definition already included these three basic ideas, but additionally emphasised the ephemeral nature ←26 | 27→of subtitles on screen as well as the ubiquitous condensation involved in subtitling.
From the temporal and spatial perspectives, a subtitle should be on screen for a minimum of 1 second and a maximum of 6 (or 7) seconds, contain two lines at most, and each line should consist of a maximum of 37 to 42 characters, including spaces. Because of these technical constraints, subtitles show traces of translated, summarised and transcribed texts, as shown in Figure 1.
This tripartite conception of subtitles means that they are governed by norms at different levels. The idea that multiple factors influence the final design of subtitles is thoroughly developed in Karamitroglou (2000). Mattsson (2006) offers a simplification of Karamitroglou’s theoretical framework by applying it to the subtitling into Swedish of discourse markers and swearwords. On the one hand, she lists ‘norms of tradition, written language norms, norms of original literature, literary translation norms, and subtitling norms’ (ibid.: 6) as influencing the decision making process in subtitling. On the other, she elaborates on further factors to which the professional subtitler is exposed and have a strong impact in the subtitling process as well, such as working conditions and broadcasting companies’ policies.
←27 | 28→
3.3.2. Syntactic features of subtitling
Structural shifts affecting the syntax of translated texts is a phenomenon long taken into consideration in TS, with authors like Nida and Taber (1982/1969: 114) stating that ‘transfer normally involves a number of shifts in coordinate and subordinate patterns’. Subtitles have also been depicted as incorporating changes in syntactic structure, if contrasted with the ST. As regards textual disposition, it is normally proposed that each subtitle should correspond to one sentence and that segmentation should answer to the phrase distribution of the sentence (Karamitroglou 1998). Scholarly works such as Díaz-Cintas and Remael’s (2007: 172–180) summarise this idea as follows: ‘an ideal subtitle is a sentence long, with the clauses of which it consists placed on separate lines’.
Thus, in accordance with the literature, syntax strongly influences segmentation (§6.2) and general textual distribution in subtitling (one sentence, one subtitle). According to two further salient ideas about the syntax of subtitling, both assumed to be factors that foster readability, subtitles should be simple in terms of syntactic structure and they should not deviate greatly from the normative constituent order.
The idea of simplicity is connected to a number of linguistic aspects:
a) simple clauses of one verbal phrase are regarded as preferable to composed clauses (Díaz-Cintas 2003; Nagel et al. 2009);
b) verbal periphrases in the ST become simple tenses in the TT (Díaz-Cintas 2003; Díaz-Cintas and Remael 2007; Bartoll 2012);
c) coordination (parataxis) is more frequently used than subordination (hypotaxis) (Díaz-Cintas 2003);
d) theme/rheme manipulation can be observed (Díaz-Cintas 2003; Vayssière 2014);
e) subtitle discourse simplifies structure by avoiding repetition (Chaume 2004a; Bartoll 2012; Vayssière 2014);
f) cohesive elements may be simplified or suppressed (Chaume 2003, 2004b).
←28 | 29→
Another idea present in the literature is posited by Díaz-Cintas (2003: 284), who explains that subtitles stick to prescriptive norms of the target language, therefore showing correct syntax even in segments which show errors or inconsistencies in the ST. This idea is shared by Chaume (2004a), when he opines that syntactic variation and deviation from the norm could result in the confusion of the audience, and authors like Díaz-Cintas and Remael (2007) highlight the idea that subtitles that follow a standard syntactic order are easier to read, which is why they are preferred in professional practice.
Gottlieb (2008: 211) recalls the idea of subtitles being constrained by different systems of norms, among which the norms of ‘the rigid written language’ stand out:
[I];n any diagonal – and thus interlingual – subtitling, one must, on top of translating utterances from one language to another, transfer the dialog from one sub-code (the seemingly unruly spoken language) to another (the more rigid written language). If this shift was not performed, as a fundamental part of the subtitling process, the audience would be taken aback by reading the oddities of spoken discourse. But as the dialog is always re-coded on the way to the bottom of the screen, people only react if the other dimension of diagonal subtitling – the translation proper – seems imperfect.
A crucial thread seems to underlie these contributions: subtitles present unmarked syntactic structures, which makes them different from the ST, irrespective of the ST nature.
The analysis reported in Chapter 5 contrasts syntactic features of subtitling with syntactic features of the ST, as a comparison with the source dialogue is deemed indispensable when characterising the so-called neutralisation in subtitling (Arias-Badia 2015).
3.3.3. Lexical features of subtitling
The principle of simplicity in subtitling is a recurrent topic in the literature about the lexicon used in this translation practice. Scholars have stressed the idea that time and space constraints lead to content reduction ←29 | 30→in subtitling, which typically affects lexical selection in the form of omissions or of general shifts.
When it comes to ommissions, de Linde and Kay (1999: 29) claim that ‘the overriding demand to reduce the amount of dialogue means that many omissions have to be made on a selective basis’. In order to tease out the strategies that lead to reduction in subtitling, relevance theory, originally posited by Sperber and Wilson (1986), has been applied to subtitling in Slovene by Kovačič (1994), who concludes that adverbials are highly likely to be removed from subtitles. Yet, while the tendency towards omission and condensation has been the focus of many scholarly publications, Díaz-Cintas (2003: 285) states that, in Spain, professional subtitlers tend to ‘overtranslate’, thus implying that subtitles are also prone