Clause complexes as the basis for construing the narrative point of view in translation context

This article briefl y discusses the role of expansions for construing the characters’ identities and personality traits in Oscar Wilde’s novel The Picture of Dorian Gray and its translation, made by João do Rio, and adaptations, made by Clarice Lispector and Claudia Lopes, into Brazilian Portuguese. The discursive confi guration or narrative axis that unveils the narrative point of view is the central linguistic realization of the discussion. The paper also problematizes the concept of omission and rewriting in the literary translation context, asserting that these kinds of (re)textualizations may create negative cultural impacts on the target-text reader. The discussion is informed by the principles of systemic-functional grammar and its contributions to literary translation, especially the concepts of projection and clause complex.


Aim of the paper 1
In this article, I aim to discuss the importance of clause complexes as the basis for construing the characters' identities and consequently their personality traits in prose texts, specifi cally novels.For this brief discussion, I take the novel The Picture of Dorian Gray as an exemplar of this stylistic phenomenon to compare it to possible (re) textualizations in the Brazilian Portuguese language.One of these (re) textualizations is the full translation made by João do Rio (1919); the other two are the adaptations of the novel made by Clarice Lispector (1974) and Cláudia Lopes (1997).In order to pursue the specifi c aim of this paper, the discussion is informed by some of the categories of transitivity and clause complexes of the hallidayan systemic functional grammar (Halliday & Matthiessen 2014).As to translation per se, this article is informed by the concepts of rewriting (Munday 2008) and omissions (Munday 2012), phenomena that are usually at work when text manipulation is at the front scene.In the following, I start by presenting the object of discussion, its boundaries and contentions.After that, I describe the corpora and then bring some examples of the phenomenon under discussion.In the end, I offer some brief remarks on the issue arisen here and open future possibilities for encouraging further discussion on the theme addressed in this article.Halliday & Matthiessen (2014:428-9) explain that "in narrative text, the fl ow of events is construed as a series of episodes.Each episode is typically developed step by step as sequences of fi gures 2 […] that are linked by means of temporal relators".The authors go on informing that clause complexes play an important role in the unfolding of narratives, because a text based on a functionally oriented systemic grammar does a great deal of work in the construction of semantic fi gures by means of clause complex structures (Halliday & Matthiessen 2014:363ff).As an example of what I am trying to say, let us analyse briefl y an extract from chapter one of the novel The Picture of Dorian Gray:

Setting the scene
Extract 1: From the corner of the divan of Persian saddle-bags on which he was lying, smoking, as was his custom, innumerable cigarettes, Lord Henry Wotton could just catch the gleam of the honey-sweet and honey-coloured blossoms of a laburnum… In extract 1, we have one major fi gure formed by the participant ("Lord Henry Wotton"), what he was doing ("could just catch the gleam of the honey-sweet and honey-coloured blossoms of a laburnum…") and where he was doing that ("From the corner of the divan of Persian saddle-bags").This fi gure construes the core of the narrative 2. Experientially, the word fi gure means the way the fl ow of events construes (expresses and simultaneously creates) registerial changes within a text.According to Halliday & Hasan (1985), texts, so to say, are the "materialization" of contexts and, although texts should be considered unifi ed wholes, their cohesive parts work as blocks of meaning making information.For the text to unfold, these blocks of information are supposed to be united by cohesive devices which consequently set a range of fi gures in terms of who is participating, by doing (acting) this or that, and in what circumstances (see also Halliday & Matthiessen 2014:ch. 5).description, as Halliday & Matthiessen (2014) suggested; likewise, clause boundaries, which place clause complexes 3 within the whole fi gure, show that the main participant, Lord Henry, is not just observing something ("blossoms"), but also behaving in a very peculiar manner ("lying" [on the divan]) as well as doing something else behaviourally ("smoking, as was his custom").
From a translational perspective, this very scene becomes more complicated still, regarding that "translation is [supposedly] not a passive refl ection of the original; it is a creative act of reconstructing the meanings of the original as meanings in the 'target'" (Matthiessen 2001:64).I have added the modal adjunct "supposedly" to Matthiessen's conclusion because not all translations (translators indeed) seem to be concerned with the principle of context into text and vice-versa, as a continuum in the fl ow of semantic manifestations of registerial variables.Matthiessen has likely proposed this as a bold hypothesis for multiple semiotic representations within translations, since his aim was to analyse how pictorial representations were linguistically textualized.His concern was with systemic parallelism, given that, according to his contribution to translation theory, translations should be interpreted as textual manifestations of systemic inter-relationships; in other words, translations should be informed fi rstly by context and secondly by linguistic choices appropriate to the target culture, without losing sight of the overlapping phenomenon of source and target texts' cultures.
As an example, let us see how the same scene was translated by João do Rio into Brazilian Portuguese: Extract 2: De um canto do divã entre almofadas persas, onde habitualmente se estirava, fumando inúmeros cigarros, lorde Henry Wotton percebia perfeitamente o brilho das doces fl ores cor de mel...In terms of clause complex, there are actually two changes in the translation that reinterpret the scene presented in the original text: the expansion (hypotactic enhancement) "as was his custom" rendered as the circumstance of extent (frequency) "habitualmente" (habitually; 3.For an ecology of clause combining, see Matthiessen (2002).often); and the adverb "just", functioning as a linguistic choice for limiting the perceptive ("catch") action of the character, rendered as the adjunct "perfeitamente" (perfectly).These changes, far from being considered just slight modifi cations in the translation, depict a quite different linguistic picture when compared to the original, for they not only modify the semantic construction of the text (e.g. while the adverb "just" limits the character's perception, the adjunct "perfeitamente" amplifi es the character's cognitive capability), but rather construes a different picture in terms of language (e.g. the expansion "as was his custom" enhances the semantic contours of the scenario, impinging on it an idea of continuity to the action carried by Lord Henry, whereas the circumstance "habitualmente" just expresses how often he does something).That is, the expansion adds more important information in terms of custom, whilst the circumstance only informs frequency in time, suppressing additional details that clearly display the behaviours assumed by the characters in the original text.
When we turn the discussion to the complex phenomenon of adaptation, often characterized by linguistic omissions and rewritings (see, for instance, Rodrigues-Junior & Barbara 2013), one aspect that changes considerably is the omission of clause complexes, a potential linguistic resource to support the semantic constructions of the narrative axes.That is, the fl ow of events, as described in the original text, is somehow missed, probably given to the need for more restricted information in terms of plot development.
Furthermore, there is one very important point that must be considered: according to Bastin (1998:8), "the controversy surrounding the supposed opposition between adaptation and translation is often fuelled by ideological issues".Although it is not my interest to pursue this issue in the discussion held in this paper, ideology is undoubtedly a motivation for choosing linguistic structures in any text, be it fi ctional, journalistic, scientifi c, etc.In the case of adaptations, omissions and rewritings are an instance of this phenomenon, for these linguistic manoeuvres inform the translator's communicative intentions according to the readership and the target language context (see Bastin 1998:87-8).
Bearing this discussion in mind, from a translational perspective, a complex question, given its answer (if there is any) is far from simple, One hypothesis that this discussion suggests is that translators re-evaluate the uses of clause complexes in order to simplify the linguistic constructions of characters' identities and behaviours, a kind of linguistic economy that informs the style of this specifi c mechanism of text rewriting 4 .If we take the example above, we will see that most the verbal groups, which shape the characters' behaviour, are widely simplifi ed in Lispector's adaptation, whereas in Lopes' text this part was completely omitted.
Unquestionably, lay readers expect a kind of "fi delity" between original and translation, as if a linear, one-to-one equivalence could actually be possible.Equivalence has been considered by researchers on translation a controversial concept, since it is arguably that either linear correspondence between cultures is not possible, on the one hand, or that cultural manifestations are not easily realised in texts, on the other hand (see Kenny 1998 andYallop 2001 for discussions on equivalence).In this respect, the idea of texts as both artefact (culturally valued) and specimen (systemically formed), as discussed by Halliday & Matthiessen (2014:3), obviously grounded on the work of Halliday & Hasan (1985), offers a signifi cant contribution to how translations are regarded as they should be in the light of how originals are culturally valued as they positively are in their (source) contexts.Even having in mind this very important and intricate aspect, translators should 4. Of course, this hypothesis does not consider the linguistic manoeuvre of complete omission, one strategy often utilized by Lopes.In the case of this paper, the aim is to provoke some discussion on how clause complexes are taken as a reliable linguistic space for making visible the translators' decision making.
consider that what is valued in a source culture, in terms of textual artefact, may not be valued likely in the same way by a target culture readership (e.g.Munday 2008Munday , 2012)).
Stemming from Matthiessen's (2001) contribution to translation studies, especially his discussion on systemic inter-relationships between original and translation, in what follows I explore how narrative point of view works as a linguistic basis for typifying characters (cf.Halliday 2002 5 ; Montgomery 1993; Simpson 1993; Rodrigues-Junior  & Barbara 2013).This discussion will give a brief idea of what I am exploring by demonstrating that clause complexes serve a twofold primary purpose: fi rstly, they keep the fl ow of events in a quantum of narrative fi gures, as discussed by Halliday & Matthiessen (2014:428ff); secondly, they typify characters as agents in the fl ow of events as a narrative text unfolds.In this second purpose, narrative point of view plays an important role in describing as well as functionally 6 placing the characters in the story.This characterization is claimed to be one fundamental basis for identifying culture (text as artefact) in text.The two purposes are intermingled and the starting point for conducting an analysis in this comparative and literary terrain is to look at clauses within what is taken here as the primary linguistic basis for unveiling the characters' behaviours and identities: the narrative axis.Narrative axis is confi gured by the verbal process and its projected clause, which may be also surrounded by a multitude of clauses, most expansions and adjuncts, whose function is to give additional information on the characters.Narrative axes are, so to speak, the powerhouse of narrators' points of view.
In what follows, I briefl y explain how the data were organized and the importance of The Picture of Dorian Gray as a textual artefact in both cultures (section 3).Secondly, I select three extracts from the original and its complete translation and adaptations, in order to exemplify how Oscar Wilde's literary masterpiece was linguistically transposed from the English-speaking cultures to the Brazilian 5. Text originally published in 1971.6.The idea that lies behind the adverb "functionally" is this: "By a functional theory of language I mean one which attempts to explain linguistic structure, and linguistic phenomena, by reference to the notion that language plays a certain part in our lives, that it is required to serve certain universal types of demand."(Halliday 2002:89) Portuguese-speaking context within two registerial possibilities: a complete translation intended to reach a major readership and two distinct adaptations aiming at a more specialized readership: the juvenile (section 4).Finally, I discuss the main topics that came up in the analysis and offer the reader some possibilities for conducting further research in this fi eld (section 4).

Corpora description and its organization 7
Oscar Wilde's The Picture of Dorian Gray, written throughout his intense and rich literary life, became famous for its 'fi ctitious' context and for revealing his homosexual tendencies at the closing period of the Victorian age in England and Ireland (cf.Pires 2005;Purchase 2006; Schiffer 2010).As expressed in Mighall (2003: ix), Characteristically, [the novel] was a succès de scandale.His novel provoked, at least in Britain, an outraged response from many reviewers, providing a foretaste of the treatment he would receive fi ve years later when what some believed they had found represented in Dorian Gray (outlawed passions and 'unspeakable' acts) were revealed to be part of its author's life.Indeed, Wilde's novel, or at least the more 'candid' fi rst version, was used by opposing counsel in the fi rst two of his three trials in an attempt to prove that he was guilty of 'a certain tendency' believed to be represented in Dorian Gray.
Distinguishing features of the novel are Wilde's linguistic choices, represented mostly by clause complexes and circumstances that helped forming the narrative axis.These linguistic features were deemed a kind of semiotic representation that made room for a moral critique of an age of restrictions and prejudices that pervaded most of the social thinking of the time.
In 1919, aiming to represent the exquisite beauty of the novel, João do Rio translated into Brazilian Portuguese The Picture of Dorian Gray.His intention was to make visible to the reader a literary masterpiece of the British belle époque, full of subtle nuances of dandyism and extravagant lifestyles (cf.Figari 2007).João do Rio was also 7. Part of this description appears at Rodrigues-Júnior & Barbara (2013).responsible for imprinting in Rio de Janeiro's society, between 1915 and 1921, the practice of dandyism, a particular aesthetic lifestyle common to aristocratic British and French people.Dandyism recognizes Oscar Wilde as its perfect icon, and João do Rio explicitly incorporated all the great elegance of this way of life.João do Rio's translation has been acclaimed the novel's best rendering ever made in the Brazilian Portuguese language.
Clarice Lispector's (1974) and Claudia Lopes' (1997) adaptations were chosen as the other versions because both can be viewed as authorial texts, given its free re-instantiation, even though both had as point of departure the source text.It means that both Lispector and Lopes created new interpretive possibilities for the characters, giving rise to distinctive fi ctional aspects.Their writing styles made new scenes of interpretation visible, especially because the adapted versions had the juvenile as their main readers, a kind of readership leading for the need of reconfi gurations as well as free versions of the original.Consequently, the adapted versions recreated the characters by omitting important details and behaviours belonging to them.

Examples
This brief discussion focuses the investigation on the linguistic features of narrative point of view of the original novel comparatively with the same extracts of the translation and adaptations.Given that I am also problematizing deliberate omissions in terms of full paragraphs in the adapted versions, a common aspect concerning adaptations, I have looked for clauses that to a certain extent appeared in the data as a whole, or at least in most of them.The selection then generated 92 narrative axes of a sample, comprised of projecting processes, projected clauses and surrounding linguistic elements, such as clause complexes, adjuncts and circumstances.After having analysed the 92 axes, for the purposes of this discussion, three narratives axes were chosen as a representative sample of the general linguistic patterns of the clause complexity as a means for construing the characters' identities in the original, compared with the same patterns in João do Rio's translation and Lispector's and Lopes' adaptations.The clauses were displayed in parallel format, one below the other, in the following order: the fi rst clause is an excerpt from the original novel; the second clause is an excerpt from the complete translated version; and the last two clauses are the excerpts from Lispector's and Lopes' adaptations, respectively.The ordering format followed the year in which each book was published, identifi ed by the initials of the original's author and of the translators.So, OW was used for Oscar Wilde, which refers to the original, published in 1891; JR was adopted for representing João do Rio, referring to the complete translation, published in 1919; and LIS and LOP were assigned as the initials for Clarice Lispector and Claudia Lopes, respectively, indicating the adapted versions, published in 1974 and 1997.The three examples below demonstrate the linguistic patterns present in most of the narrative axes examined in the sample, which are constituted by the Sayer, the verbal process, the projected clause -or the content of the saying -, and the expansions surrounding the linguistic confi guration of the narrative axes.The expansion of enhancement, formed by the non-fi nite verb tossing his head back…, was rendered in João do Rio's translation, but omitted from both Lispector's and Lopes' adaptations.In example 1, the narrator construes the character not only as someone who is sure of his decision of not sending the portrait for an exposition, but as someone who is anxious and nervous, demonstrating these feelings in a peculiar behaviour expressed by the expansion.The complete translation has tried to capture the idea (sacudindo a cabeça…), but the adaptations have reconfi gured the semiotic construing of this part by omitting the information.In example 2, the expansion of enhancement, also constituted by the non-fi nite verb, laying his hand… typifi es the scene of two men who are intimate.Both the complete translation and the adaptations have omitted this part, which raises the following question: in the case of this analysis, to what extent can we consider adaptation as deviation (cf.Bastin 1998), if the complete translated version has also omitted a detail, realized by the expansion that informed the whole picture?According to Halliday & Matthiessen (2014:442), "a typical clause complex is a mixture of paratactic and hypotactic sequences, either of which may be nested inside the other".In example 3, there is a clause complex with nesting in the expansion shaking … and strolling.In this case, there is internal bracketing, i.e. a sub-complex group initiated by the projecting clause said Basil Hallward.That is, the projecting axis (said Basil Hallward) is indeed a simple paratactic clause (1α) that has left room for the narrator to construe the character's effeminate behaviour, realized by the two hypotactic expansions in nesting position ([β'] shaking his hand off … and [β"] strolling…).This whole scene has been omitted from the complete translation and from the adaptations, which undoubtedly construes a quite different picture of the character, if compared to the original, with considerable impact, in terms of character construing, on the Brazilian reader.

Final remarks
The brief discussion held in this article has tentatively suggested that clause complexes play a significant role in construing the characters' identities from a narrative point of view.Simpson (1993:10) explains that Psychological point of view refers to the ways in which narrative events are mediated through the consciousness of the 'teller' of the story.It will encompass the means by which a fi ctional world is slanted in a particular way or the means by which narrator's construct, in linguistic terms, their own view of the story they tell.(My emphasis) In the examples analysed, it has become clear that the narrative axis itself is not the only linguistic resource used by the narrator to, on the one hand, keep the unfolding of the plot and, on the other hand, build the characters' identities; likewise, clause complexes are important linguistic mechanisms to characterize the plot and their fi ctional personas.From a translational perspective, these linguistic resources become more complicated still, because clause complexes are one part of the functional systemic grammar that gives rise to semantic constructions of texts, especially narratives, which could not be possible without taking original and translation contexts into account.That is the point that deserves careful attention when adaptation is brought into the discussion, given that the translator is supposed to both rewrite the novel and give to it precise information, as it is textualized in the original text, for the sake of completeness.
Further analyses are supposed to mix quantitative and qualitative investigation.For this purpose, I give some suggestions to explore more deeply, in literary texts and their translations, what I have discussed so far: (i) to select all the narrative axes of the corpora, (ii) organize them in types, (iii) count them, separating each type to display them in groups, and (iv) analyse the types, in order to map the linguistic resources of clause complexes used by the narrator to build the characters' identities in the original novel in comparison with the complete translation and the adaptations.By doing this, the intention is to construct a map of occurrences of clause complexes usage in the comparative literary corpus, aiming to show (i) how clause complexes helped to inform the effeminate behaviour of the characters; (ii) how the translation and the adaptations have (re)textualized these linguistic resources; and (iii) how clause complexes function as spaces for manoeuvre when translators are to decide which types of clause to choose when translating the narrative axes.