School of Foreign Languages and Literature, Shandong University, Jinan, China
1 Introduction
Since the 17th century, the development of French Sinology has evolved from missionary Sinology to academic Sinology, with the translation of Chinese documents consistently serving as the vanguard of research. The French Jesuit missionary Joachim Bouvet (1656-1730), a central figure of 17th-century missionary Sinology, took the Latin translation and systematic annotation of the Yijing (Book of Changes) as the starting point for his research, fostering an in-depth intellectual dialogue between the Yijing and the Bible and seeking cultural and doctrinal anchors for the spread of Christianity in China (Von Collani, 2009). Emmanuel-Édouard Chavannes (1865-1918), the founder of modern French Sinology, gave equal weight to the French translation, annotation, and historical research of the Shiji (Records of the Historian), producing a landmark achievement in modern French Sinology (Chavannes, 2014). His student Marcel Granet (1884-1940) expanded the application of translation-based research by translating over seventy poems from the Shijing (The Book of Poetry) and conducting sociological studies on them (Li & Bai, 2024). Rémi Mathieu (1948- ) is a key representative of this paradigm in contemporary French Sinology. He long taught at Université Paris Diderot (Paris VII), which merged into Université Paris Cité in 2019, and served as a sinology researcher at the Centre national de la recherche scientifique (CNRS). For his outstanding contributions to the study and translation of Chinese classics, he was awarded the Ordre des Palmes Académiques by the French Ministry of Education and the 11th Special Book Award of China (Mathieu & Lu, 2021). His early translation and research focused particularly on texts containing Chinese myths and legends, including the Mutianzi zhuan (The Travels of King Mu), Shanhaijing (Classic of Mountains and Rivers), Huainanzi (Book of Prince of Huainan), Liezi (The Book of Lieh-tzu), and Soushen Ji (In Search of the Supernatural). He regards translation, annotation, and research as a continuous academic process, arguing that translation is not merely linguistic reproduction but the reconstruction of cultural ideas, and that annotation goes beyond explanation to reveal the internal logic and symbolic system of the text. The Eastern Jin dynasty supernatural tale collection Soushen Ji represents one of Mathieu’s significant achievements in the study of Chinese folk literature.
Soushen Ji, compiled by the Eastern Jin historian and writer Gan Bao, is the definitive anthology of zhiguai (supernatural anomaly accounts) literature from the Wei, Jin, and Southern and Northern Dynasties, and a milestone text in the history of classical Chinese fiction. The original 30-Vol. The work was completed in the early Eastern Jin, but was scattered and lost after the Song dynasty; Hu Yinglin compiled the 20-Vol. edition available today in the Ming dynasty from surviving fragments. The book extensively records matters of immortals, divination, ghosts, anomalies, extraordinary people, and folk legends from high antiquity to the Wei and Jin periods. Guided by the core creative principle set forth in Gan Bao’s “Self-Preface” - “to demonstrate that the divine way is not unfounded” (Gan, 2009)—it systematically preserves source materials on the folk beliefs, religious concepts, and social customs of the Han, Wei, and Six Dynasties. At the same time, with its mature narrative techniques and classic story motifs (such as Gan Jiang and Mo Ye, Li Ji Slays the Serpent, and Dong Yong), it provided core paradigms and source materials for later narrative literature, making it an indispensable classic for the study of ancient Chinese literature, folklore, and the social history of the Wei and Jin dynasties.
Focusing on Soushen Ji, Mathieu successively published a translation, À la recherche des esprits: Récits tirés du Soushenji (Mathieu, 1992), and a scholarly monograph, Démons et merveilles dans la littérature chinoise des Six Dynasties (Mathieu, 2000). The former centers on translation and textual reconstruction, while the latter aims at textual analysis and the construction of a coherent cultural system, with the goal of deepening global readers’ understanding of Chinese cultural connotations—specifically, deepening non-Chinese readers’ understanding of Chinese zhiguai literature. In these two works, Mathieu used translation to deepen textual interpretation and provide empirical evidence, forming a continuous “translation-research” mechanism. In this sense, he is not merely a translator, but a researcher and disseminator of Chinese culture.
As Yang (2024) notes, existing domestic and international studies of Rémi Mathieu have concentrated primarily on his work on the Shanhaijing, Huainanzi, and Mutianzi zhuan, with comparatively limited attention paid to his two books on Soushen Ji and even less comparative analysis of their internal connections and academic significance. Through a comparative analysis of Rémi Mathieu’s two books on Soushen Ji, this paper will demonstrate that he not only treats translation as the starting point and integral component of academic research but also, through detailed annotations, achieves a close reading of the text and the construction of a cultural context, thereby providing a solid foundation for subsequent theoretical analysis. Subsequently, he integrates scattered zhiguai narratives into an internally logical and coherent cultural system, revealing the deep structure of Six Dynasties folk thought.
Rémi Mathieu’s French translation of Soushen Ji allowed French readers systematic access to this classic of Six Dynasties supernatural tales for the first time. By examining his translation principles and the functions of the paratexts, we will explore how translation undertakes the dual role of presenting cultural semantics and generating academic questions in this work.
Rémi Mathieu stated, “The original text of the Soushen Ji is in classical Chinese, a style that is concise yet rich, far removed from the elegance of classical French. We must remain close to the original text while avoiding word-for-word translation” (Mathieu, 1992). He does not strive for literary rewriting but seeks fidelity to the context, diligently preserving the original narrative rhythm and cultural imagery.
2.1.1.1 Fidelity to Context, Closeness to the Original Text
At the lexical level, Mathieu’s approach demonstrates French Sinology’s high sensitivity to proper nouns. When faced with culturally specific terms lacking Western equivalents, such as kui (夔), wangliang (魍魎), and yu (蜮), he adopts a strategy of transliteration combined with annotation. For instance, he renders yu as “yu [tortues tripodes]” (Mathieu, 1992), preserving the phonetic form in the main text while providing a functional definition in square brackets. The deeper logic behind this practice is that the cultural information carried by a proper name cannot be exhausted by any Western translated term; transliteration is an acknowledgment of its alterity, while the annotation prevents this alterity from becoming unreadable. The annotation for yu, which traces its source back to the Shijing and cites the Lunheng as evidence (Mathieu, 1992), precisely embodies his scholarly habit of investigating word by word, entry by entry, by means of annotation.
At the syntactic level, Mathieu’s consideration lies in how to reproduce the original rhythm within the syntactic norms of the target language. For example, the parallelism linking the five phases with the five virtues in the original— “When wood qi is perfect and flawless, the virtue of benevolence prevails; when fire qi is pure, conduct follows the rites...”—is rendered in French as “Quand le bois est sans imperfections, la vertu d’humanité règne; quand le feu est pur, on agit selon les rites”, preserving the parallel structure of the original while adjusting the order of clauses and main clauses to conform to idiomatic French (Mathieu, 1992). This treatment neither sacrifices the argumentative momentum of the original text nor causes awkwardness in the translation through forced accommodation of rhythm. In other words, Mathieu does not pursue formal word-for-word equivalence, but rather seeks a dynamic balance between the two syntactic logics.
At the paragraph level, Mathieu’s translation demonstrates a grasp of the narrative rhythm characteristic of the zhiguai genre. In the passage, Niu Ai transforms into a tiger; the original text uses a concise chronological progression to complete the transformation from man to tiger. The French translation, “Ses formes se modifiaient et son corps se transforma” [His forms were altering and his body transformed], reproduces this process with a progressive expression (Mathieu, 1992). Similarly, the line “Their eyes shed tears that can yield pearls” is translated literally as “Leurs yeux versent des larmes qui peuvent donner des perles”, without any interpretive additions (Mathieu, 1992). This restrained approach is precisely a tribute to the defining feature of zhiguai literature, “recording the strange”—the translator’s duty is to present the strangeness, not to dissolve it.
2.1.1.2 Selective Translation
In Mathieu’s case, the essence of selective translation is not deletion, but the hierarchical distribution of information through the annotation system. Although the main text of story 300 is brief, the annotations number up to seventeen, listing sources from the Liji (The Book of Rites), Huainanzi, and Taiping Yulan (The Imperial Reader of the Taiping Era) to the Taiping Guangji (Extensive Records of the Era of Grand Peace), with successive documentary origins cited one by one (Mathieu, 1992). The main text serves the reading experience of the general reader, while the annotations meet the research needs of the specialist. This practice of placing textual research outside the main text allows the translation to possess both readability and academic depth—the former enabling zhiguai stories to circulate in the Francophone world, the latter ensuring that this circulation is not detached from the historical context of the original texts.
From this perspective, it is evident that he did not regard translation as the sole objective, but rather used it as a means to selectively translate source materials, thereby offering a systematic interpretation of and enlightenment on how people in ancient China—particularly during the Six Dynasties period—perceived the supernatural or spirits (Mathieu, 1992). In terms of material selection, to match the length constraints of the Connaissance de l’Orient series and avoid redundancy, the translator made choices, ultimately presenting roughly half the stories of a full translation (Mathieu, 1992). Mathieu selected only classic, narratively strong, and complete stories, omitting a large number of brief prophecies, repetitive anomaly accounts, and obscure fragments of five-phase calamities, thus making the text easier for readers from other cultures to understand and accept.
The direct purpose of this approach was to provide readers in the Francophone world with a reliable and rich basis in the original text, fulfilling the initial goals of textual reconstruction and cultural semantic comprehension.
In terms of form, the annotations in this book follow the entry-by-entry arrangement of the original text. Rémi Mathieu constructed a rigorous and systematic annotation style, the layout of which fully embodies the meticulous textual research tradition of French Sinology. The annotations strictly adhere to the original text’s entry-by-entry arrangement. Each note is independently numbered, corresponding to superscript numbers in the main text, forming a clear “text-annotation” structure. This arrangement allows readers to conveniently consult relevant explanations while reading the story, greatly enhancing reading continuity and the convenience of academic cross-reference.
In terms of content, Mathieu’s annotations serve the functions of factual verification and semantic explanation. After the translation of each story, he provided detailed notes addressing the content, including not only detailed introductions of figures, places, and so on, but also extensive supplementary explanations citing folk activities, cultural concepts, and sources of information, thus giving the annotated translation the function of an encyclopedia of ancient Chinese cultural knowledge.
At the level of factual verification, Mathieu identifies and annotates the historical figures, geographical evolution, and documentary sources involved. These verifications endow the translation with a reliable philological foundation, providing a trustworthy base text for specialist readers.
At the level of semantic explanation, Mathieu’s annotations delve into the cultural texture, elucidating objects, customs, and concepts.
Furthermore, Mathieu’s annotations are by no means an isolated listing of knowledge points; rather, through the combined effect of a cross-reference system and an appended “See also” list of documents, the entire work forms an internal academic reference network. General readers can thereby deepen their impressions of specific cultural representations, while specialist researchers can delve further along the leads provided by the annotations.
Take the story Panhu—The Origin of the Barbarians as an example. The notes following the translation read (Mathieu, 1992):
[4] Di Ku, see Section 2 above.
[5] “Pan” means vessel, and “hu” means gourd, hence the dog’s name. The motif of a creature being taken from the ear is the same as in Section 92.
[6] The “five colors” can be observed in the clothing customs of these barbarians.
[7] The mountains of Yunnan, Sichuan, and Guizhou are famous for their flora and are also refuges for southern ethnic minorities. Bizarre legends of apes abducting women circulate in these mountains (Section 308).
[8] Until the 20th century, there were records of the Yao people keeping a dog’s tail (W. Eberhard, China’s Minorities: Yesterday and Today, Wadsworth Publishing, Belmont, 1982, p. 81).
[9] Note the connection between rain and incest. In East Asian myth, sibling incest is almost always accompanied by floods or torrential rains.
[10] The Han Chinese have, since ancient times, called the indigenous people of ancient Indochina “Jiaozhi.” See Shanhaijing, Vol. 6, p. 3a.
[11] On famous mountains requiring sacrifices, see Shanhaijing, Vol. 5, p. 44b.
[12] “Man” is a general term for the southern aborigines (Section 307). “Yi” generally refers to barbarians, but in antiquity specifically denoted the ethnic minorities of eastern China (Section 117).
[13] Like the Han Chinese, they revere older people and ancestral customs.
[14] Liang is located southwest of present-day Xiayi County, Henan (Section 53). There was never a Liang-Han commandery in the Han dynasty; this possibly refers to Hanzhong in Shaanxi. Ba corresponds to the northern part of present-day Chongqing. Shu corresponds to Sichuan Province. Wuling is in Changde County, Hunan. Changsha corresponds to eastern Hunan. The Six Rivers are in Anhui Province (Section 21).
[15] This custom seems to have persisted at least until the early 20th century.
It can be seen that Mathieu’s annotations are rich in content, providing textual research and explanation of relevant cultural figures, objects, place names, customs, and documentary sources, and cross-referencing other works. Through this precise, entry-by-entry annotation, the whole book transforms a medieval collection of the strange into an academic text amenable to interdisciplinary study, achieving a rare balance between readability and reliability.
This translation of Soushen Ji was one of the projects supported by the UNESCO Representative Works series. This translation program was active for approximately 57 years, from 1948 to around 2005, with the intent of fostering broad international understanding and appreciation of the world’s literary heritage (Qiu, 1991). As part of the Connaissance de l’Orient series, its primary purpose was not translation per se, but the dissemination of Chinese culture to enable the world to better understand China (Wang & Cao, 2026).1 Therefore, everything from the selection of the translation to the design of the paratexts serves this core knowledge-dissemination purpose.
Mathieu has mentioned his thoughts on the translation and study of Chinese culture on numerous occasions, such as in interviews and lectures, which become important materials for deepening our understanding of him and enriching our knowledge of the text’s background. He emphasizes the subjectivity of the translator; in the practice of translation, the translator should facilitate intercultural exchange beyond the conversion of linguistic signs. Therefore, textual translation must be premised on a deep cognitive understanding of the culture, which is the very principle he applied in translating Soushen Ji.
In a conversation with Chen Wensheng, he remarked: I believe that translation must simultaneously respect this requirement: to possess both thought and literary grace. Because every sentence, every work, every author... has their own different requirements: some prioritize thought, others prioritize literary grace. I know the losses incurred by prioritizing a text’s thought: the words carefully chosen by the original author will become unrecognizable, ultimately turning into a lifeless word-for-word translation devoid of elegance, vitality, and inspiration. I can state categorically that one must equally respect the rigor of the source language, the beauty of the target language, fidelity to the author’s original meaning, and especially the pleasure of reading the text (for the French reader). ... A translator in the truest sense is an author, a creator of a work, if they can complete the conversion and transition between two sets of signs, two languages, and two cultures, taking into account both the original and the reader, whose readership was previously ignorant of the mysteries of his research field. This is why I remain wary of those grandiose claims that cannot be established on the basis of scientific analysis, as required by modern translation studies (Chen, 2016).
In an interview with China Writers Network, he also stated: Whether the characters in ancient Chinese myths are real or fantastical, none of that matters. ... What is important is to understand their origins and why the common people of the time revered them. Many of these myths are of great benefit for understanding the popular thought of their era and even the thought of the ancient sages, because they were cited by numerous great Confucian or Daoist thinkers, such as Zhuangzi, Liezi, Liu An, and even Confucius, Mencius, and Xunzi. Myths also tell people about the origin of the world and how humanity found its place among the spirits of heaven and earth. Mythology is an indispensable and important field for interpreting ancient and modern Chinese culture (China Writers Network, 2022).
2.2.2.1 Map and Chronology
Mathieu uses a map and a historical chronology in the translation to provide Western readers with a historical context for understanding Soushen Ji. The volume opens with a “Map of China during the Eastern Jin Dynasty,” marking the distribution areas of various ethnic minorities, thereby enabling readers to intuitively grasp the geographical space of the stories (Mathieu, 1992). Furthermore, the historical chronology in the book not only sorts out the imperial reigns from the Han to the Jin, but also places the historical events mentioned in Soushen Ji into their corresponding temporal coordinates, giving the narratives a concrete historical dimension.
2.2.2.2 Index System
The author also compiled an index system that serves as a guide to knowledge. At the end of the volume, a “Thematic Index” and an “Index of Proper Names” are appended, comprising over a thousand entries, and classifying various concepts such as historical figures, mythological figures, and geographical names. Readers can use the indices to quickly locate all passages related to a specific theme, integrating fragmented information into a searchable knowledge network, thereby greatly enhancing the reference value of the translation.
2.2.2.3 Annotation System
(1) Documentary Knowledge Annotations
In Rémi Mathieu’s French translation of Soushen Ji, the annotation system exhibits an extremely broad range of documentary references, encompassing zhiguai fiction, works of various Masters, successive dynastic histories, and large-scale encyclopedias (leishu), forming a multi-layered network of textual corroboration.
In terms of cross-referencing zhiguai texts, Mathieu often takes an account in Soushen Ji as a starting point and cites similar accounts from other zhiguai texts to reveal the transformation of motifs. In terms of tracing the sources of intellectual concepts, he broadly cites the Masters and Confucian classics to explain the cosmology and ritual systems behind the stories. In terms of dual support from official histories and encyclopedias, Mathieu extensively quotes dynastic histories and large encyclopedias to provide a historical context for the supernatural tales.
For example, following the entry-by-entry notes on “Panhu—The Origin of the Barbarians,” the author writes (Mathieu, 1992):
See also Fengsu Tongyi, lost text, pp. 489-490; Hou Hanshu, Vol. 86, p. 2829; Taiping Yulan, Vol. 758, p. 7a-b, Vol. 905, p. 4a; Yiwen Leiju, Vol. 94, p. 1638; Fayuan Zhulin, Vol. 11, p. 4a; Chuxueji, Vol. 29, p. 712; Weilüe and Jinji (Taiping Yulan, Vol. 785, p. 4a-b).
Another example is the author’s even richer documentary references for the passage “On Transformation” (Mathieu, 1992):
See Zheng Xuan’s commentary on the Liji, Vol. 52, p. 1625B, and Baihu Tong, Vol. 3, p. 10a-b.
This passage can be compared with Da Dai Liji, Vol. 81, p. 475; Huainanzi, Vol. 4, p. 6a; Kongzi Jiayu, Vol. 6, p. 4a; Bowu Zhi, 5-189, p. 64 (notes 28-33); and Baopuzi, Vol. 15, p. 65.
See also Lüshi Chunqiu, Vol. 9, p. 8b note, which cites an old text of the Huainanzi. Compare with Bowu Zhi, 4-150, p. 48.
On the divine turtle, see also Zhuangzi, Vol. 26, p. 402 (Liu Jiakui’s translation, p. 299); Shiji, Vol. 128, p. 3225. As for the fox, see Xuanzhong ji (in Gu Xiaoshuo Goushen, p. 492) and Mongol and Siberian Studies.
See also Taiping Yulan, Vol. 742, p. 4a; Vol. 888, p. 4a; Vol. 934, p. 6a; Fayuan Zhulin, Vol. 43, p. 12a; Yiwen Leiju, Vol. 82, p. 1410; Chuxueji, Vol. 30, p. 750; Taiping Guangji, Vol. 457, p. 3736-3737; .
Mathieu demonstrates a high degree of philological expertise in his annotations. By invoking a wide range of classical Chinese documents, he constructs a rigorous and rich knowledge network of ancient Chinese culture, linking different concepts and theories, and enabling readers to understand the extensive characteristics of Chinese culture from a more macroscopic perspective.
(2) Cultural Knowledge Annotations
Rémi Mathieu points out that translation is not only a process of linguistic conversion but also the reproduction of knowledge, a gateway into the internal logic of ancient Chinese texts. This translation philosophy leads him to supplement the translation of the original text with a wealth of information to facilitate Western readers’ understanding of the text, turning the translated work into a field for the dissemination of ancient Chinese knowledge.
The dual function of the annotation system is the cornerstone of Mathieu’s paratextual strategy. His annotations serve the dual functions of factual verification and semantic explanation. He verifies and interprets, one by one, the historical figures, geographical evolution, and related cultural knowledge involved in the stories.
Take the story The City Transformed from a Turtle (Tortoise-City) as an example. The notes following the translation read (Mathieu, 1992):
[8] That is, 311 BC. Some encyclopedias give the year as twelve or thirteen years [of King Huiwen’s reign].
[9] Zhang Yi was the prime minister of King Huiwen of Qin and later returned to his home state of Wei. See the Shiji.
[10] The capital of Sichuan Province is located on the banks of the Yangtze River.
[11] The practice of building houses or other structures on living beings (human or animal) has existed in China since ancient times and continued into the early 20th century. Here, the turtle once again symbolizes solidity and longevity. Lu Shiqiu, in his Chinese Folk Worship, points out that during the Guangxu reign (1875-1908), people still venerated the turtle as the guardian deity of dykes.
In summary, through this system of paratexts integrating annotations, maps, a chronology, indices, and specialized studies, Mathieu’s French translation of Soushen Ji becomes a veritable encyclopedia of ancient Chinese culture. It clears reading obstacles for the general reader while providing a reliable documentary basis and rich academic leads for the specialist researcher.
Rémi Mathieu’s study of Soushen Ji underwent a deepening process from textual translation to scholarly interpretation. In 1992, he published a complete French translation that primarily addressed the “what” questions—exegesis of words, verification of names and objects, and reconstruction of historical background, aimed at general Western readers. Eight years later, his monograph placed the same text within a larger interpretive framework, answering the “why” questions—why these stories were told this way, why they were recorded as such, and what they reveal. The aim shifted from selecting texts for translation to provide readers access to the original canonical work, to drawing materials from that work to write a scholarly monograph that further deciphers the characteristics of Six Dynasties zhiguai literature.
In his monograph, Mathieu employs a distinctive tripartite text organization: “thesis - original textual example - analysis.” This format is not a simple accumulation of materials, but a concrete presentation of his methodology: he first proposes a clear thesis (e.g., “the core function of the motif of abnormal birth is portent”), then cites relevant stories from Soushen Ji as case evidence, and finally, through detailed textual analysis, places these cases within a broader cultural context for commentary, revealing the deeper logic behind them.
Take Chapter 6, “Abnormal Births and Conceptions,” as an example. The author assembles and places seemingly disconnected entries—such as No. 252 (Lady Wu dreams of the sun and moon entering her bosom and gives birth to Sun Ce and Sun Quan), No. 53 (Guan Lu predicts the death of an abandoned infant born to a “lascivious woman” upon entering a stove), No. 342 (the oviparous birth myth of the king of Fuyu), No. 347 (a woman gives birth to a snake), and No. 349 (a woman encounters a feathered being and gives birth to a snake)—under the same analytical framework. This thematic reorganization is itself a structuralist operation: it extracts these stories from their respective isolated contexts and treats them as different “variants” of the same mythical motif. First, the author points out that the core of such narratives is not the birth event itself, but its prophetic function—abnormal conception and abnormal birth are often not the end of the story, but its starting point, a symbolic portent of political change, social upheaval, or an individual’s extraordinary destiny. Second, he traces the cultural origins of this thinking, attributing it to the deeply rooted cognitive logic of “interaction between Heaven and man” (tianren ganying) among Six Dynasties literati: an inductive relationship exists between celestial phenomena and human affairs, and abnormal natural occurrences (including abnormal human births) are “warnings” or “portents” from Heaven concerning worldly matters. And Gan Bao, as a historian, systematically presented in his Soushen Ji this fundamental human need to fill the cognitive gaps of the world (Mathieu, 2000). Finally, Mathieu concludes that in these narratives, the born hero “is often relegated to being the humble instrument of the destiny he transcends” (Mathieu, 2000). Through this combination of translation and analysis, Mathieu reveals the concept of the Mandate of Heaven reflected in Six Dynasties literary narratives.
The core pathway of this turn is indebted to the methodology of structuralist mythology. Rémi Mathieu believes that Soushen Ji is not the product of Gan Bao’s personal imagination, but rather “the crystallization of a vocabulary drawn by an eminent man of letters from diverse and sometimes distant sources—written (classics, Masters, dynastic archives) and oral (folk tales collected by officials)” (Mathieu, 2000). In his monograph, he draws on the structuralist mythological methodology of Claude Lévi-Strauss to reconstruct the seemingly scattered and heterogeneous zhiguai narratives of the Six Dynasties into a coherent cultural system possessing internal generative logic.
The core operation of structuralism lies in treating the manifold surface narratives as variants of a deep structure, and revealing the stable grammar hidden beneath the text by extracting recurring “binary opposition” pairs. Mathieu’s treatment of Soushen Ji is precisely the practice of this method. He does not treat each story as an isolated wonder but positions it within a coordinate network constituted by oppositional pairs. These core oppositional pairs include: mundane world/altered world, order/disorder, life/death, good/evil, human/non-human (gods, ghosts, monsters), nature/culture, dream/waking, and past life/later life (Mathieu, 2000). In doing so, he reconstructs Soushen Ji from a collection of literary texts into a symbolic system through which Six Dynasties literati understood the world and situated themselves.
In Mathieu’s narrative, dreams are often illusory, standing in opposition to reality. Dreams possess omniscience and a prophetic nature; they can reach the blind spots inaccessible to waking human cognition and interact with reality through this medium, conveying the unknown to the waking person. This unknown encompasses both the future in time and the uninhabited in space. Moreover, this unknown is not easily or necessarily established; people usually must follow the instructions of spirits seen in dreams and bear the corresponding consequences (mostly negative), with very few able to maintain a nihilistic attitude towards ghosts and gods in this process.
In Chapter 4, “The Meaning of Dreams”, Mathieu systematically sorts through nearly thirty dream-portent stories in Soushen Ji. He selects story No. 380, Jiang Ji’s Son Entrusts a Dream to Seek Office, as a typical case: Jiang Ji’s deceased son appears in a dream to his mother, complaining of his lowly position in the netherworld (Mount Tai) and begging his father to secure an official post for him there. Upon first hearing this, Jiang Ji flatly refuses, claiming that “dreams are illusory and not worth wondering at.” Yet the deceased son appears in a dream again, providing extremely precise real-world details, ultimately forcing Jiang Ji to believe (Mathieu, 2000).
Mathieu’s analysis proceeds on two levels. First, Jiang Ji represents the skeptical rational force in society at the time regarding the immortality of the soul. At the same time, the deceased spirit gradually refutes this skeptical force by providing concrete evidence. Second, the story’s core appeal—the deceased son’s longing for an official post in the netherworld—reveals a deep interweaving of the belief in the immortality of the soul and the social hierarchy of the Six Dynasties: death did not dissolve the fixation on status and position; the netherworld is but a mirror projection of the social structure of the living world. On this level, a dream-entrustment story deeply carries the social imagination of the posthumous world held by Six Dynasties literati, and the zhiguai fiction transcends the traditional label of “strange forces and chaos” to become a significant textual carrier for decoding the spiritual structure of medieval society (Peng, 2025).
Mathieu’s analysis of animal spirits is a typical case study of the “human/non-human” binary opposition. Animal images, as cultural symbols, have deeply penetrated and shaped the spiritual world of the Chinese people (Hu, 2025). Animal spirit stories are precisely a symbolic expression of this anxious sentiment.
Man and animal are both different, even opposed, and alike, because they share the same world: the same territory, similar behaviors, and especially the emotions frequently expressed in these types of short stories. In what are called zhiguai tales, the animal is endowed with reason, because it can master human language and its logic, and even when it does not speak, it can make itself heard and understood. Amorous relationships exist within this framework, just as in dreams, where two adjacent worlds touch each other and sometimes even merge (Mathieu, 2000).
He takes the story of “Azi” (the fox woman bewitching a man) as an example: a fox transforms into a stunningly beautiful woman to seduce a man, who ultimately indulges to the point of himself acquiring a fox’s physiognomy. Mathieu cites the Xuanzhong ji account— “a fox at fifty can transform into a woman, at one hundred into a beautiful woman... causing men to lose their minds”—pointing out that behind this narrative lies a complete theory of spirit metamorphosis (Mathieu, 2000).
He extends this analysis to cases such as an otter transforming into a beautiful woman to seduce a man, and a ghost transforming into a white dog to rape a woman, thereby distilling the deep structure shared by these stories, namely, the motif2: an animal spirit, through metamorphosis, disguises itself as a human, intrudes into the social order, and poses a threat to humanity, but is ultimately defeated and expelled by human power. The repeated enactment of this narrative pattern reflects the complex mentality of the Six Dynasties literati towards their social reality—the disappearance of order amidst the turmoil of the times allowed people to freely return to nature. Yet, the chaos brought by the absence of order simultaneously caused anxiety. During this period, the personal sentiments of life and ideal ambitions of the literati were tightly intertwined with the upheavals of the era; a melancholy over the brevity of life merged with a longing to establish merit and achieve greatness (Guo, 2026). The union of human and non-human is essentially a breaking of order; physical illness and familial chaos metaphorically represent the consequences of social disorder, thereby embodying the true psychological state of the people of that time.
Furthermore, Mathieu turns his attention to the depictions of distant, strange lands and their bizarre inhabitants found in zhiguai tales, such as the “Penetrating-Chest Country” (Chuanxiong guo) in the Bowu zhi and the “Country of Women” (Nü’er guo) in the Xuanzhong ji (Commentary on the Water Classic). He notes that such descriptions adhere to a strict spatial hierarchy: the farther from the center (the Central Plains), the more abnormal the inhabitants’ appearance, customs, and lifespan. He analyzes that the deep logic of this descriptive mode of “distance generating heterogeneity” is the theory of qi: the more distant the place, the more prevalent the “strange qi” (yiqi), thus giving birth to “monsters” (guaiwu) (Mathieu, 2000).
The fact that geography can be “strange” is itself of great significance for understanding the place that the ancient Chinese assigned to organized space. The earliest relevant treatises known to us, such as the Shanhaijing and later the Shuijing zhu, all reveal the close connection between territory, spirits, and world order. Strangeness resides in the incongruity between these usually harmonious elements, and especially in the feelings aroused by this incongruity. Chinese thought abhors disorder; geographical strangeness is first and foremost an order where spatial references are subverted, where gods dominate over humans, and where man is merely an uneasy spectator. (Mathieu, 2000)
From this, Mathieu points out that the fantastical geography in Six Dynasties zhiguai tales is not pure nonsense but rather an imagination of the world order by the ancients. By depicting strange realms as worlds opposed to the mundane - where lifespans are extremely long, humans and beasts are hybrid, and genders are undifferentiated—the people of the Six Dynasties were able to inversely define what they perceived as “normal,” “Chinese,” and “civilized.” Writing about fantastical geography was, in essence, a spatial expression of the Other used by the ancients to construct their cultural identity.
Although this paradigm is illuminating, it also places extremely high demands on the translator’s scholarly competence. Without sufficient philological grounding and theoretical vision, translation can easily degenerate into a mere accumulation of materials rather than a starting point for research.
Rémi Mathieu’s translation and research surrounding Soushen Ji provide a concrete case for understanding the research paradigm in French Sinology that takes translation as its academic starting point. From the annotated French translation in 1992 to the research monograph in 2000, Mathieu’s work reveals a clear progressive logic: the translation phase, through faithful rendering and systematic annotation, accomplishes textual reconstruction and the presentation of cultural semantics; the research phase, drawing on structuralist methods, reconstructs the scattered zhiguai narratives into an encoded and coherent cultural system, revealing the concepts of the Mandate of Heaven and social psychology behind them. In this process, translation is not an ancillary tool of research; rather, through the practice of translation and annotation, it undertakes the functions of close reading and question-generation, providing a solid textual basis for subsequent interpretation.
This approach is not Mathieu’s own invention but a contemporary continuation of the French Sinological academic tradition. From Chavannes’ empirical research grounded in the translation and annotation of the Shiji, to Granet’s socio-cultural study initiated by the textual analysis of the Shijing, French Sinology gradually formed an academic paradigm that starts with translation and deepens into research. Mathieu’s work on Soushen Ji is a quintessential contemporary practice of this paradigm.
There is, of course, room for discussion in Mathieu’s research. For instance, while his structuralist method reveals deep logic, it inevitably tends to simplify complex texts into binary oppositions; his holistic reconstruction of Soushen Ji also, to some extent, obscures the heterogeneity within the text. These issues remind us that when drawing on the achievements of Western Sinology, we should not only see the inspiration of its methodology, but also maintain critical reflection.
Mathieu’s work demonstrates that translation is not merely linguistic conversion, but a process of reconstructing cultural thought; research is not merely theoretical interpretation, but must be rooted in a meticulous reading of the original texts. Against the backdrop of contemporary academia increasingly emphasizing interdisciplinary dialogue, Mathieu’s research approach, “originating in translation,” may offer some useful reference for considering how Chinese literature can “go global”.
[1] Chavannes, E. (2014). Selected Sinological Works of Chavannes. Zhonghua Book Company.
[2] Chen, R. (2016). Interview with Rémi Mathieu. In Cross-cultural Dialogues (Vol. 35, pp. 37-60). SDX Joint Publishing Company.
[3] China Writers Network. (2022, May 27). Rémi Mathieu: My favorite Chinese Confucianism and Taoism. https://www.chinawriter.com.cn/n1/2022/0527/c431803-32431691.html
[4] Gan, B. (2009). Soushen Ji (Y. Ma & G. Zhou, Trans. & Annot.). Zhonghua Book Company.
[5] Guo, J. (2026). From The Book of Songs to Tang poetry: An analysis of the evolution of classical poetic aesthetic tradition. Appreciation of Masterpieces, (6), 100-102.
[6] Hu, Y. (2025). A Study on “Animal Metaphors” in Calligraphy Theory of Wei, Jin, Southern and Northern Dynasties (Master’s thesis). Sichuan Normal University.
[7] Li, X., F., & Bai, L. (2024). Journey Of Translation: Sinologists On Translation (Vol. 2). China Social Sciences Press.
[8] Mathieu, R. (1992). À La Recherche Des Esprits: Récits Tirés Du Soushenji. Gallimard.
[9] Mathieu, R. (2000). Démons Et Merveilles Dans La Littérature Chinoise Des Six Dynasties. Éditions You-Feng.
[10] Mathieu, R., & Lu, M. (2021). Translation and research of pre-Qin literature: An interview with French sinologist Rémi Mathieu. International Sinology, (3), 1-6.
[11] Peng, Y. (2025). Folklore narrative and cultural interpretation in supernatural tales of the Wei, Jin, Southern and Northern Dynasties. Journal of Yangzhou University of Education, 43(3), 1-6.
[12] Qiu, K. (1991). Introduction to the UNESCO representative works series. Chinese Translators Journal, (2), 53-55.
[13] Thompson, S. (1946). The Folktale. Holt, Rinehart and Winston.
[14] Von Collani, C. (2009). P. Joachim Bouvet S.J.: Sein Leben und sein Werk (L. Yan, Trans.). Elephant Press. (Original work published 1985)
[15] Wang, J., & Cao, D. (2026). The paratextual functions of the complete French translation of Strange Tales from a Chinese Studio by a French sinologist. Theory and Practice of Foreign Language Teaching, (1), 87-95.
[16] Yang, J. (2024). On French sinologist Rémi Mathieu’s study of The Classic of Mountains and Seas (Master’s thesis). Shandong University.
1 The concept of the “paratext” was first proposed by the French narratologist Gérard Genette, referring to various ancillary elements surrounding the main body of a literary work, and is divided into “peritext” and “epitext”. These different types of peritexts overlap or complement each other in function, interworking with the main text.
2 The American scholar Stith Thompson, in his 1946 book The Folktale, provided a definition of the motif: “A motif is the smallest element in a tale having a power to persist in tradition. ... It is evident that by far the largest number of traditional tale types are composed of these single motifs.”(Thompson, 1946)