Abstract
-
This paper takes as its point of departure the persistent difficulty in reading Hanmun (Korean Literary Sinitic) texts despite prior knowledge of grammatical rules, and reexamines this problem from cognitive and pedagogical perspectives. It argues that reading failure in Hanmun texts lacking punctuation should not be attributed to an insufficient understanding of individual grammatical items, but rather to a lack of syntactic unit recognition and reading stability. From this perspective, the article focuses on toksong 讀誦, a core practice in premodern Hanmun education, and interprets it not merely as recitation or memorization but as a mode of reading through which syntactic structures are embodied via rhythm and repetition.
Drawing on theories of embodied cognition and skill acquisition, the article demonstrates that toksong functioned as a cognitive mechanism that fostered syntactic familiarity and stabilized grammatical segmentation prior to semantic interpretation. It further examines the place of toksong in modern Hanmun pedagogy and repositions it not as a substitute for grammatical analysis, but as a preparatory reading practice that enables such analysis. By reconceptualizing the reading of Hanmun as a structural experience that precedes meaning-centered interpretation, this study offers pedagogical and theoretical implications for contemporary Hanmun education and research on classical literacy.
-
Keywords: Hanmun, Hanmun Education, Reading Practices, Pronunciation-Based Training, Syntactic Recognition
Introduction
In the educational context of Korean Studies at Leiden University, reading Hanmun (Korean Literary Sinitic)
1 is typically encountered after students have completed a basic introductory course. Students majoring in Korean Studies are first introduced to Literary Sinitic through a basic Classical Chinese course offered by the Department of Chinese Studies, and subsequently take Hanmun courses offered within the Korean Studies curriculum as electives. This instructional structure is primarily designed around the acquisition of grammatical systems and basic vocabulary, and it has clear educational merits in that it provides students with a shared foundation in East Asian written culture.
However, observations from actual teaching practice suggest that such instruction does not automatically or smoothly translate into the ability to read Korean classical texts. The primary sources that students encounter in the Korean Studies curriculum span a wide range of fields, including history, literature, thought, and culture, and many of these materials are presented in diverse textual forms, including mixed-script texts combining Chinese characters and Hangul, as well as texts written in Hanmun. Within this textual environment, characterized by diverse contexts and genres, prior learning experience often fails to provide learners with the level of reading stability that might be expected.
Korean classical texts written in Hanmun present foreign learners with complex barriers that operate not only at the linguistic level but also across sentence structure and overall reading practices. Traditional Hanmun text is characterized by the continuous arrangement of sentences without punctuation, requiring learners to identify sentence units, phrase and clause boundaries, and proper nouns based on contextual cues. In this process, learners frequently find themselves unable to fully mobilize the vocabulary and grammatical knowledge they have already acquired, and instead return to what is effectively a beginner’s position when confronted with the text.
These difficulties cannot be adequately explained as a simple lack of knowledge of Chinese characters. Introductory Classical Chinese courses typically focus on making Literary Sinitic into an “understandable text” through standardized grammatical explanations and meaning-centered interpretation. While meaning-centered interpretation and grammatical explanation are essential for textual understanding, they alone do not sufficiently guarantee the ability to follow sentences by grasping syntactic structure in real time. As a result, learners may be able to explain sentences analytically, yet continue to experience persistent difficulty in reading original texts without the aid of annotations or translations.
However, it is important to note that although Literary Sinitic has long functioned as a shared textual tradition across East Asia, it is difficult to regard it as a language system premised on a specific speech community. This suggests that reading Literary Sinitic cannot be reduced to any single pronunciation system, but instead depends on the reader’s ability to construct syntactic structure during reading. Rather, Literary Sinitic has operated as a textual tradition that shares certain stylistic norms and reading conventions, and its modes of reception have varied according to historical period and educational context.
What is crucial is that Literary Sinitic was transmitted in a form in which characters were continuously arranged without punctuation, such that textual meaning and structure were not immediately given to the reader but had to be constituted through the act of reading. In response to these concerns, this paper aims to shed light on ‘syntactic awareness’ as an ability required prior to meaning-based explanation in the reading of Hanmun texts, and to explore alternative approaches to reading such texts.
In particular, the fact that Literary Sinitic has historically been received through diverse reading practices merits attention. The use of recitation and repeated reading as auxiliary means in premodern educational settings suggests that the reading of Literary Sinitic has not relied exclusively on meaning explanation or grammatical analysis.
Starting from this educational situation, this article reconsiders what kinds of abilities are required for students majoring in Korean Studies prior to meaning-based explanation in Classical Hanmun reading. For these learners, reading Literary Sinitic through Sino-Korean pronunciation is not merely a matter of adopting a traditional convention, but a pedagogically appropriate practice grounded in their disciplinary context. Given that Literary Sinitic has historically been read through locally embedded linguistic practices rather than a single unified pronunciation system, the use of Korean-based reading aligns with both the historical reception of texts and the educational needs of Korean Studies students. In this sense, Korean-based pronunciation functions not only as a medium of reading, but as a means of sustaining syntactic structure and facilitating comprehension.
Therefore, this study examines the educational potential of syntactic, structure-oriented training that employs pronunciation and recitation, focusing on Hanmun instruction for students majoring in Korean Studies.This reading-aloud–centered approach is aimed not only at basic instruction focused on memorization of meanings or character acquisition, but also at supporting advanced students who must directly engage with premodern texts densely packed with Chinese characters, enabling them to recognize and sustain sentence structure during reading.
Accordingly, this study theoretically recalls an important traditional layer of Literary Sinitic, in which texts were learned and internalized through recitation and repeated vocalization in specific historical contexts. This is not an attempt to generalize recitation and repeated vocalization—one historical mode of reception—as a universal precondition for reading Literary Sinitic. Instead, it seeks to reexamine these practices at the level of methodology, as one possible means of supplementing syntactic recognition training in Korean Studies courses devoted to reading Hanmun texts.
Conceptual Framework: Why Structure Before Meaning?
1) From Oral Mediation to Structural Internalization
Building up knowledge of Chinese characters is essential for reading Hanmun, but it is not uncommon that one still cannot follow the sentence even while knowing the meanings of individual characters. This is because in Hanmun meaning is not formed simply at the level of each character as a unit, but rather through referential relations, context, and syntactic cohesion. In addition, it is also necessary to consider that Hanmun functions as a textual tradition in which meaning is organized through literary/standardized idiomatic expressions, deictic words, and omitted elements.
Hanmun has functioned as a textual tradition that shares certain stylistic norms and reading conventions, and that could operate without presupposing translation. Against the background of theSinitic
tongmunsegye 同文世界 “shared script world,” multiple traditions of literary Sinitic were established across East Asia despite the absence of a shared spoken language. In premodern East Asia, literary Sinitic enabled transregional communication through script-based modes of reading and gave rise to specialized reading practices, such as
kundoku 訓讀 in Japan and
kugyŏl 口訣-based reading practices—including
sŏkto 釋讀 and sundok—in Korea, as well as brush talk
p’iltam 筆談.
2 In such a tradition, reading would not have been limited to simply extracting meaning. This is because readers in each region had to read texts by interacting with the script itself in an independent way. In short, a “world without translation” does not mean that a shared writing system guaranteed the same reading in each region; rather, it refers to the historical condition in which each region made texts work through its own linguistic environment and reading conventions.
Under these conditions, reading could not remain a matter of merely finding meaning. It necessarily involved structural operations: setting boundaries of sentences in texts where punctuation and explicit grammatical markers were provided only in a limited way, and then re-binding the elements that had been segmented. Indeed, in premodern East Asia, major Sinitic texts did not circulate as a spoken language; they were transmitted in written form, and within each region’s linguistic system they became objects of reading aloud (
nangdok), chanting or voiced reading (
songdok, often associated with
sŏngdok), vernacular reading (
hundok), and lectured reading practices (
kangdok).
3
Accordingly, in such an environment, structural operations often preceded semantic interpretation,
4 and learners came to carry out training that involved maintaining and tracking sentence structure in their engagement with texts.
Within such reading practices, meaning in reading literary Sinitic was not given as something complete within the text itself but was gradually constructed in the act of reading, and repeated voicing functioned as a key device for securing syntactic continuity.
5
In this sense, Sinitic texts was less a language completed within a fixed mode of speech than a textual tradition whose form and use were shaped through repeated practices of reading and voicing in different regions of premodern East Asia. Within such reading practices, recognizing sentence structure before meaning functioned as a natural prerequisite for reading. In the context of Korean Studies, this process necessarily takes place through Korean-based pronunciation, whereby Hanmun texts are vocalized within the phonological system familiar to learners. Through such Korean-based voiced reading, syntactic structure can be stabilized and internalized prior to semantic interpretation.
Previous studies on hanmun education in Korea have approached reading practices from multiple perspectives. Curriculum-based research, in particular, has conceptualized reading as a set of decomposable skills, including reading aloud, spacing, and syntactic segmentation. Through analyses of the national curriculum, reading has been defined as a performative process centered on vocalization and structural differentiation. These studies have emphasized the importance of segmentation practices such as spacing, thereby demonstrating that reading in hanmun involves not merely the extraction of meaning but also the processing of syntactic structure.
6
Research on traditional hanmun learning, on the other hand, has highlighted “voiced reading” as a central pedagogical method, in which repetition and vocalization are regarded as essential conditions for achieving deeper comprehension. At the same time, studies on traditional
sŏdang-style education have often adopted a normative perspective, emphasizing moral cultivation and educational values rather than examining the cognitive mechanisms underlying such reading practices.
7
Taken together, previous studies have consistently recognized the importance of reading and repetition in hanmun learning. However, these discussions have largely remained at a descriptive level. They have not sufficiently explained why such reading practices function in the process of sentence comprehension, nor how they relate to the recognition of syntactic structure. In other words, although repetition-, rhythm-, and vocalization-based reading has been acknowledged as effective, the question of why it is effective—and what kind of cognitive role it plays in the reading process— remains to be examined at a more theoretical level.
This issue becomes particularly significant in the context of overseas learners. While such traditional reading practices have been to some extent experientially shared within the Korean educational environment, they cannot be assumed as part of the ordinary learning experience for learners outside Korea. Therefore, without a theoretical account of how repetition- and vocalization-based reading leads to the recognition of syntactic structure and comprehension, such pedagogical approaches may have difficulty establishing their educational validity. From this perspective, it becomes necessary to reconceptualize repetition- and vocalization-based reading as a cognitive process related to syntactic recognition.
2) The Problem of Syntactic Segmentation in Hanmun Reading
While diverse pedagogical approaches to Hanmun are being explored across Europe and North America, instruction in Hanmun in many Western universities remains largely organized within the framework of Chinese Studies, a configuration that can limit its responsiveness to the needs of Korean Studies students.
8 Under these conditions, the reading of Hanmun by Korean Studies students can be reconsidered from the perspective of research-oriented reading purposes and instructional design. Problems in reading are often explained in terms of individual learners’ lack of ability or insufficient knowledge of Chinese characters.
However, the difficulty most frequently encountered by learners engaging with Hanmun does not lie in their inability to recall the meanings of individual words or to apply grammatical rules. Rather, it lies in their difficulty in segmenting sentences into stable syntactic units. In unpunctuated Hanmun texts, where subjects and objects are frequently omitted and word order is flexible, learners are required to infer sentence boundaries, phrase and clause divisions, and predicative relations on their own.
9
This difficulty can be observed in the following example:
生無盜賊之名而居有妻室之樂行無逐捕之患而長享衣食之饒乎
Although the sentence follows conventional structural patterns of Classical Chinese, its structure does not become immediately apparent to learners who have only acquired the meanings of individual characters. In such texts, subjects and objects are frequently omitted and word order is flexible, requiring learners to infer sentence boundaries, phrase and clause divisions, and predicative relations on their own. Even when learners know the meanings of individual words, they often struggle to determine how the sentence is structured. Without clear recognition of syntactic boundaries, learners tend to process the sentence character by character, which often results in fragmented or unstable interpretation.
In addition, learners must construct, within the process of reading itself, the referential function of certain lexical units—such as proper names or culturally specific referents. Under these conditions, meaning is often grasped only partially or reconstructed according to local context, and as a result, the coherence of reading is easily disrupted.
Returning to the example above, expressions such as 生無, 居有, 行無, 長享, as well as nominal constructions such as 盜賊之名, 妻室之樂, 逐捕之患, 衣食之饒, and functional elements like 而 and 乎—are not processed as isolated lexical items, but rather as cohesive syntactic units. These expressions do not function merely as independent grammatical items or frequently occurring verbs; instead, they operate as structural signals that organize syntactic relations within the sentence. This point will be further illustrated through detailed analysis of parallel and rhythmic structures in the following section.
Accordingly, when learners fail to recognize these expressions as unified units and instead begin to interpret the meaning of each individual character separately, the continuity of reference collapses, and the semantic integrity of the sentence is compromised. By contrast, through repeated reading, these units become rhythmically patterned and phonologically stabilized, facilitating cognitive chunking. This process enables learners to maintain syntactic structure in real time during reading.
One effective way to support this process is through recitation, which employs repetition, rhythm, and vocalization to reinforce the recognition of syntactic units.
This is evident in the use of expressions such as Tan’gun 檀君 (mytho-historical proper name), Hŏsaeng 許生 and Pak Saeng 朴生 (personal names used as narrative referents), Yedŏk sŏnsaeng 穢德先生 and Ch’ŏn’gyulja 蟬橘子 (honorific or descriptive epithets), Hamyang Pakssi 咸陽朴氏 (compound identity marker combining place and lineage), and functional or institutional titles such as P’ungbaek 風伯, Usa 雨師, and Unsa 雲師 (role-based designations within a cosmological or administrative order). These expressions are conventionally used within the textual tradition and do not function as isolated lexical items, but as stable referential units that must be recognized and maintained throughout the text. When learners begin to interpret the meaning of each individual character separately, the continuity of reference collapses, and the meaning of the sentence itself is compromised. Under these conditions, meaning is often grasped only partially or reconstructed according to local context, and as a result, the coherence of reading is easily disrupted.
These difficulties should therefore be understood not simply as individual deficits, but as a consequence of insufficiently designed training in syntactic recognition prior to semantic interpretation. In the early stages of Hanmun reading, it is thus necessary to suspend the assumption that “understanding the meaning of characters” must constitute the first step, and instead to reconsider the conditions under which structural recognition can be established first. What is required is not only the memorization of lexical meanings or grammatical rules, but the ability to maintain and track sentence structure. When syntactic segmentation remains unstable, grammatical knowledge cannot be effectively transferred to actual reading practice.
Under such reading conditions, traditional practices of “voiced reading” (
toksong 讀誦) can serve as a viable method. Traditional
toksong was, in essence, a practice grounded in repetition, and this repetition did not aim at memorizing meaning but functioned as a means of stabilizing and internalizing sentence structure. Repeated reading in this manner can therefore be reinterpreted not as a mnemonic technique, but as a cognitive device that maintains structural continuity through rhythm, repetition, and vocalization.
10
However, in the present study, toksong 讀誦 is not understood as entirely identical to traditional sŏngdok 聲讀, Buddhist toksong, or mukdok默讀. It does not necessarily mean reading aloud in a loud voice, nor does it refer to silent reading that completely excludes sound. While it fundamentally emphasizes reading with sound, toksong in this study is understood to include semi-vocalized reading processes, such as murmuring or low-level vocalization, with the aim of embodying and sustaining sentence structure through repetition and rhythm. In this process, sound functions as a crucial medium for forming and maintaining rhythm and repetition, while also triggering the recall of meaning. Moreover, its goal is not the complete memorization of meaning, but the development of structural awareness, whereby recurring patterns—such as names, phrases, and sentence structures—become automatically recognizable during reading.
Accordingly, as a method, toksong is not simply a reproduction of traditional reading practices, but a reconceptualized mode of reading centered on rhythm-based structural recognition. In this sense, repetition should not be understood as leading directly to memorization; rather, it gradually reduces the cognitive load involved in reading. In unpunctuated Hanmun texts, where explicit grammatical markers are scarce, rhythm and repeated reading can function as auxiliary cues that help readers follow and maintain sentence structure.
From this perspective, learners must be able to recognize syntactic chunks prior to assigning meaning. Here, “chunking” does not refer to a specific teaching strategy or learning technique, but to a cognitive function by which continuous linguistic input is perceived as discrete processing units that are no longer decomposed.
11 Only when such unit recognition becomes possible can learners identify relations among units and engage in stable semantic interpretation. Syntactic chunking does not occur arbitrarily, but is guided by various structural cues internal to the language. In Classical Chinese, although formal markers are limited, function words such as 之 (
chi), 者 (
cha), and 所 (
so), sentence-final particles, and connective or logical markers such 則 (
chŭk), 乃 (
nae), 故 (
ko), and 而 (
i) serve to suggest internal sentence structure. Individually, these elements do not complete meaning, but within the flow of a sentence they function as cues that make syntactic boundaries perceptible and provide minimal structural signals that enable chunking.
What is crucial is that these structural cues do not require learners to have already understood the sentence semantically in order to function. On the contrary, learners must be able to detect and respond to these cues prior to semantic interpretation in order for interpretation to take place at all.
12 While meaning-centered approaches tend to treat such cues as the result of interpretation, at the initial stage of learning these cues must be recognized in advance.
It is now examined whether these conditions amount merely to a theoretical assumption, or whether a mode of reading centered on pre-interpretive repetition and structural familiarization constituted a historically plausible reading practice. The following section briefly considers the historical conditions that made such a mode of reading possible.
3) Historical Precedent as Pedagogical Evidence
Syntactic recognition prior to semantic interpretation is not merely a theoretical assumption, but a mode of learning that has been repeatedly realized in actual educational practice, and this can be sufficiently confirmed through historical cases. Confucian oral recitation and Buddhist scriptural chanting constitute representative examples of such practices. However, these cases should not be regarded as models to be directly reproduced in contemporary education; rather, they are more appropriately understood as empirical evidence that learning conditions once existed in which sentence structure could be stably maintained through repetition and rhythm prior to meaning-based interpretation.
Confucian Oral Recitation: Repetition Prior to Interpretation
In the Confucian educational traditions of East Asia, including those of the Chosŏn period, the starting point of learning Literary Sinitic was not interpretation or grammatical explanation, but the repeated voiced reading of texts through
toksong 讀 誦.
13 Canonical texts such as the
Xiaoxue,
Analects,
Mencius, and
Doctrine of the Mean were first learned through repeated reading aloud and memorization, prior to the study of commentary or semantic analysis. In this process, learners physically experienced the rhythm and structure of sentences even when they did not yet fully understand their meaning.
What is important here is that such practices of recitation did not function merely as techniques of memorization or training of memory. Repeated reading worked as a device for maintaining word order, sentence endings, and the cohesion of syntactic units, guiding learners to grasp sentences within a stable formal framework. In other words, oral recitation functioned to “hold sentences together” without presupposing semantic interpretation, and can therefore also be understood as a condition that enabled syntactic recognition prior to interpretation.
This recitation-centered learning structure can be observed in a more institutionalized form in Chosŏn-period village schools (sŏdang). In sŏdang education, kangdok 講讀 centered on practices such as paegang 背講 and myŏn’gang 面講, in which learners repeatedly vocalized classical texts and internalized their rhythm and structure through embodied repetition. Such recitation was designed to enable learners to cognitively engage with written signs through the combined modalities of vocalization, hearing, and rhythm prior to semantic interpretation, and it provided a cognitive foundation for subsequent commentary study and question-based instruction. Previous studies likewise point out that sŏdang education emphasized repetitive training aimed at stabilizing textual coherence munni 文理 and syntactic sensitivity prior to textual interpretation, and characterize recitation as a core pedagogical device in Confucian education.
Zhu Xi’s Reading Theory: shúdú 熟讀 as a Procedural Stage
These practices of recitation were theoretically formalized in the reading theory of Zhu Xi 朱熹(1130-1200), a central figure of Song-dynasty (960–1279) Neo-Confucianism. Zhu Xi’s approach to reading has often been understood primarily as a theory of moral cultivation, but when examined from the perspective of learning procedures, its core structure consists of the sequence
shúdú 熟讀 “thorough reading” →
jīngsī 精思 “refined reflection”. Shúdú does not refer to the immediate clarification of meaning or the formation of judgments, but also includes a preparatory stage in which texts are repeatedly read in order to stabilize their formal structure and flow.
14
For Zhu Xi, repeated reading functioned as a process through which texts were “embodied” prior to conceptual thinking, while jīngsī was positioned as a stage that becomes possible only after such structural familiarity has been established. The well-known exhortation by Chosŏn Confucian scholars to “read a hundred times” should likewise be understood not as a call to delay interpretation for its own sake, but as an attempt to secure the conditions under which interpretation could eventually take place. In this sense, Zhu Xi’s reading theory provides a theoretical foundation for understanding repetition prior to meaning as an independent procedural stage in learning.
It can be said that shúdú 熟讀, as understood by Zhu Xi and other Confucian scholars, ultimately aims at a deep understanding of the text and, more fundamentally, at moral cultivation and reflective self-cultivation. However, within this process, it is possible to isolate, as one of its operative components, a procedural function that stabilizes syntactic structure prior to semantic interpretation.
Buddhist Chanting: Rhythm and Structural Sensitivity
A similar learning mechanism can be found in the Buddhist tradition of scripture chanting tokkyŏng 讀經. Sutras such as the Heart Sutra, the Diamond Sutra, and the Lotus Sutra were commonly learned through repeated chanting prior to doctrinal explanation or interpretive instruction. In this process, practitioners became familiar with structural patterns of sentences through recurring phrases and rhythmic repetition, even without fully understanding the semantic content of the text.
What mattered in Buddhist chanting was not the immediate comprehension of doctrine, but the sensory recognition of sentence boundaries and textual continuity through rhythm and repetition. Short and repetitive sentence patterns and the regular rhythmic structure of sutra texts naturallysupported segmentation through chanting, providing conditions under which learners could intuitively grasp sentence structure. In this respect, Buddhist chanting also serves as a case demonstrating that structural familiarization prior to semantic interpretation was possible in educational practice.
Moreover, in Japanese, Tibetan, and Korean Buddhist traditions, scriptural chanting was institutionalized not merely as ritual recitation, but as an educational practice through which texts were internalized as embodied and vocal patterns via memorization and repeated vocalization. In Tibetan monastic education in particular, memorization of scriptures functioned as a prerequisite for commentary study and dialectical training, while Korean Buddhist traditions such as pŏmp’ae 梵唄 and pŏmŭm chanting likewise formed transmission systems in which the vocal structure of texts was repeatedly embodied.
Rather than being treated as a culturally specific or unique practice, Buddhist chanting is understood here as one historical instance of a broader learning mechanism observable across multiple traditions. Comparable forms of recitative or chant-based textual transmission can be found in other religious traditions, including Judaism and Christianity, suggesting that rhythm, repetition, and vocalization can facilitate structural familiarization independently of immediate semantic comprehension. In this respect, Buddhist chanting provides empirical support for the pedagogical possibility that syntactic structure can be stabilized through embodied and rhythmic engagement prior to meaning-based analysis. Taken together, these examples suggest that repetition-based reading practices across different intellectual and religious traditions share a common procedural logic: the stabilization of textual structure prior to semantic interpretation.
Structure → Embodiment → Reflection
Although Confucian oral recitation and Buddhist chanting developed within different purposes and contexts, they share a common cognitive structure at the initial stage of learning. In both traditions, it can be observed that repetition and rhythm operate in ways that both enable the attainment of semantic understanding and maintain sentence structure, and through this, one can discern a tradition that makes it possible for learners to first embody structure and subsequently strive toward interpretation and reflection.
This learning process can be summarized as follows. First, structural recognition precedes semantic understanding. Second, through repetition and vocalization, this structure becomes embodied. Third, as such embodiment accumulates, structure is internalized, making stable interpretation and reflective thinking possible.
This sequence of “structure → embodiment → reflection” should not be understood as a normative model to be directly reproduced from specific traditions, but as historical evidence that structural recognition prior to meaning was in fact achievable in educational practice. In this sense, traditional recitation and chanting do not provide ready-made methods for contemporary classrooms, but rather offer theoretical support for the idea that embodied repetition can function as a mechanism enabling early-stage syntactic parsing.
Approaches to Hanmun reading that emphasize repetition and vocalization therefore concern not merely instructional techniques, but more fundamentally the preconditions of language processing itself. Because classical written texts are presented with minimal punctuation and morphological marking, learners must be able to segment sentences into manageable units before applying lexical meanings or grammatical rules. Under such conditions, semantic interpretation becomes not the starting point but the result of reading, and initial syntactic parsing must precede interpretation.
From this perspective, repetition should be distinguished from memorization or mechanical drilling. Repetition does not primarily serve to store information, but can function as a process that reduces processing load and automates structural recognition. In particular, repetition mediated by rhythm and vocalization helps maintain sentences as syntactic units rather than semantic ones, thereby forming a cognitive foundation upon which subsequent interpretation and analysis become possible. In this sense, repetition should be reconceptualized not as synonymous with memorization, but as a precondition that enables structural recognition. In this sense, the pedagogical effectiveness of repetition in Hanmun reading does not derive from reinforcing meaning, but from facilitating the recognition and stabilization of syntactic units under conditions where structural cues are limited. Repetition-based approaches are therefore particularly effective in reading environments that require active construction of sentence structure.
Embodied Repetition and Pedagogical Design
1) Mechanism: Embodied Repetition as a Cognitive Scaffold
In the preceding section, it was shown—through cases of recitation and chanting in premodern East Asia—that pedagogical conditions enabling the stable recognition of sentence structure prior to semantic interpretation had been realized in actual educational practice. Building on these historical and theoretical discussions, it is now possible to reframe toksong as a principle of instructional design for the teaching of classical Hanmun texts to students of Korean Studies. The repetition-based approach observed in historical cases resonates with discussions in embodied cognition theory, which emphasize the role of bodily experience in the formation of cognitive structures.
The recitation-based approach adopted in classroom instruction, as envisioned in this study, does not exclude beginner learners in need of grammatical knowledge. However, this study originates from the process of teaching students who, despite having acquired basic grammatical categories, continue to experience difficulty in reading actual texts. Accordingly, it places particular emphasis on learners for whom knowledge of Chinese characters and Hanmun does not readily operate in real-time reading.
A recurrent problem observed in the teaching of Hanmun text to Korean Studies students is that, although learners are able to explain the grammatical functions of particular characters, they fail to use these as immediate segmentation cues when confronted with unpunctuated texts. Their difficulty in reading arises not from a lack of grammatical knowledge or understanding, but from the failure of previously learned structural cues to be activated at the appropriate moment in real-time reading. In other words, the core problem lies not in a lack of knowledge, but in the failure to automatically recognize sentence structure.
Under these conditions, reading places a heavy cognitive burden on learners, as each sentence must be reanalyzed anew, making it difficult to maintain the overall structure of the sentence while reading. If learners wish to segment a sentence into syntactic units without strain before applying lexical meanings or grammatical rules, recitation can be employed at this stage. What functions critically here is not sound itself, but the rhythm formed through repeated exposure, which sensorially highlights points of closure and continuation, pause and flow, within the sentence.
15 This function may be defined as 'a set of structural cues provided by repetition and rhythm', through which learners come to intuitively recognize the syntactic boundaries of a sentence.
From this perspective,
toksong cannot be understood as a mere pronunciation exercise or a memorization technique. Rather than serving as a method for teaching new grammatical knowledge,
toksong should be understood as a procedural device that enables already learned grammatical elements—such as particles or conventional syntactic patterns—to function as immediate cues for syntactic segmentation in the actual process of reading. One might argue that grammatical knowledge of a language can be sufficiently recognized through an understanding of rules. However, through repeated performance, it becomes possible to expect a shift in reading of the following kind: a transition from ‘reading that consciously applies grammatical rules’ to ‘reading that can process sentence structure automatically’.
16 Repetitive utterance or rhythmbased exposure allows sentences to be maintained not at the level of meaning units but at the level of syntactic units, thereby forming a precondition that makes semantic interpretation possible at a subsequent stage. In this sense, repetition needs to be reconceptualized not within a framework that understands
toksong memorization or mechanical training, but as a process through which structural recognition is automated.
What is delayed in toksong is not the recognition of signs themselves, but the immediate assignment of meaning to those signs. In both Confucian and Buddhist traditions, the object of recitation was always the canonical text as a written, characterbased text, that is, a linguistic object endowed with a clear system of signs and a syntactic organization.
In other words,
toksong is not an act that suspends symbolic processing; rather, it constitutes a specific cognitive mode in which signs are processed prior to their reduction to meaning.
Toksong should thus be understood not as a ‘pre-signifying, nonsymbolic stage,’ but as a form of ‘sign recognition’ that precedes semantic interpretation.
17 Through repeated
toksong, signs are no longer apprehended merely as abstract units of meaning but become embodied as rhythmic and segmentable structures. In particular, function words (such as 09), sentence-final markers (such as 也, 矣, 焉), and structural markers indicating negation, interrogation, or origin (such as 不, 何, 自) operate within such repetition as rhythmic and phonetic cues, thereby enabling learners to recognize sentence structure prior to semantic interpretation.
Likewise, conventional combinations of sinographs can be immediately retrieved during the reading process only when they are recognized as structural chunks through rhythm and repetition. While syntax-oriented toksong training takes structural recognition as its primary objective, classroom practice simultaneously exposes learners to vocabulary items and proper nouns within the same syntactic environments. Through this process, learners gradually come to recognize specific expressions and names with increasing stability, without relying on deliberate memorization. In particular, personal names, place names, institutional terms, and idiomatic expressions that frequently appear in Korean Studies texts must be grasped as ‘chunks’ prior to translation in order for reading to become stable; toksong thus also facilitates the sensory attunement of naming conventions and formulaic expressions within syntactic structure.
Such embodied repetition can be expected to produce different effects for learners at different levels. In the case of beginner learners, recitation can contribute to lowering the entry barrier to sentence reading by repeatedly exposing learners to signs and syntactic structures, thereby fostering a sense of familiarity with written texts. By contrast, for learners who have already acquired basic grammatical categories and knowledge of the functions of sinographs, recitation can yield more clearly discernible effects. In this case, it can provide conditions under which grammatical elements that are already known come to operate automatically and unconsciously in the actual process of reading. Put differently, when basic semantic knowledge and grammatical information are already in place, recitation can more effectively activate a cognitive pathway through which signs are embodied at the level of structure, prior to their interpretation as units of meaning—not as a form of mechanical training, but as a process through which structural recognition is automated. This is why the present study emphasizes this approach as an intermediate pedagogical strategy for transferring learned syntactic knowledge into actual reading performance.
2) Design Principles and Structure
For students of Korean Studies who are required to treat classical texts as research materials, the central challenge of reading does not ultimately lie in whether grammatical rules are understood, but in whether syntactic structures can be maintained and navigated in a stable manner during reading. In other words, what is required is not simply a student who knows grammar, but a researcher who is able to hold on to structure while reading. Toksong can function as a procedural device that stabilizes the cognitive conditions necessary for meaning-making—namely, the ability to preserve and segment sentences at the level of syntactic units—and, in doing so, supports the recognition of overall structural coherence. When this redefinition is translated into concrete instructional design, it is articulated through the following principles.
Chunking Before Interpretation
First, this design does not treat interpretation as the starting point of reading. In classical Hanmun texts, semantic understanding is positioned not as a prerequisite, but as an outcome that becomes possible only after syntactic processing has been stabilized. Accordingly, sentences are first presented not as units of translation, but as minimal syntactic chunks that learners are cognitively capable of processing. This principle is not intended to replace meaning-oriented approaches; rather, it represents a design decision that explicitly foregrounds the cognitive conditions that must be secured before meaningful interpretation can occur. These syntactic chunks are presented not as objects of analysis, but as formal patterns repeatedly exposed through auditory and visual modalities. In this respect, the form in which texts are presented may vary according to proficiency level. For beginner learners, texts may be provided with punctuation or other visual segmentation cues—including traditional devices such as t’o (吐)
18—that support initial syntactic recognition. By contrast, for more advanced learners, unpunctuated or facsimile-style texts are introduced in order to encourage the active construction of syntactic boundaries. In both cases, however, chunking is not treated as externally fixed, but as a process to be gradually internalized through repetition and exposure.
Rhythmic Consistency as a Structural Cue
Second, for syntactic chunking to operate in a stable manner, learners must be provided with a consistent system of signals that allows them to detect structural boundaries. In this study, that role is fulfilled not by the precision of individual sounds, but by rhythm and pattern maintained through repetition. When sentences are presented in uniform rhythmic units, learners are able—prior to analytical processing—to perceptually identify points of syntactic cohesion and transition. Through fine-grained formal markers such as function words and sentence-final endings, recurrent structural patterns that appear in the same positions, and sentence sequences that share parallel rhythmic patterns, the conditions for structuring can be acquired.
Optional Vocalization
Third, vocalization should be understood as a flexible mode. The essential point in this study lies in revealing structure through repetition and rhythm. Depending on the learner’s situation, not only full vocalization, but also low-level vocalization or shadowing the teacher’s voice can be regarded as equally valid pathways toward structural embodiment. This approach also enhances accessibility and accommodates learner diversity, particularly for those who experience psychological discomfort associated with vocalization
Delayed Explicit Explanation and Structural Reconfiguration
Fourth, explicit grammatical explanation should be delayed until after structural familiarization has taken place. Even when learners do not yet grasp the full meaning of a sentence, repeated exposure allows them to develop a sensory awareness of where the sentence segments and how it continues. Explanations of grammatical rules and meaning operate most effectively when they are introduced only after structural recognition has reached a certain degree of st Classical Chineseability. This represents a proposal to adjust the cognitive timing at which explanation is made to operate. When explanation is provided at a stage where learners have already come to perceive the overall contour of sentence structure through repetition, it functions not as the memorization of rules but as the verbalization of structural intuition.
Subsequently, learners engage in reading unsegmented, continuous texts in which no prior syntactic chunking is provided. At this stage, learners reconstruct structure by autonomously setting syntactic units within the sentence, drawing on the structural sensibility embodied through earlier repetition and rhythm. This process does not constitute analysis aimed at interpretation or translation, but rather practice in autonomous structuring for the purpose of maintaining structure while reading.
Simultaneous Reinforcement of Syntactic Structure and Lexical Familiarity
Fifth, the dual effect whereby syntactic structure learning and familiarity with Sino-Korean vocabulary are simultaneously reinforced should be taken into consideration. In the context of Hanmun education for students of Korean Studies, repeated exposure is not confined to grammatical instruction alone. Personal names, place names, institutional terms, and idiomatic expressions that frequently appear in classical texts must be recognized as cohesive units prior to translation in order for reading to become stable. When such elements are naturally exposed through repetition at the level of syntactic units, they are more likely to be perceived as conventionalized combinations before any individual lexical analysis takes place. In this respect, toksong extends beyond the embodiment of function-word grammar to encompass the sensorial familiarization of naming conventions and expression units characteristic of Korean Studies texts.
Lexical units that occur frequently in classical texts—such as 富, 貴, 財, 賤, 道, 職, 君, 賢人, 小人輩, 大人, and 居士—belong to a core lexical stratum that must allow for the immediate retrieval of meaning simultaneously with phonological recognition, rather than remaining at the level of mere visual or phonetic identification. While only a limited selection has been presented here for reasons of space, the actual scope of such vocabulary is far more extensive. Moreover, these items are closely connected to layers of Sino-Korean vocabulary that remain in active use in modern Korean, allowing learning effects to extend beyond the classroom into broader language use. When these Sino-Korean pronunciations are naturally exposed through repetition at the level of syntactic units, lexical familiarity is reinforced in tandem with the stabilization of structural processing. In this sense, the present approach possesses a design advantage in that it does not separate syntactic structure from lexical sensibility, but rather addresses both in an integrated manner within the same repetitive process.
3) Classroom Implementation of the Design Principles
For the design principles outlined in the previous section to function effectively, actual classroom design must first clarify the units in which syntactic boundaries become stably perceptible during learners’ processing of texts.
Accordingly, the present study sets 'relative clause' structures as the basic learning unit for the purpose of syntactic training in Hanmun education. The term 'relative clause' used here does not apply the strict typological category of modern linguistics as such, but rather serves as a working definition established for the convenience of instructional design and analysis. In this study, a relative clause refers to a structural unit that can be recognized as a closed syntactic chunk through the combination of specific function words, and that can be extended to serve various sentence constituents such as subject, object, or complement.
For example, in a sentence such as “君之所欲者富也, 君之所欲” contains the verb 欲; however, it has already been nominalized by 所, and through its combination with 者 it is fixed as a closed noun phrase. This noun phrase is structurally separated from the subsequent 富也. Even without interpreting the meaning, learners can structurally recognize that the sentence consists of two cohesive syntactic units, and that the relative clause functions as the central constituent of the sentence. In this sense, the relative clause does not operate as a single grammatical item, but rather as a unit that provides a learning pathway from structural recognition, through constituent formation, to the comprehension of the sentence as a whole.
The following subsection therefore presents how such relative-clause units are realized in recurrently observable forms in actual texts, through core patterns that frequently appear in Park Ji-won朴趾源’s Heosaengjeon 許生傳, Yi Yong-hyu李用 休’s Chagŏgi 此居記, and poems by Chŏng In-bo鄭寅普, which are used as materials in the author’s Hanmun class. The presentation of these patterns does not consist in listing grammatical items, but functions as applied examples that demonstrate how the theoretical framework established in the previous section—namely, structural recognition and syntactic unit processing prior to meaning interpretation—can be secured in real textual contexts. Moreover, the core patterns presented here are not independent grammatical items, but applied examples that show how relative clause structures are formed, bound, and extended as sentence constituents.
One of the most basic ways in which relative clauses operate is through a sequence in which a structure containing a verb is first converted into a nominal unit by the combination of 所 and a verb, and is then bound within the sentence through the marker 之. Although these two operations can be distinguished at the level of grammatical description, in actual Hanmun texts they do not appear as separate steps. Rather, they function together as a continuous mechanism through which actions or events are recognized as stable syntactic chunks.
The 所 + V structure first forms a processable unit within the sentence by nominalizing an action or event. What is important here is that learners do not need to analyze this expression as a verb-centered predicative structure. Through repeated exposure, the expression is instead recognized as a single chunk. In this sense, 所 + V functions as a minimal unit for the formation of relative clauses, providing a formal basis for structural recognition prior to semantic interpretation. In Heosaengjeon, such 所 + V structures occur with relatively high frequency and offer typical examples of relative clause formation. Expressions such as 所欲 (soyuk 소욕), 所爲 (sowi 소위), and 所見 (sogyeon 소견) all contain verbs (欲, 爲, 見), yet in actual reading they are processed not as descriptions of actions, but as nominal units.
These structures are immediately extended by combining with a higher noun, as in 君之所欲 (gunji soyu 군지소욕)or 人之所爲 (inji sowi 인지소위), or by modifying a following noun, as in 所賣之財 (somaeji jae 소매지재) or 所見之事 (sogyeonji sa 소견지사). Through this extension, the entire structure is bound into a single noun phrase. At this point, 之 functions as a rhythmic and structural binding signal that fixes the previously formed structure as a single unit. Upon encountering 之, learners can intuitively recognize—before interpreting meaning—that the preceding structure has been grouped into one noun phrase.
This binding mechanism becomes even more explicit in Yi Yong-hyu’s Chagŏgi. The text can be regarded as a textbook example of A 之 B–type structures, particularly through the repeated use of demonstrative-based noun groups centered on 此, which greatly reinforces the effect of syntactic binding. In sentences such as 此人居此所也, 此國此州此里, and 此人……奇士也, the whole expression is first recognized as a single referential unit, even though verbs or predicative elements are present. The following paired sentences illustrate this effect especially clearly:
此居 (ch’agŏ 차거), 此人居此所也 (ch’ain kŏ ch’aso ya 차인거차소야).
此所 (ch’aso 차소), 卽此國此州此里 (chŭk ch’aguk ch’ajŏu ch’ari 즉차국차주차리).
Here, the repeated pattern of 此 + noun group strengthens the effect of grouping the preceding structure into a single unit. Without analyzing each sentence in detail, learners can recognize stable syntactic boundaries through the repetition of the same structural frame, and are prepared to process the information that follows in the same way.
A similar pattern of structural recognition is found in sentences such as 如欲求之 (yŏyok kuji 여욕구지), 當於此記 (tangŏ ch’agi 당어차기). The sentence-initial 如 functions as a structural signal that opens a conditional premise, while 欲求 is processed not as a sequence of individual actions, but as a compressed unit representing the content of that condition. The following 之 then serves as a binding marker that temporarily closes this premise, allowing 如欲求之 as a whole to be recognized as a stable syntactic chunk prior to semantic interpretation.
From Syntactic Chunks to Judgment Units: 者, 也, and 則
The next step concerns the process by which these secured syntactic units are confirmed as the central constituents of a sentence and come to function as units of judgment. In this process, functional particles such as 者, 也, and 則 play a crucial role.
First, 者 functions as a signal that closes a previously formed relative clause into a single noun phrase. In sentences from Hŏ Saeng-jŏn such as 君之所欲者 (kunji soyokcha 군지소욕자) 富也 (puya 부야) and 人之所爲者 (in’ji sowicha 인지소위자) 天也 (ch’ŏnya 천야), structures like 君之所欲 and 人之所爲have already been recognized as unified chunks at the structural level. The appearance of 者 fixes these chunks as the subject of the sentence. Even before receiving any grammatical explanation, learners perceive—through rhythm—that the structure up to 者 has reached a point of closure. These functional markers are first recognized structurally through repeated rhythmic exposure, which naturally leads into the discussion in the next section on structural automatization in environments of parallelism, antithesis, and repetition.
The following particle 也 functions as a marker that brings judgment on this noun phrase to completion. In expressions such as 富也 and 天也, 也 does not introduce new information; rather, it signals that the sentence has reached a structural endpoint. Through chanting and repeated reading, the 者–也 combination becomes embodied as a stable judgment rhythm, allowing learners to acquire a clear structural sense that a complete judgment has been formed, even prior to semantic interpretation.
A similar rhythm of judgmental closure can be observed in expressions such as 奇 士也 (기사야) in Yi Yong-hyu’s Chagŏgi, or in phrases like 竟體是蘭薰 (경체시나훈) found in Chŏng In-bo’s poetry. In these cases, 也 or 是 functions less as a carrier of propositional meaning than as a closing signal that binds the preceding structure into a single unit of judgment.
Meanwhile, 則 plays the role of placing these judgment units within a broader logical flow. As a marker that introduces conditions, contrasts, or conclusions, 則 forms a structural expectation—prior to semantic interpretation—that a result or logical development is about to follow.
Repetitive Structural Environments : Parallelism and Rhythm
In actual Classical Chinese texts, such structures accumulate within environments characterized by the repetition of function words, parallel arrangement, antithetical pairing, and rhythmic similarity. These environments form conditions under which learners can automatically recognize structure even while reading long sentences. In other words, relative clauses do not function as isolated grammatical items; rather, they operate most stably as core syntactic units within repetitive structural environments. In the prose passages of Hŏsaengjŏn, parallel structures repeatedly appear in which two syntactic units are perceived as forming a clear pair through the rhythm of recitation, prior to any interpretation of meaning. The following example illustrates particularly well how such structural recognition operates automatically:
有田有妻, 何苦爲盜.
生無盜賊之名, 而居有妻室之樂.
行無逐捕之患, 而長享衣食之饒.
The key feature of this passage lies in the fact that, even without interpreting the meaning of each sentence, the structure itself is first perceived as forming a “pair.” The two sentences share the following identical structural frame:
生無 A, 而居有 B
行無 C, 而長享 D
Here, the alternation between 無 and 有, the shift marked by 而, and the almost identical sentence length together produce a strong parallel rhythm in the process of recitation. As a result, learners are able to immediately recognize that the two sentences repeat the same structure and stand in a corresponding relationship, even without knowing their meanings.
In particular, expressions such as 盜賊之名, 妻室之樂, 逐捕之患, and 衣食之饒 are each presented as closed syntactic chunks composed of the pattern “noun + 之 + noun.” Within the rhythm of recitation, these expressions are stably maintained as single units. At this stage, learners structurally recognize 無 X and 有 Y as individual state units, and perceptually grasp 而 as a transitional signal that contrasts and connects the two states. When read aloud, these sentences are naturally segmented as follows:
生無 / 盜賊之名 / 而 / 居有 / 妻室之樂
行無 / 逐捕之患 / 而 / 長享 / 衣食之饒
These points of segmentation are not the result of semantic analysis, but structural boundaries that are first fixed through rhythm. What is crucial here is that such parallel and antithetical structures facilitate reading not because they make meaning “easier,” but because the two syntactic chunks that constitute the sentence are repeatedly presented within an identical structural frame. This directly supports the central concern of this paper: that the core task of reading lies not in whether rules are understood, but in whether sentence structure can be maintained during processing.
As discussed earlier, Yi Yong-hyu’s Ch’ageogi can be regarded as a text in which such repetitive structural environments are most overtly manifested. Similar environments are also strongly realized within the poetic rhythm of Chŏng In-bo’s poetry. Because poetry tends to maintain relatively stable line length and rhythm, once a basic rhythm is grasped, the task of identifying syntactic boundaries becomes comparatively easier. However, beyond this initial segmentation, it remains necessary to identify further layers of parallel arrangement within the text. In consecutive verses such as the following from Chŏng Inbo’s poetry:
忽遌伊人喜滿襟, 毫端如意道幽深.
秋山似解琴中趣, 一陣淸風自遠林.
Each line shares a similar syntactic length and rhythm and is arranged in parallel. In this context, function words such as 忽, 似, and 自 operate as signals that reveal structural correspondence between lines. In addition, idiomatic expressions such as 如 意, 秋山, 淸風, 遠林, and 幽深 can also be acquired through repetition within this rhythmic and parallel structure.
Conclusion
This paper has sought to reconsider what kind of reading competence is required prior to meaning interpretation in Hanmun education for students of Korean Studies, and to identify its core in the ability of syntactic recognition. The central problem of Hanmun reading addressed in this study is closely related to learners’ prior learning experiences concerning how Hanmun texts have been processed in terms of phonetic and structural units. In Hanmun texts, which are presented as continuous strings without punctuation, learners must identify sentence boundaries and structures on their own. In actual reading practice, however, it is frequently observed that grammatical rules and lexical knowledge already acquired by learners fail to connect to the recognition of sentence structure. This suggests that the core difficulty of Classical Chinese reading lies not so much in whether learners know basic grammar and character meanings, but rather in whether they are able to maintain sentence structure while reading.
Against this backdrop, this paper has attempted to reposition the role historically played by regionally differentiated pronunciation-based reading practices within the East Asian textual tradition in the context of Classical Hanmun course design for Korean Studies. Reading Hanmun texts through different phonetic systems functions as a cognitive mechanism that enables learners to maintain sentences as syntactic units rather than as units of meaning. When specific patterns of pronunciation, rhythm, and intonation are accumulated through repeated chanting, learners eventually become able to automatically segment Hanmun texts into syntactic units and proper-name units prior to semantic interpretation.
From this perspective, the toksong proposed in this paper is not a reproduction of traditional learning practices, nor a supplementary tool for memorizing meanings. Rather, toksong is redefined as a form of cognitive training that enables learners to follow Hanmun texts while maintaining stable syntactic units. In particular, toksong based on Korean pronunciation should not be regarded as a matter of convenience, but as a rational pedagogical strategy that allows learners to sustain sentence structure during reading. Based on this line of inquiry, this paper has proposed toksong-centered syntactic recognition training as one possible approach to Hanmun course design for students who have completed basic-level instruction.
Furthermore, if such an approach were to be reconstructed within the same conceptual framework at the introductory level, it could potentially yield meaningful effects across the entire learning trajectory from the elementary to the intermediate stages. At Leiden University, Korean Studies students currently enter departmental Hanmun courses after completing introductory Classical Chinese classes offered by the Department of Chinese Studies. However, discussions are now underway to reconsider this instructional structure and to reorganize it into a stepwise Hanmun curriculum better aligned with the research and educational context of Korean Studies. The proposal advanced in this paper directly engages with this ongoing institutional transition.
That said, sufficient empirical data have not yet been accumulated to assess the actual learning effects of chanting-centered syntactic recognition training in classroom settings. Accordingly, the present proposal is offered as an educational possibility, and its effectiveness will need to be examined more precisely through future research, including classroom observation, analysis of learner responses, and assessment of changes in reading performance.
Notes
Reference List
- Anderson, John R. The Architecture of Cognition. Cambridge, MA: Harvard University Press, 1983.
- Barsalou, Lawrence W. “Perceptual Symbol Systems.” Behavioral and Brain Sciences 22, no. 4 (1999): 577-660. ____. “Grounded Cognition.” Annual Review of Psychology 59 (2008): 617-645.
- DeKeyser, Robert. “Skill Acquisition Theory.” In Theories in Second Language Acquisition, edited by Bill VanPatten and Jessica Williams, 97-113. Mahwah, NJ: Lawrence Erlbaum, 2007.
- Denecke, Wiebke. “Translation.” In The Oxford Handbook of Classical Chinese Literature, edited by Wiebke Denecke, Wai-yee Li, and Xiaofei Tian, 560-578. Oxford: Oxford University Press, 2014. ____. “Worlds without Translation: Premodern East Asia and the Power of Character Scripts.” In A Companion to Translation Studies, edited by Sandra Bermann and Catherine Porter, 204-216. Oxford: Wiley-Blackwell, 2014.
- Goody, Jack. The Logic of Writing and the Organization of Society. Cambridge and New York: Cambridge University Press, 1986.
- Graham, Angus. C. Studies in Chinese Philosophy and Philosophical Literature. Albany: State University of New York Press, 1987. ____. Disputers of the Tao: Philosophical Argument in Ancient China. La Salle, IL: Open Court, 1989.
- Graham, William A. Beyond the Written Word: Oral Aspects of Scripture in the History of Religion. Cambridge: Cambridge University Press, 1987.
- Jang Kyŏngjun 장경준. “Beobeung sojang ‘Beophwagyeong’ gugyeol-ui eosun pyosi법응스님 소장 법화경 구결의 어순,” Gugyeol yeongu구결연구 47 (2021): 41-78.
- Jang, Yoo-seung 장유승. “Hanmun eumdok-ui silsang-gwa byeonhwa: Joseon hugi gyeongyeon-ui eumdok-eul jungsimeuro 한문 음독의 실상과 변화: 조선후기 경연의 음독을 중심으로,” Hanguk hanmunhak yeongu 한국한문학연구 67 (2017): 251-273.
- Jin Jae Kyo 진재교. “Hanguk hanmunhak yŏn’gu wa ‘tong Asia’: Tong Asia hanmunhak ŭi kŭnŭngsŏng 한국한문학연구와 동아시아: 동아시아 한문학의 가능성,” Hanmunhakpo한문학보27, no. 1(2012): 3-29.
- Keum Jang-tae 금장태. “Sŏngjae yu junggyo ŭi ǔmaknon 省齋 柳重敎의 樂론,” Chonggyo wa Munhwa 종교와 문화 13.（2007）：173-210.
- Kornicki, Peter F. Languages, Scripts, and Chinese Texts in East Asia. Oxford: Oxford University Press, 2018. ____. “Reading Sinitic Texts in the Vernaculars.” In Languages, Scripts, and Chinese Texts in East Asia, chap. 6. Oxford: Oxford University Press, 2018.
- Kwon, Moon Bong 권문봉. “Jeontongjeok seodang gyoyukgwa hyeondaeui hanmun gyoyuk-e daehayeo 전통적 서당교육과 현대의 한문교육에 대하여,” Hanja-Hanmun gyoyuk 한자한문교육11 (2003): 167-187.
- Kim, Eun Kyung 김은경, “Joseon Sidae Dokseo bangbeopron yeongu 조선시대 독서 방법론 연구,” Hanmun gogeon yeongu 한문고전연구 13 (2006): 303-346.
- Kim, Jin-bong 김진봉, Kim, Dong-gyu 김동규, Shim, Eui Shick 심의식, and Jeong, Hyo-young 정효영. “2022 Gaejeong hanmun-gwa gyoyuk gwajeong ‘gyosuhakseup’ui juyo naeyong bunseok yeongu 2022 개정 한문과 교육과정 ‘교수·학습’의 주요 내용 분석 연구,” Hanmun gyoyuk nonjip 한문교육논집 63 (2024): 109-142.
- Lakoff, George, and Mark Johnson. Metaphors We Live By. Chicago and London: University of Chicago Press, 1980. ____. Philosophy in the Flesh: The Embodied Mind and Its Challenge to Western Thought. New York: Basic Books, 1999.
- Miller, George A. “The Magical Number Seven, Plus or Minus Two: Some Limits on Our Capacity for Processing Information.” Psychological Review 63, no.2 (1956): 81-97.
- Mun Hyŏnsu 문현수. “Goryeo sidae ‘Hwaeomgyeong’ jeomto seokdok gugyeol gwa ‘Yugasajiron’ jeomto seokdok gugyeol-ui hyeonto bangsik bigyo yeongu 고려시대 화엄경 점토석독구결과 유가사지론 점토석독구결의 현토방식 비교연구,” Hangukhak Yeongu 한국학연구 78 (2025): 81-116.
- Ott, Margus. Embodiment Theory and Chinese Philosophy: Contextualization and Decontextualization of Thought. London: Bloomsbury Academic, 2024.
- Park, Gyun Yeol 박균열, and Lee, Sang Ho 이상호. “Hanmun dokseoseong-ui uiuiwa teukjing: Jirisan Gyeongnam jiyok-eul jungsimeuro 한문 독서성의 의의와 특징: 지리산 경남 지역을 중심으로.” Yulli gyoyuk yeongu 윤리교육연구 20 (2009): 257-274.
- Ryoo Jeong-min 류정민. “Taja ŭi sisonŭro pon Koripŏ ŭi tachŭngsŏng: yŏngmigwŏn hanmun munŏnmun kyoyuk ŭi p’edaegoji t’amgu.” 타자의 시선으로 본 고립어의 다층성-영미권 한문 문언문교육의 페다고지 탐구. Hanja-Hanmun kyoyuk 한자한문교육 51 (2021): 115-138.
- Shin, Doo Hwan신두환. “Seongdok gwa hanmun gyoyuk 성독과 한문교육.”Hanja-Hanmun gyoyuk 한자한문교육 21 (2008): 153-179.
- Song, Pyung Nyul 송병렬. “2009 gaejeong hanmun-gwa gyoyuk gwajeong ‘ilgi’ yeongyeok-ui naeyong bunseok-gwa hyanghu gwaje 2009 개정 한문과 교육과정 ‘읽기’ 영역의 내용분석과 향후 과제.” Hanmun Gyoyuk Nonjip 한문교육논집 39 (2012): 95-124.
- Yoon, Jae Min 윤재민. “2007-nyeon gaejeong hanmun-gwa gyoyuk gwajeong-ui ilgi yeongyeok bunseok 2007년 개정 한문과 교육과정의 읽기 영역 분석.” Hanja-Hanmun gyoyuk 한자한문교육 21 (2008): 23-47.
- Yoon, Jae-min 윤재민. “2015 Gaejeong gyoyuk gwajeong-e ttareun hanmun-gwa gyogwaseo gaebal banghyang-e gwanhan yeongu-ui pillyoseong 2015 개정 교육과정에 따른 한문과 교과서 개발 방향에 관한 연구의 필요성.” Hanmun gyoyuk nonjip 한문교육논집 47 (2016): 5-20.
- Zhu Xi 朱熹. Zhuzi yulei 朱子語類. Seoul: Sonamu, 2001.
News
