翻譯：一個符號系統的探討

全文

(1)國立臺灣師範大學翻譯研究所博士論文 Graduate Institute of Translation and Interpretation National Taiwan Normal University Doctoral Thesis. 翻譯：一個符號系統的探討 Translation Studies as Sign Systems Studies 吳碩禹 Shuoyu Charlotte Wu 指導教授：張漢良博士 Advisor: Han-liang Chang, Ph.D. 中華民國 103 年 1 月 January 2014.

(2)

(3) Acknowledgements It has been in my dreams, in my imagination, in daily exchanges with my colleagues and family and also in the discussions with my advisor. As a PhD student, I have pictured all kinds of scenarios in which I put the final full stop in the thesis and declare the completion of the task. Yet, when the day finally came, none of the pre-devised scenarios could depict what it really felt like. The full stop signals not merely the completion of the doctoral thesis but also the end to a wonderful stage in my life in which I came to experience the best and worst of life, to learn the wisdom from others and foolishness of mine, to challenge myself with belief and incredulity, and to embrace the hope and despair in life. For the past four and half years, the thesis has been a developing organism that initiates incessant dialogues with me. But such dialogues would have never been provoked without the marvelous people that inspired me. First of all, I would like to express my gratitude to my advisor, Prof. Han-liang Chang. On a path that is less traveled by, I am fortunate enough to be deeply inspired, encouraged and guided by Prof. Chang. From the very beginning of this doctoral project, I have benefited from his generosity in imparting his wisdom. Without his patient and insightful instruction, the completion of the thesis would have been impossible. And without being inspired by his painstaking scholarship and his devotion to academy, I would have never discovered the joy in research work. Next, I shall express the deepest appreciation to the thesis defense committee members—Prof. Shang-Kuan Chang, Dr. Ken-Fang Lee, Prof. Te-Hsing Shan and Prof. Hsiu-Chih Tsai—for their insightful comments and advice on the improvement of the thesis. My heartfelt thanks are due to the teachers in GITI who have also inspired me in their lectures in many different ways—Ken-Fang Lee, Posen Liao, Sher-Shiueh Li and Daniel Hu. I also owe a debt of gratitude to my colleagues in GITI, Oscar Lin, Hsinhsin Tu, Hungshu Chen, Ami Liu, Huifeng Cheng, Yumin Chang, Yaling Chen, Chien-lin Wu, Yung-Chieh Chiang, Guey-Lan Liao and Shutsai Chang who have been great friends and interlocutors. From the study groups to the water cooler talks, I gained much inspiration from the exchanges with these brilliant colleagues. Last but not least, I would like to thank my parents (especially my mother, Rebecca), my brother Andre and my husband Richard. Without their full support, I could never survive the hardship and accomplish the challenging task..

(4)

(5) 摘要本研究旨在以符號學作為後設語言，探究並描述翻譯文本生成之相關問題。本研究分為正文與附錄兩部分。正文共分為四章，附錄則有文本分析兩篇。首章由亞里斯多德《詩學》之模擬與創造兩概念出發，探討翻譯詩學的性質，確立翻譯文本生成乃基於後設溝通。第二章由生物符號學的自體生成機制探討翻譯如何藉由結構重組促成個別文本的生成以及意義的層層衍生。第三章探討翻譯如何由個別語言活動之連結，轉而引發系統互動。第四章則由洛德曼的對話結構出發，以翻譯為自我與他者對話之角度，探討系統互動對互動中的個別系統有何意義。附錄兩個文本分析，一為伍光建所譯之〈會揭露秘密的心臟〉，此文譯自愛倫坡之短篇小說，文本分析以格雷馬斯之符號矩陣以及敘事模型探討譯者的操弄如何促成翻譯的結構重組。第二篇文本分析以話語模態探討白芝在漢學家以及譯家兩個身分之間的翻譯作為有何不同，以及其作為與翻譯對話結構之關聯。. 關鍵字：翻譯文本生成、詩學、後設溝通、互文、自體生成、結構重組、爆炸性翻譯、對話結構、語義圈、形塑系統、符號矩陣、伍光建、白芝.

(6)

(7) Abstract. With the aim to provide a priori description of translation production, the thesis investigates translation production in four different aspects: translation as metacommunication, translation as autopoietic restructuring, translation as explosion and then the relation between translation and dialogism. Diverse as the focuses are, they can be joined by a main thread of thought—text production. The thesis starts from an epistemological inquiry into the nature of translation production, which is followed by a biosemiotic investigation of how text ontogeny takes place in translative semiosis. The discussion of text ontogeny then leads to the consideration of translation production on systemic level, which intends to answer the following two questions: How may individual linguistic acts contribute to inter-systemic interactions in translation and what does the interaction ‗mean‘ to the systems involved translation interactions? The theorizations in the four chapters are followed by two texts analyses in the Appendix, which aim to provide operational procedures to investigate translation production. With the models adopted from biosemiotics, cultural semiotics and social semiotics, the thesis is intended not only to re-examine ‗translation‘ as an object of research, but also to provide a meta-language to describe translation under the framework of general semiotics.. Keywords: translation production, poetics, metacommunication, intertextuality, autopoiesis, autopoietic restructuring, translation as explosion, dialogism, semiosphere, modeling system, semiotic square, Woo Kwang-Kien, Cyril Birch.

(8)

(9) Table of Contents Foreword. 1. Chapter One. Intertextuality in the „Poetics‟ of translation 1. Introduction 2. Translation and Poetics: Metacommunication 3. Intertextuality: From the Kristevian viewpoint 4. Intertextuality in translation: Reference or Interpretation? 5. Conclusion. 9 9 10 21 31 41. Chapter Two. Autopoiesis and Interpretive Semiosis: Translation as a Biological Phenomenon 1. Introduction 2. What is translation all about? 3. Translative semiosis vs. iconicity 4. Autopoiesis and translative semiosis 4.1 Autopoiesis and language 4.2 The autopoietic mechanism of translation 4.3 What is ‗translation‘? 5. Conclusion. 43 43 45 48 54 55 60 71 76. Chapter Three. From Languaging to Semiosphere: Translation as a 79 Bridging Mechanism 1. Introduction 79 2. Maturana on language 83 3. Lotman on language and translation 86 4. From languaging to semiosphere 90 4.1 Text and translative semiosis 90 4.2 Autopoietic translation in the semiosphere 93 5. Conclusion 98. Chapter Four. Dialogue and Translation: A Lotmanian Perspective 1. Introduction 2. Benjamin and Derrida on translation dynamics: equivalence and difference as preconditions 3. Translation: a Lotmanian perspective of multiplicity in unity. 101 101 103 108.

(10)

(11) Appendix One. Appendix Two. References. 4. ‗Multiplicity in Unity‘ vs. Dialogism 5. Conclusion. 116 121. Text analysis: The Greimassian model and Its Application 1. Introduction 2. The Greimassian model 2.1 The discoursive level 2.2 The semio-narrative level 2.3 The deep level: the semiotic square. 123. 3. Autopoietic restructuring in Woo‘s translation. 123 124 125 128 131 134. 4. Conclusion ‗The tell-tale heart‘ by Poe Woo‘s translation of ‗The tell-tale heart‘ Woo‘s abridgement of Poe‘s ‗The tell-tale heart‘. 144 147 151 155. Modality and application: A case study of Cyril Birch‟s translation 1. Introduction 2. Reflections of a working translator: Birch‘s meta-description of ‗translation‘ 3. Description of translation 4. Translation modality in actual translation 5. Conclusion 考城隍. 159. 牡丹亭第三齣：訓女. 181. 牡丹亭第七齣：閨塾. 183. 159 159 165 173 177 179. 185.

(12)

(13) Foreword Peeter Torop, one of the Moscow-Tartu semioticians, in the article ‗Methodological remarks on the study of translation and translating‘ (2007:347) states:. Disciplinary identity depends on how the disciplinary research object is conceived of. If translation studies is a discipline studying translation and translating, it is natural that it can define its identity at the intersection between translation and translating. This intersection is translation process. The need for an epistemological description of translation process arises from the need to make the process describable or to find a possibility of its optimum description.. For Torop, the epistemological description of translation process may ‗contribute to the inner dialogue within translation studies‘ and allow it to be ‗a genuine multidiscipline and to process at the same time towards a disciplinary unity‘ (p.361). Torop‘s concern is not without reason. In his review, the ‗description‘ (i.e. theorization) of translation are mainly a posteriori, which ‗offers a set of characteristics to choose from in describing a particular translation‘ (ibid.:351). With an eye to translation studies as a unity, he urges for an a priori description of translation. The above statement is the best description of what this thesis attempts to do—to formulate an epistemological description of translation production by semiotic approaches. With the aim to describe translation production1 , a process that includes both ‗translating‘ and ‗translation‘, the thesis is divided into four chapters, accompanied by two text analyses in the appendix. The four chapters are: ‗Intertextuality in the ―poetics‖ of translation‘, ‗Autopoiesis and interpretative semiosis: Translation as a biological In the above quotation, Torop sees ‗translation process‘ as the ‗intersection‘ of translation and translating, which describes both the translation product and process. However, as the word ‗process‘ may have a strong implication of being opposed to ‗product‘, I choose to use ‗translation production‘ in the thesis to refer to the realization of translation, namely, the process that goes from translating to translation. When translation production is seen as the realization of translation, the word ‗translation‘ in this thesis may refer to both ‗translation‘ and ‗translating‘ in Torop‘s sense. 1 1.

(14) phenomenon‘, ‗From languaging to semiosphere: Translation as a bridging mechanism‘ and ‗Dialogue and translation: A Lotmanian perspective‘. The two text analyses in the appendix are ‗Text analysis: The Greimassian model and its application‘ and ‗Modality and application: A case study of Cyril Birch‘s translation‘. The analyses aim to illustrate semiotic approaches to text analysis in transaltion. They may, but are not specifically devised to, cover issues discussed in the previous four chapters. Such a connection between the chapters and the appendix is possible as all the discussions made in this thesis are related to translation production. Despite the various issues and theoretical models discussed and adopted in the thesis, the four chapters and two text analyses can be joined by a main thread of thought—text production in translation. With the help from semiotic models, the thesis attempts to provide a priori descriptions on the different aspects of translation production by the discussion of issues such as intertextuality, similarity and difference, translatability and untranslatability, and dialogism in translation. Chapter One focuses on the nature of translation studies as a systemic knowledge, namely, a ‗poetics‘ of translation. By revisiting the Aristotelian notions of ‗poiesis‘ and ‗mimesis‘, I propose that the intertextual and interdiscursive nature of mimesis is the key to translation production. On this basis, it is argued that translation production is the production of intertextuality between systems, which leads to the second major issue dealt in Chapter One—the models of intertextuality from Anton Popovič and from Julia Kristeva. The comparison of the two models indicates that Popovič‘s model is the pertinent one to our investigation of translation production. In addition, the comparison also helps to clarify a current methodological confusion in translation studies that arises from the misuse of the two models. In Chapter Two, the biosemiotic model from Humberto Maturana is used to examine the nature of translation semiosis. The chapter first focuses on the distinction 2.

(15) between translative semiosis and the general semiosis of natural language. This is made by a review of features of translation. Then a critique to Susan Petrilli‘s theorization of translation production is presented to help clarify why translation semiosis does not hinge on iconicity between signs as Petrilli suggests but on the consensuality built in translation interactions. Maturana‘s theorizations on ‗autopoiesis‘ provides a solid ground for us to examine the autopoietic restructuring in translation; namely, how texts may be generated, develop and evolve in translative semiosis. The discussion of individual observer ‘s construction of translation act in Chapter Two leads to the disscussion of how individual translation acts of translation may contribute to intersystemic interaction. This brings me to examine how autopoietic restructuring in translation formulates composite systems of consensuality that may further enable ‗translation as explosion‘. From languaging to semiosphere, translation links the individual and the collective by creating both textuality and intertextuality (in Popovič‘s sense) between systems. Thusly construed, translation is not confined to interlingual, intralingual or intersemiotic transactions. Instead, any interactions (i.e. the forming of limited translatability) between two heterogeneous structures can be seen as translation. In this regard, translative semiosis may be the most fundamental mechanism for semiosis. Chapter Four explores the relation between translation and dialogue. Following Chapter Three, dialogue is regarded as the interaction between heterogeneous structures. Hence, the relation between translation and dialogue lies in what Lotman terms as ‗multiplicity in unity‘. The chapter first goes back to Walter Benjamin‘s and Jacque Derrida‘s formulations of translation dynamics that arises from the intersection of heterogeneous structures. My review suggests that their formulations of translation dynamics are not based on the viewpoints of the systems involved in interactions. That is to say, they address translation interactions from the viewpoint of an outsider. 3.

(16) The brings me to investigate ‗multiplicity in unity‘ from the perspective of the insider, which intends to answer what ‗multiplicity in unity‘ means to ‗Self‘ (i.e. the ‗I‘ system in translation interaction). To answer the question, I reformulate Lotman‘s model of dialogism, which demonstrates how translation production is a process that shapes the unknown into the known to the receiving system. With regard to the two text analyses, in Appendix One ‗Text analysis: The Greimassian model and its application‘ I use the semiotician Algirdas Julien Greimas‘s model to examine Woo Kwang-Kien‘s 伍光建 translation of Edgar Allan Poe‘s short story ‗The tell-tale heart‘. The examples in Appendix One illustrate Woo‘s text production. They can also be seen as illustration of the autopoietic restructuring of translation. In Appendix Two, ‗Modality and application: A case study of Cyril Birch‘s translation‘, the sinologist Cyril Birch‘s translation of Mudanting 牡丹亭 and his commentary article on the translation of the Chinese literature ‗Reflections of a working translator‘ are examined to see how ‗translation‘ is invested with value content in Birch‘s translation practice and translation critique. By investigating the modalities in Birch‘s texts, we find that they are underpinned by his two different identities (i.e. a sinologist and a translator). As a researcher, Birch values whether the target readers may understand the source text. As a translator, Birch endeavors to demonstrate his understanding of the source texts. The contrast between the two identities, when examined under the Greimassian model, demonstrates two different kinds of dialogues in translation: one with Self and one with Other, which may also be seen as an example of the dialogism mentioned in Chapter Four. The brief outline above not only demonstrates the major topics in the thesis but also reveals a potential methodological flaw in this thesis. As mentioned, several semiotic models have been adopted, including Popovič‘s formulation of intertextual relations, Maturana‘s biosemiotic model, Lotman‘s theorizations of semiosphere and 4.

(17) dialogism and then Greimas‘s model of text analysis. To describe translation production with these diverse models may bring in qualitative change to the models as their theoretical stances may not be on the same level. However, these models are not joined together by idiosyncratic use. Different and diverse as they are, there are still some shared perspectives, which we can build our discussions on. To begin with, both Popovič and Lotman are influenced by a strong Slavic tradition of literature and philology, whose focus of research lies in systemic relations. With regard to Maturana, even though his research focus is based on the philosophy of biology, the considerations he has given to observers‘ active constructions in language provide a niche for us to use it as a mechanism for text production and to incorporate it with Popovič‘s and Lotman‘s formulations. As to the fourth model, the Greimassian model, as both the Greimassian and Lotmanian models have been influenced by Hjelmslev, it can be said that, under the framework of general semiotics, the two models share a similar stance on signification (i.e. signification arises from the signifying whole rather than the adding up of signs). In addition to the stance on signification, another reason that makes the Greimassian model the pertinent model of text analysis is its consideration of the observer. As Greimas indicates, the subject of speech act can be identified as the speaking subject because the speaking subject is always transcoded into the speech act by being endowed with a ‗presence‘ in the instance of discourse. In this sense, the three-tiered Greimassian model provides the operational procedures to examine how observers (in this case, translators) take position in the text (i.e. schematize the texts and realities they observe into the texts). The practical and rigid procedures in the Greimassian model are of great importance to the present thesis. On one hand, it compensates Lotman‘s model, whose lack of operational procedure is often under criticism. On the other hand, it also strengthens our use of Popovič‘s and Maturana‘s models as our concern for observer ‘s active construction may be 5.

(18) examined by operational procedures in texts. The above description is intended to provide not only an overview of the adopted models but also an explanation of how and why these models can be incorporated to form a meta-language to describe translation. The diversity of the models adopted also suggests the various layers of perspectives embedded in the thesis. For clarification, Richard Macksey‘s (1974:xxv) diagram of critical coordinates (please refer to the following chart) is used as a heuristic tool to illustrate the positions I take in the four chapters. Contemplating on the nature of literature criticism, Macksey orients different kinds of literary criticisms in the coordinates form by four poles: Work, World, Author and Audience.. The issues in Chapter One and Chapter Two may be placed in the third and second quadrants in the Cartesian coordinates. The discussions of the ‗poetics‘ of translation and translation as metacommunication are both related to the MAKING of translation. Therefore, it can be inferred that the stance I take lies on the intersection between the Author and Work. In Chapter Two, Maturana‘s biosemiotic model indicates the concern for the relation between the Author and the World (i.e. how an observer 6.

(19) constructs his description of the external world). This allows me to orient the discussions in Chapter Two in the second quadrant in which OBSERVING is the underlying structure for the discussions. For the third chapter, as its focus falls on how individual linguistic acts may contribute to inter-systemic interactions, my position moves back and forth between Work and World as the interactions between works would result in inter-systemic intersections (i.e. interactions between the worlds). In the fourth chapter, the discussion of dialogism from the viewpoint of the receiving system is built on the crisscrossed relation between the Audience (i.e. receiving system as addressee) and the Worlds. Hence, Chapter Four can be positioned in the first quadrant where recognizing the distinction between the ‗unknown‘ and the ‗known‘ is the basis for translation dialogism. From making, observing, interaction to recognition, the thesis attempts to look at translation production from four different levels: text production in translation (Chapter One), individual text production (Chapter Two), text production between systems (Chapter Three) and then to text production in the eyes of the ‗I‘-system in interaction (Chapter Four). The explorations and investigations summarized above are crude attempts. There is no intention of producing any grand theories about translation. It, instead, is hoped that the use of semiotic models may provoke some, if any, reflections on the a priori descriptions of translation studies, which may then allow self-descriptions to assign ‗clear boundaries and a considerably higher degree of unification‘ (Lotman, 2009: 173) in the dynamic semiotic space constituted by translation studies.. 7.

(20) 8.

(21) Chapter One Intertextuality in the „Poetics‟ of Translation‟. 1. Introduction The declaration of methodological independence from comparative literature and applied linguistics defines translation studies as ‗interdiscipline‘ (Snell-Hornby, Pochhacker, and Kaindl, 1992). Yet along with the interdisciplinary characteristics of translation studies, the term ‗translation‘ has expanded to cover a wide range of phenomena. ‗Translation‘ may be defined as rewriting, adaptation, representation, transformation or simply, ‗translation as such‘. The expansion, however, poses a new challenge to translation scholars: how to define the territory of translation? When we say translation is representation, which is the more fundamental mechanism, translation or representation? How about translation and rewriting? And then, how about translation and transformation? The blurred boundary between translation and all the ‗definitions‘ suggests a methodological bewilderment: an unspecified causality. When we claim translation is representation, are we saying translation contributes to representation or the other way around? The vagueness suggests that what is often construed as the ‗definitions‘ of translation are more of metaphorical than ‗scientific‘ descriptions of translation2. The vague descriptions reflect that translation as an object in question roams in territories that are subjectively-defined by different researchers, which, I suggest, may result from the lack of consideration of the nature of translation studies as a systemic knowledge. A revisit to the nature of translation as a systemic knowledge may shed light to our inquiries into translation, which hence becomes one of the aims of the ‗Scientific‘ is used here in Hjelmslev‘s sense, which is a description that obeys the ‗empirical principle‘. Namely, such a description ‗shall be free of contradiction (self-consistent), exhaustive and as simple as possible‘ (Hjelmslev, 1963:11). 9 2.

(22) present chapter. To investigate the nature of translation studies, I trace back to Aristotle‘s Poetics. When translation is observed under the context of the Aristotelian corpus, translation production can be construed as a process that goes from possibility to realization, building up an inter-discursive relation between texts. The inter-discursiveness points to us the core of translation production—intertextuality. In 1.2, two models of intertextuality, one from Kristeva and one from Popovič, are compared in order to determine the pertinent model for the thesis. In 1.3, Basil Hatim‘s and Lawrence Venuti‘s misreadings of intertextuality are illustrated to demonstrate how the misuse of Kristeva‘s and Popovi č ‘s models may cause methodological confusion in translation studies. This responds to the very first argument at the beginning of this paragraph: the uncertainty of the nature of translation studies as a systemic knowledge is of consequential influence on the research of translation.. 2. Translation and poetics: Metacommunication The tower of Babel, the story often interpreted as the origin of the multiplicity of tongues and hence the origin of translation (Derrida, 1992 [1985]; Steiner, 1992 [1973]; Niranjana, 1992 3 ), signals men‘s endless quest: to create in order to In Jacque Derrida‘s Des Tours de Babel, ‗Babel‘ as a proper name conjures the ‗origin of myth, the metaphor of metaphor, the narrative of narrative, the translation of translation‘(Derrida,1992 [1985]: 218) as such a name expresses the incompleteness of language, which arises from the multiplicity of tongues. Derrida‘s questioning of how to translate ‗Babel‘ signifies a distrust of the common language or common tongue that the Biblical story describes. If the potential of common language is ‗totalizing‘, then the fall of Babel, in a Derridean sense, paves the path to the confusion of tongues, which denies any act aiming to the completion or totalizing in language. Translation exists in-between not natural languages systems, but the totalizing of language and multiplicity of tongues. The Biblical story is further read by George Steiner from a different perspective. The fall of Babel, as Steiner sees it, signals the multiplicity of tongues that keeps mankind ‗vital and creative‘ by their ‗being scattered among tongues‘ (1992:244). ‗Totalizing‘ and ‗multiplicity‘ in Derrida‘s argument is turned into ‗the uncertainties of relation between formal and substantive universality‘ in Steiner (ibid.:111). As the relation between the substantive universal and the ‗particularized realities of natural language‘ are dynamic, ‗language‘ is also dynamic and transformative; translation is hence the ‗acid test‘ that aims not to the exchange of modes of meaning but to understanding. In a different vein, based on the notions of istoria and episteme (in the Derridean sense), Tejaswini Niranjana further depicts the fall of Babel not merely as the origin of the multiplicity of languages but as the heterogeneity of representations. 10 3.

(23) understand and to be understood. The task, after the fall of Babel, i.e. the signal of the fall of common language/tongue/lip4, is passed onto translation. Interestingly, not only the origin of translation is depicted in the form of an allegory, before or even after the establishment of translation studies 5 , translation is also frequently described by metaphors6, such as ‗[t]he translator considers thought content a prisoner which he transplants into his own language with the prerogative of a conqueror (St. Jerome, cited in Friedrich, 1992:11 [1965])‘; translation as ‗the transplantation of written works‘ (Schleiermacher, 1992; 37 [1813]); translation is like ruminating the rice for others (Kumārajīva, cited from Chen, 2000: 17); translation as ‗the trial of the foreign‘ (Berman, 2000:285 [1985]); translating as ‗penetration‘ (Steiner, 1992), to name just a few. The use of metaphor, in an Aristotelian sense, is to bridge between the unknown and the known in order to make the unknown into something knowable (Aristotle,. Translation in this sense is the re-writing of the pre-established and favored representation systems (i.e. history), which is ‗speculative, provisional and interventionist‘ (Niranjana, 1992: 173). From the above, it can be seen that the Biblical story is more than an allegory used to describe the origin of translation (i.e. the fall of common language). Rather, it embodies a dialogic context that inspires scholars to contemplate the nature of translation through the ‗dialogues‘ with or responses to the previously established ‗translation of translation‘ (Hermans, 2002). ‗Translation‘ in this sense is materialized through the discussions/reflections on the story. 4 In the Biblical story, the Tower of Babel is intended to be built after the whole world speaks one single language. In the Septuagint Old Testament, the story of Babel begins with ‗Καὶ ἦν πᾶζα ἡ γῆ σεῖλορ ἕν, καὶ θωνὴ μία πᾶζιν‘ (And all the earth was one lip, and there was one language to all). The one ‗lip‘ enables mankind to understand each other and thus build the Tower of Babel. And it is the confusion of the tongue that causes the fall of Babel as well as the communication barriers between mankind: δεῦηε καὶ καηαβάνηερ ζςγσέωμεν ἐκεῖ αὐηῶν ηὴν γλῶζζαν, ἵνα μὴ ἀκούζωζιν ἕκαζηορ ηὴν θωνὴν ηοῦ πληζίον. καὶ διέζπειπεν αὐηοὺρ κύπιορ ἐκεῖθεν ἐπὶ ππόζωπον πάζηρ ηῆρ γῆρ, καὶ ἐπαύζανηο οἰκοδομοῦνηερ ηὴν πόλιν καὶ ηὸν πύπγον. διὰ ηοῦηο ἐκλήθη ηὸ ὄνομα αὐηῆρ Σύγσςζιρ, ὅηι ἐκεῖ ζςνέσεεν κύπιορ ηὰ σείλη πάζηρ ηῆρ γῆρ, καὶ ἐκεῖθεν διέζπειπεν αὐηοὺρ κύπιορ ὁ θεὸρ ἐπὶ ππόζωπον πάζηρ ηῆρ γῆρ (Come, and having gone down let us there confound their tongue, that they may not understand each the voice of his neighbor. And the Lord scattered them thence over the face of all the earth, and they left off building the city and the tower. On this account its name was called Confusion, because there the Lord confounded the languages of all the earth, and thence the Lord scattered them upon the face of all the earth). It can be inferred that in the Septuagint, the rise and fall of Babel both hinge on σειλόρ/σείλη (one lip, one mouth) and the θωνὴ (one voice) that it creates. In this sense, σειλόρ stands for a lip/mouth not merely as a physical organ but as a mechanism for communication shared by all mankind on earth; namely, a common language. The fall of Babel in this sense also stands for the fall of the common language. 5 James Holmes‘s paper ‗The Name and Nature of Translation Studies‘ (2004 [1988]) is usually considered as the manifesto and the landmark of the establishment of translation studies. 6 Chamberlain‘s article ‗Gender and the metaphorics of translation‘ (2004) has demonstrated a thorough discussion about metaphors of translation, and points out how translation is shaped by the hidden models in the metaphors. 11.

(24) Rhetoric, 1410b10). The diversity of the metaphors for translation shows how slippery the notion is when being defined. Yet, diverse as the metaphors may be, a common feature of translation can still be traced in these metaphors: like metaphor in an Aristotelian sense, translation also bridges the gap, the gap between two textual entities. The metaphors above do imply this common feature, but they do not answer how the gap is bridged in translation. To answer such a question brings us to the systemic investigation of translation production, namely, the ―poetics‖ of translation. As a matter of fact, when it comes to translation, poetics is the last thing that should be overlooked. Poetics, construed in the Aristotelian sense of ―productive episteme‖ (επιζηήμη ποιητικῆς), hinges on two important notions, poiesis and mimesis, as it is stated at the beginning of Poetics: πεπὶ ποιηηικῆρ αὐηῆρ ηε καὶ ηῶν εἰδῶν αὐηῆρ, ἥν ηινα δύναμιν ἕκαζηον ἔσει, καὶ πῶρ δεῖ ζςνίζηαζθαι ηοὺρ μύθοςρ εἰ μέλλει καλῶρ ἕξειν ἡ ποίηζιρ,ἔηι δὲ ἐκ πόζων καὶ ποί ων ἐζηὶ μοπίων, ὁμοίωρ δὲ καὶ πεπὶ ηῶν ἄλλων ὅζα ηῆρ αὐηῆρ ἐζηι μεθόδος, λέγωμεν ἀπξάμενοι καηὰ θύζινππῶηον ἀπὸ ππώηων. ἐποποιία δὴ καὶ ἡ ηῆρ ηπαγῳδίαρ ποίηζιρ ἔηι δὲ κωμῳδία καὶ ἡ διθςπαμβοποιηηικὴ καὶ ηῆρ αὐληηικῆρ ἡ πλείζηη καὶ κιθαπιζηικῆρ ηςγσάνοςζιν οὖζαι μιμήζειρ ηὸ ζύνολον: διαθέποςζι δὲ ἀλλήλων ηπιζίν, ἢ γὰπ ηῷ ἐν ἑηέποιρ μιμεῖζθαι ἢ ηῷ ἕηεπα ἢ ηῷ ἑηέπωρ καὶ μὴ ηὸναὐηὸν ηπόπον7 (1447a 8-24. Emphasis mine) I do not intend to present a thorough discussion of poiesis and mimesis because the two notions can be adequately dealt with only in book-length studies, which is certainly beyond the scope of the present thesis. What I would like to do is to point. 7. The Greek citation and the French translation below are both based on Roselyne Dupont-Roc and Jean Lallot‘s (1980) translation and annotation on Poetics. The Greek text they used for translation is mostly based on Parisinus 1741 (‗Notre texte est fondé sur le ms. A, dont les leçons primitives (sauf lapsus évidents‘, Dupont-Roc and Lallot, ibid. :24.). The Dupont-Roc and Lallot translation is an early entry to the book series of Poétique launched by Editions du Seuil under the joint editorship of Tzvetan Todorov, Gérard Genette and Hélène Cixous. This new translation is noted, among other things, for its reorientation of Aristotle‘s text to contemporary structuralist poetics and semiotics. 12.

(25) out how and in which sense can we assume there is a poetics of translation. To specify what poetics is in an Aristotelian sense, I shall make a detour to specify the relation between poiesis and mimesis. What follows are the French and Chinese translations of the above Greek paragraphs. Roselyne Dupont-Roc and Jean Lallot‘s (1980) translation in French and Han-liang Chang‘s Chinese translation are adopted as my understanding of poiesis and mimesis is majorly based on their viewpoints. Nous allons traiter de l'art poétique en lui-même, de ses espèces, considérées chacune dans sa finalité propre, de la façon dont il faut composer les histoires si l'on veut que la poésie soit réussie, en outre du nombre et de la nature des parties qui la constituent, et également de toutes les autres questions qui relèvent de la même recherche. Suivant l'ordre naturel, traitons én premier ce qui est premier. L'épopée et la poésie tragique, comme aussi la comédie, l‘art du dithyrambe, et, pour la plus grande partie, celui de la flûte et de la cithare ont tous ceci de commun qu'ils sont des représentation. Mais il y a entre eux des différences de trois sortes : ou bien ils représentent par des moyens autres, ou bien ils représentent des objets autres, ou bien ils représentent autrement, c'est-à-dire selon des modes qui ne sont pas les mêmes. (Dupont-Roc and Lallot, 1980: 33, Emphasis mine) 我們要討論的是『創作』本身（peri poietikes autes），它的各個「分類」（eidon autes）和各類的特質，使得作品成功的情節結構的正確方法，組成成分的數目和性質，以及其他有關的事物。依照自然的順序，我們先談首要的事情，史詩、悲劇、喜劇、酒神祭歌、管笛樂、弦樂，這一切都是「摹擬」的變形。區分它們的有三方面：媒介、對象和模式。有時人們根據技術規則或能力，透過狀寫事物的方式，創造出各種「摹擬」，媒介包括色彩和形象；有時媒介是聲音；同樣的，就我們剛才所指出的藝術總體說來，「摹擬」的媒介有：韻律、語言、音樂，或者單獨使用，或者結合為一體。」(Chang, 2012 :6. Emphasis mine). From the above, it can be seen that the notion ‗poiesis‘ is related to the two Greek words : ποιηηικῆρ (poietikes) and ποίηζιρ (poiesis), with the first one referred to as ‗the art of making (‗l'art poétique‘)‘or ‗making or producing (of work) (‗創作‘)‘ and the second one as ‗the making‘ or the ‗production‘. Mimesis, on the other hand, is related to the noun μιμήζειρ (mimeseis) and the verb μιμεῖζθαι (mimeisthai), with the 13.

(26) former read as ‗representation (représentation)‘ or ‗imitation or representation (‗摹擬‘)‘and the latter as ‗to represent (‗représenter‘)‘ or ‗to imitate or represent (‗摹擬‘)‘ though the transliterated Greek word ‗mimesis‘ had entered the English language in the sixteenth-century and has been preferred by many classicists to its English equivalents. From the opening statement of Poetics cited above, poiesis and mimesis can be regarded as the two sides of the same coin, to rehearse a Saussurian commonplace of the linguistic sign. The epic, tragedy, comedy, music with flute and lyre etc., these all come into being because of mimesis. Meanwhile, they also stand for the various species of poiesis. That is to say, through the mechanism of mimesis poiesis as production may take place. Hence, mimesis enables the realization of poiesis. This viewpoint is better elaborated in both Dupont-Roc and Lallot (1980) and Chang (2012) in which the relation between poiesis and mimesis is talked about with reference to two other Aristotelian notions, dynamis and potentiality8. Identifying the relation between poiesis and mimesis as that of production brings us one step further to ask ‗what does this production aim at?‘ As without mimesis poetic production would never take place, then it is reasonable to assume that the aims of poetic production may lie in the aims of mimesis. To illustrate what mimesis aims at, Chang (2012), following Gerald F. Else (1957) and others, proposes to recontextualize ‗mimesis‘ not merely in the immediate context of Poetics, but also in the larger context of Aristotelian corpus in which the notion of mimesis is referred to. In Meteorology, for instance, Aristotle observes ‗[n]ow broiling and boiling are artificial processes, but the same general kind of thing, as we In Dupont-Roc and Lallot (1980:143), they see poetics as an oriented activity (‗une activité orientée‘) that each kind of production is guided by its own dynamis (they translated as ‗finalité‘). Dynamis enables production to pass from potentiality to realization (what Dupont-Roc and Lallot term as ‗acte/energeia‘). That is to say, poetics is the production of the realization of potentiality. In a similar yet more elaborated way, Chang (2012:6) points out that mimesis is the mechanism that induces the dynamis in poiesis and thus embodies poiesis as a signification process that goes from potentiality and probability to actuality. 14 8.

(27) said, is found in nature too. The affections produced are similar though they lack a name; for art imitates nature9‘ (381b6). In Physics, mimesis is depicted as 1) ‗if on the other hand art imitates nature, and it is the part of the same discipline to know the form and the matter up to a point…: if this is so, it would be the part of natural science also to know nature in both its senses‘ (194a21); 2) ‗generally art in some cases completes what nature cannot bring to a finish, and in others imitates nature‘ (199a15). Here the Aristotelian notion of ‗art‘ (ηέσνη) refers to artificial making, as opposed to but paralleled by the operations of Nature (θύζιρ). Hence, mimesis can be seen as standing in-between Nature as given and the world of artificial production, which induces the dynamis and completes production (poiesis) through the imitation of the natural world. Based on the function that mimesis plays in-between the world of poiesis (i.e. harnessing production) and the world of reality (i.e. imitating or completing the natural world), we can assume that Aristotle‘s Poetics is a systemic knowledge about production that connects the world of artificial making and that of reality. Namely, it is an epistemology of ―production‖. And it is based on the assumption that mimesis and poiesis join to create a textual reality by representing and negotiating a previously existing textual reality that I propose poetics as a domain essential to translation studies. Surely, this kind of mimesis is no longer restricted to the Aristotelian correspondence between Nature (θύζιρ) and art (ηέσνη), but expanded to cover the relationship between texts, as the greater translator of Homer Alexander Pope (1688-1744) aptly summarizes: ―To imitate Homer is to imitate Nature‖ (An Essay on Criticism 1711). This intertextual and interdiscursive nature of mimesis is particularly true to translation. Finally, according to Aristotle‘s own classification of knowledge (Metaphysics,1064a10-14), the third branch of episteme consists of. For the English translations of Aristotle‘s Meteorology and Physics, Jonathan Barnes‘s translations are used. 15 9.

(28) ―productive science‖ (ποιηηικῆρ [―poetics‖ in the broad sense]) which includes two subjects: rhetoric and poetics (in the narrow sense of what is dealt with in his Poetics, such as the ―making‖ of tragedy and the epic). From our historical vantage point, it may not be inappropriate to incorporate Translation Studies into the larger framework of poetics as ―productive science‖ 10. To illustrate what I mean by the poetics of translation, I shall invoke a relatively ignored theoretician of translation studies of the recent past, Anton Popovič (1976), whose formulation of intertexual relationships may provide us with food for thought. According to Popovi č , translation is one kind of meta-communications, which includes ‗all types of processing (manipulation) of the original text, whether it is done by other authors, readers, critics, translators, etc.‘ (Popovi č ,1976:226). In meta-communication, the processor of the text manipulates an object text (i.e. prototext) so as to create a text (i.e. metatext) that becomes ‗a model of the prototext‘ (ibid.:226). The relation between prototext and metatext allows us to define meta-communication in another way: the production of metatext by its modeling on prototext. What meta-communication creates in this regard is text continuity, or to be more. specific,. inter-relation. between. texts.. The. text. continuity. in. meta-communication, however, does not simply arise from the ‗the reflection of the original‘ (Popovič, ibid.:233). As Popovič illustrates, metatext is embodied by two ontological aspects: a textual one and an extra-textual one. The extra-textual one comes from text processor ‗as creator to reality‘ (Popovič, ibid.:233). The production of 10. text. continuity. in. metacommunication. is. hence. the. production. of. According to Aristotle, five kinds of knowledge can be acquired by man: art, scientific knowledge, practical wisdom, philosophical wisdom and intuitive reason (Dolezel, 1990). It can be inferred that Aristotle has his own system of knowledge/epistemes in which different episteme has different functions. In fact, Chang (2013) classifies the Aristotelian corpus into four different kinds of epistemes: Organon, theoretical sciences, practical sciences and productive sciences. Chang categorizes both Rhetoric and Poetics to the category of productive sciences as they both share the principle that Aristotle depicts: ‗For in the case of productive science the principle of production is in the producer and not in the product, and is either an art or some other capacity‘ (Metaphysics, 1064a 11-13). 16.

(29) inter-relations—the inter-relation between prototext, metatext and reality, which is illustrated as follows by Popovič: Reality Author1→Text1 →Receiver1. Reality Author2→Text2 →Receiver2 Popovič‘s scheme (1976:234). In the (meta)text production, the text processor decides the modes of text continuity (e.g. quotation, allusion, translation or plagiarism) by his/her manipulation of the degree of variance and invariance between Text 1 and Text2. By manipulation Popovič refers to two operations of text processors: ‗he conveys information about invariants of the original and, at the same time, discovers in the original further virtual or concealed meanings‘ (Popovič, ibid.:233). Text processor ‘s being creator to reality and his manipulating of textual variance imply a semiotic view of texts. If we borrow Charles Sander Peirce‘s model of sign, the text processor, reality and the text constitute the basic elements for a sign relation: the text processor, upon his encountering with the reality, forms an interpretant of the reality in his mind and then is able to produce a text of the reality. Hence, the text processor serves as the Mind that generates the Interpretant of the Object (reality) and further formulates the Sign (the text). A text in this sense is a unity formed by the observed reality, the text processor‘s observation of the reality and the realization of the observation. Text production is hence considered the transcoding of realities through observation. With texts thus construed, it implies that the text continuity in metacommunication, namely, the inter-relation between prototext, metatext and reality is a special kind of ‗intertextuality‘. The ‗intertextuality‘ here should not be interpreted in Kristeva‘s sense (which will be specified later). Instead, it refers to the inter-textual ‗contacts‘ 17.

(30) that ‗are reflected in texts themselves‘ (Popovič, ibid.:225). As the poetics of translation belongs to the domain of productive epistemology, its object of research would fall on the production of translation. The intertextual-relation of metacommunication above would help us to get a clearer picture on translation production since translation is a type of metacommunication. That is to say, the inter-textual relation in translation is the key to the production and hence the poetics of translation. This seems to bring us back to the age-old question of textual relation in translation. Yet with a new perspective of texts being tripartite unities of texts, text processor and realities, I will redefine translation production (i.e. the production of intertextual relation) through my reformulation of Popovi č ‘s formulation. To describe the inter-textual relation in metacommunication, Popovič (1976) suggests the ratio prototext: metatext reveals metatext‘s modeling on prototext as it shows the variance and invariance between the two texts. The ratio hence indicates the nature of meta-communication acts, be it quotation, parody, imitation, commentary, annotation or translation. Popovič suggests that the ‗principal aspects‘ of intertextual relations, such as ‗semantic, axiological and those associated with the author‘s strategy‘ (Popovič, ibid.:227) can all be examined by the comparison of variance and invariance between texts. Yet, in this fairly reasonable argument, there is one important link missing, which can be spotted when the intertextual relation above is described by a formula:. r (ratio)=metatext/prototext= variance/invariance. The ratio certainly changes when different pairs of metatexts and prototexts are put into the formula. However, even though the value of r fluctuates with the texts, it is valid or meaningful only when one describes the modes of inter-relations or the 18.

(31) inter-relation between one prototext to its various metatexts. For instance, to translate Aristotle‘s Poetics and to write a commentary on Aristotle‘s Poetics form two different sets of ratio value between metatext and prototext, with the ratio for translation being lower (i.e. invariance outweighs variance) and that for commentary being higher (i.e. variance outweighs invariance). However, it would be meaningless to compare the ratio of stylistic or semantic variance between Else‘s translation of Poetics and Eileen Chang‘s translation of The Old Man and the Sea as they are not evaluated on the same basis. Hence, Popovič‘s model is valid only in describing either the categorical difference between different modes of inter-textual relations (e.g. quotation vs. parody) or the lineal tendency of the ‗contacts‘ that a prototext reflects in various metatexts. If we want to use it to describe the production of inter-textual relations, a few changes have to be made. Popovič‘s formulation is not valid in describing the production of intertextual relation as it is based on a static view of text. Texts, in Popovič‘s formulations of shifts, are implied as fixed end products of metacommunication. For instance, it is possible to compare the shifts in Else ‘s and Dupont-Roc and Lallot‘s translation of Poetics only when the prototext is regarded as a fixed entity. That is to say, the prototext Poetics is assumed to be the same to its different readers as well as processors, and so are the metatexts. Text processors as a variable are bracketed in Popovič‘s formulation the object of his observation is the categorical differences between modes of inter-literary relations. For my present purpose in representing the production of intertextual relation, text processors are included into the scheme because it is text processors that enable the realization of text as a unity (of the observed reality, the text processors‘ observation of the reality and the realization of the observation), which inter-textual relation is built upon. When text processor is taken into consideration, the formula above can be rewritten as: 19.

(32) Metatext= r(P proto-text )x prototext. In this formula, prototext and metatext are not static combination of signs, but how text processors process the two texts. In the case of translation production, prototext implies the author‘s manipulation of the reality he observed. P proto-text stands for the ‗perspective of prototext‘, which allows the translator‘s observation and manipulation to be taken into consideration. r is a coefficient function of proto-text. By stating r, the ratio of shift in translation production, as the coefficient function of proto -text suggests r is a context dependent coefficient whose value is dependent on the perspective toward proto-text (Pproto-text ). With the inclusion of text processors‘ perspectives, the formula depicts the dynamic interplay between prototext and metatext and explains how inter-textual relation is built between them (i.e. the realization of the text processor ‘s perspective on prototext). What does this formula tell us about the poetics of translation? If we place this formula in the context of metacommunication production, it shows a unique feature of translation production. In metacommunication, the metatext may at times be seen as a secondary text as it is ‗a meta-sign of a work which is already in existence‘ (Popovič, 1976: 233). In this sense, the inter-textual relation that metacommunication generates seems to be what Dionýz Ďurišin (1974), following in the footsteps of Victor Zhirmunsky, terms as ‗genetic contact‘, which is based on a causal, syntagmatic relation. However, if we take into consideration the dynamic, transformative inter-relation between prototext and metatext, the modeling of metatext on prototext brings in a second layer of relationship into their ‗genetic contacts‘. Metatext‘s modeling suggests a one-way presupposition (i.e. a relation of selection) between the texts (e.g. Eileen Chang‘s translation of Ernest Hemingway‘s The Old Man and the Sea presupposes the source 20.

(33) text, but Hemingway‘s text does not presuppose Chang‘s translation). In this respect, the inter-textual relation is built on the axis of selection. It is between the two kinds of inter-textual relations (i.e. text continuity and the selection that comes from modeling) that lies the most unique feature of translation/metacommunication production: the production of intertextuality being based on the imposition of the axis of selection (paradigmatic relation) on that of combination (syntagmatic relation). And, to quote Jakobson (1987:71) again, when ‗the principle of equivalence from the axis of selection‘ is projected into ‗the axis of combination‘, the poetic function of a message arises. In this sense, each realization of inter-textual relation in translation is always poetic in nature. From the above, the discussion of intertextuality in translation helps us to define the territory of our investigation of the poetics of translation within the range of intertextuality production. But before going into any further exploratio n, it is necessary to make another detour to Kristeva‘s notion of ‗intertextuality‘ in order to specify the distinction between ‗intertextuality‘ in Popovič‘s sense and that in Kristeva‘s sense.. 3. Intertextuality: from the Kristevian viewpoint When ‗intertextuality‘ is mentioned, it often conjures the poststructuralist concept made known by Kristeva. Kristeva‘s model is very different from Popovič‘s focus on literary processes. Different as they are, the shared term ‗intertextuality‘ may lead to a methodological confusion in translation studies (see section 4). A clarification is thus indispensable. How prototexts and metatexts formulate intertextual relations has been clearly specified in section 2. In order to demonstrate the dissimilitude in Kristeva‘s model, I shall focus on the following aspects: 1) what does ‗text‘ stand for, 2) how do texts formulate ‗intertextuality‘ and 3) what is the ‗intertextuality‘ in Kristeva‘s 21.

(34) model? Kristeva‘s notion of intertextuality is inspired by and adopted from Mikhail Bakhtin‘s ‗intersubjectivity‘. Sharing a similar enthusiasm on the novel, both Bakhtin and Kristeva endeavor to re-define the genre not in terms of themes but of the structural features of novels, which lead them to their reflections on language use. It is their formulations on language use, rather than those on the poetics of novels, that will be discussed in the following section with regard to the pertinence to our discussion of intertextuality. In ‗The Problem of Speech Genres‘ (Bakhtin, 1986), with reference to the different ‗spheres of human activity‘ (ibid.:65), Bakhtin makes distinctions between speech genres. Such a distinction results from two kinds of utterances: primary and secondary utterances 11 . Bakhtin makes efforts on clarifying the Saussurian dichotomy between langue and parole, stating that they are not in a relation between system and realization. Rather, to Bakhtin, they belong to the different spheres of human language. While langue specifies the systemic structure of human language, parole/speech falls in the category of communication. And in communication, it is speech, according to Bakhtin, that produces the ‗senses‘ beyond linguistic meanings. Speech, in this regard, is not the ‗free combination of forms of language‘ (ibid.:81), but is bound to different speaking subjects: ‗[f]or speech can exist in reality only in the form of concrete utterances of individual speaking people, speech subjects‘ (ibid.:71). It implies that each utterance is reified by a speaking subject and hence the concatenation of utterances is formed by the change of speaking subjects, ‗which creates clear-cut boundaries of the utterance‘ (ibid.:72). This allows Bakhtin not only to designate utterances as the ‗unit of speech communication‘. 11. Bakhtin specifies the difference between primary (simple) and secondary (complex) speech genres as follows. Secondary genres consist of ‗novels, dramas, all kinds of scientific research, major genres of commentary‘ (1986:62). Primary genres, on the other hand, take ‗form in unmediated speech communion‘ and have an ‗immediate relation to actual reality‘ (ibid.). 22.

(35) (ibid.:73), but also to bring forth the following statement: with regard to the concatenation of utterances, each utterance‘s ‗beginning is preceded by the utterances of others, and its end is followed by the responsive utterances of others‘ (ibid.: 71). The statement suggests that the concatenation of utterances always implies a hidden dialogic relation as it takes the presupposition between utterances to form the chain. But this is not the only dialogic relation embedded in utterances. In each individual utterance there has to be a speaking subject and an addressee of the utterance12, which forms another layer of dialogic relation. The above features of utterance indicate the uniqueness of speech communication: two layers of dialogues are always inherent in each utterance, with one between the speaking subject and its addressee and the other between the speaking subject of one utterance to other speaking subjects of other utterances13. Speech in this sense, no longer belongs merely to synchrony. It is the interface between history (i.e. other‘s utterances) and text (as utterance) and also between the ‗inner and outer verbal life 14 ‘ (Volosinov, 1994 [1973]:60 15 ). By re-positioning speech as utterances embodied by speaking subjects and by seeing utterances as layers of dialogues (i.e. between subject and addressee & subject and other subjects), Bakhtin‘s reflection on language use (i.e. intersubjectivity) provides the basis for Kristeva‘s theorization. Following Baktin‘s. interests in utterances, Kristeva‘s. formulation of. To quote Bakhtin: ‗Any utterance always has an addressee, whose responsive understanding the author of the speech work seeks and surpasses‘ (Bakhtin, 1986: 126). 13 Bakhtin describes it as: ‗The utterance is addressed not only to its own object, but also to other‘s speech about it‘ (ibid.:94). 14 Bakhtin sees utterances/texts not as a container of message. Rather he sees it as a production process of dialogues that forges understanding. In addition to the reader/addressee of the utterance that would respond to and try to understand the utterance (which he terms as the ‗second party‘), there is a ‗third party‘ in the dialogic utterance. It is the ‗superaddressee‘ presupposed by the author, ‗whose absolutely just responsive understanding is presumed, either in some metaphysical distance or in distant historical time‘ (Bakhtin, 1986: 126). This superaddressee can be seen as the ‗double‘ identity of the author, whose inner verbal actions is objectified when realized as outer verbal action. 15 Although the authorship of the Bakhtin Circle, especially that between Volos inov and Bakhtin, remains under dispute, I am not concerned with the identities and mis-identities of the Circle‘s members. Following Morris‘s principle of anthologizing the reader of Bakhtin, I also see Volos inov‘s discourse as part of the Bakhtinian discourse. 23 12.

(36) intertextuality is built upon the articulation of words, as in Bakhtin‘s viewpoint, words are ‗abbreviation or representative of the utterance‘ (Baktin, 1986:120) and also because ‗[t]he possibilities and perspectives embedded in the word‘ are ‗essentially infinite‘ (Baktin, ibid.). The focus on words enables Kristeva to move from Bakhtin‘s intersubjective relation that results from dialogues (i.e. between a subject and an addressee & between a subject and other subjects) to the crisscrossed space/relation between texts: ‗each word (text) is an intersection of word (texts) where at least one other word (text) can be read‘ (Kristeva,1986: 37). By the shift from ‗utterance as text‘ to ‗word as text‘, Kristeva diverts from Bakhtin‘s model of speech communication to a topological model of signification. Such a change is brought about by the way she associates the notions of word, text and intertextuality. A word, in Kristeva‘s model, does not have a fixed meaning. A word gets its significance when it is articulated, or in Kristeva‘s words, when it gets its ‗status‘. The so-called ‗status of the word‘ is formed by two coordinate axes: one is the word‘s position in the subject-addressee relation, the other is the word‘s relation to its historical context. The word‘s status is decided by where the synchronic and diachronic axes meet. Hence, a word only signifies when it enters a dialogic relation with both the synchronic and diachronic axes16 . And the process for a ‗word‘ to become ‗a word with status‘ designates a process of signification. This is what allows Kristeva to assert that the status of the word is the ‗minimal structural unit‘ (Kristeva, 1986: 36). Kristeva‘s reinterpretation of the Bakhtinian concept ‗word‘ moves the ‗inter-subject‘ relation in Bakhtin‘s model seamlessly from the foreground to the 16. We can trace this back to Bakhtin. For Bakhtin, a word or even a sentence in a language system ‗belongs to nobody‘ as it ‗has no author‘ (Bakhtin, 1986:84). And a word gets its meaning when it ‗pertains to a particular actual reality and particular real conditions of speech communication‘ (ibid.: 86). Kristeva‘s emphasis on the status of word is an attempt to merge Bakhtin‘s distinction on ‗word‘ and ‗utterance‘ and to replace them with ‗word‘ and ‗status of word‘ (articulation of word). In this case, ‗word‘ is the potentiality and ‗status of word‘ is the realization of this potentiality. This is an indispensable transition when she re-positions the dialogic model on intertextual relations. 24.

(37) background and brings ‗text‘ to the spotlight. The ‗inter-subject‘ relation is tuned down to be the precondition for the word to gain its status. It is inherent in the realization of the word. Hence, the ‗inter-subject‘ relation is always implied. Texts, on the other hand, is the new center of focus. It is a twin concept to ‗word‘ as they are both based on the same mechanism 17. A quick review of Kristeva‘s definition of text shows it. Text, defined as ‗a trans-linguistic apparatus that redistributes the order of language by relating communicative speech, which aims to inform directly, to different kinds of anterior or synchronic utterances‘ (Kristeva, 1981: 36), designates a process of realization. Just as a word becomes articulated when it turns ‗inter-word‘ (i.e. gaining its status/relation to other words), a text is realized when it informs ‗anterior or synchronic utterances‘; that is, when it becomes ‗inter-textual‘. In this sense, intertextuality describes not text-to-text relation, but the ‗articulation‘ of texts. From word, text to intertextuality, Kristeva stretches Bakhtin‘s model not only to illustrate a process of signification, but more importantly, to help such signification escape the reign of language, or the signification of language. This brings us to Kristeva‘s discussion of poetic language and the topological model underlying her formulation. In Kristeva‘s formulation, intertextuality is the foundation of poetic language, with dialogue being their mediating mechanism. As mentioned, text is realized only when it is in a dialogic relation (i.e. intertextual relation) with other texts. In-between two texts, their relation may arise from their affirming, negating, non-affirming or non-negating each other. For Kristeva, both affirming and negating other texts designate the signification of natural language as natural language follows a logic of ‗0-1‘, which means for each word or each text, there is an ultimate meaning endowed 17. In Kristeva‘s formulation, the two concepts are often used interchangeably. It can be seen from the following quotation: ‗each word (text) is an intersection of word (texts) where at least one other word (text) can be read‘ (Kristeva,1986a: 37). 25.

(38) in them. No matter whether one text affirms or negates other texts, the phenomenon implies the acknowledgement of the objective value or meaning pre-assigned to them. It is when the intertextual relation is built to neither affirm nor negate an ultimate objective meaning that they create signification out of ambivalence. And Kristeva terms this ambivalence as poetic language, which follows a logic of ‗0-2‘, a logic that is built upon the transgression of the ‗0-1‘ logic18. But how may the transgression of law take place? According to Kristeva, it hinges on two kinds of operations of signification: symbols and signs. When intertextuality is constituted by a relation of symbol, as Kristeva quotes Peirce, it is governed ‗by virtue of a law, usually an association of general ideas‘ (cited from Kristeva,1986b: 64). That is to say, the symbolized and the symbolizer signify only because they are caused to do so by a law. The relation between the symbolized (the universals) and the symbolizer (the marking) is a restrictive affirmation of the universals. To put it in another way, with a governing law, it does not matter which symbolizer (the marking) is put into the sign relation, as the result would always be the same, namely, pointing to the same symbolized (universals). From here, it can be posited that the relation between the symbolized and the symbolizer is the one based on causality. This causality/law/restriction enables Kristeva to compare a symbolic operation of signification to monologue, which ‗is a belief in the literal sense of the word‘ (Kristeva,1986a: 48). In this sense, Kristeva sees signification based on symbolic relation as the one obeying the ‗0-1‘ logic, the one underlying the signification of natural languages. Signification based on sign relation, on the other hand, designates that of poetic language. A sign relation is referred to in a Saussurean sense, as is alluded to by Kristeva. A sign signifies by its relation to other signs. It ‗evokes a collection of To quote Kristeva: ‗A literary semiotics must be developed on the basis of a poetic logic where the concept of the power of the continuum would embody the 0-2 interval, a continuity where 0 denotes and 1 is implicitly transgressed‘(1986a: 41). 26 18.

(39) associated images and ideas‘ from which their meanings arise and under which lies ‗a principle of transformation: within its field, new structures are forever generated and transformed‘ (Kristeva,1986b: 72). From this perspective, a sign is only articulated when in relation to other signs, which makes it ‗polyvalent and multi-determined‘ (Kristeva,1986a: 36). It is not restricted to the ‗0-1‘ logic or the law governing meanings in natural languages. Instead, it introduces ‗a logic exceeding that of codified discourse‘ (Kristeva,1986a: 36). The exceeding logic allows the signification of signs to challenge ‗God, authority and social law‘, but it does not intend to replace the law with a new one. Rather, it merely transgresses the law so as to negate the universality and certainty by creating polyvalence (i.e. for the two to ‗contradict and relativize each other‘, Kristeva, ibid.:49) and by keeping the text ‗dialogical‘ and ‗rebellious‘ (Kristeva, ibid.). The signification of sign, in this regard, is the production of polyvalence that introduces the ‗double‘ into intertextual relation, which further becomes the minimal unit of poetic language19 . The difference between symbol and sign, from the above argument, lies in their production. Symbol generates monovalence; sign polyvalence. As Kristeva sees monovalecne as the ‗1‘ of the ‗0-1‘ logic (i.e. the God, the law, the censorship), the embodiment of this monovalence is the mono-logic discourse. Polyvalence, on the other hand, is reified as the ambivalent, the double, the dia-logic discourse. Monologue and dialogue, symbol and sign, natural language and poetic language, these point to us not merely how texts signify but also the fundamental mechanism that distinguishes Kristeva‘s model from Popovič‘s, which will be illustrated in the following paragraphs. As mentioned, in the Kristevean model the relation between ‗text‘ and ‗inter-text‘ should be regarded as the potentiality and the realization of texts. Intertextuality hence. 19. Kristeva: ‗[T]he minimal unit of poetic language is at least a double‘ (Kristeva,1986a: 40). 27.

(40) describes texts‘ productivity in signification20. But how do texts enter an intertextual relation? The ‗permutation of texts‘ (1986a: 36) is Kristeva‘s answer. Kristeva suggests that the permutation of text enables ‗several utterances, taken from other texts, [to] intersect and neutralize one another ‘ (Kristeva,1986a: 36). But what is not specified in her response is the underlying mechanism that allows permutation of texts to put intertextuality into realization. To answer this for Kristeva, it takes us to examine the organizing principle of permutation. According to Kristeva, ‗[m]onological discourse corresponds to Jakobson‘s systematic axis of language‘ (Kristeva,1986a: 47). It is not hard to see why Kristeva makes such an assertion. On the systematic axis of language (which should be a Hjelmslevian term21), or, to return to Jakobson, on the paradigmatic axis of language, sign relations are built by the organizing principle of equivalence. It thus is fair to suggest that equivalence is seen as a pre-existing relation (or ‗law‘, to use Kristeva‘s words) before the realization of texts, which explains why Kristeva designates monologue, a discourse obeying the ‗0-1‘ logic, to the systematic axis. Following this vein, we can assume that dialogue, the discourse where ‗two texts meet, contradict and relativize each other ‘ (Kristeva,1986a: 49), falls on the syntagmatic axis whose organizing principle is based on contiguity—as contiguity reifies a new sign relation beyond the ‗0-1‘ logic of natural language between texts. The organizing principle here provides answers to our enquiry above. Permutation of texts is bound to be organized by contiguity as the syntactical relation formed by contiguity enables the intersection and neutralization of utterances. And the organizing principle of permutation of text (and thus of Kristeva: ‗The text is therefore a productivity‘ (Kristeva,1986a: 36). The ‗systematic axis of language‘ comes from Louis Hjelmslev (Hjelmslev, 1963:39). Hjelmslev reformulates the Saussurean ‗in presentia‘ and ‗in absentia‘ relations of language as ‗process‘ and ‗system‘, in which ‗the process can here be called a text, and the system a language‘. However, the most conventional way of naming the relations ‗in presentia‘ and ‗in absentia‘ is Jakobson‘s ‗syntagm‘ and ‗paradigm‘. What Kristeva terms as ‗Jakobson‘s systematic axis of language‘ should be referred to as Jakobson‘s ‗paradigmatic‘ axis or Hjelmslev‘s ‗systemic‘ axis, both of which describe a semiotic relation based on selection. 28 20 21.