Article body

When scholars or critics identify ‘changes,’ ‘adaptations’ or ‘manipulations’ in translations of children’s literature, they often rightly describe and analyse them in terms of the differing social, educational or literary norms prevailing in the source and the target languages, cultures and literatures at that given time.[1]A rich source of such observations are the many translations of Astrid Lindgren’s Pippi Långstrump (1945), which give a good indication of what was perceived by the target cultures, at the time of translation, to be inacceptable for child readers. In a scene in the novel Pippi, Tommy and Annika are playing in the attic when Pippi finds some pistols in a chest. She fires them in the air and then offers them to her friends who delightedly accept. In the German translation Pippi doesn’t give the pistols to her friends, instead she instructs them – and the readers – by changing her mind, putting them back in the chest and declaring “Das ist nichts für Kinder!” (Lindgren 1965,205) (“that’s not right for children”), a sentiment totally out of character. She herself had made fun of such moralising just a few moments previously when firing the pistols. A possible explanation for this change in the German translation could be that post-war Germany didn’t want its children to be encouraged to use weapons.[2]

The point of focus of this article will not be changes and manipulations in translated children’s literature and the reasons motivating them, whether social, educational or aesthetic. Instead I want to concentrate on the agency of such changes, the translator, in order to identify her/his presence in the translated text.

The translator’s visibility has been a much discussed issue in translation studies since Lawrence Venuti used the term ‘invisibility’ to describe both the illusionistic effect of the translator’s discourse and the practice by publishers, reviewers, readers, etc. in contemporary Anglo-American culture of judging translations acceptable when they read fluently.[3] His ‘call to action’ to translators has been for visibility by use of nonfluent, nonstandard and heterogeneous language, by producing foreignized rather than domesticated texts. He rightly insists on talking about translators as real people in geopolitical situations and about the politics of translation and ethical criteria (cf. Venuti 1995). But the translator’s discursive presence can, I submit, also be identified in texts which aren’t nonfluent, nonstandard and ‘foreignized’; it can be located on a theoretical level in a model of narrative communication as shown by Giuliana Schiavi in 1996, and on the level of analysis of the text based on such a model, where the translator’s presence is evident in the strategies chosen, in the way s/he positions her/himself in relation to the translated narrative.

My guiding questions are: What kind of translator is making her/himself felt in the text? Where can s/he be located in the act of communication which is the narrative text? How does the implied reader of her/his translation (the target text) differ from that of the ‘original’ (the source text)? To do this, I will present a theoretical and analytical tool, a communicative model of translation which links the theoretical fields of narratology and translation studies. Before doing so, however, I have to emphasise two points. Firstly, the model applies to all fictional literature in translation. Due to the asymmetrical nature of the communication in and around children’s literature where adults act on behalf of children at every turn, the translator as s/he becomes visible or audible as a narrator is often more tangible in translated children’s literature than in literature for adults (where Venuti likes to talk about visibility, my preferred metaphor is audibility, the voice that is heard in the text[4]). I will, secondly, be talking about narrative texts only. A model for drama or poetry would call for appropriate modifications.

Narrative communication: a model

The point of departure for the model to be presented in three steps is the basic narrative structure proposed by Seymour Chapman in Story and Discourse (1978) (Fig. 1).

Fig. 1

Six part model of narrative communication based on Chatman 1978

Six part model of narrative communication based on Chatman 1978

-> See the list of figures

In this well-known and commonly applied model six different parties form three pairs. The narrative text – indicated by the box in the middle – is the message transmitted from the real author to the real reader, from the one who physically wrote the text of the book to the one who holds it in her/his hands and reads it (or has it read to her/him). These parties are not to be found within the book itself, nor does the real author communicate directly with the real reader, the communication takes place between the constructed pairs within the narrative text. The first of these pairs is the implied author and the implied reader.

The real author, according to Chatman (1990,75), “retires from the text as soon as the book is printed and sold,” what remains in the text are “the principles of invention and intent” (ibid.). The source of the work’s invention, the locus of its intent is the implied author, whom Chatman calls a silent instructor, the “agency within the narrative fiction itself which guides any reading of it” (ibid.). The implied author, an agency contained in every fiction, is the all-informing authorial presence, the idea of the author carried away by the real reader after reading the book. The implied reader is the implied author’s counterpart, “the audience presupposed by the narrative itself” (Chatman 1978,149f), the reader generated by the implied author and inscribed in the text.

The asymmetrical nature of the communication in children’s literature is reflected in this model as follows: an adult implied author creates an implied reader based on her/his (culturally determined) presuppositions as to the interests, propensities and capabilities of readers at a certain stage of their development. The implied author is thus the agency in children’s literature which has to bridge the distance between ‘adult’ and ‘child.’[5]

The next, innermost, pair in the model is the narrator and the narratee. The narrator is the one who tells the story, hers/his is the voice audible when a story is being told. The narratee, in the words of Barbara Wall, is “the more or less shadowy being within the story whom … the narrator addresses” (1991,4). The narrator is not always sensed as a persona in the text; Chatman distinguishes between the ‘overt’ and ‘covert’ type. Overt are the narrators who feature as figures in the narrative, an example being Oswald Bastable in Edith Nesbit’s The Treasure Seekers,

“There are some things I must tell before I begin to tell about the treasure-seeking, because I have read books myself, and I know how beastly it is when a story begins, “‘Alas!’ said Hildegarde with a deep sigh, ‘we must look our last on this ancestral home’” …”,[6]

Nesbit 1899,3f

or Christopher Robin’s father who tells the stories and features in the frame in Milne’s Winnie-the-Pooh (1926). Equally overt are narrators who don’t feature as characters but as an authorial presence in the text, such as the one who declares in Nesbit’s The Enchanted Castle: “the sensible habit of having boys and girls in the same school is not yet as common as I hope it will be some day” (Nesbit 1909,7). The overt narrator has become less common in children’s literature over the past few decades, but even without saying ‘I,’ s/he can be no less revealing of character and attitude.[7]

The narratee, too, can be a character in the novel – Christopher Robin in the frame of Milne’s Winnie-the-Pooh is an obvious example, or the social worker for whom Hal writes the account of his story in Aidan Chambers’ Dance on my Grave (1982); more often s/he isn’t actually portrayed but evoked. The overt first-person, authorial narrator occasionally addresses her/his narratees with questions or appeals like “You know the kind of house, don’t you?” or “You may imagine their feelings” (both examples from Nesbit’s The Enchanted Castle).

The narrator is created by the implied author and is not to be confused with that agency. Similarly the narratee should not be identified with the implied reader. In some cases there will be some overlap. If we again take Winnie-the-Pooh as our example: Christopher Robin of the frame is the narratee, but the implied reader or rather implied readers include older children and adults reading the story to children. There are elements in the text which appeal to and can be understood only by them, indeed which are written specifically with an older audience in mind. In this case we can speak of a text with dual or even multiple address.[8]

Translation and narrative communication

The second of the three steps, moving from general narrative theory to the specifics of translation, looks at translation in terms of narrative communication. The model in Fig. 1 applies to an original (non-translated) text and its readers. Taking Erich Kästners Emil und die Detektive (1929) as an example: the real author is Kästner, the real reader is someone who reads the original novel in German. Only those who read German can be real readers of that specific text. Where does this leave those who read Kästner’s Emil in Spanish, Swahili or Swedish? They aren’t accounted for in Chatman’s model, which can only represent an original text.[9] In the case of a translated text, however, the message transmitted by the real author in the source language is read by the real reader in the target language. Kästner wrote Emil in German but a Spaniard reads it in Spanish. To account for what has happened in translation the model has to be expanded.

Translation is depicted in Fig. 2 as two sequential processes of communication.

Fig. 2

Translation in narrative communication, incorporating the implied translator and the implied reader of the translation [‘…’ denotes narrator and narrate]

Translation in narrative communication, incorporating the implied translator and the implied reader of the translation [‘…’ denotes narrator and narrate]

-> See the list of figures

On the left side of Fig. 2 is the source text (the ‘original’) with the already familiar parties. Where the real reader was situated at the end of the process illustrated in Fig. 1, we now have the translator. The translator acts in the first instance as the real reader of the source text. As someone familiar with the source language as well as the conventions and norms of that culture, s/he is in a position to slip into the role of the implied reader of the source text. Above and beyond that s/he tries to identify ‘the principles of invention and intent’ of the text – the implied author and the implied reader. (This is particularly significant for the process of translating children’s literature. As an adult, the translator does not belong to the primary addressees of most children’s books. S/he has to negotiate the unequal communication in the source text between adult (implied) author and child (implied) reader in order to be able to slip into the latter’s role.)

Parallel to the source text is the target text (the translation). As the creator of the translation, the translator acts, in the second half of the process shown on the right side of Fig. 2, as a counterpart to the real author of the source text; s/he is the one who creates the target text in such a way that it can be understood by readers in the target culture with language, conventions, codes and references differing from those in the source culture. However, the translator does not produce a completely new message, as Giuliana Schiavi who identified the translator’s presence in narratological terms writes, s/he “intercepts the communication and transmits it – re-processed – to the new reader who will receive the message” (1996, 15). By interpreting the original text, by following certain norms, and by adopting specific strategies and methods, the translator, according to Schavi, “builds up a new […] relationship between what we must call a ‘translated text’ and a new group of readers” (ibid., 7); in doing so s/he also creates a different implied reader to the one in the source text; the implied reader of the translation.[10] This implied reader can be equated with the implied reader of the source text to different degrees but they are not identical. The implied reader of the translation will always be a different entity from the implied reader of the source text. This statement applies to all translated fictional texts.

If the implied reader of the translation differs from her/his counterpart in the source text. then the question has to be asked: what is the agency which creates the difference? The implied reader of the source text, the reader inscribed in the text, is generated by the implied author. By the same token the implied reader of the target text is generated by a similar agency: the implied translator.

Based on these deliberations, the final, complex model of the translated narrative text[11] and all its agencies (Fig. 3) can be described as follows:

Fig. 3

Communicative model of the translated narrative text

Communicative model of the translated narrative text

(st = source text, tt = target text)

-> See the list of figures

The communication between the real author of the source text and the real reader of the translation is enabled by the real translator who is positioned outside the text. Her/his first act is that of a receptive agent, who then, still in an extratextual position, transmits the source text via the intratextual agency of the implied translator. The narrator, narratee and implied reader of the target text, all generated by the implied translator, can be roughly equivalent to their counterparts in the source text; however they can also differ greatly, as the following examples will reveal.

In translated texts, therefore, a discursive presence is to be found, the presence of the (implied) translator. It can manifest itself in a voice which is not that of the narrator of the source text. We could say that two voices are present in the narrative discourse of the translated text: the voice of the narrator of the source text and the voice of the translator.

The voice of the translator

The translator’s voice can be identified on at least two levels. One of them is that of the implied translator as author of paratextual information such as prefaces or metalinguistic explanations such as footnotes. Here ‘the translator’ can be heard most clearly. When, for example, s/he tells the readers of the German translation of Barbara Park’s My Mother Got Married (and other disasters) (1989) that Thanksgiving Day is a harvest festival which takes place on the fourth Thursday in November (Park 1991,115), it is clearly not in translation of an explanation to be found in the source text (American readers hardly need to be told what Thanksgiving Day is). It is information composed for readers of the target text by the translator and proffered in her/his own voice.

In his “companion piece” to Schiavi’s with concrete examples of the implied translator, Theo Hermans locates the translator’s voice as an “index of the Translator’s discursive presence” (Hermans 1996,27) in situations where s/he has “to come out of the shadows and directly intervene in a text which the reader had been led to believe spoke only with one voice” (ibid.). These are primarily moments of paratextual intervention where explanations are crucial, self-reflexive references to the medium of communication itself, moments where (the source) language itself is the theme and “when ‘contextual over-determination’ leaves no other option” (ibid., 23). Hermans’ Voice of the Translator is therefore primarily a metalinguistic one, in principle “wholly assimilated into the Narrator’s voice.” I would argue that the translator’s voice is not only heard in such interventions, it can also be identified on another discursive level, on the level of the narration itself as a voice “dislocated from the one it mimics” (ibid., 43), one which is not assimilated into the voice of the narrator of the source text. This specific voice, hitherto largely unrecognised by translation studies or narratology, is what I call the voice of the narrator of the translation.[12] Two examples will illustrate how this voice can manifest itself and the consequences it has for communication within the text.

The first is a passage from the classic Swiss children’s novel first published in 1881/2, Johanna Spyri’s Heidi. It occurs in the novel just after Heidi has completed the difficult task of teaching the young goatherd Peter to read, so that he can be a source of comfort to his blind grandmother by reading her beloved hymns aloud when they are snowed in during the winter and Heidi is unable to visit. The grandmother nonetheless prefers Heidi’s rendering of the hymns, and the following reason is given:

“Das kam aber daher, weil der Peter sich beim Lesen ein wenig einrichtete, daß er’s nicht zu unbequem hatte. Wenn ein Wort kam, das gar zu lang war oder sonst schlimm aussah, so ließ er es lieber ganz aus, denn er dachte, um drei oder vier Worte in einem Vers werde es der Großmutter wohl gleich sein, es kämen ja dann noch viele. So kam es, daß es fast keine Hauptwörter mehr hatte in den Liedern, die der Peter vorlas.”

Spyri 1978, 276

The manner in which Peter shortcuts his reading and how he justifies it to himself is related by a third-person narrator briefly focalising Peter’s point of view. This is reflected in the use of colloquial language (“er’s”), the particles “gar” and “wohl gleich,” the subjunctive case relating his thoughts as indirect speech (“werde,” “kämen”) and in the uncommented report of his naïve logic. The authorial description of the result is understated and mildly ironic but the narrator neither remarks upon nor judges Peter’s actions or thoughts.

In an anonymous translation published in 1949,[13] the voice of the narrator of the translation tries, largely, to emulate that of the narrator of the original, even though the colloquial tone is muted.

“The reason (…) was that Peter used to tamper with the words a little, so as to make the reading slightly less trouble. When he came to a word that was too long, or in some other way seemed difficult, he simply left it out; for he thought that Grannie wouldn’t notice the absence of two or three words in a verse – after all, there were so many of them. The result was that the hymns, as read aloud by Peter, had scarcely any nouns.”

Spyri 1949, 224

The version of the same passage in M. Rosenbaum’s popular and much issued translation is compressed, with the final authorial dictum omitted:

“The reason was, of course, that Peter was rather lazy about reading for the grandmother, and if a word were too difficult or too long he just skipped over it thinking it would not matter very much to the grandmother seeing there were so many words!”

Spyri 1955, 210

The narrator of this passage, unlike the one of the source text, passes judgment on Peter (“rather lazy”). With the phrase “of course” a bond of agreement is insinuated between the narrator of the translation and implied reader and the final exclamation mark is a comment on Peter’s thoughts, signalling to the reader that they should be regarded as hilarious. The narrator of this translation therefore appeals directly to the implied reader and makes what s/he has to say more explicit. The implied translator obviously has a reader in mind who may not grasp the subtle irony of Spyri’s narrative, whose reading of the text has to be guided with authorial asides and exclamation marks, thus transforming the laconic explanation of the source text into an overstatement.

Less subtle is the amplification to be found in a German translation of Lewis Carroll’s Alice in Wonderland, published in 1949[14] in the passage in which the ‘Mock Turtle’ is first introduced. In the original it reads as follows:

“Then the Queen left off, quite out of breath, and said to Alice, ‘Have you seen the Mock Turtle yet?’
’No,’ said Alice. ‘I don’t even know what a Mock Turtle is.’
’It’s the thing Mock Turtle Soup is made from,’ said the Queen.
’I never saw one, or heard of one.’
’Come on, then,’ said the Queen, ‘and he shall tell you his history.’”

Franz Sester, one of thirty-two translators to have produced unabridged German versions of Alice in Wonderland, apparently found this too short on explanation. What were his young readers supposed to think a Mock Turtle was? He therefore added a lengthy passage with no equivalent in the English original in which Alice is, culturally adapted, a well-behaved German girl who learns English at school. In the course of the explanation of what a Mock Turtle is, the reader is introduced to Alice’s teacher and Alice’s aunt and is given a recipe for Mock Turtle soup:

“‘Wie kann man mich zu einem Tier führen,’ so dachte Alice, ‘das es doch gar nicht gibt?’ Alice hatte schon im zweiten Jahre Englisch. Die Lehrerin hatte den Kindern bereits beigebracht, daß ‘turtle’ auf deutsch ‘Schildkröte’ bedeutet, während ‘mock’ auf deutsch ‘nachgemacht’ heißt. Mock-Turtle-Suppe war also nichts anderes als eine ‘nachgemachte’ Schildkrötensuppe, die gar nicht mit Schildkrötenfleisch zubereitet war. Die Folge des guten englischen Unterrichts, den die Lehrerin gab, war also, daß alle Mädchen der Quinta genau wußten, was eine Mock-Turtle war und aus Dankbarkeit der Lehrerin gleich den schönen Spitznamen ‘Die Mockturtle’ gaben.

  Bei der Hochzeit ihrer Tante hatte Alice auch einmal in der Küche zugesehen, wie die Mock-Turtle-Suppe zubereitet wurde. Sie erinnerte sich noch genau, daß in die Suppe ein halber Kalbskopf, ein Ochsengaumen, Suppengrün und andere Zutaten kamen. Später bei Tisch hatten der kleinen Alice die Kalbskopf- und Ochsengaumen-würfelchen in der Suppe besonders gut geschmeckt, denn, wißt ihr, wenn es sich um Essen handelte, hatte Alice immer ein besonders großes Interesse und ein ausge-zeichnetes Gedächtnis. So etwas behielt sie immer viel besser als englische oder französische Vokabeln.”

Carroll 1949, 69

(‘How can I be brought to an animal which doesn’t even exist?,’ thought Alice. This was Alice’s second year learning English in school. The teacher had already told the children that ‘turtle’ was ‘Schildkröte’ in German and that ‘mock’ meant ‘nachgemacht.’ So Mock Turtle Soup was nothing other than an imitation turtle soup which wasn’t made with turtle meat at all. Thanks to the excellent English instruction by their teacher, each girl in the sixth class knew exactly what a Mock Turtle was, and out of gratitude to their teacher they gave her the lovely nickname ‘the Mockturtle.’   

At her aunt’s wedding Alice had also seen how Mock Turtle soup was made in the kitchen. She could remember exactly that half a calf’s head, an ox’s gum, some carrots, onion, celery, leeks and parsley and other ingredients were used to make the soup. Later, during the meal, little Alice especially savoured the small pieces of calf’s head and ox’s gum in the soup because, you know, when it came to eating, Alice was always very interested and remembered anything to do with that much better than English or French vocabulary. (my translation – eo’s)

The voice of the narrator of this translation overrides that of the narrator of the source text, the tone is heavy-handed and pedantic. The work of the implied translator was informed here, as always, by the time and place in which the translation was carried out (the detailed descriptions of food and eating must have been particularly attractive to readers during the hungry post-war years in Germany) and especially by his notion of the implied reader based on assumptions as to the interests, propensities and capabilities of readers at that stage of their development. He obviously envisaged the implied reader of his translation as a child devoid of the fantasy necessary to imagine what a mock turtle might be, a child to whom everything had to be explained. His implied reader can’t cope with a ‘foreign’ setting, any action has to be transported to familiar, German, territory. It could be claimed that this implied translator missed the point of the book and produced a nanny translation devoid of nonsense. Unlike the source text, it holds absolutely no attraction for adults, any incidental humorous effect is involuntary. The implied translator of this Alicens Abenteuer im Wunderland and the implied author of Carroll’s Alice in Wonderland come from entirely different planets in the universe of children’s literature, and the implied translator is sure that his is a safer place for child readers than Carroll’s one.

The discursive presence of the translator can be located in every translated narrative text on an abstract level as the implied translator of the translation. The translator’s voice can make itself heard on a paratextual level as that of ‘the translator’ and inscribed in the narrative as what I have called ‘the voice of the narrator of the translation.’ This particular voice would seem to be more evident in children’s literature than in other bodies of literature due to the specific, asymmetrical communication structure which characterises texts which are written and published by adults for children. In these texts, contemporary and culture-specific notions of childhood play some part in determining the construction of the implied reader: what do ‘children’ want to read, what are their cognitive and linguistic capabilities, how far can/should they be stretched, what is suitable for them – these are only some of the questions implicitly answered by the assumptions evident behind the ‘child’ in children’s literature and behind the child in any specific children’s book. The same questions are asked again by the translator and by the publisher of the translation. The strategies chosen by translators for children are, as Riitta Oittinen reminds us, primarily dictated by their child image.[15] Assumptions about ‘the child’ and the ensuing construction of readers of specific children’s books and of their translations can lead to vast divergences between implied readers of source and target texts for children, as the few examples here have illustrated. Using a model of narrative communication it is possible to locate and name the place and agent of this divergence, to identify exactly where ‘changes’ and ‘manipulations’ happen and how these can be described in terms of narrative strategy, in terms of the construction of implied readers of the translation.

André Lefevere and Susan Bassnett describe translation as one of the most obvious, comprehensive, and easy to study ‘laboratory situations’ for the study of cultural interaction, because a comparison of original and translation “will not only reveal the constraints under which translators have to work at a certain time and in a certain place, but also the strategies they develop to overcome, or at least work around those constraints. This kind of comparison can, therefore, give the researcher something like a synchronic snapshot of many features of a given culture at a given time.” (1998,6). One of these features is doubtlessly child image as one of the most influential factors determing the strategies developed by translators of children’s literature. The model of narrative communication presented here offers a tool with which varieties of implied readers of source and target texts at certain times and in certain places can be identified by analysing the narrative strategies chosen by the translator as indicitive of her/his idea of the reading child and the kind of literature appropriate for that child. These strategies manifest themselves in audible form in the voice of the translator in children’s literature.