Miksi näkökulma muuttuu käännöksessä? Eksplisiittistämisen ja normaalistamisen selitysvoima ja seuraukset

Näkökulma on olennainen osa kerrontaa, sillä se vaikuttaa väistämättä siihen, miten kerrottuun suhtaudumme. Käännöksiin kerronnan näkökulma ei kuitenkaan aina välity sellaisenaan. Väitöskirjassa näkökulman muuttuminen käännöksissä selitetään seurauksena erilaisista tarkentavista ja normaalistavista kielellisistä muutoksista, joita käännöksiin on tehty. Tällaiset muutokset taas ovat käännöksissä niin yleisiä, että niiden on aiemmassa käännöstutkimuksessa ehdotettu luonnehtivan kääntämistä yleensä. Väitöskirjassa tätä oletusta kääntämiselle tyypillisestä tarkentavasta (eksplisiittistävästä) ja normaalistavasta tendenssistä hyödynnetään siis selityksenä näkökulman muuttumiselle käännöksissä. Samalla muutokset kerronnan näkökulmassa havainnollistavat sitä, minkälaisia seurauksia eksplisiittistämisellä ja normaalistamisella voi olla, miten ne voivat tekstin luentaan vaikuttaa. Tutkimuksen aineisto on kaunokirjallinen, ja se on koottu Dostojevskin romaanista Rikos ja rangaistus , sen kuudesta suomennoksesta (1889 2008) ja kahdesta englanninkielisestä käännöksestä (1914, 1992). Kerronnan näkökulma rajataan keskittymällä vapaaseen epäsuoraan esitykseen, kertomakirjallisuudessa yleisesti käytettyyn näkökulmatekniikkaan, jossa limittyvät ja kerrostuvat toisaalta kertomuksen henkilön, toisaalta sen kertojan näkökulmat kuitenkin niin, ettei eri näkökulmien yhtäaikaista läsnäoloa osoiteta tekstissä avoimesti ja suoraan. Käännöksissä toinen näkökulmista usein väistyy tai katoaa, toinen taas nousee hallitsevaksi. Tutkimuksessa tämä selitetään sillä, että kerrostuneet, epäsuorasti osoitetut näkökulmat ilmaistaan tyypillisesti kielellisen aukkoisuuden ja pintatason epäyhtenäisyyden avulla. Kun nämä piirteet kohtaavat kääntämiselle tyypillisen eksplisiittistävän ja normaalistavan tendenssin, yhteentörmäys on odotettavissa. Näkökulman muuttuminen vaikuttaa tekstin luentaan eri tavoin, tapauksesta riippuen kuten vähentämällä lukijan samastumista kertomuksen henkilöön tai muuttamalla hänelle muodostuvaa käsitystä henkilöhahmon luonteesta tai kerrotuista tapahtumista. Vaikka käännösten kielellinen tarkentaminen ja normaalistaminen onkin usein perusteltua ja tarpeen, näkökulman kannalta sen seuraukset tekstin luennalle voivat siis olla epätoivottuja ja siksi oletettavasti tahattomia. Kääntäjille väitöskirja tarjoaakin käsitteellisiä työkaluja, joiden avulla he voivat aiempaa selvemmin tiedostaa toisaalta käännöksiin tekemänsä tarkentavat ja normaalistavat muutokset, toisaalta niiden vaikutuksen kerronnan näkökulmaan.The aim of the present study is twofold. Firstly, it aims to account for shifts in the narrative point of view, repeatedly occurring in translations of fictional prose, with the notions of explicitation and normalization, posited as universals of translation. At the same time, these shifts are treated as a specific context illustrating the consequences that the above-mentioned translation universals may have for the interpretation of translated text. Therefore, with respect to the notions of explicitation and normalization, it is both their explanatory power and their effects that are in focus here: the former is tested and the latter illustrated with the narratological notion of point of view. The narrative point of view, being a notoriously fluid concept, is approached through the notion of free indirect discourse (henceforth FID). FID is a mode of discourse presentation widely used in literary fiction to merge the perspective of the narrator with that of a fictional character ? or, in other words, to evoke another voice or perspective against the background of narration. While representing the consciousness or speech of a fictional character, it is, nevertheless, coloured by the attitude of the narrator, whether ironic, empathetic or neutral. FID is typically used in fiction to create an impression of hearing the character s own voice or following his or her mental processes with less (or less obvious) mediation from the part of the narrator ? or to create irony, empathy or intentional ambiguity concerning the speaker of the discourse presented. In most cases, FID can be identified by certain linguistic indices, such as the omission of reporting clause, reference to the subject of consciousness with a third-person pronoun and, in a seeming contradiction to the latter, the inclusion of deictic, lexical and syntactic choices typical of direct discourse. However, as the markers of FID range from purely linguistic to broadly contextual ones, the identification of FID is always more or less a matter of interpretation, and the device itself can be described as an interpretive rather than a grammatical category. Earlier research on the translation of FID has shown that in translations, FID is often substituted with other modes of discourse presentation, or simply rendered less discernible than in the original; in both cases, the narrative point of view is altered. In translations, the merging or clash of intonations present in FID is subdued or replaced with the unity of voice. This is typically taken to be a consequence of grammatical and stylistic restrictions imposed by the specific language pair involved in translation. This, however, seems an unlikely explanation, as the problem has been observed in several studies conducted between several different language pairs and including translations dating from different periods of time. Even if there are certain language-specific factors complicating the translation of FID, I suggest that the problem is not, in the first place, language- or culture-specific, but essentially universal in nature. In fact, I suggest that the weakening and loss of FID in translations result from linguistic changes that explicitate and normalize the text ? that is, manifestations of explicitation and normalization, much-discussed candidates for the status of translation universals. Following from this, I propose that these alleged translation universals explain the weakening and loss of FID in translations in the sense of a more general statement explaining a more specific one (see Croft 1990 for his scalar concept of degrees of generalization). The hypothesis is tested by analyzing a corpus of texts consisting of passages with FID from Dostoevsky's novel "Prestuplenie i nakazanie" (Crime and Punishment) and its translations into Finnish (dating from 1889, 1907, 1922, 1970, 1986 and 2008) and English (1914, 1992). The main emphasis is on translations into Finnish. Translations of the same novel into two different receiving languages as well as translations into the same language dating from different periods of time are included in order to eliminate (or discern) the influence of language-specific, time-bound and idiosyncratic factors. Additionally, comparative data is provided by passages with FID from English and Finnish fiction (Jane Austen, Juhani Aho) and their respective translations into Finnish and Russian. Due to the nature of FID as an interpretive category, my approach is essentially qualitative. The method of analysis is comparative translation description, with classical narratological concepts (FID and related modes of discourse presentation) employed as tools for identifying the narrative point of view and mode of discourse presentation in both the source and the target text, and, consequently, for locating and categorizing translation shifts in point of view. Linguistic changes accompanying these shifts are then analyzed with the help of basic linguistic categories. Finally, the linguistic changes are classified according to their pragmatic function in terms of translation universals: do they explicitate or normalize the text? The analysis reveals that all of the linguistic changes accompanying the weakening or loss of FID in translations can be categorized as manifestations of explicitation or normalization. Most typical explicitating changes are the addition of a reporting clause and the replacement of the personal reference to the subject of consciousness with a more explicit one (third-person pronoun with an explicit noun reference or impersonal reference with a third-person pronoun), while normalization is typically manifested in the elimination of direct discourse constructions (such as expressive punctuation) or the replacement of the third person pronoun referring to the subject of consciousness with the first person. Consequently, linguistic changes complicating the transfer of FID into translation, although seemingly different at first sight, form a consistent pattern when treated as different manifestations of explicitation and normalization. The frequency of individual changes behind the weakening and loss of FID vary to some extent depending on languages in question, or even on individual translator, but the overall tendency remains the same. To make sense of the consequences of explicitating and normalizing changes in the context of FID, I reconceptualize the linguistic description of FID in terms of its textual implicitness and incoherence. Because of the omission of reporting clause, the nature of FID as an instance of represented speech/thought is only implicitly stated in the discourse. Implicitness characterizes also the personal reference to the subject of consciousness, for normally only a (third-person) personal pronoun is used instead of, say, any more explicit form of reference such as a proper name. Textual incoherence or seeming textual inconsistency is created by the combination of features characteristic of indirect discourse with those characteristic of direct discourse typically, a third-person reference to the experiencing character combined with exclamations, value-laden adjectives etc. obviously belonging to this character. Identifying implicitness and incoherence as essential characteristics of the linguistic form of FID is the key, I claim, to explaining the non-transfer of FID into translations. Changes that spell out the implicit references of the source text (explicitation) or level off its textual incoherencies (normalization) do, in the context of FID, affect the essential features characterizing the mode: implicitness and incoherence. This, in turn, leads to the weakening or loss of the mode and, consequently, to shifts in the narrative point of view. Therefore, if explicitation and normalization are indeed universal tendencies in translation, this might explain why the weakening and loss of FID is so common in translations. So far, research on translation universals has concentrated on proving the existence of the universals posited, while the question of their effects (narrative or other) has received little consideration. In the present study, I attempt to move on and look at the effects of explicitation and normalization in a specific context. In FID, the implicitness and the seeming incoherence of textual surface have a crucial influence on the reader s intepretation of narrative perspective and so, consequently, has the tendency of translated texts to be more explicit and textually conventional than their originals. Depending on context and on the function of FID in the source text, the weakening or loss of FID may affect the interpretation of the text in various ways. Most typically, the target reader is distanced from the fictional character, and the impression of hearing the character s inner voice is weakened. In certain cases, this may lead to a change in character traits or even the reader s understanding of the truth-value of the passage. However minor in individual cases, the accumulation of these effects leads to an overall simplification and disambiguation of the narrative structure of the text and, therefore, supports Toury s (1995) law of growing standardization also a possibly universal tendency. Therefore, the weakening or loss of FID in translations is both a manifestation and a consequence of universal tendencies in translation. Even though the intentionality of the shifts can only be speculated upon, effects such as distancing the reader from a fictional character are not likely to be intentional on the part of the translator. I suggest that even if the linguistic manifestations of explicitation and normalization are conscious and intentional, translators are not necessarily aware of the way these changes affect the narrative structure of the text: the resulting shifts in point of view may still be unconscious and unintentional. From this perspective, the posited universality of explicitation and normalization is of special interest. If the tendencies of explicitation and normalization are indeed universal in translation, they are likely to affect all translations, unless consciously controlled. Therefore, in order to avoid unintentional shifts in narrative point of view, translators would benefit from an awareness of not only explicitating and normalizing tendencies operative in translation, but also of the influence they may have on the transfer of the narrative point of view in translation