Luyken deals with the difference between literary and audiovisual translations in his chapter on semiotics of audiovisual language transfer (153 – 165). He comes to the conclusion that subtitling and dubbing certainly are a form of translation. However, there are some special features that make them different from the translation of a written text.
When translating a book, the text is transferred from one language to another. In this case the new work completely replaces the original and the two are completely independent. On the other hand, the message of a film is expressed by various elements such as image, acting, sound or language. And obviously only some of these are replaced. When a film is dubbed, the visual component stays completely the same, only the auditory component is changed. And it is even more interesting in the case of subtitling, as both visual and auditory components remain, the actual translation only being added to the original work. Thus, the transferred form of one element of a work has to ‘coexist’ with other elements of the same work. Those remain in the original form (Luyken, 153).
The particularity of audiovisual translation also lies in reduced scope of the translator. If there is something which has to be explained in the literary translation, the translator can use a footnote or he can include the explanation directly into the text. This is impossible in a film.
A big problem arises when there is a difference between the average word length of the source and target languages. Such an example is English and Czech. There are many short words or phrases in English the translation of which is much longer in Czech. (“I think so” – “Myslím, že ano”; “I did it” – “udělal jsem to”, or sometimes even “já jsem to udělal”). This leads to a necessary condensation of the translated text which has to be done by the translator.
Thus, according to Luyken, audiovisual translation
“…adds information to that contained in the original text and leaves
some out. It can never attempt to transfer every bit of information from
one language into the other. It is at one and the same time both more
and less than conventional translation. Less, because it does not translate
everything. More, because the audiovisual Translator/Writer has to make
editorial decisions all the time about omissions or condensation of the
original text, and about new information that has to be inserted into it.
This all has profound implications for the programme which is subject
to any form of Language Transfer” (154).
These were some particular features shared by both subtitling and dubbing. Needless to say, there are many others, which are specific only for subtitling or only for dubbing. These will be treated in detail in the following chapters.
3. 1. Subtitling
3. 1. 1. The Process of Subtitling
As stated by Ivarsson and Carroll, the subtitler usually works with a script or dialogue list, which are provided to him in the source language. In the best case he obtains a post-production script which includes the whole dialogue (even the parts which are very badly audible). If he is lucky enough, he also has a glossary where dialect words, slang, insider jokes etc. are explained. But it is important that every such dialogue list is checked. There are sometimes mistakes and discrepancies between this list and the actual film dialogue. It is always important to fit the subtitles to what is really being said (79 – 80).
According to the same authors, when the film is supposed to be subtitled into many language versions, a master list is available. This contains the in and out times, prepared by a technician in the country of origin. In addition, there is a transcription of the source language dialogue, but these are abridged. The subtitler’s job is then just to transform it into his language. He has to observe the previous spotting as well as the subtitle length. This can be very inconvenient as the translator is not able to adjust is to the needs of his language (54 – 57).
In case the subtitler has not got the master list, he must do the spotting himself. He uses specialised software such as Poliscript which facilitates the job significantly. All subtitling systems nowadays use time codes. Such a time code “provides an 8-digit address for every frame of a videotape (or every image of a film)” (Ivarsson and Carroll, 141). In the past, when a video recorder was used for subtitling, this was recorded on the tape and when needed, it could be displayed. But in new subtitling programmes it is not necessary to use the VHS any more, the films are imported directly into the computer and are converted into an mpeg file. In such programmes the time codes are shown permanently.
The time code shows “hours:minutes:seconds.frames” like e.g. 01:28:15.09. There are colons between hours, minutes and seconds, but between seconds and fames there is a dot. “The number of frames per second depends on the standard: 24 frames a second for film, 25 frames a second for PAL or SECAM video and some films, and 30 frames a second for NTSC video” (Ivarsson and Carrol, 141).
With the assistance of the time codes, the subtitler can spot the film and then prepare the subtitles. During the whole process, however, he has to bear in mind certain rules which should be observed, as stated by Ivarsson and Carroll:
Position on the Screen
Subtitles should be placed at the bottom of the screen in order not to interrupt the image action. The only case when the subtitle can change its position is the one when something important is shown in the part of the screen where subtitles are inserted. These can be for example captions. The subtitles can then move either above these captions or even to the top of the screen.
Number of Lines
There can be two lines at a maximum, otherwise the subtitles would cover too much of the screen, which would be very disturbing. Furthermore, three lines would be difficult to read in the short time available. If there is only one line displayed, it can be positioned either on the upper or lower line. The former is easier for the viewer to read as he is used to the beginning of the subtitle. The latter, on the other hand, interferes much less to the background image.
The subtitles can be either centred or they can appear at the left margin of the screen. In Western Europe, centred subtitles are used in the cinema while left-justified subtitles are used in the television. In the Czech Republic and some other countries, however, even the TV usually displays centred subtitles. This seems to be “due to unconscious adoption of the principle applying in the cinema, or possibly to the fact that optical film stock was used for the subtitles” (Ivarsson and Carroll, 49).
Number of Characters per Line
The majority of sources state that there should be maximally 35 – 40 characters in each line. If there were more, the characters would have to be too small and it may be difficult for the audience to read. On the other hand, if there were less, the text reduction and omissions would be unbearable.
Typeface and Distribution
According to Ivarsson and Carroll, it is advisable to use typefaces with no serifs such as Arial as their legibility is better than that of typefaces with serifs such as Times New Roman. As for the distribution, proportional types are preferable to monospace ones as the former save space (42).
3. 1. 2. Duration of a Subtitle
While spotting, it is essential to bear in mind the fact that “subtitles which remain on the screen long enough to be read more than once are just as irritating as subtitles that disappear before the audience has had time to finish reading them” (Ivarsson and Carroll, 67). That is why rules concerning the maximum and minimum duration of one subtitle have been introduced. Karamitroglou states that that the average reading speed of subtitles is 150 – 180 words per minute, which is 2.5 – 3 words per second. As a full two-line subtitle contains about 14 – 16 words, it should be projected for some 5.5 minutes. After the addition of the time necessary for the eye to notice that a subtitle has appeared at the bottom of the screen, we get to the final result – 6 seconds. Thus, a full two-line subtitle should remain on the screen for 6 seconds, but at the same time it should not exceed this time, because viewers would immediately start to reread it (Karamitroglou, http://accurapid.com/journal/04stndrd.htm).
As for a single-line subtitle, Karamitroglou argues that the optimum time for which it should remain on the screen is about 3.5 seconds. The half second is added because it has been proven that viewers tend to read two-line subtitles a little bit faster than separate one-line subtitles. These 3.5 seconds are again not only minimum time, but also maximum time.
Karamitroglou also introduces a rule for the minimum time of a single-word subtitle, however short or simple this word may be. He fixes this time at 1.5 seconds. If it disappeared sooner, the viewer’s eye would not be able to recognise it, and it would seem just like an irritating flash.
If there are two consecutive subtitles, Karamitroglou maintains, there has to be at least a ¼ second gap between them. This is necessary for the eye to notice that there has been a change in the subtitles. If this break was not maintained, the viewer would not be able to distinguish the two different subtitles and would think it is still the previous one, especially if the two are approximately of the same length.
All these are of course recommendations rather than hard and fast rules and the times can vary according to the audience it is designated to (as the reading speed of the elderly is certainly lower than that of teenagers). Nevertheless, the more these rules are observed, the higher the quality of the subtitles.
Likewise, it is important to stick to the speed consistency. As the viewers very quickly get used to the reading rhythm, it should stay the same during the whole film and there should not be any deviations such as sudden acceleration or deceleration of the subtitles (Ivarsson and Carroll, 69).
3. 1. 3. Synchronisation
As stated by Ivarsson and Carroll, in the early days the basic principle was to synchronise perfectly the subtitles with the speech. The subtitle appeared exactly at the same moment when the utterance started and disappeared the moment it ended. But the attitude soon changed and this practice has been abandoned (72).
The leading-in time should stick to the beginning of the utterance. Nevertheless, Karamitroglou argues that it is better to make a delay of ¼ of a second, as this is approximately the time the brain needs to “process the advent of spoken linguistic material and guide the eye towards the bottom of the screen anticipating the subtitle” (Karamitroglou, http://accurapid.com/journal/04stndrd.htm).
On the other hand, according to Karamitroglou, the lagging-out time does not have to stick so closely to the end of the utterance and the subtitle can remain on the screen even after the utterance has been pronounced. Nevertheless, he argues that this should not exceed 2 seconds, because if it remains longer, the viewer has the impression that the subtitles do not correspond exactly to what is being said.
As for the synchronisation with camera takes and cuts, Karamitroglou claims that the subtitles should observe the major ones which signify thematic change and thus the subtitles should disappear before them. In case the cut is not particularly significant and does not mean any substantial change, the subtitle can remain on the screen.
3. 1. 4. Punctuation
Punctuation in subtitles is usually the same as in any other written language form. There are, however, some cases, stated e.g. by Ivarsson and Carroll, when particular rules are applied for subtitles in order to facilitate their reading to the audience. These rules have to be consistent and logic, otherwise they could confuse the viewers rather than help them to understand the film translation.
Two punctuation marks essential for subtitling seem to be hyphens and suspension dots. These have to follow particular rules.
According to Ivarsson and Carroll, hyphens are used at the beginning of subtitles translating dialogue. As this is considered to be their main role in subtitling, all other usage of hyphens should be avoided (111):
-He had a car accident.
However, the same authors stress that it is important not to mix hyphens with dashes, which are slightly longer and have a different meaning. They are used in subtitling for distinguishing between sentences uttered by a single person which are addressed to different people:
Did you prepare the meal? – You can
tell us what you wanted to, Frank.
Suspension dots are also treated in detail by Ivarsson and Carroll. If they are used without spaces in the middle of a sentence, it means that the character is hesitating (113 – 114):
He said…that he doesn’t love her any more?
On the other hand, if a space follows, they mean the character was interrupted:
He said… Yes, that he doesn’t love her.
It is necessary to repeat the suspension dots if they coincide with a subtitle break:
It is necessary…
…that we pay a visit to the parents
and ask if they don’t need anything.
As for other punctuation, the quotation marks, exclamation and question marks and commas usually observe the same rules as in any other written text, while underlying, semicolons and parentheses should be completely avoided (Ivarsson and Carroll, 114 – 116).
3. 1. 5. Other Conventions
There are certain other rules which effect the subtitles. Although these may vary in different countries, there is a tendency to bring them into line. These include the use of upper case and italics as well as some other particularities (subtitling of songs or letters).
According to Ivarsson and Carroll, upper case can be used in the subtitles to translate shouting or exclamations which are very loud (116):
CATCH HIM !
The same authors also state that the use of italics is quite common in subtitling. They usually signify 1) speech which is not uttered by the character on screen – this can be for example a voice from telephone, radio or TV, 2) interior monologue and dream scenes, 3) narrator’s voice, 4) songs, and 5) everything which would be written in italics in a normal text such as words in a foreign language or titles and names (Ivarsson and Carroll, 118).
A specific case are songs. If the lyrics are translated, Ivarsson and Carroll argue, it is advisible to put them down in italics. However, these authors stress that not every song in a film needs to be translated. It only has to be subtitled if it is important for the story or there is another serious reason. It is not necessary that the song translation rhymes, nevertheless it should stick to the rhythm of the song so that it is easy to read while listening to the original. Sometimes it is enough to subtitle only the beginning of the song in order that the viewer has some idea what it is about (Ivarsson and Carroll, 120 – 122).
Very interesting is the subtitling of letters. It happens very often that a letter is read aloud in a film. The way this is subtitled differs according to various factors. Ivarsson and Carroll state six possibilities (119):
1) The writer of the letter thinks to himself while writing the text = interior monologue -
2) The writer thinks aloud while writing the text = audible speech – normal style
3) The writer reads the letter aloud after writing it = quotation – normal style + quotation
5) The recipient reads the letter aloud – normal style + quotation marks
6) The voice of the recipient is heard while he reads the letter without moving his lips
= interior monologue – italics + quotation marks
There are, of course, other things which might cause some problems. These are for example currency, units of measurement, abbreviations or proper names and brand names. It is not necessary, however, to treat these here in detail, as the film translation of these usually observe exactly the same rules as the literary translation. The same applies for strong language used in a film.
3. 1. 6. Translation and adaptation of the subtitles
It is evident that if the above mentioned rules and time constraints are to be observed, the subtitles have to be considerably shortened in comparison to the original dialogue. It is the subtitler’s job to decide what will be kept and what will be omitted.
According to Ivarsson and Carroll, the subtitler sometimes has to decide between omission and paraphrase. Omission means that a whole part of the dialogue is left out, paraphrase, on the other hand, stands for modification of the dialogue in order to make it shorter. Out of these two, Ivarsson and Carroll argue, the former is more convenient as it seems to be less annoying for those who understand the original. Thus, the sentence “Well, I just want to tell you that yesterday Jane came to our room and told us that she has heard that everybody will have to sit the exam again some time in September.” should be subtitled in this way:
Yesterday Jane told us
that everybody will have to sit the exam again.
Nevertheless, subtitlers usually take advantage of both these methods and use them simultaneously (Ivarsson and Carroll, 87).
3. 1. 6. 1. Omission
There are certain words and expressions which can be omitted without lowering the quality of the translation. As stated by Karamitroglou, these include particularly two groups:
1) Padding expressions such as “you know”, “well” etc. can be omitted as they usually do not contribute to the semantic meaning of the utterance.
2) Words that are presumably known to the target language audience and are pronounced separately do not have to be necessarily subtitled. Such words are for example “yes”, “no”, “sorry”, “please”, “thanks” or “ok”. Similarly, according to Karamitroglou, names of the characters can be omitted if they are called out. But on the other hand, if these expressions are included in a larger context, they always have to be subtitled.
3. 1. 6. 2. Use of simple vocabulary
Ivarsson and Carroll claim that, if possible, common and simple words should be preferred to unusual ones. When the subtitler has the possibility of choosing between several words with the same or similar meaning, he should always bear in mind that the more common and simple the word, the faster it is read by the viewer. The only exception from this rule is, of course, the case when florid language plays a vital part either in the whole film or in the speech of a certain character (Ivarsson and Carroll, 89).
3. 1. 6. 3. Simple syntax
Syntactic structures should also be simplified as much as possible. There are very often cases, when reduction of a complex sentence is possible without changing the meaning. Karamitroglou states seven such particular cases:
Active for passive constructions:
“It is believed by many people.” => “Many people believe.”
Positive for negative expressions:
“We went to a place we hadn’t been before.” => “We went to a new place.”
Temporal prepositional phrases for temporal subordinate clauses:
“I’ll study when I finish watching this movie.” => “I’ll study after this movie.”
Modified nouns for the referring relative clauses:
“What I’d like is a cup of coffee.” => “I’d like a cup of coffee.”
Gapping for double verb insertion:
“John would like to work in Germany and Bill would like to work in France.” => “John would like to work in Germany and Bill in France.”
Straightforward question sentences for indicative pragmatic requests:
“I would like to know if you are coming.” => “Are you coming?”
Straightforward imperative sentences for indicative pragmatic requests:
“I would like you to give me my keys back.” => “Give me my keys back.”
3. 1. 7. Subtitles Editing
As one line usually is not enough, the translator often has to split the subtitle into two lines. This has to be done very carefully and the breaks should not be in contradiction with the sense of what is being said – each line should be logically and grammatically compact. Moreover, according to Ivarsson and Carroll, one subtitle should correspond to one phrase or sentence or should be composed of several short phrases or sentences. On the other hand, Ivarsson and Carroll argue, one long sentence spreading across several subtitles should be avoided (93 – 96).
When the dialogue is too fast, two line subtitles are used. Each line then corresponds to what is being said by one character. The beginning of such a subtitle is always hyphenated and if it is a question and an answer, the former should always be on the top line and the latter on the second line. The only exception is the case when the answer cannot appear in the subtitle before it is actually uttered (Ivarsson and Carroll, 94).
-Have you seen that film? But: -I couldn’t believe what he did. .
-Yes, I have. -What did he do?
-He killed her.
3. 1. 7. 1. Subtitling of written texts
Not only speech, but also written texts which appear in the production and are important for the understanding of the film, have to be translated. These are not only titles, but also captions (small texts indicating when and where the scene is taking place in feature films or names of speaking persons in documents) as well as various newspaper headlines, letters etc.
Translation of all these texts should “appear and disappear simultaneously with the original and they should correspond typographically as closely to the original as the subtitling system allows. Capitals should be written as capitals and handwritten texts as italics” (Ivarsson, 97).
3. 1. 8. ESIST
Although all the above mentioned rules are recommendations, their observing is essential for high quality of the translation. Subtitling suffers by the non-existence of hard and fast rules which all subtitle producers would be obliged to observe. An important effort to change this situation seems to be the Code of Good Subtitling Practice published by ESIST (European Association for Studies in Screen Translation) which can be found on this association’s web pages (www.esist.org). The Code states some basic principles a high quality subtitler should always stick to.