(Deutsche Forschungsgemeinschaft) since 1999. We gratefully ... includes an isolation of occurrences of vulnerable items and manual annotations to locate ...... Grammatik und darüber hinaus ... Ein internationales Handbuch, Vol. I, G. Helbig ...
Viktor Becher, Juliane House & Svenja Kranich
use of “cultural filtering”, i.e. adapting the original text in accordance with conventionalized expectation norms of the target audience. If, however, the influence of global hegemonic English is pervasive enough, it will effectively prevent cultural filtering and instead impose Anglophone norms on texts in other languages. To test this hypothesis we relied on the results of House’s (1996; 2006) and Clyne’s (1987; 1994) empirical contrastive pragmatic work and its postulation of certain dimensions of communicative preferences along which English and German text production tend to differ. Particularly relevant to the case studies presented in this article are the two dimensions ‘directness vs. indirectness’ and ‘content-orientation vs. addressee-orientation’. In general, English discourse in many genres tends to be more indirect and oriented towards the addressee, while German discourse may be described as more direct and oriented towards content (House 1996; 2006).2 Given the absence of comparable contrastive pragmatic research using different language pairs, we concentrated in this project on the language pair English and German. To test the project hypothesis of global English influence on textual norms we constructed a multilingual diachronic corpus of texts in two genres which we assumed were particularly prone to global Anglophone influence: popular science and economic texts. The bulk of the corpus consists of English originals, their German translations as well as comparable, i.e. non-translated German texts.3 For an analysis of the translation relation, source and target texts were manually aligned sentence by sentence, so that an electronic search of the originals automatically retrieves the relevant segments in the translations, and vice versa. The popular science texts which we will examine in the present study were mostly taken from the journal Scientific American and its German daughter publication, Spektrum der Wissenschaft. The total number of words in the popular science part of our corpus is about 500,000. The texts were sampled from two distinct time frames: 1978–1982 and 1999–2002. Given this corpus design, we are in a position to conduct comparative research into language-specific and genre-specific conventions and into the nature of the translation relations. However, in order to make valid statements about whether it is in fact Anglophone texts that have led to variations and changes of German textual norms over the past 25 years and not any independent development of norms in the two languages involved, we also have to draw
Convergence and divergence of communicative norms through language contact in translation Viktor Becher, Juliane House & Svenja Kranich
Hamburg, University and Research Center on Multilingualism This contribution addresses the question of whether and how translation as a classic case of language contact can act as a trigger for convergence and divergence phenomena between two languages. We present two studies which indicate that translation-induced convergence does not occur unconditionally: while we found no signs of English-German convergence in the use of modal verbs (study 1), the use of sentence-initial concessive conjunctions in translated and comparable German texts shows convergence with Anglophone usage patterns (study 2). Explaining these disparate results, we hypothesize that divergence occurs when bilinguals perceive profound differences between source and target language (as is the case in English and German lexicogrammatical means for expressing modality), while convergence takes place when bilinguals perceive items as equivalent in form and function (as is the case in English and German concessive conjunctions).
1. Introduction This paper presents some recent results from the project Covert Translation.1 The general assumption underlying this project is that the dominant status of the English language in global communication influences and changes communicative norms in other languages through language contact in covert translation and comparable text production. Covert translations (House 1977; 1997) are translations in which the communicative purpose of the original text is upheld via the . The project is directed by Juliane House, current research fellows are Viktor Becher and Svenja Kranich. Claudia Böttger, Julia Probst, Nicole Baumgarten and Demet Özçetin were members of the research team in previous project phases. The project is part of the Research Center on Multilingualism (Sonderforschungsbereich “Mehrsprachigkeit”) at the University of Hamburg. It has been funded by the German Science Foundation (Deutsche Forschungsgemeinschaft) since 1999. We gratefully acknowledge this ongoing generous support.
2. Of course, the characteristics of English and German discourse summarized here should not be viewed as absolute dichotomies but rather as points on a cline. 3. Additionally, the corpus contains translations in the opposite direction (German → English) as well as translations from English into French and Spanish for validation purposes.
2nd proofs
Convergence and divergence through language contact in translation
Viktor Becher, Juliane House & Svenja Kranich
on larger German and English reference corpora like for instance the DWDSKerncorpus4 and the British National Corpus. The methodology used in this project is a combination of qualitative and quantitative methods. The qualitative part of the research consists of detailed contrastive analyses of some 70 texts, the quantitative part combines frequency counts with statistical operations, and the re-contextualized qualitative analysis includes an isolation of occurrences of vulnerable items and manual annotations to locate collocations and co-occurrences with several pragma-linguistic categories. We can thus divide our project work into three phases: in a first qualitative phase, texts were analyzed and compared in their entirety on the basis of House’s systemic-functional (register-theory based) translation evaluation model (1997), such that textual profiles were established reflecting whether and how over time cultural filtering has changed or has indeed been completely abandoned. Results of the analyses in this phase show that changes did indeed occur in the realization of subjectivity and addressee orientation. The concept of subjectivity is related in systemic-functional theory to Stance and captured along the register dimension of Tenor (Halliday & Matthiessen 2004). Similar concepts discussed in the literature are point of view and perspective (Smith 2003), metadiscourse (Crismore & Farnsworth 1990), evaluation (Hunston & Thompson 2000) and emotive prosody (Bublitz 2003). Subjectivity is also clearly related to the function certain linguistic elements have in influencing hearers/writers (intersubjectivity or addressee-orientation). For our quantitative work, the complex and rather fuzzy concepts subjectivity and addressee orientation were operationalized to include phenomena such as speaker-hearer deixis, mental processes (in the sense of Halliday), commenting parentheses, and expressions of modality and connectivity. Results of the second, quantitative project phase essentially confirmed the changes over time in regard to the frequency with which functional categories such as speaker-hearer deixis or coordinate conjunctions as means of expressing subjectivity and addressee orientation were realized. In the current third re-contextualized qualitative project phase, where we focus on the translation relation and isolate complex collocation and co-occurrence patterns, we are examining whether equivalent elements occur in the same linguistic environments, and whether they are used for the same communicative functions. Our findings concerning adaptations of German communicative norms in the areas of subjectivity and addressee orientation suggest that in the space of 25
years there has indeed been an adaptation of popular science and economic texts to Anglophone norms with regard to the use of linguistic means of expressing speaker-hearer deixis (cf. Baumgarten & Probst 2004; Baumgarten & Özçetin 2008; Baumgarten 2008), as well as coherence (cf. Baumgarten’s 2007 detailed study of sentence-initial conjunctions And/Und ). What these phenomena have in common is the fact that they constitute formal and functional equivalents in the two languages involved such that a bilingual speaker (or a translator) is capable of “interlingual identification” (Weinreich 1953). Interlingual identification, in turn, facilitates transfer (cf. also Heine & Kuteva 2005: 4, 219–234). Conversely, in cases where there is no clearly discernible formal and functional equivalence in markers of subjectivity and addressee-orientation, Anglophone influence could not be established in our analyses. Cases in point are the use of mental processes (Özçetin 2008), the realization of explicitness in the description of events and states of affairs (House 2008; Baumgarten et al. 2008), and the employment of linking constructions (Bührig & House 2004; 2007). These findings have given rise to a more specific project hypothesis, namely that the impact of English communicative norms on German communicative norms is facilitated in regard to those linguistic expressions where form-functionequivalence can easily be established by the bilingual individual. In the following, two recent project case studies which further examine this hypothesis will be presented. The first study investigates the use of modal markers – as a case of non-equivalence of form and function in German and English discourse – while the second study investigates the use of sentence-initial concessive conjunctions as a case of formal and functional equivalence. Both case studies were carried out on the popular science part of our corpus described above.
2. The use of epistemic modal markers 2.1 The textual function of epistemic modal markers Broadly speaking, epistemic modal markers function as “a speaker’s comment on the status of information in a proposition. They can mark certainty (or doubt), actuality, precision, or limitation […]” (Biber et al. 1999: 972). Speakers can use these linguistic elements to “express[…] their lack of confidence in the propositions expressed in these utterances” (Coates 1995: 59). This can be seen in the following examples from our popular science corpus (English original and German translation plus English gloss):
4. The DWDS-Kerncorpus is a diachronically balanced corpus of written German covering the entire 20th century, each decade being represented by approximately 10 million words of text (Geyken 2007). The corpus can be queried free of charge at http://www.dwds.de.
(1) a.
Thus some individuals may have a defective natural interferon system.
b. Somit haben wahrscheinlich einige Patienten ein defektes Interferonsystem. ‘Thus some patients probably have a defective interferon system.’
2nd proofs
Convergence and divergence through language contact in translation
Viktor Becher, Juliane House & Svenja Kranich
In the above examples, the speaker (i.e. the author of the text), makes it clear that he is not one hundred percent certain of the truth of the proposition. In the English Example (1a), may indicates a relatively low degree of certainty, while wahrscheinlich in the German translation (1b) marks a rather high degree of conviction. That is, in both the English original and the German translation, the speaker modifies the statement, but the modal strength of the epistemic modal element used is different. Clearly, epistemic elements with high and low modal strength (i.e. marking both high and low certainty as to the truth of the proposition) are found in both the English and the German system. However, in accordance with the typical communicative preferences in English and German texts referred to in Section 1 above, we may assume that epistemic modal markers in general and those of low modal strength in particular will be more commonly used in the English texts, because they contribute to making texts more indirect as well as more addressee-oriented. The use of epistemic modal markers is thus not always motivated by a real lack of conviction in the truth of the proposition, but may also be used as a particular conventionalized communicative strategy. The following example illustrates this:
of German texts. If we look at the translation of (2b), we actually find that the German text again makes a somewhat stronger statement: (2) b. Planetenringe sind nicht nur schöne und spektakuläre Gebilde: Sie könnten geradezu der Schlüssel für die Erkenntnis sein, wie Planeten entstehen. ‘Planetary rings are not only beautiful and spectacular formations: they could in fact be the key to the understanding of how planets are born.’
This can be understood as an adaptation to German textual preferences, i.e. the use of a cultural filter: the translator chooses to strengthen the proposition by adding the modal particle geradezu, thus making the passage better conform to a preference for a direct and content-oriented communicative style rather than a more indirect and addressee-oriented style typical of English texts. 2.2 German translation strategies of English epistemic modals In order to investigate how common such adaptations in the translations are, four English modals – may, might, can and must – were searched for in the corpus and their epistemic uses were isolated. All instances in which the proposition containing the modal was translated were analyzed.5 The first and crucial criterion concerns the modal strength of the translation choice,6 i.e. the degree of certainty expressed concerning the truth of the proposition. Five different translation types can be distinguished in this respect. The first one is a translation using a modally unmarked proposition. Here complete conviction in the truth of what is asserted in the translation is expressed (as in 3b):
(2) a. Planetary rings are not only striking, exquisite structures; they may be the Rosetta stones to deciphering how planets are born.
One might suppose that the author has more conviction in the correctness of the proposition than the use of may betrays, and that he strives to conceal this in order to make the statement less open to criticism. Such a use of an epistemic modal expression as a hedging device may be motivated by a wish to state matters less directly and leave more room for non-face-threatening intervention (such as disagreement) on the part of the addressee. Hyland (1998: 351) notes that “deference, humility, and respect for colleagues’ views” are conveyed through the hedging use of epistemic modal markers (cf. also White 2003; Graefen 2007). White and Sano (2006) stress that the use of such devices produces a more ‘dialogic’ text which allows for alternative positions and voices. They explain that hedging devices such as epistemic modal markers
5. Since we are looking at covert translations (cf. House 1997), not all passages of the English original are present in the translations. Sometimes passages are simply left out, possibly because they are not deemed interesting for the German readership. 6. It is important to keep in mind that epistemic modal markers are not always chosen in order to express a lack of conviction in the truth of the proposition on the part of the speaker, but may just as well serve as a hedge. Using an expression of low modal strength can thus either be motivated by a low conviction in the truth of the proposition, or it can function as a communicative strategy which leaves a great deal of room for non-face-threatening disagreement. Markers of high modal strength, on the other hand, either mark a high degree of conviction in the truth of the proposition, or, as communicative strategy, only leave a certain amount of room for a different opinion. They ‘hedge less’, as they give clear preference to the opinion expressed. Such markers would thus seem to be more in line with typical German than with English communicative preferences.
… signal a recognition that, in the current communicative context, these are contentious matters and thereby […] signal recognition that those being addressed may query, reject or at least find such propositions novel or otherwise problematic. By this mechanism, the interpersonal cost to any who might advance alternative views is lowered as their position is recognised as a valid one in the current ongoing colloquy. (White & Sano 2006: 194)
Thus we come back to the communicative preferences summarized above: the hedging use of an epistemic modal makes the text more interpersonal, more addressee-oriented, more indirect, and therefore more typical of Anglophone than
2nd proofs
Convergence and divergence through language contact in translation
Viktor Becher, Juliane House & Svenja Kranich
Here again the test allows us to perceive that the English modal may is of low modal strength (This may or may not be true is possible), while the German modal element has high modal strength (it is not possible to say *Das ist wohl wahr, oder es ist wohl nicht wahr). The categories ‘same modal strength in translation’ and ‘lower modal strength in translation’ are evidenced by Examples (5) and (6) respectively:
(3) a. But the vast majority of the matter in the universe may not emit photons of any wavelength.
b. Doch der größte Teil der Materie im Kosmos strahlt gar keine Photonen ab. ‘But the greatest part of the matter in the universe does not emit any photons.’
In the majority of cases, however, the English modal element is translated by a German modal element. These instances were classified according to the relation between the modal strength in the English original and in the German translation. Three types were distinguished: higher modal strength, same modal strength, and lower modal strength in the translation. First a more fine-grained classification was attempted. But then a basic classification was found to be more intersubjectively reliable. For the two-way distinction between high modal strength and low modal strength, we applied an easy test: expressions conveying low modal strength can be combined with the negated expression using the same modal element without creating a contradiction. One can thus say: This may or may not be the case. Maybe it is true, maybe it is not true. Expressions of high modal strength or probability, where the speaker clearly favors one possibility over all alternative possibilities, on the other hand, disallow such a combination, e.g. *This should or should not be the case. *Probably it is true, probably it is not true. This test was used in the present study to decide how to classify the modal strength of the translations. The present classification is thus based on a two-way distinction between markers of high and of low modal strength. Additionally, cases where the German translation adds an element that upgrades the strength of the assertion, e.g. when English might is translated by könnte tatsächlich ~ durchaus ~ geradezu (all roughly translatable as ‘indeed, in fact’) were always classified as having ‘higher modal strength in the translation’. Similarly, when another modal element was added which downgraded the overall modal strength, we included it in the category ‘lower modal strength in translation’ (an example of this is found in (6) below): The translation type ‘higher modal strength’ has already been evidenced in Examples (1) and (2). A further example can be seen in (4):
(5) a. Imaging studies over the past decade have indicated which brain regions might malfunction in patients with ADHD.
b. Welche Gehirnregionen beteiligt sein könnten, lassen in den letzten zehn Jahren angefertigte Gehirnaufnahmen vermuten. ‘Which brain regions might be involved is indicated by brain images produced in the last ten years.’
(6) a. Such models might then be used to understand in greater detail the coevolution of memes.
b. Gegebenenfalls könnten die Modelle helfen, die Koevolution von Memen und Genen gründlicher zu verstehen. ‘Possibly the models might help to understand the coevolution of memes and genes more thoroughly.’
Finally we find translations where the modally marked proposition is rendered by a particular lexical choice rather than through recourse to a grammatical marker of modality, as in the following example:
(7) a.
… because mammary cells grow vigorously at this stage of pregnancy, indicating that they might do well in culture.
b. … zu diesem Zeitpunkt wachsen die Zellen des Euters lebhaft, was gute Kultivierbarkeit versprach. ‘… at this point the cells of the udder grow vigorously, which promised good culturability.’
In instances of the last type, the expression of less than full conviction regarding the truth of the proposition via a modal marker in the English original is rendered lexically in German, often in quite creative ways. These cases were very difficult to classify with regard to the modal strength they express, so that they have been put into a category called ‘other’. Concerning the four types where the modal strength of the translation choice could be determined, we may consider the types evidenced in (3) and in (4) as adaptations to German textual preferences, marking full or higher conviction and thus being more direct and more content-oriented, while translations of the types evidenced in (5) and in (6) remain close to the (more indirect, more addressee-oriented) English original.
(4) a. The cause of these losses remains unknown, but it may reflect the complexity of the genetic reprogramming needed if a healthy offspring is to be born.
b. Der Grund für diese hohen Verluste ist unbekannt, liegt aber wohl darin, daß die Eizelle das Erbgut des untergeschobenen “älteren” Zellkerns reprogrammieren muß […]. ‘The reason for these high losses is unknown, but probably lies in the fact that the egg needs to reprogram the introduced “older” cell nucleus […]’
2nd proofs
Convergence and divergence through language contact in translation
Viktor Becher, Juliane House & Svenja Kranich
2.6 (1999–2002) occurrences per 10,000 words. Taken individually, the findings do not allow strong generalizations. However, considered together, we can certainly see a general trend, as Figure 3 shows:
The results for the two different time-windows included in the corpus are shown in Figures 1 and 2: 100%
100%
unmarked
80%
higher
60% 40%
0% may
can
1978–1982
* The absolute numbers for this time-span are as follows: might n = 14, may n = 57, can n = 105, must n = 20. Total n = 196.
Thus one can say that when translators introduce changes regarding the strength of an assertion, there is a clear tendency to increase the modal force compared to the English expressions containing might, may, can or must, i.e. to use a modal marker with a higher modal strength or translate the proposition using a modally unmarked indicative. Translations with a lower modal strength, on the other hand, are much less frequent. We can regard this as an adaptation of the text to German communicative preferences, i.e. as the use of a cultural filter. Instead of a decreasing use of cultural filtering in the English-German translations due to the prestige associated with the English language, which the project hypothesis would have led us to expect, we find, in fact, an increasing divergence from the English model. We shall now see how this could be explained.
higher same
40%
lower
20%
other
0% might
may
can
1999–2002
Figure 3. Modal strength in German translation of English modals: 1978–1982 vs. 1999–2002
unmarked
60%
other
0%
Figure 1. Modal strength in German translations of English modals 1978–1982*
80%
lower
20%
must
100%
same
40%
other might
higher
60%
lower
20%
unmarked
80%
same
must
Figure 2. Modal strength in German translations of English modals 1999–2002* * The absolute numbers for this time-span are as follows: might n = 148, may n =184, can n = 327, must n = 29. Total n = 688. The reason for the elevated numbers compared to the earlier time-span mainly lie in the corpus make-up, which contains 42,500 words of English text for the 1978–1982 time-span, but 122,866 words for the 1999–2002 time-span. This has to do with the fact that for the earlier time-span popular scientific English texts which have covert translations into German are hard to come by, as the genre was only getting established in German.
2.3 Possible reasons for increasing divergence Looking at Figure 3 we could see that quite a remarkable proportion of the translations do not choose an expression of the same modal strength in German. This is true of almost 30% of the translation choices of the four epistemic modals analyzed in the earlier translations, and of a little over 40% of the translation choices in the later translations. We may thus ask why the German texts diverge so commonly from the model present in the English originals. The reasons may lie in a lack of a form-and-function equivalence between English and German in the field of modality. Other studies conducted in our project which have shown significant influence of English communicative styles on German styles in the German translations have dealt with linguistic elements showing a high degree of formal and functional similarity between the two languages (see for instance Baumgarten 2007
Already from these two figures it becomes apparent that no drastic diachronic change can be observed between the two time-spans. We may note that, overall, translators seem to choose a ‘freer’, more creative translation (as exemplified in (7) above) more often in the latter time-span. This is, however, not immediately relevant for the present investigation. What seems more interesting in the context of the present study is that apparently, translators use modally unmarked propositions as translations of must more often in the first time-span than in the second one. However, one should note that epistemic uses of must are, overall, not all that frequent in our corpus, having an average frequency of only 5.3 (1978–1982) and
2nd proofs
Convergence and divergence through language contact in translation
Viktor Becher, Juliane House & Svenja Kranich
on sentence-initial And and Und, and Baumgarten & Özçetin 2008 on the use of we and wir in business communication, as well as the following case study on sentence-initial But and Aber/Doch). These linguistic elements are easily identifiable as equivalents by any bilingual speaker and translator, and they thus provide a gateway for an influence on the frequency and/or functions of these elements. In the field of modality, on the other hand, such obvious equivalences are difficult to find in the language pair English-German. One of the major differences lies in the kind of categories predominantly recruited for the expression of modal meanings in the two languages. While in present-day English the modals are a clear-cut class of their own (cf. e.g. Denison 1993), having grammaticalized to a high degree, the German modal verbs still share many more characteristics with ordinary verbs, as is evident for instance in the existence of non-finite forms (Example 8), the ability to take a direct (Example 9) or a prepositional object (Example 10):
Furthermore, German texts show a certain predilection for lexical expressions rather than grammatical elements. Thus, phrases such as unseres Erachtens (‘in our opinion’) and so weit bisher bekannt (‘as far as it is known so far’) fulfil the same function as epistemic modal markers in that they modify the certainty with which the truth of the proposition is asserted, but they add further information (e.g. the grounds on which it is asserted), thus satisfying the preference for explicitness characteristic of German texts (cf. House 2004a, 2004b). In other words, the linguistic expressions predominantly chosen for the epistemic modification of the proposition are clearly different in the two languages. The translations of the English modal verbs also reflect these systemic differences. The realization of English might, may, can and must in the translations can be categorized into the following types: a. b. c. d. e.
(8) a. Maria hat das Gedicht am besten gekonnt. b. *Mary has could [past participle] the poem best. (9) a. Peter darf noch ein paar Kekse. b. *Peter may some more cookies. (10) a. Julia muss nach London. b. *Julia must to London.
zero translation modal verb modal adjective/adverb/particle a combination of several modal elements (e.g. modal verb + modal particle) a creative translation which renders the modal meaning via other means (typically lexical choice)
For reasons of space, we shall just look at the translations of all four modal verbs together. This gives us a good impression of the general tendencies:
A further difference, which can be seen as a consequence of a higher degree of grammaticalization, is that in English the modals are the predominant means of expressing modal meaning. In German, however, a variety of linguistic elements are used to this end. Among the grammatical elements expressing epistemic modal meanings, German commonly uses modal adjectives and adverbs (e.g. möglich, wahrscheinlich), as well as modal particles (e.g. wohl), a category not present in the English inventory (cf. Nehls 1989; König 2001; Teich 2003: 111). From the present state of further investigation, in which we compare the use of epistemic modal elements in the English originals, their German translations and the original texts in our corpus, these other categories seem to be clearly preferred over modal verbs in the untranslated, comparable German texts.7
100% 80%
a) zero
60%
b) modal verb c) modal particle/adv. /adj.
40%
d) multiple modal marking
20%
e) other
0% 1978–1982
1999–2002
Figure 4. Formal categories of German translations of English modals 1978–1982 vs. 1999–2002
7. The current work in progress compares expressions of epistemic modality in English originals, German translations and German originals, taking the functional side of these expressions as our departure point. In a manual analysis of a subpart of the corpus, all expressions that function as a modification of the speaker’s conviction in the truth of the proposition will therefore be isolated. These expressions are then categorized according to modal strength and linguistic
category of the element used. This allows us to gain both a more detailed view of the contrasts between English and German and the potential differences between German translations and German originals in the genre of popular scientific texts. Additionally, further studies of translation relations, comparable to the present one, will be conducted in the field of modal expressions.
2nd proofs
Convergence and divergence through language contact in translation
Viktor Becher, Juliane House & Svenja Kranich
Figure 4 shows the following diachronic development: there is a certain rise in the use of modal particles, adverbs or adjectives (c) as well as in the use of combinations of several modal elements (d). Also, the creative choices grouped together here under (e) become more frequent. The rise in categories (c) through (e) happens at the expense of the more straightforward translation of an English modal verb by a German modal verb. As far as the formal properties of the modal expressions are concerned, there is also an increasing divergence from the English original expression. Summarizing, we can conclude that as far as both the functional and the formal sides are concerned, German translations of popular science texts do not seem to be converging with the English model – on the contrary, they show increasing divergence from it (cf. also Kranich 2009).
(11) a.
But this process is not automatic.
b. Eine solche Entwicklung vollzieht sich allerdings nicht automatisch. ‘Such a development, however, does not take place automatically.’ (12) a.
But that scenario is now thought to contradict observations […].
b. Dieses Szenario scheint jedoch den Beobachtungen zu widersprechen. ‘This scenario, however, seems to contradict the observations.’
In both examples, the respective translator most probably preferred to place aller dings and jedoch (both are roughly equivalent to ‘however’) sentence-internally in order to leave the first position in the sentence to the subject, which presents given information. In some cases, an English-German translator may even achieve an adequate translation of sentence-initial But without using a connective at all: (13) a. But the newfound organism not only appeared to poison fish – it ate them as well!
3. The use of sentence-initial concessive conjunctions
b. Nicht so unser Neuling: Er vergiftet die Fische nicht nur – er frisst sie auch! ‘Not so our newcomer: it not only poisons fish – it eats them as well!’
Sentence-initial concessive conjunctions (SICCs) were chosen for analysis because their different use in English and German is indicative of two interesting contrasts between the two languages. These contrasts account for the fact that SICCs (Engl. But, German Aber and Doch,8 both meaning ‘But’) are much rarer in German than in English (cf. Table 1 below). Accordingly, in English-German translations it is often the case that a German equivalent of sentence-initial But such as Aber or Doch is avoided. Specifically, English-German differences in the use of SICCs are due to contrasts in terms of (1) word order and (2) textual norms. In the following, the two contrasts will be discussed in turn. I. Word order. In contrast to the largely grammatically determined word order of the English sentence (SVO), German word order is pragmatically controlled. That is, German observes a stricter organization of the sentence in terms of given and new information than English (Hawkins 1986; Doherty 2001; 2005). And connectives like but “compete with the propositional parts for the initial position in a sentence” (Doherty 2001: 223). This is why German prefers to use a sentenceinternal connective in many situations where English would use a sentence-initial one. Cf. the following two examples from our corpus.
In (13), the contrast expressed by But has been translated into German solely by means of word order. The topicalization of the negated adverb so has placed new information at the beginning of the clause. This departure from the default given-new ordering of the German sentence gives rise to a strong implicature of contrast (cf. Büring 1999, 2006). The example shows that in the case of sentenceinitial But, omission does not necessarily equal non-translation: the informationstructural expectations of German readers are so strong that a departure from the canonical given-new ordering of the German sentence will suffice to reproduce the effect of sentence-initial But. The use of a direct equivalent like Aber or Doch is often unnecessary. II. Textual norms. Another contrast which accounts for the finding that SICCs are more common in English than in German is located in the domain of culturally determined discourse norms: while English (spoken and written) discourse may be characterized as interactional,9 dialogous and addressee-oriented, German discourse has been described as ‘transactional’, monologous and content-oriented (cf. the literature quoted in Section 1 above and Byrnes 1986; Kotthoff 1989;
9. We follow Thompson and Thetela (1995) in distinguishing interactive from interactional properties of text. While every text is interactive, e.g. in the sense that an author needs to anticipate reader expectations to communicate successfully, a text is only interactional if its interactive quality is explicitly verbalized; for example, possible reader expectations may be ‘hard-wired’ into the text as questions. The term interactional as we use it can therefore be paraphrased as ‘overtly interactive’.
8. But, aber and doch are often called “contrastive” conjunctions. This is somewhat misleading, as only in a minority of cases these conjunctions are actually used to convey a purely contrastive relation. Their default use seems to be the signalling of a concessive relation (Becher 2007; cf. also the following examples, of which only (13) can be analyzed as contrastive).
2nd proofs
Convergence and divergence through language contact in translation
Viktor Becher, Juliane House & Svenja Kranich
Baumgarten & Probst 2004). Among the most important devices that authors can use for “arguing with the reader” (Thompson 2001) are concessive conjunctions (cf. Grote et al. 1997). In our corpus data, we find that especially in sentenceinitial position But is used for the purpose of simulating an interaction between author and reader, an expository strategy which is often not reproduced in German texts by English-German translators, as it is not in line with German textual norms. Specifically, sentence-initial But regularly occurs in two highly interactional discourse patterns (Thompson 2001: 65f), Question–Answer and Claim–Response (cf. Hoey 2001).10 The following two examples illustrate the use of But in the Question part of a Question–Answer pattern: (14) a.
The translator of (14) has decided to retain the question, but has weakened its ‘interactionality’ by refraining from translating the objection marker But. In this way, the translator has shown his or her concern for conforming to German textual norms. The translator of (15) has gone even further; not only has s/he dropped the SICC, but also dispensed with the question of the English original by explicitly verbalizing the lack of knowledge it implies (… is not clear yet). We can say that, in this case, the translator has applied a cultural filter that has completely converted the interactional orientation of the source text into a transactional orientation of the target text. The following two examples illustrate the use of But as a signal of the discourse pattern Claim–Response. While in the Question–Answer pattern But introduces the Question, i.e. the first part of the pattern, in the Claim–Response pattern it usually signals the transition to the second part of the pattern:
But what caused these calamities in the first place […]?
b. Wie kam es zu der erstaunlichen Klima-Instabilität […]? ‘How did the astonishing climate instability come about ?’ (15) a.
(16) a. The public perception is that we have looked extensively for signs of life elsewhere. But in reality, we have hardly begun to search.
But why do these inhibitory events occur during waking in narcoleptics?
b. Für die Öffentlichkeit scheint es, als hätten wir unsere Fühler bereits weit ins All ausgestreckt. Tatsächlich hat die Suche nach außerirdischem Leben aber kaum erst begonnen. ‘To the public, it often seems as if we had already extended our feelers far into the universe. In reality the search for extraterrestrial life has however just begun.’
b. Warum die Bewegungshemmung bei Narkoleptikern zur falschen Zeit auftritt, ist noch nicht klar. ‘Why the movement inhibition occurs in narcoleptics at the wrong time is not clear yet.’
While it is not made explicit whether the question in (14) and (15) is attributable to the author, the reader or even to both parties, it is clear that this way of introducing a new subtopic evokes the impression of an interaction between author and reader. If we take the view that rhetorical questions are attributable to the ‘reader-in-thetext’ (Thompson & Thetela 1995; Thompson 2001) – which need not be identical with any real reader – the use of sentence-initial But makes sense here, as it makes the question seem like an objection to what was said before and thus even heightens the interactional appearance of the text. Another effect of But is that it signals the beginning of the Question–Answer pattern. (The Answer part of the pattern is omitted in (14) and (15).) Finally, the use of sentence-initial (or rather: utteranceinitial) But in this context is strongly reminiscent of face-to-face conversation, where it occurs more frequently than in written texts (Biber et al. 1999: 84).
(17) a. Still, for some it may seem disturbing that life, certainly in its physical incarnation, must come to an end. But to us, it is remarkable that even with our limited knowledge, we can draw conclusions about such grand issues. b. Manchen mag die Vorstellung befremden, daß jedes Leben irgendwann einmal zu Ende geht. Für uns ist es jedoch bemerkenswert, wenn wir überhaupt solche Schlußfolgerungen ziehen können. ‘To some it may be disturbing that every life has its end. To us, however, it is remarkable that we can draw such conclusions at all.’
As in the Question–Answer pattern, the use of But in the Claim–Response pattern of (16) and (17) is reminiscent of spoken conversation, where the discourse marker is commonly used by speakers for claiming the floor even in the absence of a transition-relevance place (Gumperz et al. 1984: 11; Schiffrin 1987). This, in combination with the plural speaker deictic we/us, gives an impression of a high degree of author involvement: the authors in the two examples present the opinion of a more (example. 16) or less (example 17) closely specified group of people as a Claim, with which they contrast their own opinion as a Response. Again, we can observe the tendency of English-German translators to mitigate the interpersonal/ interactional quality of the source text: instead of using sentence-initial Aber,
10. Note that the use of SICCs is not invariably interactional. Rather, it appears that SICCs are put to use very differently in different registers. For example, Bell (2007) finds sentenceinitial But in academic writing predominantly used as “a tool which allows speakers to subtract unintended meanings away from an existing meaning” (Bell 2007: 197). On the other hand, he quotes an example from his corpus where But appears as part of a Question–Answer pattern (no. 28, Bell 2007: 196). A comparative study of the use of sentence-initial But in academic and popular scientific texts seems worthwhile.
2nd proofs
Convergence and divergence through language contact in translation
Viktor Becher, Juliane House & Svenja Kranich
which shares the discourse function of But as a “turn-getter” (Schlobinski 1992), the translators of (16) and (17) have resorted to sentence-internal connectives. Independently of the discourse pattern in which sentence-initial But occurs, it regularly appears in short and pointed sentences. Such sentences typically occur in spoken rather than written discourse and are thus particularly suitable for creating the illusion of spontaneous and interactional speech:
examples 3 and 20 indicate, German aber and doch do occur sentence-initially. However, this happens less frequently than in English texts, and usually entails a strong rhetorical effect. The examples discussed above show that an adequate translation of sentence-initial But into German often means its omission (But → 〈null〉) or its replacement by a sentence-internal connective (But → aber, doch, jedoch, or the like). In line with our project hypothesis, it is hypothesized that
(18) a. If some singular, unavoidable flaw caused every cell to fail eventually, no animal would escape aging. But some do.
1. sentence-initial Aber and Doch occur more frequently in the English-German translations than in the non-translated comparable German texts of our popular science corpus (cf. Section 1), 2. this over-use of SICCs in the English-German translations is due to an interference of the English source texts and 3. the comparable, i.e. non-translated German texts ‘inherit’ over time the more frequent use of SICCs from the translations. 4. It is further hypothesized that the driving force behind these changes is the adoption of the interaction-oriented textual norms of English, i.e. EnglishGerman translators and German authors start using SICCs more frequently in order to imitate the author-reader interaction characteristic of (prestigious) English popular science texts.
b. Gäbe es ihn, würde kein Tier dem Altern entkommen. Solche Lebewesen existieren aber. ‘If [such a flaw] existed, no animal would escape aging. Such creatures exist, however.’ (19) a. In 1990 a new Peruvian “environmental code” finally made the eradication practice illegal. But pesticides remain pervasive. b. Erst 1990 verbot ein neuer peruanischer “Umweltkodex” die Ausrottungs praxis, doch Pestizide beherrschen nach wie vor die Landwirtschaft. ‘Not until 1990 did a new Peruvian “environmental code” forbid the eradication practice, but pesticides are still pervasive in agriculture.’ (20) a. Scientists’ search for life beyond Earth has been less thorough than commonly thought. But that is about to change. b. Die wissenschaftliche Suche danach ist bislang allerdings auch weniger gründlich gewesen als gemeinhin angenommen. Doch das wird sich bald ändern. ‘The scientific search for it has so far however been less thorough than commonly assumed. But that will soon change.’
The remainder of this section will present some evidence in support of these hypotheses. 3.1 Quantitative findings
Of the three Examples, (19) is especially interesting. The translator has merged the two short sentences of the English original into a single sentence, in this way abandoning the colloquial and incremental style of the English original in favor of a more formal and content-oriented style of exposition. (Cf. Fabricius-Hansen 1998, who suggests that English texts tend to ‘build up’ the discourse in an incremental manner while German generally favors a hierarchical and syntactically more compressed mode of writing.) Summing up, the discussion of the above examples has shown that the use of SICCs in German texts is generally dispreferred for (1) information-structural and (2) stylistic reasons.11 We do not want to overstate the case, though: as
Table 1 shows the frequency of sentence-initial But and Aber/Doch respectively in the English source texts, their German translations and the non-translated comparable German texts of the popular science corpus. Since the number of sentences – and hence the number of opportunities to use a SICC – differs between the individual subcorpora, the frequency counts given in the table are normalized on the basis of 1000 sentences.
and has subsequently stayed on this low level until the end of the 20th century. (On the other hand, Doch shows no comparable development.) It would be an interesting topic for further study to assess whether this rather abrupt decline in the use of Aber is part of a more extensive change in German communicative norms around the middle of the 20th century or whether it is an isolated phenomenon (e.g. induced by an emergence of certain prescriptive norms).
11. It is interesting to note that German textual norms regarding the use of SICCs as they are presented here seem to be a fairly recent development. A query in the DWDS-Kerncorpus reveals that the frequency of Aber has decreased by more than 50 percent around the end of the 1950s,
2nd proofs
Convergence and divergence through language contact in translation
Viktor Becher, Juliane House & Svenja Kranich
3.2 Translation analysis
Table 1. Frequency of sentence-inital But in English texts and Aber/Doch in their German translations and comparable German texts (per 1000 sentences; n = 621)
English source texts (But) English-German translations (Aber/Doch) Non-translated German texts (Aber/Doch)
1978–1982
1999–2002
development
32.6 22.7 9.0
32.6 30.1 19.8
– +32.5% +119.8%
Table 2 compares the translational equivalents of sentence-initial But chosen by translators in the earlier time-frame to those chosen in the later time-frame.12 Table 2. Translations of sentence-initial But into German (n = 208)
Let us first consider only the figures for the earlier time-frame (1978–1982). Here, we see that the frequency of But in the English texts is indeed considerably higher than the combined frequency of Aber and Doch in the non-translated German texts (32.6 vs. 9.0 occurrences per 1,000 sentences). The English-German translations are almost exactly in the middle, which could mean that they are (1) affected by a “shining-through” (Teich 2003) of the English source texts and (2) only partly follow the textual conventions of German (at least as far as the use of SICCs is concerned). In any case, hypothesis 1 is supported by the frequency counts. Taking the later time-frame (1999–2002) into account, we can observe two remarkable diachronic developments. First, the number of SICCs in the EnglishGerman translations has risen by 32.5 percent, which indicates that the degree of source language shining-through has increased. Second, the non-translated texts seem to have followed suit: here, the frequency of SICCs has more than doubled (+119.8%), almost reaching the frequency level of the translations in the earlier time-frame. Summing up the results obtained from the frequency counts, it appears to be the case that the English-German translations have – at least in terms of frequency – aligned their use of Aber and Doch to the use of But in their English source texts. And the non-translated German texts, in turn, appear to have followed the lead of the English-German translations, which indeed seem to function as role models for German popular science texts (cf. hypothesis 4 in the previous section, and cf. also Baumgarten 2008). Thus, the frequency counts presented in Table 1 make an English influence on German textual norms via translation seem very likely. However, two questions must be answered before we can think about accepting this explanation:
1978–1982
1999–2002
development
But → Aber, Doch
26.8%
47.4%
+20.6
But → aber, (je)doch, allerdings
46.4%
36.2%
–10.2
But → 〈null〉
17.9%
7.2%
–10.7
But → 〈other〉
8.9%
9.2%
+ 0.3
total
100%
100%
A definite trend is evident: while the number of ‘literal’ translations (But → Aber, Doch) has almost doubled (+20.6 percentage points), the number of ‘free’ translations (But → aber, (je)doch, allerdings and But → 〈null〉) has decreased by an equivalent amount (–10.2 and –10.7 percentage points respectively). The results indicate that the first question posed at the end of the last subsection is to be answered affirmatively: the frequency rise of Aber and Doch in the EnglishGerman translations is due to a change in translation behavior. The table shows that over time, translators hesitate less and less to adopt sentence-initial But as Aber or Doch. And as the following examples illustrate, the changed translation behavior seems indeed to result from – quite successful – attempts to imitate the interactional style of the English source texts: (21) a. Something must have drained away its [sc. the sun’s] angular momentum. But what? b. Irgendwann muß sie den Drehimpuls verloren haben. Aber wie? ‘At some point it must have lost the angular momentum. But how?’ (22) a. … the entire endeavor of robotics has failed rather completely to live up to the predictions of the 1950s […] It is true that industrial robots have transformed the manufacture of automobiles, among other products. But that kind of automation is a far cry from the versatile, mobile, autonomous creations that so many scientists and engineers have hoped for.
1. Is the frequency rise of Aber and Doch in the English-German translations actually due to an increased shining-through of the English source texts and an accompanying adoption of their textual norms? 2. Do the non-translated German texts in their use of Aber and Doch actually follow the example of the English-German translations or have they developed idiosyncratic uses of the two conjunctions?
12. Occurrences of But in citations and in sentences that were left out in the German translation were excluded from the analysis (cf. Note 5).
The questions will be answered in the following two subsections.
2nd proofs
Convergence and divergence through language contact in translation
Viktor Becher, Juliane House & Svenja Kranich
b. Gewiß, Industrieroboter spielen in der Produktion von Autos und anderen Gütern inzwischen eine wesentliche Rolle. Aber die sind weit entfernt von dem, was so viele Forscher und Ingenieure sich erträumt hatten. ‘Certainly, industrial robots play an essential role by now in the production of cars and other goods. But they are far away removed what so many researchers and engineers have dreamed of.’
(24) Keine Region auf unserem Planeten ist so entlegen und fern jeder menschlichen Zivilisation wie die Polargebiete. Nirgendwo sonst sollte also die Luft so rein und der Himmel so klar sein. Doch weit gefehlt! ‘No region on our planet is as remote and far from every human civilization as the pole areas. Nowhere else should the air be so clean and the sky so clear. But far from it!’
In (21) the short, elliptical question prefaced by the objection marker But gives rise to a high degree of interactionality and author-reader involvement. The author creates the illusion of following a train of thought together with the reader. It is interesting to note that the translator did not mechanically transfer the two sentences into German: he has changed the subject of the first sentence (something → sie [the sun]) and found an equivalent metaphor for expressing the event described (drained away → verloren). This shows that the retention of But as Aber is the result of a conscious decision, namely not to apply a cultural filter. (22) exemplifies what Antaki and Wetherell have aptly described as “making a show of conceding” (1999: 7). The author pretends to concede a possible objection (Claim) by the reader, even emphasizing its truth (It is true that …). In the immediately following Response (introduced by But), however, the author completely marginalizes the ‘conceded’ objection by showing that it is not at all relevant to her argument that “the entire endeavor of robotics has failed rather completely to live up to the predictions of the 1950s”. What we observe here is thus a hardly concealed attempt to persuade the reader, who is pictured as (possibly) objecting to the argument of the author. As in (21), the translator apparently does not feel the need to mitigate the overtly expressed author-reader-interaction and almost literally renders both It is true that … (as gewiß ‘certainly’) and the objection marker But (as Aber) in German.
(25) … eine Folge von sehr vielen Einzelanweisungen […] Erst deren Ausführung in ihrer Gesamtheit erweckt den geschilderten Eindruck. Aber wer ist es, der diese Einzelanweisungen ausführt? ‘… a sequence of very many single orders. Only their execution in their entirety gives rise to the described impression. But who is it that carries out these single orders?’
Like (22) above, (23) and (24) may be described as ‘show concessions’ following a Claim–Response pattern. Note that the use of gewiß (‘certainly’) is exactly parallel to the use of the adverb in the translation of (22). This is remarkable because gewiß in this function is extremely rare in contemporary German, as a search for this adverb in the 1990–2000 time-frame of the DWDS-Kerncorpus confirms. (24) is a particularly interesting case of a show concession. The author makes use of the discourse pattern Hypothetical–Real (Winter 1994), a variant of the Claim– Response pattern (Hoey 2001: 179, 188): he deliberately leads the reader up the garden path by incrementally drawing the false picture of unpolluted pole areas. The short, elliptical Doch-sentence then destroys this illusion. Finally, (25) features Aber as an objection marker prefacing a rhetorical question (cf. Examples 14, 15 and 21 above). The interactional (and incremental) mode of presenting information achieved by means of sentence-initial Aber and Doch evidenced in Examples 23 through 25 is (or used to be?) highly untypical of German texts. The findings of the qualitative analysis point to an influence of English-German translations on original German popular science texts. We can thus argue that the increase in frequency of Aber and Doch in these texts is the result of an adoption of Anglophone communicative norms, which have been introduced to the genre of popular science via translations from English.
3.3 Qualitative analysis of non-translated German texts The question that remains to be answered is whether the frequency increase of Aber and Doch in the non-translated German popular science texts in fact results from an adoption of textual norms introduced by the English-German translations. A qualitative analysis of all occurrences of sentence-initial Aber and Doch in the non-translated texts indicates that this is indeed the case. The following examples are representative of the findings.
4. Conclusion
(23) Gewiss sind unter den Lockstoffen für bestäubende Insekten auch Monoterpene. Aber das kann nur ein Nebeneffekt sein… ‘Certainly, among the attractants of pollinating insects are also monoterpens. But that can only be a side-effect…’
Both case studies presented in this paper confirm the findings of previous contrastive pragmatic studies regarding different communicative preferences in English and German texts by House (cf. e.g. 1996; 1997; 2006). The use of epistemic modal
2nd proofs
Convergence and divergence through language contact in translation
Viktor Becher, Juliane House & Svenja Kranich
While in regard to the former linguistic elements, bilingual speakers, and thus also translators, can easily establish the elements as equivalents of each other, the fundamental differences in the linguistic inventory of expressions of epistemic modality in the two languages effectively block such a process in this domain. We could see that English mainly uses a set of highly grammaticalized modals, while German tends to mark epistemic modality with a variety of different elements, some of which (the modal particles) do not exist in English at all. One can therefore assume that no clear one-to-one equivalence relations are apparent to the bilingual speaker/ translator. Such a lack of a transparent equivalence relation can be assumed to prevent convergence towards source language conventions. This would confirm Teich’s (2003: 211–218) findings that typological characteristics of the languages involved have a clear impact on the kind and degree of source-language induced interference phenomena in translations. Of course, this general hypothesis will profit from further investigations into the question of convergence and divergence of textual conventions in translation.
markers, particularly those of low modal strength, which can be associated with less direct and more addressee-oriented rather than content-oriented discourse styles, was shown to be more typical of English than of German popular scientific texts: German translators tend to omit these markers or replace expressions of low modal strength with markers of high modal strength. And the use of the SICCs But, Aber and Doch, which serve to make a text more interactional, was also shown to be originally more associated with English textual conventions. However, as regards our hypothesis that English text conventions have an increasing influence on German communicative preferences, the results of the two case studies presented here differ greatly. In the case of modal markers, the German covert translations show a continuous adaptation to German communicative preferences: epistemic modal markers are left out so that unmodified propositions can take the place of modified ones, and markers of low modal strength (which make the text most indirect and least content-oriented) are frequently replaced by modal markers of higher strength. And both these tendencies are apparent in both time-spans, 1978–1982 and 1999–2002, without any clear indication of changing preferences. In regard to the SICCs investigated in the present study, on the other hand, we found a clear trend towards a decreasing application of a cultural filter. In the second time-span, the German translations of sentence-initial But quite commonly also employ a SICC (Aber and Doch), while translations in which adaptations to German communicative preferences, such as the rendering of the contrast through word order inversions, were undertaken, are used much less frequently compared to the earlier time-span. Thus, we were able to establish an increasing convergence of German covert translations towards the model represented by the (prestigious) English source texts. Moreover, we found that the Anglophone text conventions introduced by the English-German translations are also adopted by authors of originally German texts. Anglophone text norms were thus shown to find their way into the German genre of popular scientific writing. Other studies carried out within the project ‘Covert Translation’ have pointed to similar convergence phenomena, as pointed out in the introduction. For instance, in the use of the personal pronouns we-wir a certain impact of the English preferences on the German covert translations was established (Baumgarten & Özçetin 2008). And similar to what we reported here on the use of SICCs, in the use of sentenceinitial And-Und, convergence towards the English model also extended to German comparable texts produced without an English model (Baumgarten 2007). The main difference between elements such as we-wir, And-Und and ButAber/Doch and the epistemic modal verbs, where we found no signs of increasing convergence, lies in the different degrees of perceivable form-function equivalence.
Acknowledgements This paper has profited from many stimulating discussions we had with Volker Gast while he was visiting scholar at our research centre. Furthermore, we would like to thank the two anonymous reviewers for their helpful comments.
References Antaki, C. & Wetherell, M. 1999. Show concessions. Discourse Studies 1(1): 7–27. Baumgarten, N. 2007. Converging conventions? Macrosyntactic conjunction with English and and German und. Text & Talk 27: 139–170. Baumgarten, N. 2008. Writer construction in English and German popularized academic discourse: The uses of we and wir. Multilingua 27(4): 409–438. Baumgarten, N. & Probst, J. 2004. The interaction of spokenness and writtenness in audience design. In Multilingual Communication [Hamburg Studies on Multilingualism 3], J. House & J. Rehbein (eds), 63–86. Amsterdam: John Benjamins. Baumgarten, N. & Özçetin, D. 2008. Linguistic variation through language contact in translation. In Language Contact and Contact Languages [Hamburg Studies on Multilingualism 7], P. Siemund & N. Kintana (eds), 293–316. Amsterdam: John Benjamins. Baumgarten, N., Meyer, B. & Özçetin, D. 2008. Explicitness in translation and interpreting: A critical review and some empirical evidence (of an elusive concept). Across Languages and Cultures 9(2): 177–203. Becher, V. 2007. Die pragmatischen Funktionen von BUT und ABER im Vergleich. Eine Untersuchung anhand von englischen populärwissenschaftlichen Texten, ihrer deutschen Übersetzung und deutschen Vergleichstexten. MA thesis, University of Hamburg. Bell, D. 2007. Sentence-initial And and But in Academic writing. Pragmatics 17(2): 183–201.
2nd proofs
Convergence and divergence through language contact in translation
Viktor Becher, Juliane House & Svenja Kranich
Biber, D., Johansson, S., Leech, G., Conrad, S. & Finegan, E. 1999. Longman Grammar of Spoken and Written English. London: Pearson Education. Bublitz, W. 2003. Emotive Prosody: How attitudinal frames help construct context. In Anglistentag 2002. Proceedings, E. Mengel, H.-J. Schmidt & M. Steppat (eds), 381–391. Trier: WVT. Bührig, K. & House, J. 2004. Connectivity in translation: Transitions from orality to literacy. In Multilingual Communication, J. House & J. Rehbein (eds), 87–114. Amsterdam: John Benjamins. Bührig, K. & House, J. 2007. Linking constructions in discourse across languages. In Connectivity in Grammar and Discourse, J. Rehbein, C. Hohenstein & L. Pietsch (eds), 345–366. Amsterdam: John Benjamins. Büring, D. 1999. Topic. In Focus – Linguistic, Cognitive, and Computational Perspectives, P. Bosch & R. van der Sandt (eds), 142–165. Cambridge: CUP. Büring, D. 2006. Intonation und Informationsstruktur. In Text – Verstehen. Grammatik und darüber hinaus, H. Blühdorn, E. Breindl & U.H. Waßner (eds). Berlin: de Gruyter. Byrnes, H. 1986. Interactional style in German and English conversations. Text 6: 89–106. Clyne, M. 1987. Cultural differences in the organization of academic texts. Journal of Pragmatics 11: 211–47. Clyne, M. 1994. Inter-Cultural Communication at Work: Discourse Structures across Cultures. Cambridge: CUP. Coates, J. 1995. The expression of root and epistemic possibility in English. In Modality in Grammar and Discourse, J. Bybee & S. Fleischmann (eds), 55–66. Amsterdam: John Benjamins. Crismore, A. & Farnsworth, R. 1990. Metadiscourse in popular and professional science discourse. In The Writing Scholar: Studies in Academic Discourse, W. Nash (ed.), 118–136. Newbury Park CA: Sage. Denison, D. 1993/2004. English Historical Syntax: Verbal Constructions. London: Longman. Diewald, G. 1999. Die Modalverben im Deutschen. Grammatikalisierung und Polyfunktionalität. Tübingen: Niemeyer. Doherty, M. 2001. Discourse relators and the beginnings of sentences in English and German. Languages in Contrast 3 (2): 223–251. Doherty, M. 2005. Topic-worthiness in German and English. Linguistics 43: 181–206. Fabricius-Hansen, C. 1998. Informational density and translation, with special reference to German – Norwegian – English. In Corpora and Cross-Linguistic Research: Theory, Method, and Case Studies, S. Johansson & S. Oksefjell (eds), 197–234. Amsterdam: Rodopi. Geyken, A. 2007. The DWDS corpus: A reference corpus for the German language of the 20th century. In Idioms and Collocations: Corpus-Based Linguistic and Lexicographic Studies, C. Fellbaum (ed.), 23–40. Birmingham: Continuum. Graefen, G. 2007. ‘Hedging’ als neue Kategorie? Ein Beitrag zur Diskussion. . (accessed March 2008) Grote, B., Lenke, N. & Stede, M. 1997. Ma(r)king concessions in English and German. Discourse Processes 24: 87–118. Gumperz, J.J., Kaltman, H. & O’Connor, M.C. 1984. Cohesion in spoken and written discourse: ethnic style and the transition to literacy. In Coherence in Spoken and Written Discourse, D. Tannen (ed.), 3–20. Norwood NJ: Ablex. Halliday, M.A.K. & Matthiessen, C.M.I.M. 2004. An Introduction to Functional Grammar, 3rd edn. London: Arnold. Hawkins, J. 1986. A Comparative Typology of English and German: Unifying the Contrasts. London: Croom Helm. Heine, B. & Kuteva, T. 2005. Language Contact and Grammatical Change. Cambridge: CUP.
Hoey, M. 2001. Textual Interaction: An Introduction to Written Discourse Analysis. London: Routledge. House, J. 1977. Translation Quality Assessment. A Model for Analyis. Tübingen: Narr. House, J. 1996. Contrastive discourse analysis and misunderstanding: The case of German and English. In Contrastive Sociolinguistics, M. Hellinger & U. Ammon (eds), 345–361. Berlin: Mouton de Gruyter. House, J. 1997. Translation Quality Assessment. A Model Re-visited. Tübingen: Narr. House, J. 2004a. Explicitness in discourse across languages. In Neue Perspektiven in der Übersetzungs- und Dolmetschwissenschaft, W. Koller, J. House & K. Schubert (eds), 185–208. Bochum: AKS. House, J. 2004b. Linguistic aspects of the translation of children’s books. In Übersetzung –Translation – Traduction: An International Encyclopedia of Translation Studies, H. Kittel, A.P. Frank & N. Greiner (eds), 683–697. Berlin: de Gruyter. House, J. 2005. Politeness in Germany – Politeness in Germany?. In Politeness in Europe, L. Hickey & M. Stewart (eds), 13–29. Clevedon: Multilingual Matters. House, J. 2006. Communicative styles in English and German. European Journal of English Studies 10(3): 249–267. House, J. 2008. Beyond intervention: Universals in translation? trans-kom. Zeitschrift für Translationswissenschaft und Fachkommunikation 1(1): 6–19. Hunston, S. & Thompson, G. (eds). 2000. Evaluation in Text: Authorial Stance and the Construction of Discourse. Oxford: OUP. Hyland, K. 1998. Boosting, hedging, and the negotiation of academic knowledge. Text 18: 349–382. König, E. 2001 Kontrastive Analysen Deutsch-Englisch: Eine Übersicht. In Deutsch als Fremdsprache: Ein internationales Handbuch, Vol. I, G. Helbig et al. (eds), 324–330. Berlin: de Gruyter. Kranich, S. 2009. Epistemic modality in English popular scientific texts and their German translations. trans-kom. Zeitschrift für Translationswissenschaft und Fachkommunikation 2(1): 26–41. Kotthoff, H. 1989. Pro und Kontra in der Fremdsprache. Frankfurt: Peter Lang. Lyons, J. 1977. Semantics. Cambridge: CUP. Nehls, D. 1989. German modal particles rendered by English auxiliary verbs. In Sprechen mit Partikeln, H. Weydt (ed.), 282–292. Berlin: de Gruyter. Özçetin, D. 2008. Die Versprachlichung mentaler Prozesse in englischen und deutschen Wirtschaftstexten. Arbeiten zur Mehrsprachigkeit/Working Papers in Multilingualism 88. (Hamburg: Sonderforschungsbereich Mehrsprachigkeit). Quirk, R., Greenbaum, S., Leech, G. & Svartvik, J. 1985. A Comprehensive Grammar of the English Language. London: Longman. Schiffrin, D. 1987. Discourse Markers. Cambridge: CUP. Schlobinski, P. 1992. Funktionale Grammatik und Sprachbeschreibung: Eine Untersuchung zum gesprochenen Deutsch sowie zum Chinesischen. Opladen: Westdeutscher Verlag. Smith, C. 2003. Modes of Discourse. Cambridge: CUP. Teich, E. 2003. Cross-linguistic Variation in System and Text. A Methodology for the Investigation of Translations and Comparable texts. Berlin: de Gruyter. Thompson, G. 2001. Interaction in academic writing: Learning to argue with the reader. Applied Linguistics 22(1): 58–78. Thompson, G. & Thetela, P. 1995. The sound of one hand clapping: The management of interaction in written discourse. Text 15(1): 103–127. Weinreich, U. 1953. Languages in Contact. The Hague: Mouton.
2nd proofs
Convergence and divergence through language contact in translation
White, P.R.R. 2003. Beyond modality and hedging: A dialogic view of the language of intersubjective stance. Text 23: 259–284. White, P.R.R. & Sano, M. 2006. Dialogistic positions and anticipated audiences: A framework for stylistic comparisons. In Pragmatic Markers in Contrast, K. Aijmer & A.-M. Simon-Vandenbergen (eds), 189–214. Amsterdam: Elsevier. Winter, E. 1994. Clause relations as information structure: Two basic text structures in English. In Advances in Written Text Analysis, M. Coulthard (ed.), 46–68. London: Routledge. Zifonun, G., Hoffmann, L. & Strecker, B. 1997. Grammatik der deutschen Sprache, Vol. 3. Berlin: de Gruyter.
2nd proofs
