Abstract

In order to find varieties in a target language corresponding to dialects in the source language, I suggest methods of perceptual dialectology. First of all, the target group has to prove its knowledge of the dialects in question by mental mapping. Furthermore, it is important, which social schemes are triggered by using the different varieties. At least, the speech forms evoke holistic impressions which can be visualized by drawings. Along with this, feedback to the pictures of a movie is given. The case study on “Manitou’s Shoe” (Germany 2001) with its synchronization in English and Russian will demonstrate how these methods work[url=#n1][1][/url].

Keywords

German varieties, Russian varieties, English varieties, Linguistic stereotype, social stereotype, perceptual dialectology, mental map

1. Introduction

Usually, varieties of any language employed in movies are not dubbed. The main question is whether it is necessary - and if necessary, whether it is in general possible - to translate, or better, to transpose (cf. R. Jacobson 1959: 197 “creative transposition”) a dialect from the original language into a variety of the target language. The most evident reason may be that very little is known about the varieties of both, the target language (language index 2) and the original variety (language index 1), except perhaps a vague impression, and therefore it seems to be an unsolvable problem to find an adequate transposition apart from a translation into standard variety with perhaps a few characteristic lexical items. Therefore, to make a virtue of necessity, the use of standard variety goes along with best comprehensibility, that is to say, speech can be recognized easily, clearly and fast, and a production gains the widest average audience, albeit the widespread use of standard language is like cleaning up the language (cf. Nadiani 2004: 59). Even the attempt to find actors of a certain variety is not as easy as one might think; the directors of the famous German TV-serial “Tatort” took great effort to employ characters speaking the local variety1 of Cologne, but they were not successful (cf. Koch 2008). Because of that, the problem of transposition is not solved by “just” finding an adequate variety - this should be difficult enough - furthermore, when this first hurdle is cleared, one has to engage suitable actors/speakers for the variety2 desired. 

To cope with these difficulties, I suggest to use evidence given by folk linguistics, that is to say, to interpret attitudes of ordinary speakers towards dialects and their knowledge about dialects in general with a set of methods elaborated in perceptual dialectology (drawing of mental maps and componential analysis). But first of all I want to justify the indispensable necessity for transposing varieties by discussing the social implication of dialects and their correlation to stereotypes. A case study (Manitou’s Shoe, Germany 2001) shall show, how these methods are applied in finding Russian varieties2 matching the German varieties1.

2. Social and linguistic stereotypes

The most evident argument in favour of varieties was outlined by Th. Herbst (1994: 92):

Der Akzent von Sprechern kann [...] Klischeevorstellungen auslösen, die sich auf ihren Charakter beziehen.

However, the thesis, that accents evoke clichés which modify the character of a figure does not mean that a character has to correspond to the cliché in every case, but there may be a play with clichés to build up a complex figure. Considering this point of view, the existence of varieties shows far-reaching consequences for dubbing:

Für die Synchronisation ergibt sich daraus, daß sie [...] der Tatsache Rechnung tragen muß, daß im Originalfilm Akzent bzw. Dialekt nicht nur Aufschluß über regionale Herkunft, soziale Stellung und Rasse gibt, sondern daß mit bestimmten Varietäten auch Stereotypen verbunden werden, was auch filmisch genutzt werden kann. (Herbst 1994: 93)

The question is, whether the implications of varieties employed in movies are intended to design characters, and it has to be taken into account that there will result a modified character by using different varieties. At this point translation works hand in glove with interpretation.

The lexeme “stereotype” as used in the common language refers to a one-sided view of a subject, primarily drawing attention to the negative features involved. As a scientific term, “stereotype” was coined in the social sciences, first by W. Lippmann 1922:

For the most part we do not first see, and then define, we define first and then see. In the great blooming, buzzing confusion of the outer world we pick out what our culture has already defined for us, and we tend to perceive that which we have picked out in the form stereotyped for us by our culture. (Lippmann 1946: 61)

A stereotype works as a cognitive scheme, reducing the complexity of the perceptual recognized reality to socially relevant facts. In this way, it results in order by abstraction and allows identity where still multitudinous differences at a lower level exist, and this is essential in contrast to other social groups with different stereotypical features. This is due to the conventional, relatively fixed structure of a stereotype. This definition by no means refers to primarily negative features, it is in general a cognitive categorization of complex entities. In linguistics as well as in social sciences, this term often is used in the negative sense: a stereotype is a remarkably simple categorization, generalizing in an inadequate way, unjustly using negative values (cf. Konerding 2001: 153). To distinguish this negative stereotype from the neutral one it is called prejudice[url=#n2][2][/url], or to follow the terms suggested by Konerding (2001), limitation-stereotype (Limitationsstereotyp), in contrast to basic-stereotype (Basis-Stereotyp). Limitation-stereotypes are simplifying evaluative categorizations, they cause identity in in-groups by excluding outgroups. Additionally, the members of outgroups seem to be more similar than the members of the in-group (outgroup-homogeneity effect, cf. Kristiansen 2003: 81). Basic-stereotypes have a quite different function: they allow a quick orientation of speakers in different situations. For my further argumentation two main characteristics will be important: basic-stereotypes are useful in evaluating complex, but conventional situations very quickly, and stereotypes are holistic entities, the features are amalgamated and therefore not recognized as single features. Furthermore, limitation-stereotypes are essential for in-group and outgroup differentiation.
In linguistics, stereotype is a term mainly used in semantics and pragmatics, but Kristiansen (2003) draws attention to stereotypes on a phonetic level, referring to a relationship between accents and society, for which terms like “indexical feature” (Abercrombie 1967) or “social deixis” (Levinson 1979) are in use (cf. Kristiansen 2003: 71). Typical phonetic features of a dialect are known as shibboleths, and these shibboleths are evoking linguistic stereotypes, which Honey defines as

popular and conscious but imprecise general characterizations of the speech forms of particular social groups (Honey 1997: 99).

In this way, linguistic stereotypes operate like basic-stereotypes, as

cognitive reference points, as abstractions relative to which speaker and hearer may locate each other and position themselves (Kristiansen 2003: 82).

Limitation-stereotypes result when speakers of other varieties exaggerate conspicuous features to caricature a dialect.

Concerning the topic of this essay, dialects and accents have social implications, and these implications are not triggered by lexical items only, but mainly by phonetic features. The non-translation of these features effect a social change of the role model. Nadiani (2004: 57-58) shows three ways of employing dialects in movies: the imitation of dialect, the imitation of dialect with clichés, and the use of genuine dialects. The first two possibilities are artificial creations of dialects and therefore refused by Nadiani. However, linguistic stereotypes as shown in media also have advantages: the few linguistic features allow a quick categorization of the speakers social origin, they act as a trigger. A dialectal speech form does not follow the formula “the more dialectal, the more typical”, even a few features are sufficient to recognize a dialect (cf. Hundt 1992:80). Artificial dialects in multimedia, i.e. multimedialects (cf. Koch 2008), can on the one hand result from reducing genuine dialects, they are then normalized speech forms, or on the other hand they are standard language enriched with some frequent dialectal features. Because the main interest in communication is focused on the meaning of speech, not on the form, a few, even one or two very frequent features are sufficient to indicate regional and social origin, for instance

Missingsch: combination of the consonants st, sp
Franconian: spirantization of g to j
Upper Saxon: centralization of the full vowels, lenition of the fortis consonants
Swabian: s-palatalization in the inflection of the 2sg., reduction of unstressed syllables to schwa
Bavarian: back a, voiceless s. 

Linguistic stereotypes are easy to create, have a wide range and are, in most cases, recognized as artificial only by speakers, who know this dialect in its genuine variety. A contrast to a non-dialect speaking counterpart enforces the impression - as a holistic entity - that the other part speaks at least a kind of dialect related to a well-known dialect scheme.

3. Perceptual dialectology

When confronted with the problem of dialect dubbing, the question of which variety to choose is the first to arise: a genuine dialect with many features on all linguistic levels differing from standard language is most true to create authentic characters who fit into the setting, but this dialect will be hardly understood by non-members of this speech community. Therefore, the primary function of speech, the transfer of information and acting by speech is jeopardized and especially in movies the single scenes have to be tightened by the spoken word . And returning to the thesis that schemes are perceived as a whole with the single features amalgamated, the question is whether emotions, attitudes towards the dialects are included in this entirety, triggered by the acoustic character of the linguistic features. Therefore, I think that two factors are crucial and object to perceptual linguistics:

1. What varieties are well known by ordinary speakers, can they be allocated to geographical and social dimensions?

2. How are the different varieties perceived, what emotional evocations are triggered?

The two methods to find an answer are to draw a mental map and to connect an auditory impression with a picture. Additionally, the amalgamated features of a speech impression may be identified by using a componential analysis.

3.1. Drawing mental Maps

When a speech form is recognized, especially a genuine dialect, the hearer will allocate it in geographic dimensions. This allocation implies social categorizations. To test the knowledge about dialects, hearers of different geographic origin are asked to draw the region where they think the dialect in question is spoken on a map which only shows geographic features and towns but no politic or linguistic borderlines (cf. Kelle 2000, Auer 2004, Preston 2002). P. Auer describes the result in a centre-periphery-model, leaving undefined gaps on the map, “white spots”. The concentric circles show in most cases a well-known city of this area in its centre (cf. Auer 2004: 152-153), and adding all drawings of the interviewees in one overlay, a picture emerges which shows the region thought as typical for the speech form by all interviewees. I think that for dubbing those dialects are most suitable for which ordinary speakers from different regions are able to establish a clear correlation between the dialect and a region.

3.2. Evocation of pictures by speech forms and component analysis

The idea that speech forms may correlate with pictures arises from psychological sciences, a discipline called Gestalt psychology. It was Köhler in 1929 who first established a relation between certain sound patterns and visual form. Famous is the resemblance of the nonsense words “takete” and “maluma” with angular and round forms: each sound evokes a visual impression, and interviewees asked to assign the sounds to a given form do not hesitate to relate “takete” with the jagged lines, “maluma” with the bended, rounded lines (cf. Köhler [1929] 1947: 224-225). Berthele (2006) transfers this knowledge to German dialects, and he asks the question: “Wie sieht Berndeutsch so ungefähr aus?” (What would Bernese German approximately look like?). He offers twelve different drawings made by the Swiss artist Andreas Gerber (Bern, Swiss) with the aim to create a broad selection of visual patterns to evoke speech stimuli, but not with a certain variety in mind. Then, a test with 45 students originating from all regions in Switzerland, showed, that certain Swiss dialects, standard German, other well-known German varieties (Bavarian, Swabian) and even other national languages (English, French, Italian) show - more or less - clear distinctions when assigned to a drawing. For example, standard German is related by 43% to a sharp angular picture, by 17 % to a picture showing sharp arrows. This may lead to the conclusion - by using the jagged and rounded drawings of Köhler as prototypes - that standard German resembles more the impression of “takete” than Bavarian, which is closer assigned to forms similar to “maluma”.

Additionally, the speakers can describe the attitudes towards the speech forms by a choice of adjectives like ‘soft, smooth’ or ‘sharp, cold’, and by this, the amalgamated features of the holistic impression are visualized. In the case study of “Manitou’s Shoe” I want to show how this data can be used in a component analysis.

4. Case study: Manitou’s Shoe (Germany 2001)

In 2001, a western-parody called “Manitou’s Shoe” was created by Michael Bully Herbig, and this movie was to become the most successful German movie ever, awarded three times with the “Deutscher Filmpreis”. In this comedy, different varieties of the German language are employed and distributed to the characters, and it is obvious that the speech forms are essential parts to the design of the figures. The main characters are presented in a black and white manner: The “bad guy” Santa Maria speaks very clear standard German whereas the two good guys, the cowboy Ranger and the Indian Abahachi, are portrayed by actors speaking a kind of reduced Bavarian regional vernacular - of course, this is part of the parody. Additionally, the twin brother of the Indian called Winnetouch is gay and speaks the same vernacular with an emphatic intonation and nasalized voice. The friend of the protagonists, a Greek, speaks a kind of artificial ethnolect used by migrants of the first generation. These varieties trigger stereotypes, holistic schemes together with the appearance of the figures: the bad guy wears a black suit, in the first scene his eyes are covered by black sun glasses and heavy rock music underlies the action. The gay Indian Winnetouch wears pink clothes, his gestures are campy. Abahachi and Ranger are dressed like a typical Indian and cowboy, the Bavarian dialect associates them with the countryside, they speak and act in a casual way. In a holistic view, Santa Maria evokes a cold and hostile feeling, the other actors leave a warm and friendly impression, and this is confirmed throughout the whole movie by their actions. By choosing these speech forms, the schemes are kept alive, in case of the Bavarian variety the movie as a parody is kept in mind. Therefore, in dubbing, one has to consider suitable equivalents in the target language with regard to “translating” the function of triggering stereotypes, in order to sustain the designed character.

4.1 The English Synchronization

Taking a look at the English synchronization, the varieties employed are indeed parallels to the German varieties concerning the function transported: The bad guy is speaking with a British English accent (Received Pronunciation) while the good guys are speaking General American English, or the so called Network English. The effects are the same as caused by the German varieties, standard English sounds - especially in contrast to the other varieties - cold and clear, emotionless, but the Network English works along with a casual behaviour. Some linguistic stereotypes are (cf. Trudgill/Hannah 2002:35-42):

General American English in contrast to Received Pronunciation (RP):

Consonants:

[r]: greater retroflexion than in RP
rhoticity: pronunciation of postvocalic /r/, (card, bird, car); RP is non-rhotic
[t]: flapped consonants: [t] , [nt] → [d̯ ]
[l]: ‘dark’ / l / in all positions: [ɫ]; RP: allophonic differentiation [l] - [ɫ]

Vowels:

/a/ → [æ]: before voiceless fricatives (path, grass), before /nt/, /ns/, /nd/, /ntʃ/, /mp/

The possibility to distinguish, recognize and allocate these two varieties is given, and therefore the English synchronization compared to the German original works in the intended way. However, at least no regional dialects are employed, but national varieties, and the question arises, whether no genuine dialects can be found. One scene plays with a parallel in location of dialects: The barkeeper says “Ihr müsst Ranger sein! Der Mann mit dem Südstaatenslang!” (‘You must be Ranger, the man with the slang from the Southern States’; synchronized version: “You gotta be Ranger, right? I was hoping to meet you one day.” Min. 39:46). The Southern States of the USA, where the action is located, is paralleled to the “German” South due to the use of Bavarian dialects, which means Bavaria and Austria. This parodic play is not translated, obviously because Ranger actually does not speak this slang but Network English. Here, the results from perceptual dialectology can be brought in: As Preston (2002) shows with the method of drawing mental maps, the south-eastern part of the United States is the best known area associated with a different dialect, and this dialect is perceived as the most dialectal one, it corresponds the least with a correct standard variety. This Lower Southern accent bears special linguistic features:

Special Features of the Lower Southern Accent:

Lower Southern accents are non-rhotic
‘Breaking’: [ə] – offglide in stressed monosyllables (bad [bæijəd])
/ai/ as monophthong [a:]  (high [ha:])
/ɪ/ = /ɛ/ before a nasal consonant (pin = pen)
verbforms wasn’t, isn’t: [z] → [d]  ([ɪdnt] , [idn])

These few but frequent features could be used to enrich the speech of Ranger; then an allocation to the South is possible. But does it make sense to trigger socio-cultural stereotypes of the Southern States with linguistic features? One can achieve an alienation as well as with the Bavarian dialect: the Afro-American population of the Lower South is, in its ethnicity, as exceptional as the Indian population to which Ranger is associated by sharing the same speech form with Abahachi. A parodic play could be initiated like Ali G does, when pretending to be black (speech form, clothes), but obviously being a white man (cf. Sebba 2003: 281), which is as strange as speaking Bavarian in the Wild West, but on a different level. But there is no evidence that the dialect of the Lower South will be understood, recognized and then trigger social schemes in other English speaking countries, e.g. Australia, and, perhaps to ensure a wide range, this alienation effect was avoided - or quite simply, nobody considered this solution. On the whole, the English synchronization is well done, even Santa Maria’s accomplice was carefully supplied with a Mexican English variety.

4.2 The Russian synchronization

In the Russian synchronization no dialects are employed, all characters are speaking standard Russian. This means that no schemes are evoked by different varieties, so that a fundamental process, the transfer of the Indians into another culture, does not occur: the alienation by speech variation, which is a fundamental requirement to create the parody, is neither used to provoke a first laugh nor to assign different schemes to the characters maintained throughout the whole movie.

The question is, whether there was no other way to translate the speech forms, or whether there were no methods to find adequate varieties. I assume that there are Russian varieties to be employed in the movie and that the methods to find suitable vernaculars are on the one hand mental mapping and on the other hand the evocation of holistic impressions by pictures, supported by a description in words.

The Russian people are conscious of different regional speech forms. One of the most prominent examples is the former president of the Soviet Union, Mikhail Gorbachev, whose speech is coloured by features belonging to the South of Russia, and this way of speaking was successfully used to parody Gorbachev [url=#n3][3][/url]. In order to find vernaculars to be employed in the synchronization of “Manitou’s Shoe”, one has not only to look out for different vernaculars, but for vernaculars well known by the target group. Because this movie firstly aims at young people, it was the students knowledge about Russian dialects which I considered for this investigation [url=#n4][4][/url]. Therefore, a blank map, only showing the borders of western Russia, some Rivers, the Ural, and some important towns, was provided to 11 Russian students (female, age 18-30) at the German department of the University of Nizhny Novgorod, Russia [url=#n5][5][/url]. They had to draw the borders of dialect regions they know into the map. Then they were asked whether they speak a dialect themselves. The different lines were either demarcation lines throughout the map (Fig. 1)

or circles surrounding a town (Fig. 2).

All lines put together in one overlay resulted in map Fig. 3.

At first glance, one gains the impression of a confused tangle of lines, but in building up the overlay step by step a large degree of convergence becomes apparent. At least four regions are emerging, designating four dialect regions: the North (Arkhangelsk), the North-West (Petersburg), the East (Kirov), the Middle (Moscow) and the South (Voronezh, Volgograd). This differentiation has its correspondence in linguistic research, which divides the Russian language into three large regions, the North, the Middle around Moscow, and the South (cf. Berger 1999: 559). The demarcation lines, taken from this map and projected into the overlay of the mental maps (Fig. 4), show basically that general knowledge is - approximately - mirrored by scientific research.

Now, to return to the movie, at least three dialects - and this is sufficient - can be used to design the characters even in the Russian synchronization. It does not matter whether genuine dialect speakers are engaged, but there is also the possibility to enrich standard language by very frequent dialect features:

However, the question is, whether the dialects are evoking approximately the same schemes as the dialects in the source language do. In order to solve this problem, the students were asked to describe the dialects marked in the map with a short component analysis, that is to say, whether they are rather perceived as hard or smooth, cold or warm, muffled or clear, monotonous or a singsong, round or sharp, pleasant or curt, and calm or hectic. This analysis concentrates on adjectives of perception, in a further research social implications (rural - urban, antiquated - modern, non-elaborated - elaborated etc.) have to be taken into account as well. The results are shown in Fig. 5 [url=#n6][6][/url].

The predominant characteristics are:

Northern Russian: hard, cold, muffled, singsong, round, curt, calm
Middle Russian: smooth, warm, clear, monotonous, round, pleasant, calm
Southern Russian: smooth, warm, muffled, singsong, round, pleasant, calm

Whereas the Middle and the Southern Russian variety are perceived as quite similar - only the speech melody [url=#n7][7][/url] makes a clear distinction possible -, the Northern Russian variety seems to be a perfect counterpart to the other two varieties. Why not use a Northern Russian dialect, which is hard, cold, curt for the bad guy Santa Maria, and a Southern Russian variety - smooth, warm, singsong - for the Indians and Ranger? Regarding the environment where these dialects are spoken in reality, the Northern Russian variety is more suitable to rural, natural, traditional figures like the Indians, the Middle Russian dialect around Moscow, which is close to standard Russian, aims in its modernity and urban way of life at characters like Santa Maria. Mainly the clear distinction of the varieties and their social implications are important, the perception of the speech can be modified, in addition to the character of a dialect itself, by the individual way of speech. To prove this thesis a sample of the voices2 was offered to the 11 Russian students2, and the same sample1 [url=#n8][8][/url] was offered to 46 German students1 [url=#n9][9][/url], with the task to assign them to drawings, taken from Berthele (2006). I selected six drawings, five of them were most often chosen for standard German and Bavarian in the test of R. Berthele, picture 6 was neither selected for standard German nor for Bavarian dialect. The results are given in Fig. 6 and 7 [url=#n10][10][/url]:

Abahachi1 is assigned to drawing 1 with clear, sharp arrows, two students even chose drawing 5 with its sharp angular blocks. Ranger2 gathered the most votes on drawing 5. In contrast to this, Santa Maria2’s voice was perceived as a ‘round’ impression, assigned to drawings 2 and 3.  As for Berthele’s results on German and Bavarian, this is just the other way round. Therefore, I decided to give the same samples but in German to German students, and the result was not too different from the Russian samples: Santa Maria1 is assigned to the ‘round’ drawings as well, but the more fuzzy variant 6 was most often chosen, next to the flowery variant; Ranger1 was assigned to the sharp and clear drawings 1 and 5, and only Abahachi1 was most often named with drawing 2, a round and clear form, which comes close to Berthele’s results for the Bavarian dialect. Therefore, the impression of the German students on the German samples is quite similar to the impression of the Russian students on the Russian samples. So, why do both samples evoke impressions different from Berthele’s results? One reason may be the artificial character of the speech Abahachi1 and Ranger1 are using, a kind of reduced dialect which is closer to the standard. Another factor may be the quality of voice itself, Santa Maria has a very dark, sonorous voice; Abahachi and Ranger, compared to this, a normal to high-pitched voice. According to Herbst (1994: 84-86), biological features are important to create character equivalence, that is to say, the quality of voice implies character traits. Therefore, with designing Santa Maria1’s voice by using standard German with a dark, sonorous touch, the producers of this figure play with contrasts and clichés. The influence of the voice quality on a character is the chance to meet the character evoked by a dialect, and is a useful instrument to design a character if a transposition of a dialect does not seem to be possible. In the case of the synchronization of Ranger1 and Abahachi1 into Russian dialects, the Northern Russian dialects with its rather hard, curt, muffled qualities seem to be suitable to design a equivalent character, and the triggered rural implications fit as well. Even the play with the Southern States can be transposed: then, Ranger2 is the character with the slang from the Northern States. And in case of Santa Maria2, the modern, urban dialect of Moscow, with warm, round, pleasant features, would make a good contrast to the appearance and behaviour of this figure.

The most striking argument for finding a suitable dialect in synchronization is that dialects, together with the quality of voice, trigger a holistic impression, which is reflected in the pictures of the movie - or it is not, in case the designer of the character plays with the difference of implications and the behaviour and appearance of the figure. Then the ‘mere translation’ is to become an interpretation.

5. Conclusion

The purpose of this investigation was to demonstrate that folk linguistic knowledge can be used to select adequate dialects in translation. These dialects can be normalized dialects or even artificial dialects, multimedialects, which show only a few but frequent and very characteristic features of these dialects. They will be recognized by the viewers, and the characters will be associated with cultural and social stereotypes, of which the viewer is reminded each time the character opens his mouth. Especially in a comedy, when one joke follows the other, leaving no time to reconsider all details, this is a crucial item. Linguistic stereotypes work like basic-stereotypes and enable quickly the reference to social schemes. Finally, the triggering of clichés and schemes by linguistic stereotypes does not exclude an artificial design of characters, quite the reverse, it allows a sensitive play with - apparently - contrasting features. 

About the author(s)

Günter Koch was born 1972 and has a PhD in German Linguistics. He studied German linguistics, literature and history at the University of Passau, Germany, and at the University of Stirling, U.K. His main research interests are dialectology and morphology, the ongoing projects deal with intonation of varieties. He is teaching German Linguistics at the University of Passau.

Email: [please login or register to view author's email address]

•  Previous article
• Next article