What is „translation‟?

From the above discussion, it can be seen that the interpretative function of translation arises not from the properties of individual elements, but from the interaction between the Translated and Translatant. Hence, I would propose here that the features of translation I summarized in section 2 are the realization of the autopoietic interaction between the Translated and Translatant. In the following paragraphs, I will try to

72

examine what these features of translation means (listed in Table 2.1) from the viewpoint of the autopoietic mechanism of translation.

Table 2.1 Features of translation and translative semiosis

Features of translation Features of translative semiosis 1) the source text can potentially induce form a directional relationship.

First, in terms of the ‗similarity‘ between the Translated and Translatant, I think it is the indispensable result of the self-reproduction of the translated. What is similar between them is the organization rather than the structure. Thus, the corresponding relationship between the organizations of the signs (Translated, Translatant) is much more important than whether their structure is identical or not. This may provide a potential answer to the question of ‗what is a translation‘. For some researchers (Pym, 2010; Blum-Kulka, 1986/2004), the criteria to distinguish translation from non-translation lies in textual features such as lexical simplification, explicitation, adaptation or equalizing of usages in the target text. However, if there is no similarity in the organizations between Texts, even if they do possess these textual features, the translational relationship cannot be established. Therefore, I suggest that the fundamental criterion for translation is this similarity in organization. Interestingly, it is also this kind of similarity in organization brought by self-production that restricts the potential ‗reaction‘ range of the Translatant in the translative semiosis, because in

73

general semiosis any structurally plastic unities (i.e. unities that do not necessarily have similar organization as the original unity) can potentially interact with the original unity without losing its identity.

Secondly, the difference between the Translated and the Translatant results from the mutual perturbation between them. Once these two entities start to interact with each other, according to Maturana, there certainly will be some compensatory changes induced by the perturbation to avoid the disintegration of the autopoiesis. Two kinds of compensatory changes may occur under perturbation:

(1) ‗conservative changes‘: happens only when the structures between the components changes. This means that the deformation caused by the perturbation would not influence the whole autopoietic space.

(2) ‗innovative changes‘: happens when components themselves change. This means ‗a displacement of the system in the autopoietic space because its components changed‘ (Maturana and Varela, 1980:99).

This change implies that ‗total equivalence‘ in translation is a myth. In the autopoietic mechanism of translation, whenever there is translation taking place, there is interaction. Whenever there is interaction, there is change, and this involves difference in structures. Furthermore, the notions of conservative change and innovative change also suggest that the degree of the ‗difference‘ between the Translated and Translatant would lie in the degree of perturbation and cannot be

‗predicted‘ or ‗regulated‘ before perturbations taking place.

Thirdly, the mutual informing relation between the Translated and the Translatant is actually established by the consensual domain. I contend that it is the interlocking interactions in the consensual domain that brings in the recursive interpretative semiosis in both language and translation. And consensual domain can also form ‗a metadomain of consensual distinctions‘ (Maturana and Varela, 1980: 121) that enables

74

translative semiosis to be established on the autopoietic language mechanism.

Fourthly, the directionality that exists between the Translated and the Translatant also results from the self-reproduction process. What I would like to stress here is that the directionality is only an indication of the chronic order of the generation of Translatants in a consensual domain. This does not suggest a hierarchy of one -way imposition (i.e. source to target; translated to translatant). As illustrated in the previous paragraphs, once the Translatant is formed, it can maintain a parallel interaction with the Translated. I take this point as a crucial explanation of the relation between the source and target texts in translation studies. In the development of translation practice and theory, the source text, as the ‗origin‘ of the translation, has been highly valued. Therefore, several theorists (Chamberlain, 1988; Flowtow, 1997) currently see source text as the wielding of power, and thus try to derogate the idea that source text can generate the target text. In fact, the ‗source sign‘, as a sign in the interpretative system of language, is only a temporal semiotic space that realizes the sign, which depends on further generation of interpretants to ‗define‘ its relative position in the language system. Therefore, I agree that the notion of ‗source text‘ is uncertain. However, for the translative semiosis to take place, there must be some kind of sign/signs that induces Translatant through the Mind. Hence, saying that there is no ‗source text‘ can only be a metaphor for the uncertainty of the ‗original‘. This questioning, nevertheless, can never eliminate the directionality existing in the self-reproduction of the Translated to the Translatant. This directionality, on the other hand, does not suggest superiority of the Translated, as it takes the interaction with the Translatant to form a consensual domain for the translative semiosis. Hence, it should be clearly stated that the directionality between Translated and Translatant in the autopoietic mechanism is based on the biological phenomenon (i.e. an autopoietic production), whereas the power relationship between source text and the target text is

75

resulting from social conventions. And we should be careful not to deny this fundamental biological mechanism of translation production when challenging the social status of source texts.

From the discussions above, it can be seen that the translation mechanism is a kind of autopoietic mechanism based on the autopoiesis of language. What distinguishes it is the self-reproductive nature in translation, which also explains the unique features of translation and translative semiosis. By looking into the interactions between translated and translatant in translation, it is evident that the translation mechanism is not based on ‗code semiosis' (i.e. code functions as the ‗conventional link‘ between two signs. See Barbieri, (2010). That is, translation is not some kind of one-to-one or fixed correspondence between two sign systems. Rather, it is an autopoiesis that not only ‗transfers‘ but also ‗reconstructs‘ the original sign from one sign system to another by interpretive semiosis, which takes place in the process of structural coupling and in consensual domains.

And the autopoietic nature of translation mechanism may also provide some clue to the prosyntactic translation of biotranslation (i.e. translation between systems without syntax) in Kull and Torop‘s discussion. I suggest that the concept of

‗organization‘ in the autopoiesis of translation might explain why translation can take place in prosyntactic system in biotranslation. Take Kull and Torop‘s illustration (2003) of the translation between Parus cristatus and Parus montanus for example.

Even if there is no syntax (a mechanism that can generate unlimited recursive signs) involved in Parus montanus’s recognition and translation of the alarm call of Parus cristatus, the organization of the sounds that constitutes the alarm call (i.e. the relationship between individual sound in the alarm call) could still serve as the potential basis for translation, and thus make prosyntactic translation possible. In this sense, it is possible that both syntactic translation and prosyntactic translation are

76

based on the organization of the Translated. The difference between them falls on the recursion of signs. While syntax can generate unlimited number of recursive signs, prosyntactic systems can only generate a limited number of signs in the system.

However, I must stress that this argument is mainly based on my current understanding of the autopoiesis of translation and may still need further exploration.

Nevertheless, the ‗organization‘ of signs in the autopoiesis of translation, which could be seen as the potential element underpinning both syntactic and prosyntactic translation, may reflect the strong biological root of translation.

5. Conclusion

In this chapter, I try to approach translation from the periphery to the core. By seeing translation as an autopoietic mechanism, the interaction between translated and translatant is no longer the airy ‗dialog‘ that is difficult to describe, as in Petrilli‘s model. Rather, interaction is the most fundamental link in translative semiosis, which influences the most important features of translation: whether the Translated is translatable or untranslatable; whether the ever-generating interpretative semiosis would take place; and why there is similarity and difference in translation. This implies that the autopoietic translation mechanism is a bottom-up approach.

Nevertheless, this bottom-up tendency does not mean that translative semiosis can only take place within the bottom level of the semiotic systems. In fact, by consensuality, translative semiosis forms a rigid system that is built up by layers and layers of consensual domains, which allows perturbations (i.e. interactions) between systems to take place at different levels, both high and low. By this autopoietic self-productive mechanism, I would suggest that translation is certainly not the equivalent exchanges between systems but a more fundamental mechanism in human communication: if language is the recursive generating of new components in the

77

consensual domain as in Maturana, then translation would be the recursive generation of new inter-connections among semiotics systems.

78

79

Chapter Three

From Languaging to Semiosphere: Translation as a Bridging Mechanism

1. Introduction

In the discussions of Popovič‘s and Maturana‘s models in Chapter 1 and 2, translation production can be construed as construction: the construction of inter-textual relation between two realities and that of observer ‘s description. With an eye to these two models, translation production depends on the text processor ‘s (i.e. translator ‘s) active construction/modeling of both the texts and realities that s/he observes. This transformative view of translation in the previous two chapters is mostly addressed from the perspective of individual text processor ‘s modeling in the process of text ontogeny. However, as is implied in Popovič‘s model, when a translator produces a metatext by his modeling of prototext, s/he is also producing interdiscursive and intertextual relations between two realities. That is to say, the production of metatext or translatant³⁹ is not merely an individual linguistic act. Rather, it always implies inter-systemic interactions, which brings us to a new domain of inquiry about translation production: how does a translator ‘s (as a text processor and as an observer) individual linguistic act (i.e. producing metatext/translatant) contribute to inter-systemic interaction in translation? Also, where does the dynamics of inter-systemic interaction come from? How does it enable translation to function as a semiotic mechanism, namely, a mechanism that creates semiosis on the systemic level?

39 Both the terms ‗metatext‘ and ‗translatant‘ refer to the product of translation. The difference between the two terms lies in the different aspects of textual relations that are emphasized. While ‗metatext‘

places greater importance on the modeling of prototext, ‗translatant‘ focuses on the interpretative function of the translation product.

80

To answer the above questions leads us to a reconsideration of the nature of

‗texts‘ in translation. As the observer of prototext and realities between systems, translators translate by formulating ‗descriptions‘ about what they observe. These

‗descriptions‘ can both be seen as the product of individual linguistic acts and also that of inter-systemic interactions. By examining this duality of text (i.e. as both individual and systemic text production), we may gain more clues about how inter-systemic interactions take place in translation production. The dynamics of translation as individual linguistic acts has been addressed in Chapter 2. In order to answer the questions raised above, I will try to situate the model in Chapter 2 into Lotman‘s theorization of semiosphere, a model that concerns how individual texts can be put ‗into contact with the general cultural processes of creation and reception‘

(Torop, 2009:xxvii). That is to say, what I intend to unveil in the present chapter is how the Maturanian ‗languaging‘ may contribute to the Lotmanian ‗semiosphere‘ in translation production.

The seemingly different methodological stances of Maturana and Lotman—one from biology and the other from culture—would not impede such an attempt as the main architect of their discussion of language (i.e. the bases of ‗text‘ production) falls on one similar issue: how language is constructed through the recursive interactions between elements of the system (i.e. ‗linguistic acts‘ or ‗texts‘) rather than on the static sum total of these elements. And it is based on their viewpoints of language that we may be able to find the convergence in these two models which answers the question raised above. Before introducing their respective theorizations about language, I will first give a brief outline of their overall stances to meaning generation (i.e. semiosis).

As mentioned in Chapter 2, in the 1970s, Maturana‘s early interest of research lies in the relationship and the interaction between the observer and the unity. Although

81

the term ‗autopoiesis‘ is not used in Maturana (1970), he has defined cognition as ‗the actual acting or behaving‘ (ibid.: 8) of a living system. This argument builds up the ground for his later discussion on the autopoietic nature of living systems. By seeing cognition as the realization of acts or behaviors that enable the unity to interact dynamically with other unities without losing its own organization, cognition is thus no longer the monopoly of human beings but a feature of all living systems. Based on the formula of ‗cognition equals interaction, interaction equals living‘, Varela, Maturana and Uribe (1974) specify the six criteria of autopoiesis, which overturns the vitalistic tradition in biology. Whether a unity can be considered a living system does not lie in its reproduction potentials, but in its autopoietic nature to realize ‗the autopoiesis of the old‘ unity (ibid.: 189). Later in Maturana (1975), Maturana (1978), Maturana and Varela (1980), Maturana (1985) and in Maturana (1988), autopoiesis is further developed to describe a homeostatic system organized ‗as a network of processes of production‘ (Maturana and Varela, 1980: 79), which allows a unity/system to change its structure while interacting with other systems without losing its organization. Maturana‘s discussion on autopoiesis has revealed a constructivist stance of meaning generation (i.e. forming description/behavior to the outer world is part of the realization of autopoiesis), which is also shown in his discussion on language that will be specified in section 2.

On the other hand, Lotman develops his consideration of cultural semiotics (i.e.

the ‗creation and reception‘ processes in culture systems, as Torop suggests above) on the distinction between primary and secondary modeling systems as well on the notion of semiosphere. Lotman (1977) re-interprets Zaliznjak, Ivanov and Toporov‘s

‗primary and secondary modeling systems⁴⁰‘ as the ‗act of exchange‘ between the

40 In their formulation, primary and secondary modeling systems exist in a gradation, which is

‗determined by the degree of abstraction of the sign system S from the totality of object W‘ (Zaliznjak, Ivanov and Toporov, 1977:47). Lotman‘s reinterpretation turns the gradational difference into a

82

individual and the collective, which brings in the central issues in most of his later publications: tension and dynamics in structural systems. In Lotman and Uspensky (1978), language is discussed as two different levels. On the primary level, language is the isolated realization of each ‗specific communicative function‘; whereas on the secondary level, language is the collective abstraction—it becomes the ‗presumption of communicability‘ (ibid.: 229). With this distinction, Lotman further develops a holistic viewpoint of semiosis. In Lotman (2005), he borrows the notion of

‗biosphere‘ from Vernadsky and then proposes a new approach to semiotics, i.e. the semiosphere, to replenish what has been lacking in Saussure‘s and Peirce‘s semiology/semiotics. Lotman argues that both Saussure‘s and Peirce‘s considerations of signs fall on the act of individual sign exchange, which is then inferred as the model of natural language or the model of signs. For Lotman, the prerequisite for semiosis does not lie in ‗summarizing the separate semiotic acts‘ but in a semiotic universe which ‗makes the specific signatory act real‘ (Lotman, 2005: 208). By changing the target from individual sign exchange to the dynamics/tension between sign systems, Lotman introduces to us a holistic viewpoint of signification, which he later formulates as ‗semiosphere‘. For Lotman, signification comes from semiosphere as it is:

the semiotic space necessary for the existence and functioning of languages, not the sum total of different languages; in a sense the semiosphere has a prior existence and is in constant interaction with languages. (Lotman, 1990:123)

Semiosphere, in this regard, should not be taken as the static territorial demarcation of language repertoire/products. Rather, it is both the diachronical and the synchronical development of languages. It is the formed and the forming of the ‗semiotic space‘ of

hierarchical one.

83

cultures, which consists of not only the semiotic experiences but also the semiotic acts in a certain culture.

From the above, I would say both Maturana and Lotman take a constructivist approach to semiosis, even though the target of their arguments are different (i.e. one on the development of single unity and the other on the development of semiosphere). Their underlying focuses are similar. For both of them, semiosis comes from the dynamics derived from structural interactions, rather than the combination of signs. In addition to how interactions result in text production (i.e.

creating ‗description‘ or ‗information‘), they both express their concern for the identity and the development of the unities/systems involved in communication (i.e.

how unities/systems can interact and hence develop without losing their own identities in interactions). It is the constructive approach they both take that provides us the ground to reach for their shared viewpoints in language in the following two sections, which will be the basis for the rejoinder of the two models in section 4.