as a simple and rather mechanical reinscribing, from one script into another, of the same text within a broader shared inscriptional ecology—it was perceived as intralingual.

Summary

To sum up this section, then, I agree with Yi Kyŏngha (2013: 47) that intralingual translation is not only a feature of the much-celebrated turn-of-the-twentieth-century parallel texts in hanmun, kukhanmun honyongch'e (國漢文混用體, “mixed script”) and sun kungmun (純國文, “pure Korean”) but has been a feature of the Korean inscriptional landscape for many centuries. This also suggests that we need to reassess Wakabayashi's (2005: 57) comment concerning Japanese kanbun kundoku that “a similar practice did indeed exist in Korea but it had far less impact than in Japan.” The inscriptional technologies in Korea were quite varied and long-lived and participated in complex practices of reinscription that were perceived as intralingual translation. If Toury (1986) finds Jakobson's three-way typology of translation too linguistic in its orientation, I would add, on the basis of the premodern Korean phenomena, that it is also too modern and too Eurocentric.

Terms for Translation, Types of Translation

If, as in the preceding section, an examination of actual translation practices affords one vantage point on ideas about translation, a study of the various terms used for translation activities affords yet another. Eva Hung (2005: 68) makes the following interesting point in her discussion of the various ways in which 譯 (Ch. yi) has been used in Chinese: “In China's historical documents the word ‘yi’ (commonly taken as ‘translate’) signifies several different things: 1) a government post held by someone responsible for inter-lingual communication work, and by extension all persons engaged in inter-lingual communication; 2) written translation; 3) oral translation, i.e., interpreting; 4) transliteration.”¹²

It is Hung's fourth point that I wish to focus on here—the fact that “translation” in premodern East Asian contexts frequently had a graphic or inscriptional understanding, including the rewriting or reinscribing of texts from one form of writing into another. In this regard, Hung and Wakabayashi (2005: 13) are quite right to emphasize the “importance of scripts in the Asian context, with the shared use of the Chinese script in East Asia having vital implications for translation and culture”; they use the term “inter-script translations.” Our focus here is Korea, but Yi Yŏnghun (2011: 133) points out that in Edo Japan, the word hon'yaku “translation” was also used to include the (re-)transcribing of Japanese written in Roman characters into a mixture of sinographs and kana, as well as the rewriting of sinographs into kana.

Moving now to traditional Korea, we find a wealth of terms for translation, as well as a wealth of evidence that “translation” frequently meant “rewriting in another script.” Chang Yusŭng (2005: 195), for example, points out that the term pŏndŭng 翻謄 could mean both chŏnsa (轉寫, “transliterate; transcribe”) and pŏnyŏk (飜譯, “translate”).¹³ Both Dormels (1997) and Chŏng Kwang (2010) make the interesting claim that the Korean vernacular script was invented in the first instance as a kind of phonetic transcription system (akin to our modern-day IPA). In this very first function, it was used to transcribe the Chinese pronunciations of sinographs, especially the “orthodox/correct sounds” required for the civil service examinations in China. In this usage, the script was referred to as chŏngŭm (正音, “orthodox/correct vocalizations”). But these Chinese pronunciations were so different from the contemporary Chosŏn Korean pronunciations of sinographs that Sejong and his scholars devised a new set of normative Sino-Korean readings called the Tongguk chŏngun (東國正韻, “Correct Sounds of the Eastern Country”) in 1447. In this particular usage, the script was called Hunmin chŏngŭm (訓民正音, “The Correct Sounds for the Instruction of the People”). But before long, the script also came to be used to inscribe vernacular Korean in its entirety, at which point it came to be called ŏnmun (諺文, “vernacular script”).¹⁴ Thus the vernacular script had three different names, depending on what exactly was being recorded and for what purposes. Chŏng Kwang also makes the point in his article that the term pŏnyŏk (飜譯, “translation”) was used to refer to the recording in vernacular script of the pronunciations of sinographs, while the term ŏnhae 諺解 was used to refer to “resolving” or “unraveling” LS into vernacular Korean.

In his useful article on ŏnhae and pŏnyŏk in premodern Korea, Yi Hyŏnhŭi (2013: 3) stresses that the term pŏnyŏk, in addition to words, texts, sentences, and phrases, could be used with respect to the sinograph pronunciations appended to words in annotational/exegetical style. He also presents a number of useful examples from the Annals of the Chosŏn Dynasty (Chosŏn wangjo sillok 朝鮮王朝實錄) illustrating how “translation” could be script-oriented. For example, one passage from the Sejong Sillok 世宗實錄, kwŏn 13, 26 nyŏn [1444] kapcha relates a royal command to the effect “Go to the Ŭisach’ŏng 議事廳 and have them translate the Yunhui 韻會 into ŏnmun (“以諺文譯‘韻會’”). As Yi notes, this was Sejong's project to “translate” the Collected Rhymes, Ancient and Modern (Gujin yunhui 古今韻會, 1292)—that is, to provide with the new alphabet the readings (togŭmpŏp 讀音法) or “countertomy”¹⁵ spellings (fanqie 反切) as annotations. Another example:

Yi Hyŏnhŭi (2013: 4) goes on to cite An Pyŏnghŭi as having been the first to point out that the term yŏkhun 譯訓 (lit. “translate and gloss”) in Chosŏn could also mean yŏgŭm 譯音 or ŭmyŏk 音譯, that is, “provide with a sound translation  =  alphabetically written sound gloss.” By the same token, the character hun 訓 used in terms like hyanghun (鄕訓, “local [Korean] gloss”), hanhun (漢訓, “Han [spoken Chinese] gloss”), monghun (蒙訓, “Mongol gloss”), waehun (倭訓, “Japanese gloss”), and so on had the same valence as ŭm (音, “sound; sound gloss/translation”). Yi supposes that this particular usage of 訓 was unique to Chosŏn.

Yi Hyŏnhŭi (2013: 4) also gives an example from idu that is closely related to the example cited above from Sŏ Minjŏng. The following is from the Treatments for Infectious Diseases in Cows, Horses, Sheep, and Pigs (Uma yangjŏ yŏmbyŏng ch'iryobang 牛馬羊猪染病治療方, 1541). The text provides hanmun along with an idu translation and a vernacular Korean translation. Idu forms are underlined, and the texts in square brackets are glosses provided in both “borrowed graph orthography” (ch'aja p'yogi 借字表記) and vernacular script for taryuk (獺肉, “raccoon flesh”) and talsi (獺屎, “raccoon feces”).

Commenting on this example, Yi (2013: 4) claims, “This shows that translation from hanmun was possible without the mediation of kugyŏl,” but given Chang Yusŭng's discussion above, one could just as well presume that the kugyŏl phase indeed took place (either elsewhere on paper or simply in the writer's head) but was omitted.

Another scholar who has made welcome contributions to the history of translation in premodern Korea is Yi Yŏnghun (2011, 2012) with his two studies of the various terms related to translation found in the Annals of the Chosŏn Dynasty. According to Yi Yŏnghun (2011: 132), the expression pŏnyŏk 飜譯 was widely used already during the Chosŏn dynasty, and whether in the guise of 飜譯 or 翻譯, is attested even before the invention of the Korean alphabet (147). In his follow-up study, Yi Yŏnghun (2012: 168) adds that pŏnyŏk can be found in the Annals rendered as 翻繹, 繙譯, and 繙繹 in addition to the more common 飜譯, and that 譯 on its own is attested already as early as T'aejo's time at the end of the fourteenth century in reference to “translating” Mongolian writing into hanmun (172). The following entry was submitted by Sŏl Changsu 偰長壽 (1341–99), chejo or commissioner in the Bureau of Interpreters, concerning examination standards in the Bureau:

The next example (cited in Yi Yŏnghun 2012: 176) is from the interrogation of Hwang Chŏnguk:

Sŏnjo 41 kwŏn, 26 nyŏn (1593 kyesa = Ming Wanli 萬曆 21 nyŏn) 8 wŏl 2 il (kyemi) 4th entry

Here is one more example from Yi Yŏnghun (2012: 182) showing how “translate” could include the notion of “transliterate; translate into another script.” The text being translated here is an invocation to frugality:

One of the main points of Yi Yŏnghun's follow-up article is to demonstrate the wide variety of terms besides 飜譯 used for “translation.” Yi (2012: 179–89) lists no less than nineteen: 譯解, 譯說, 譯審, 譯書, 譯成, 反譯, 象譯, 重譯, 諺譯, 傳譯, 飜寫, 飜錄, 飜出, 飜校, 飜解, 飜鮮, 飜書, 飜謄, and 飜讀. He concludes by claiming that this rich variety of terms for translation set Korea apart from China and Japan at this time.

Concluding Remarks

In the absence of any explicit metalinguistic statements about translation, it is ultimately difficult to know for certain what Koreans in prior centuries thought about translation, whether or not they viewed translation back and forth from LS and various registers of vernacular Korean as interlingual or intralingual, or whether or not this was even a valid or relevant distinction for them. The larger point here seems to be this: that LS (unlike Mongolian, Manchu, or Japanese) was not exactly a “foreign” language for Koreans in Chosŏn, any more than it was for the Vietnamese in traditional times. As Keith Taylor (2005: 173) notes, the Vietnamese “did not consider [classical Chinese] as something alien; on the contrary, they viewed it as the highest expression of their civilization.” Much the same applied to traditional Korea. It is hard to find evidence that Koreans conceived of LS as a “foreign” language in anything approaching our modern sense of the term, in which case interlingual translation is beside the point. Or perhaps it would be more appropriate to simply admit that the alleged distinction between inter- and intralingual translation is much less useful in traditional Korea. Let me be clear here (especially in light of the provocative title of Denecke 2014) that I am not questioning the existence of translation per se in premodern Korea between LS and vernacular Korean. Rather, I am concerned that modern-day (especially modern-day Korean) understandings of the relationship between hanmun and the vernacular in Korea are oversimplified and too often succumb to the scripto-nationalist othering of everything “Chinese” in modern Korea. This understanding forces an assumption of interlingual translation upon a complex reality that was often more akin to intralingual translation.¹⁶

What I have tried to do here is bring together a series of examples that demonstrates the wide range of inscriptional practices in premodern Korea, all of which are connected in more than tangential ways to translation. This inscriptional (and translational) spectrum was nothing as simple as a binary of cosmopolitan orthodox hanmun versus vernacular Korean in the form of ŏnhae translations, but was a range of inscriptional styles that included idu and kugyŏl.¹⁷ The ways in which texts were inscribed, reinscribed, and transliterated between these different inscriptional styles challenge modern-day notions of translation, on the one hand, but also invite an understanding of them as rather more intralingual than interlingual—and definitely “intracultural,” to borrow Hwang's (2012) term, but not “intraliterary,” to follow Gbur's (1993) terminology—provided we understand that LS was not a “foreign” language for literate Koreans in Chosŏn. Moreover, the ways in which language and writing (script) were so tightly interwoven, along with the ways in which “translation” could include sound translation and transliteration, go some way toward explaining why Koreans even today think nothing of speaking of “translating into Chinese letters.”

In her useful discussion of intralingual translation, Karen Zethsen (2009: 795) concludes that “the differences between intralingual and interlingual translation seem to be more a question of degree than of kind,” and she goes on to list four main factors that seem to be influential in intralingual translation: knowledge, time, culture, and space (805). But our survey of some of the relevant premodern Korean facts strongly suggests that there is a fifth main factor to be taken into account, that of script (or inscriptional repertoire, including notions of orthography, notational system, munch'e 文體∼écriture, etc.).

The fact that Chosŏn Koreans included what to us looks like transliteration in their notion of “translation” and subsumed under that notion (and could switch back and forth between) versions of the same notional text or semantic content in orthodox LS, kugyŏl-glossed LS, idu, mixed-script Korean, or vernacular Korean in han'gŭl suggests an ideology of language and writing whereby an underlying and independent semantic core is simply remapped or reinscribed using a different inscriptional mode (munch'e 文體). This is oddly reminiscent of the fixed code myth criticized by Roy Harris as lying at the heart of everything that is wrong with modern structuralist linguistics, and thus suggests that certain modern linguistic notions criticized by postmodernists as Western or Eurocentric may in fact have their origins elsewhere. For Harris (2004: 195) the impression of a “fixed code” that lies somewhere “out there” “is reinforced, at least in literate societies, by educational systems based on the preservation and circulation of authoritative texts.” Moreover, “this code-based concept of languages implies at the same time a conception of ‘meaning’ in which meanings themselves are treated as independent of the code” (Harris 2011, 87). In other words, then, it may well be that in the thought world of traditional Korean literati, steeped as they were in cosmopolitan LS texts, LS and vernacular Korean in han'gŭl were just two ends of a single scripto-linguistic spectrum and two different registers in a unified system where LS was by no means “foreign” (but Manchu, Japanese, and Mongolian most certainly were), and a transcendent plane of meaning could be cast and recast in any of a number of inscriptional options arranged across a spectrum that included multiple scripts and orthographic devices. As Sturrock (1991: 313) notes, “A philosophy of translation which presupposes meaning as transcendent entities is Platonist, or else idealist,” and “the temptations to Platonism in respect of meaning are constant and insidious, and they bedevil all thought about translation” (318).

Much is being written recently about the topic of translation between different munch'e or inscriptional styles in the context of Korea's kaehwagi period from the 1890s into the first decade of the twentieth century with its parallel texts in LS, kukhanmunch'e 國漢文體 mixed script, and “pure” kungmun 國文, but I hope to have shown here that this is already a considerable simplification of what was earlier a broader array of options. It is tempting to imagine therefore that until China and Sinitic language(s) and sinography were decentered and othered (see Schmid 2002), these sorts of facile interinscriptional flip-flops were all just intralingual and intracultural recastings, regraphicizations, or reinscriptions of the same underlying semiotic core, as far as the Koreans of the day were concerned.

This article was originally written for the workshop “Intralingual Translation, Diglossia, and the Rise of Vernaculars in East Asian Classical and Premodern Cultures” at the École Pratique des Hautes Études, Paris, January 19–20, 2017. I would like to thank the conference participants and the organizers, Rainier Lanselle and Barbara Bisetto, for their feedback. My thanks also go to the SJEAS reviewers for their advice on the manuscript.

Notes

References