38
39 errors, whereas accuracy category is similar to semantic errors. Vehmas-Lehto’s covert errors, in turn, could be included in O’Brien’s style category. There is an obvious overlap between these two translation error categorisations and for this reason it is not necessary to include both categorisations separately in the error analysis. From these two models, I have decided to employ O’Brien’s model in my analysis because the model is more recent and it is based on quality evaluation models which are in active use in the translation industry. Furthermore, O’Brien’s model allows, in my opinion, more latitude in analysing translation errors
specifically in subtitles. For instance, O’Brien’s model includes two categories which respond well to the technical conventions of subtitling. Even though the main focus of this study is on translation errors which occur because something goes wrong when the source text is
transferred into the target language, subtitling errors, i.e. violations against subtitling
conventions discussed in Chapter 2.1., will be given due attention in case they can be seen to affect the quality of the subtitles negatively or to mislead the viewer. Also, as the main purpose is to assess the usability of error analysis in the quality assessment of subtitles, it is more practical to employ a model that corresponds better from the start to the requirements of subtitling. As I do not have access to the company-specific guidelines or quality manuals of MTV or BTI, the evaluation of the content and language of subtitles is mainly based on YLE subtitling guidelines (Vertanen 2007; see Appendix 1 for YLE subtitling guidelines in Finnish). I will now discuss how O’Brien’s (2012) quality evaluation model is used in the present study and how I have interpreted and defined the error categories in the scope of my study and subtitling in general. Examples provided below are selected from the material.
The two main categories used in this study are translation errors and subtitling errors as defined in Chapter 3.4. Translation errors are further divided into four subcategories according to O’Brien’s (2012) model. These subcategories are linguistic errors,
terminological errors, accuracy errors and stylistic errors. I have compiled the errors and their expected severity levels in Table 1 below.
Table 1. Error categorisation.
Error category Severity level 1. Translation error Minor/Major
1.1. Linguistic error Minor 1.2. Terminological error Minor/Major
1.3. Accuracy error Major
1.4. Stylistic error Minor
2. Subtitling error Minor
40 The first subcategory of translation errors is linguistic error. Linguistic errors can be defined as violations against the target language grammar and the target language usage. In other words, they are instances of faulty or unconventional target language usage, such as
misspellings or typos, punctuation errors, incorrect noun and verb inflections, inappropriate verb tenses or verb forms and discrepancies in the relation between a governed and a
governing word. Example 1 is an instance of a punctuation error. In this subtitle, according to Finnish grammar, comma should be inserted to separate the main clause “sanoit itsekin” from the subordinate clause “että sait siipeesi viime kerralla”. “Että” is a subordinating conjunction which marks the beginning of a subordinate clause and in Finnish main clause and
subordinate clause are separated by a comma (Maamies 1995). The line breaks in the examples follows the line breaks of the original subtitles. In the examples, dash has been inserted to mark the part where the sentence continues on the next line.
Example 1.
Speaker Source text Subtitle
Cameron You said yourself you got pretty bruised last time.
Sanoit itsekin –
että sait siipeesi viime kerralla.
However, there is an exception to the punctuation rule. Comma can be left out if the main clause and the subordinate clause are extremely short sentences (Maamies 1995). In this study, a ‘short sentence’ is defined as a sentence that consists of less than six words. In these kinds of instances, comma can be left out without it being considered as a linguistic error.
This exception to the rule does not only apply to sentences such as the one in the example but also to sentences where, for example, a coordinating conjunction separates two main clauses.
Example 2 is an instance of misspelling. In this instance, the linguistic error arises from the misspelling of the word “enenpiä”. The correct spelling is “enempiä”.
Example 2.
Speaker Source text Subtitle
Marlon So, not wishing to be too controversial but are you still on top of Interpol’s most wanted list?
Enenpiä kiistelemättä, oletko yhä Interpolin etsityimpien listalla?
41 Example 3 is an instance of incorrect verb conjugation. The error lies in the incorrect
conjugation of the negative verb “ei”. In the source text, Lizzie is talking about herself so the negative verb should conjugate the first person pronoun as in “en” [I wouldn’t].
Example 3.
Speaker Source text Subtitle
Lizzie I wouldn’t want to put you through any trouble. Ei halua olla vaivaksi.
Furthermore, linguistic errors include cultural references which are defined as things and ideas relating to a specific culture and are thought to be understandable only by someone who comes from or understands that specific culture (Pedersen 2011). Translations of cultural references in a television programme such as Emmerdale would itself provide an extensive object of study. However, this analysis only notes cultural references if there is something
“wrong” with their translations or they are evaluated to have a negative effect on the source text. For instance, in cases where it can be suspected that the translation of a cultural reference might hinder the reading by not being easily understandable or mislead the viewer by being unfamiliar or incorrect. I have attempted to pay careful attention to words, such as ‘tea’,
‘dinner’ and ‘pub’. The word ‘tea’ is most commonly used to refer to a hot drink but in British English it is also used to refer to a cooked evening meal. The word ‘dinner’, on the other hand, is commonly used to refer to the main meal of the day taken in the evening. In British English, however, it is often used instead of ‘lunch’ to refer to the meal taken around midday. Even though cultural references are always unfamiliar on some level, this evaluator assumes that at least the regular viewers of Emmerdale have gained some knowledge through the series of the British culture and thus, are not easily spooked by unfamiliar things.
Emmerdale is so tightly set in the British culture that, for example, substituting cultural references appearing in the dialogue with references from the target culture might seem inappropriate in this cultural context or puzzle regular viewers.
The second subcategory is terminological error. Errors in this category are defined as instances of lack of adherence to industry-specific glossary and lack of consistency in term usage or inappropriate use of terms in the particular context. At first, this error category did not seem relevant to me in the scope of the present study. The Oxford Dictionary (Oxford Dictionary: http://www.oxforddictionaries.com/, accessed on 7 July 2016) defines “term” as
“a word or phrase used to describe a thing or to express a concept, especially in a particular kind of language or branch of study” and it did not seem likely that subtitles of Emmerdale
42 would entail terminology usage. Other genres, for example cooking shows such as
MasterChef Australia, might, on the other hand, be filled with different cooking terms that translators must attend to. After realising, however, that several characters of Emmerdale also have trades and businesses, such as farming, veterinary clinic and scrap yard, which may, at times, require them to use special terms, the possibility of terminology appearing in
Emmerdale did not seem very unlikely after all.
The third subcategory is accuracy error. Accuracy errors are defined as failures to convey the original content of the source text. Accuracy errors can arise from mistranslations, omitting relevant parts of the source text or adding information that affects the content of the source text. In other words, accuracy errors can be instances of factual errors, meaning errors, unnecessary omissions, unnecessary additions and inaccurate cross-references, and they might change the original content or message of the source text in some ways. Example 4 is an instance of an accuracy error. In this particular scene Natasha is discussing her marriage with Debbie when Debbie asks her how long Natasha and Mark have been married. As there is a significant age difference between Natasha and Debbie, Natasha is joking how Debbie’s question is making her feel old.
Example 4.
Speaker Source text Subtitle
Natasha Make me feel ancient, why don’t you? Kuin olisin ikivanha. Mikset sinä?
In example 4, the error lies in the mistranslation of the clause “why don’t you”. “Why don’t you” in this context is a question tag used at the end of a command and it is not a direct question from Natasha to Debbie about why Debbie has not been married.
Example 5 is an instance of an unnecessary omission. As condensation and omission are the main strategies used in subtitling, defining what constitutes as an unnecessary
omission is quite tricky to do before collecting the material. Subtitles represent the translator’s view on what they thought was relevant and important information for the viewer and
questioning this view is always context-bound and open to interpretations. So, it should be noted that what I have evaluated as an unnecessary omission represents my view of what I thought would have been relevant for the viewer to know. Also, I have tried to pay extra attention to space and time limits. Where the space or time do not require omission, the omissions are considered unnecessary. In example 5, Marlon is supposed to be meeting Rhona in the pub but she is nowhere to be seen.
43 Example 5.
Speaker Source text Subtitle
Marlon Any sign of her? Onko näkynyt?
In example 5, the object of the sentence “häntä” [her] has been omitted from the subtitle.
Even though the viewer already knows Marlon is waiting for Rhona which means that the object does not contain relevant information, the space and time would allow the object to be retained in the subtitle and thus, the example above is classified as an unnecessary omission.
The fourth subcategory is stylistic error. As O’Brien (2012: 61) notes in her article, it is difficult to reach a conclusion on what constitutes a stylistic error. In the context of subtitling, a stylistic error is even more problematic as space and time are limited. Features of slang or dialect, for example, can be easily understood when spoken but can be fairly difficult to convey to viewers in written form. Therefore, these elements are often translated in a more easily readable form, most often in standard language, so that the viewers to be able to read and understand them in the few seconds the subtitle is visible on screen. Nevertheless, in the scope of this study, a stylistic error is defined as a lack of adherence to the style and
atmosphere of the source text and being inconsistent in the use of stylistic features. Thus, stylistic errors can refer to tone, register, slang, literal translation and awkward syntax. In my analysis, tone is considered to include elements that convey the style and atmosphere of the source text. Subtitles are not merely words on screen but they also need to convey the atmosphere, humour or emotions of the source text as well as represent the characters and their personalities, and tone is one way to achieve this.
Example 6 is an instance where the subtitle, in my opinion, fails to convey the angry sarcasm of Mark’s line. Mark is angry with Debbie because she made Mark’s wife think that Mark was having an affair and now he is trying to hit a nerve with his sarcastic comment. It is a well-known fact that Debbie’s childhood has not been the easiest as her parents were not only cousins but also very young when they had her and so, she has been raised in foster care.
Example 6.
Speaker Source text Subtitle
Mark What’s wrong with you, Debbie? Broken home, was it?
Mikä sinua hiertää?
44 Register, on the other hand, is considered to mean the way a character speaks in different situations and social contexts. Register is closely linked to colloquial language and sociolect which were discussed in Chapter 4. Even though there is a variety of colourful regional and social dialects in Emmerdale, the language spoken by the characters is mainly in keeping with good taste and propriety, and there is hardly any crude cursing or slang usage. Deviations from register are taken into consideration when the translation of a line is unsuitable for a certain character or if there is inconsistencies in the use of certain register. By unsuitable register for a certain character, I mean that, for example, translations of lines of a character representing a higher social class are filled with elements of spoken language. They do not sound quite right in association with that certain character. By consistent usage of register, I mean that the speech of a certain character has been translated in a similar way in the target language from episode to episode. For example, the speech of a character who is a member of the Dingle family (see Chapter 4) contains elements of spoken language in every episode.
Evaluating what is a suitable register and what is not, I base on my expertise as a long-term viewer of Emmerdale. The lines of children have been omitted from the material because the way children speak is often more informal than adults’ way of speaking.
Example 7 is an instance of an unsuitable register for a particular character in the series. The character speaking in this example is Scarlett whose father used to own Home Farm which is the largest property in Emmerdale village with eleven bedrooms and numerous communal rooms. The owner of Home Farm is usually the wealthiest person in the village which means they represent a higher social class. Since her father died, Scarlett inherited a lot of money and is successfully running her own business. The use of “se” [it] when referring to humans makes the line sound more informal and colloquial than the source language is and does not quite sit well with a member of a higher social class.
Example 7.
Speaker Source text Subtitle
Scarlett He’s joking. Ei se ollut tosissaan.
The other main category used in this study is a subtitling error. As this study concentrates mainly on translation errors, I have decided to combine DTP and UI errors included in O’Brien’s (2012) model into one category. Subtitling errors include deviations from the technical conventions of subtitling. As I have previously mentioned, the focus of this study is not to analyse how well the subtitles follow the technical conventions governing subtitling,
45 rather the focus is on the language. However, sometimes deviations from the subtitling
conventions, for example, a subtitle that takes up three lines, can affect viewer experience by affecting readability or by confusing the viewer. Thus, I find it necessary to include this category in the analysis.
In evaluating the severity of the translation errors, the errors are divided into two levels according to O’Brien’s (2012) model. These are minor errors and major errors. Minor errors are easily noticeable from the target text but they do not change the meaning or the message of the source text. Minor errors might annoy and puzzle the viewers or affect the fluency of the target text or even make the text difficult to understand at times. Example 3 presented above and reproduced here as example 8 is a linguistic error where the error lies in the incorrect conjugation of the negative verb “ei”. In the source text, Lizzie is talking about herself so the negative verb should conjugate the first person pronoun as in “en” [I wouldn’t].
The error can be quite easily noticed from the subtitle but does not affect the meaning of the source text and for that reason is evaluated as a minor error.
Example 8.
Speaker Source text Subtitle
Lizzie I wouldn’t want to put you through any trouble. Ei halua olla vaivaksi.
Major errors, on the other hand, are considered to have a negative impact on the content of the source text. They either change the original meaning of the source text or fail to convey, in the case of subtitling, the atmosphere or humour of the source text to viewers. These types of errors can mislead or confuse the viewer and affect the viewing experience. Example 4 presented above and reproduced here as example 9 is a meaning error which is evaluated as a major error. In this example the error lies in the mistranslation of the clause “why don’t you”.
“Why don’t you” in this context is a question tag used at the end of a command and it is not a direct question from Natasha to Debbie about why Debbie has not been married. The subtitle changes the content of the source text and for this reason is evaluated as a major error.
Example 9.
Speaker Source text Subtitle
Natasha Make me feel ancient, why don’t you? Kuin olisin ikivanha. Mikset sinä?
46 As was introduced in Chapter 3.4., O’Brien’s (2012) model includes a third severity level as well. This third level is called critical and it contains errors that are considered to have serious effects on “meaning, product usability, company liability, consumer health, safety and
behaviour”. Even though Emmerdale might be a matter of life and death to some viewers, I would not evaluate any translation error occurring in the subtitles of the series as having a critical impact on anyone’s life. As Tuominen’s (2013: 302–309) study on some spontaneous comments on subtitles on the internet indicates, meaning errors and grammatical errors might cause the viewers to question translators’ professionalism and intelligence but I do not believe this is what O’Brien (2012) means by a critical error. Critical errors are more likely to occur in, for instance, manuals where it is crucial that the target text contains the same information as the source text. Thus, the third level is not taken into account in this study. In my opinion, classifying errors into minor and major errors will be enough to provide a glimpse on how different errors affect the quality of the subtitles.
Placing translation errors into categories and evaluating the severity level of each error might at times be difficult. The technical restrictions of subtitling require some serious creativity from the translator in order to fit the essence of what is said on to the screen. Thus, some leeway has to be given when assessing translation errors in subtitles as omission and condensation cannot be avoided. Some errors might be ambiguous and they could be placed in several different categories. In this kind of qualitative analysis, the allocation of errors into error categories is often a judgement call by the evaluator and based on the evaluator’s subjective view on what they consider to be an error. However, when categorising and evaluating errors, I use the ‘meta rules’ for error categorisation provided by the J2450 metric which is a standard generated by the Society of Automotive Engineers (SAE) (O’Brien 2012:
56, 63). This metric instructs that when an error is ambiguous, it is always better to choose the earliest primary category. The primary categories in this analysis are linguistic error,
terminological error, accuracy error and stylistic error. Also, when in doubt, the metric advises to “always choose major over minor”.
Translation errors will be categorised and assessed according to the translation error categorisation and severity levels introduced in this chapter. Each error is placed in an error category and the severity level of the error will be evaluated. The number of translation errors in each category is compiled to a single table. From the table, it is possible to see which translation errors occur most frequently. Every category is then discussed and analysed separately. Each error category is considered from the following perspectives: