RSS feed

Subtitling stereotyped discourse in the crime TV series Dexter (2006) and Castle (2009)

Blanca Arias Badia and Jenny Brumme, Universitat Pompeu Fabra, Spain


Recurrent communicative events in TV police procedurals, as dictated by genre conventions, lead to the use of recurrent, stereotypical language in these audiovisual products. This paper explores repeated lexicon and syntactic structures (text blocks) from corpse finding and initial forensic analysis, as well as from interrogation sequences in crime TV shows. It also reflects on the extent to which such repetition poses a challenge for subtitlers. The examples for analysis have been taken from the American TV shows Dexter (2006) and Castle (2009). The paper sheds light on how textual recurrence supports recognition of the genre and plot structures and helps to condense, up to a certain degree, the relevant elements of the message in the case of subtitling. As the genre under study is at the crossroads between fiction, specialised discourse and evocation of everyday conversation, this paper combines approaches from different research fields in order to use the most widespread concepts related to recurrence for the benefit of translation studies.


Crime fiction, genre recognition, text blocks, stereotyped discourse, police procedural, fictional orality, audiovisual translation, subtitling.

1. Introduction

Wolf (1984: 190) posited the idea that the aim of producing coherent texts, easily recognisable by audiences, causes TV series to be ruled by the inherent repetitions and thematic recurrence of each specific genre. The most conventionalised units in the genre therefore stay unaltered in order to favour such recognition. In this paper, we are interested in seeing how lexical and pragmatic recurrence is used to achieve such genre recognition and the effect it may have on translation. This phenomenon is known as text blocks (Textbausteine) in the field of the description of specialised texts1.

Our corpus of analysis is composed of the first five episodes of the first seasons of two crime TV series, namely Dexter (Showtime and Paramount, 2006) and Castle (ABC, 2009). The source text (ST) is a transcript of the English dialogue in the episodes and the target text (TT) is the Castilian Spanish subtitled version of these episodes that has been published in DVD format.

We have identified recurrent communicative events typical of this genre. For the purposes of this paper, we will focus on sequences consisting of (a) the finding of a corpse and the first forensic analysis at the initial stage of the investigation, and (b) the interrogation of witnesses and suspects. Our study is twofold in nature: we will first identify the text blocks that typically occur in each communicative event, and then consider the implications of these stereotyped text blocks on the subtitling process. We aim to show that the condensed nature of subtitles is supported by recurrence. Since they are familiar with the conventions of the genre, viewers experience less difficulty in filling in the aspects left out by the subtitles (cf. Inigo and Westall 1997: 134 or Kovačič 1994, quoted in de Linde and Kay 1999: 5).

2. Recognition and novelty in narrative genres

Wellek and Warren’s classical approach to literary genres in their Theory of Literature perfectly adapts to our object of study, i.e. audiovisual genres. We think this is also the case for scriptwriters:

Men’s pleasure in a literary work is compounded of the sense of novelty and the sense of recognition […]. The good writer partly conforms to the genre as it exists, partly stretches it. By and large, great writers are not the inventors of genres: Shakespeare and Racine, Moliere and Jonson, Dickens and Dostoevsky, enter into other men’s labors (1949: 245).

This article focuses on the extent to which the sense of recognition is satisfied linguistically in the two crime TV series under study. However, both series have been successful both in the American and in the Spanish market and that is a sign that they have offered something new to audiences. Let us briefly introduce the reader to the sense of novelty in each of them.

In Dexter we see strong evidence of typically recurring communicative events but at the same time a scenario which is novel. A medical examiner who is actually a serial killer works for the police and overturns the conventional storyline while the recurrent linguistic structures provide familiarity. There is also lexical novelty given that Dexter is a blood spatter analyst. The specialised terminology is usually connected to the semantic field of blood and introduces a distinctive linguistic element to the series.

In Castle, novelty comes from a character portraying the stereotypical role of a crime fiction writer who enters the context of criminal investigation to work closely with the police. The writer’s dialogue often adds humour to the hunt by making use of literary language. Viewers who are familiar with the genre may also enjoy the wealth of intertextual allusions, in a similar way to what Bönnemark (1997: 37) explains when she deals with the detective genre: “other detective fiction is described or alluded to as being unrealistic, and the current work, by implication, becomes realistic […]. The current story stands out as real and sensible in comparison.” Castle’s figure usually compares the fiction of his novels with the presupposed ‘reality’ of his investigations with Detective Beckett, who usually starts by rejecting his apparently ridiculous initial hypotheses and ends up by agreeing with them towards the end of each episode.

It is precisely the sense of recognition that allows scriptwriters to break open genre stereotypes and recurrent communicative events. Scholars from different fields have reflected on such recognition in narrative. In his early works (1968[1928]), Propp presented us with a proposal of recurrent motifs and functions in fairy tales. Thus, the hero in a story would recurrently react in a determined manner to some kind of misfortune to which they were exposed, they would struggle against a villain to achieve success, which typically involved social recognition or a wedding to their beloved one. Literary theories have been constructed to account for the structure of texts of a certain genre (cf. Pfister (2001[1977]); Genette 1979 and 2002). Likewise, researchers have proposed prototypical narrative sequences in crime fiction (e.g. Nusser 2003: 22–33, Bönnemark 1997, Baroni 2007: 167–224). More specifically, studies have been undertaken to account for the recurrent themes, types of characters or settings in TV series (e.g. Balló and Pérez 2005).

However, little has been said in these proposals about the recurrent linguistic structures of prototypical sequences or communicative events in crime TV series. Therefore, our article focuses on linguistic analysis and in particular on recurrent text blocks. Our proposal combines notions from the field of specialised text analysis with notions from literary analysis. It seems to us that our corpus holds features that call for such an approach: on the one hand, specialised terminology from the fields of law and medicine abounds in criminal investigation; on the other, TV police procedurals drink from the wells of literary narrative.

3. Recognition and subtitling

In subtitling, spoken film dialogue is rendered in written format. This transformation does not imply a simple transfer from the spoken medium into the written one, since the film dialogue has previously been designed in the form of a script (written medium) and then acted out (spoken medium). This type of language holds traits of the spoken interaction evoked in the written medium and is to a certain extent artificial or fictive (fictional orality). In addition, subtitling as a modality of audiovisual translation is well-known for its so-called time-and-space constraints (cf., for example, Díaz Cintas 2003: 117-119). The translation process involved in subtitling therefore leads the resulting TT to include features of summarisation, in the sense that nuances of meaning which are expressed verbally in the ST may need to be omitted or condensed in the TT. Such a verbal omission does not necessarily involve a loss of meaning, since nuances may be reconstructed through non-verbal channels of communication. For example, certain facial expressions may contribute to portraying a feeling, or suspense music may help the audience interpret a scene.

Because of the above-mentioned time-and-space constraints, the ability to detect repetition or redundancy in the ST is crucial when it comes to subtitling. Therefore, we believe that the field can benefit from the notion of text blocks. This notion is somewhat present in Chaume’s suggestion (2003: 201) that TV series hold “predictable elements, even appearing in a predictable order. The aim of the present study is to highlight these elements and connect the concept of text blocks with the task of the subtitler. It is undoubtedly helpful for the subtitler to be aware of the presence of these blocks in their ST.

4. Subtitling text blocks in recurrent communicative events

Let us start by identifying the main recurrent text blocks of our corpus and determine whether repetition is present in the Spanish subtitles. According to Nusser (2003: 22-33), the finding of a corpse and the interrogation of witnesses and suspects are among the prototypical sequences of crime fiction products. Therefore sequences depicting preliminary forensic analysis and interrogation have been selected for our analysis.

These sequences can be defined as dialogical rather than conversational, following Cattrysse and Gambier’s description (2008: 52). They demonstrate that dialogue and not conversation is generally the vehicle for plot development. This informs our interest in seeing whether subtitlers preserve, reduce or omit the main sense of the utterances in these sequences.

4.1. Text blocks in corpse finding and forensic analysis sequences

The central plot of the ten episodes studied, without exception, starts with the finding of a corpse. If we follow van Dijk’s proposed model of context categories (2001: 73), these sequences would be described as follows. First, the general domain of the discourse practice that we analyse is criminal investigation. Second, its global act, i.e. its overall purpose, is for the police to gather the first pieces of evidence to start the investigation. Third, the interlocutors taking part in this sequence of the episode — in this initial communicative event — are typically three or four police officers and a medical examiner. One of the police officers, usually the lieutenant, is in charge of the case in hand and gives orders or asks questions to their subordinates, who try to provide the lieutenant with the most accurate information about the case at this initial stage of the investigation. These exchanges involve both superiors and subordinates and the different social roles provide examples of asymmetric communication. Finally, the kind of cognition involved serves multiple purposes: to share knowledge between characters, to speculate and formulate the initial hypotheses, and to coordinate the hunt by giving orders. Figure 1 summarizes the context categories of these sequences.

Badia 1

Figure 1. Corpse finding and preliminary forensic analysis sequences: context categories

In our analysis, we have found that this recurrent kind of communicative interaction includes a wide range of repeated linguistic features. Specifically, interrogative clauses, ellipsis, pragmatic markers for directing attention, specialised lexicon, and modalisers of speculation seem to be distinctive features of these corpse-finding sequences in crime TV shows.

4.1.1. Interrogative clauses

In these initial sequences, the camera often follows the main character(s) in Castle or Dexter to the crime scene, where other police officers and medical examiners have already started the investigation. The context may make it unnecessary for anyone to ask the subordinates to give an account of what has been found up until that moment. According to van Dijk (2001: 72), speakers in general reuse existing frames when encountering a new situation and so do not need to construct completely new models. Our findings regarding the different lieutenants’ questions fit van Dijk’s argument, i.e. the questions always follow the same pattern.


ST: [L] Did we get an ID? [Dexter 1x02]




TT: ¿Lo hemos identificado? [Have we identified it?]




ST: [B] You got an ID? [Castle 1x04]




TT: ¿Identificación? [Identification?]




ST: [B] You got something? [Castle 1x04]




TT: ¿Tienes algo? [Have you got something?]




ST: [B] So, what you got for me? [Castle 1x04]




TT: ¿Qué tenemos? [What do we have?]




ST: [DO] So, what do we know? [Dexter 1x04]




TT: ¿Qué sabemos? [What do we know?]

Even other characters, not just lieutenants, resort to these formulae when they arrive at the crime scene and are also told about the news. In (6–8), we find Dexter’s questions when he is given an account of the findings made before he is called to examine the body as a blood spatter analyst.


ST: [D] So, what do we know? [Dexter 1x04]




TT: ¿Qué sabemos? [What do we know?]




ST: [D] What do we have? [Dexter 1x04]




TT: ¿Qué tenemos? [What do we have?]




ST: [D] You have an ID? [Dexter 1x02]




TT: ¿La has identificado? [Have you identified her?]

Different translation strategies are observable if we look at examples (1–8):

  • rendering of a complete, stereotyped clause, as in (1), (7) and (8);
  • condensation of the meaning through the use of an elliptical utterance, as in (2), where the noun identificación stands for the whole clause (i.e. ‘¿Lo habéis identificado?’), similar to the TT in (1) and (8);
  • neutralisation of features of fictional orality present in the ST, as in (3), (4), (5) and (6).

The reduction caused by strategies (a) and (b) does not affect meaning. Subtitlers seem to focus on the most essential parts of verbal expression and also provide stereotyped solutions within the target culture. Thus, the principle of relevance applied to subtitling plays a crucial role (cf. Kovačič 1993, Díaz Cintas 2003: 210–214).

Strategy (c) deserves further discussion. The omission of features of fictional orality affects the way in which the audience gains access to a character’s speech. Thus, the marked colloquial register of (3) (i.e. “You got something” vs. “Have you got something?”) is neutralised into standard register in Spanish. In (4), (5) and (6), where so is omitted, we find a less cohesive TT, the style of which is more telegraphic than that of the ST. Interestingly, linking words like so are among the elements listed by Chaume (2003: 167–168) in his inventory of the most frequently omitted linguistic elements in subtitling.

As previously suggested, audiences are expected to be able to grasp the nuances conveyed by linguistic features of fictional orality by listening to the soundtrack (intonation) and watching the picture and photography (facial expressions and gestures)3. However, Chaume posited the idea (2004: 200, our translation) that “the translator must thoroughly study the psychological description of each character.” Discourse markers in particular are among the linguistic elements which convey information about a character’s personality. Therefore, their inclusion or omission in the subtitled version is likely to have consequences for the reception of a character’s personality (cf. Mattsson 2009, Bartoll 2010 and Arias Badia 2014).

The device of repeating question structures has also been detected when lieutenants ask their subordinates about the progress of the investigation, as in:


ST: [B] Esposito, where are we on her cell phone? [Castle 1x02]




TT: Espósito, ¿qué sabemos de su móvil? [Esposito, what do we know about her cell phone?]




ST: [B] Where are we on the boyfriend? [Castle 1x02]




TT: ¿Qué sabes del novio? [What do you know about the boyfriend?]




ST: [B] What about her ex-boyfriend? [Castle 1x02]




TT: ¿Qué sabemos del ex novio? [What do we know about the ex-boyfriend?]




ST: [B] What about these calls? [Castle 1x02]




TT: ¿Y estas llamadas? [And these calls?]

Notice the shifts in the subtitling of (9–12): for two set expressions in the ST (“where are we on”, “what about”), the Spanish version also provides stereotyped questions within the target culture. In (9–11), the subtitler resorted to the verb saber in Spanish. Díaz Cintas (2003: 205) notes that this verb is usually preferred for condensation in subtitling, over alternative forms such as estar seguro or estar convencido. The presence of saber in the above instances supports these findings. Example (12) is the only one in which the verb is not used. Here, the formula “what about” is simplified by a questioning use of the conjunction y [‘and’]in Spanish (literally, ‘And these calls?’).

4.1.2. Ellipsis

Our corpus has provided evidence that, when subordinates give their account, their discourse typically presents situational ellipsis  (cf. Biber et al. 1999: 158), as can be observed in the following examples:


ST: [E] One victim.


Donald Kendall, 18.


He was a senior at Redding Prep.




[B] Redding Prep?


[C] Private school... on Park Avenue. [Castle 1x04]


TT: Una víctima. [One victim.]


Donald Kendall, 18. [Donald Kendall, 18.]


Estudiaba en el instituto Redding. [He studied at Redding Institute.]




¿El colegio Redding? [The Redding college?]


Es un colegio privado en Park Avenue. [It’s a private school on Park Avenue.]




ST: [D] Same guy, same pattern.


[A] Bone dry.


No blood again. [Dexter 1x01]




TT: - El mismo tipo, el mismo patrón. [The same guy, the same pattern.]


|- Completamente seca. [Completely dry.]


Nada de sangre, otra vez. [No blood, again.]


Thus, in (13), Esposito provides basic information about the victim (name and age) in telegraphic style — the verb to be is elided. Later, when he is asked by his lieutenant Beckett about the victim’s school, Castle adds information by eliding the verb again. In the subtitles the colloquial prep is rendered with the standard forms instituto and colegio. In the final sentence, the verb to be is included in the subtitle.

In (14), Dexter gives an account of his first impression with the same telegraphic style, by eliding the verb to be. Batista, the police officer in charge, agrees and adds information without resorting to verbs again. In (14) we find an example of how intrinsic features of the language of crime fiction series contribute to the use of less characters in a line when subtitling.

4.1.3. Pragmatic markers

There are two main strategies employed by forensic experts in the studied sequences to make themselves understood when describing the first pieces of evidence encountered around the crime scene, or the corpse itself. On the one hand they use similes, for example, “like using a ruler”, (Dexter 1x01);“very/almost surgical”, (Dexter 1x04). On the other, they point at the parts of the body or object to which they are referring. This gesture is usually accompanied by a pragmatic marker, such as an imperative or question clause (cf. González 2012) to make their interlocutor pay attention (look, notice, see, check out).


ST: [B] Look -- he cut the leg in four pieces, almost like using a ruler, but this leg is in three pieces. Now, look, he started to make a fourth cut but stopped. [Dexter 1x01]




TT: Mira. [Look.]


Seccionó esta pierna en cuatro partes, [He sectioned this leg in four pieces,]


|parece que las hubiera medido. [it seems as if he had measured them.]


Pero cortó la otra pierna en tres partes. [But he cut the other leg in three parts.]


Mira, empezó a hacer un cuarto corte, [Look, he started to make a fourth cut,]


|pero se detuvo. [but he stopped.]




ST: [D] Look at the blood spatter. Look at the patterns, tells a story. [Dexter 1x01]




TT: Observe las salpicaduras. [Observe the spatters.]


|Cuentan una historia. [They tell a story.]




ST: [D] Angel, look at this. I think there’s something in his mouth. [Dexter 1x02]




TT: Ángel, mira esto. [Angel, look at this.]


Creo que tiene algo en la boca. [I think he has got something in his mouth.]




ST: [D] You see this big pond of blood right there? [Dexter 1x01]




TT: ¿Ve esa enorme mancha de sangre de ahí? [Do you see that enormous stain of blood there?]




ST: [D] Notice the long, thick, heavy drips? [Dexter 1x01]




TT: ¿Se ha fijado en esas gotas [Have you noticed those drips]


|grandes y pesadas qu e cayeron ahí? [big and heavy that fell there?]


The lexical selection of the TT in translating the imperative forms calls our attention: while in (15), (17) and (18) subtitlers opted for short verbs related to the deictic gesture (mirar and ver), they used longer words or expressions in (16) and (19) (observar and fijarse en), which are less deictic. This demonstrates that subtitlers do not always resort to the shortest words available to convey a message, as the above presented example of saber could suggest (see 4.1.1.). Their rendering seems to consider the more or less deictic features of the setting. In these instances, the longer and more explanatory options also elevate the register when compared to the ST.

4.1.4. Specialised lexicon

In spite of the obvious effort on the part of the police and medical examiners to make themselves clear, specialised terminology is often used in this kind of interaction in Dexter and Castle. Some examples of terminological units present in our corpus are cerebral haemorrhaging, temple, temporal bleeding, vaginal laceration or trauma. In the subtitling process, these units undergo different translation procedures. Table 1 lists the three major procedures employed in our corpus and provides an explanation and an example for each of them.



(from Castle 1x02 and Castle 1x04)




The equivalent specialised term is provided in the TT.

cerebral haemorrhaging

hemorragia cerebral




The term in the ST is omitted. Syntactically, emphasis is laid on a different part of the sentence.

Triangulation puts it

Según parece,

What is the evidence of...

¿Cómo sabes que...


Only part of the specialised term is rendered in the TT, i.e., the TT is less informative or specific than the ST.

temporal bleeding


Single GSW to the chest

Un solo disparo en el pecho

Table 1. Major translation procedures for terminology in our corpus

Likewise, two approaches are typically adopted in the scripts of the studied series in order to make terminology clearer “to the lay person,” i.e. other characters appearing in the sequence or, rather, the series’ audience: the first approach consists of explicitly paraphrasing the term, as in example (20).


ST: [DB] And that’s important because?


[D] Sure sign of post-mortem severance.


[D] Meaning: the victim was already dead when she was relieved of her fingertips. [Dexter 1x02]




TT: - ¿Y eso qué significa? [And what does that mean?]


|- Que fue una amputación post mortem. [That it was a post-mortem amputation.]


La víctima estaba muerta [The victim was dead]


|cuando le cortaron los dedos.4 [when they cut off her fingers.]

The second approach involves a character explaining the specialised term: this happens, for example, when a character asks what the medical examiner’s suggestion is. Sometimes this explanation is expressed by means of a simple gesture and the inclusion of humour, as in (21):


ST: [B] So what’s the evidence of sex?


[L] Traces of spermicide.


[B] (Questioning facial gesture. Cf. Castle 1x02, 11:47, ABC 2009)


[L] The guy wore a condom.


[C] Boy, it really has been a long time for you, hasn’t it? [Castle 1x02]




TT: ¿Cómo sabes que tuvo relaciones? [How do you know she had relations?]


Por los rastros de espermicida. [Because of the traces of spermicide.]


(Questioning facial gesture.)


Usó condón. [He used a condom.]


Realmente hace mucho que no lo haces, ¿verdad? [It really has been a long time since you haven’t done it, right?]

If we look at (20) and (21) and focus on the translation of the elements that allow audiences to gain access to the two main terminological units present in each example — namely, post-mortem severance and spermicide —, we can see that in (20) the subtitle omits the reformulating element “Meaning: …” present in the ST. The reader of the subtitles is asked to reconstruct this semantic link to the previous sentence.

Example (21) differs from (20) in that it does not include a reformulating element. The TT omits the subject of the explanatory sentence (“Usó condón.”), which is explicit in the ST (“The guy”). However, the viewer may fill in the meaning from the context. The subtitler simply reproduces the style of the ST — and, as has been observed in other instances in this paper, their task is again made easier by such a telegraphic style.

4.1.5. Modalisers

The last linguistic feature, repeatedly present in preliminary forensic analysis sequences, that we would like to mention here are modalisers, specifically, markers of speculation. Since the investigation is only at its initial stage, medical examiners modalise their speech to a great extent, as may be observed in (22–27):


ST: [D] The cut appears to be almost delicate [Dexter 1x04]




TT: El corte [The cut]


|parece casi hecho con delicadeza, [seems almost made with delicacy.]




ST: [D] The blood flow on the hand suggests the heart was still pumping [Dexter 1x04]




TT: La sangre de la mano sugiere [The blood in the hand suggests]


|que el corazón aún latía cuando la cortaron. [that the heart was still pumping when they cut it.]




ST: [L] From the lack of blood inside the boat, I’d say he wasn’t killed at sea. [Castle 1x04]




TT: Y, dado que no hay sangre en el bote, [And, since there is no blood on the boat]


|creo que no murió dentro. [I think he did not die inside.]




ST: [D] Given the skin rupture and secondary spatter distance, had to be a 50-60 foot drop. [Dexter 1x02]




TT: Por el desgarramiento de la piel [From the skin rupture]


|y la salpicadura, cayó de unos 15-20 m. [and the spatter, he fell from 15-20 meters]




ST: [L] I mean, it could have been a rape, but my best guess is- [Castle 1x02]




TT: Pudo haber sido una violación, [It could have been a rape,]


|pero yo diría que... [but I would say that…]




ST: [A] It’s possible that he got interrupted. [Dexter 1x01]




TT: Es posible que lo interrumpieran [It is possible that they interrupted him]

With regard to the subtitles, our corpus provides samples which preserve typically omitted elements according to Chaume (2003: 167–168). Such is the case of modalisers. In examples (22–27) the context does not allow for the omission of these elements. These modalisers serve as guides for the audience: investigators could not possibly confirm a hypothesis at the start of an episode. They are bound to hedge when presenting the first findings made. Otherwise, the necessary mystery of the crime story would be immediately unravelled and the episode would lose a sense of intrigue.

Contrary to the disappearance of linking word so (see above), the subtitles here always preserve modalisation. In instance (23), the verb of speculation suggest is literally rendered in the Spanish version (sugerir). In (24), we find modalisation in the use of the conditional “I’d say”, which is translated by means of the common thought introducer verb creer in Spanish. Apart from that, the prepositional phrase that justifies the examiner’s hypothesis (“From the lack of blood”) is preserved in the subtitle by means of a subordinate clause introduced by a formal conjunction (“dado que no hay sangre en el bote”).

The only instance where modalisers have not been translated is (25). The marker of speculation “had to be” does not appear in the TT, probably for reasons of space. However, the audience may reconstruct the causal relation by means of the previous prepositional phrase “por el desgarramiento” (literally, ‘because of the rupture’).

These findings suggest that the description of subtitling strategies is far from being consistent and only more in-depth studies of different genres would allow to draw conclusions about the linguistic features of this type of translation.

In their chapter on genres, McCarthy and Carter tell us: “The underlying pattern will, naturally, have different surface realizations which will create different registers, but it is the underlying pattern itself which genre analysts are keen to capture” (1994: 25). In the next section we focus on another basic communicative event that is part and parcel of the underlying pattern of police procedurals, namely the interrogation of witnesses and suspects.

4.2. Text blocks in interrogation sequences

Interrogation has been studied before from different points of view (cf., for example, Kuyumcuyan (2012), on the polyphony of interrogation in fiction). In this section, we would like to describe its main linguistic features and see whether they are preserved in the translated version of our corpus.

Referring once again to van Dijk’s model (2001: 73), we argue that the general domain of interrogation is again criminal investigation and its global act is still for the police to gather information about the victim or suspects. The participants of this discourse practice are the police in the role of interrogator and the suspect or witness in the role of those being interrogated. The gathering of information is achieved by means of questions whereby the police aim to gain access to the knowledge and opinion of suspects, sometimes by also sharing their own opinion in the form of hypotheses. Figure 3 summarizes the context categories of these sequences.

Badia 2

Figure 2. Interrogation sequences: context categories

Since we are dealing with interrogation in this section, we would like to describe the function of the questions included in these sequences. In essence, their purpose is to fill in the gaps in knowledge and to help the plot develop. At the same time, they represent the conventions of the genre and serve as a guide for audiences, since they fulfil their expectation to find these sequences in the product they are consuming. In addition, as dialogical sequences, the fictional orality5 present in interrogation provides useful information to the audience about the characters’ personality, because these are sequences where characters are under intense pressure. Figure 4 depicts the aforementioned functions of questions in interrogation sequences.

Badia 3

Figure 3. Functions of questions in interrogation sequences

What follows is an account of some major linguistic features present in the speech of interrogator and interrogated characters. We shall cover issues which range from the semantics and syntax of information gathering to pragmatic aspects in the reaction of characters being interrogated.

4.2.1. Interrogators’ questions

As in preliminary forensic analysis sequences, question structure is often repeated in the episodes we analysed. We detected (a) general questions, the aim of which is to gather information about the victim or about other suspects; (b) questions trying to place the victim or suspect physically in a determined spot at a determined time; and (c) questions trying to pinpoint the emotional state or the victim’s ideas.

(a) Examples (28) and (29) provide us with general questions, the aim of which is to gather information about the victim or about other suspects. They illustrate how police officers typically ask for information about names and addresses. As it can be seen, these questions do not pose a challenge for subtitlers, probably due to their stereotypical nature. Only small structural shifts can be observed, as in (29), where the colloquial question in the ST is rendered as a modalised, yet affirmative clause in the TT.


ST: [B] Do you know his last name? [Castle 1x02]




TT: ¿Saben cómo se apellida? [Do you know what his last name is?]




ST: [B] I’m guessing she lived here? [Castle 1x02]




TT: Supongo que vivía aquí. [I suppose that she lived here.]

(b) Instances (30) and (31) show questions trying to place the victim or suspect physically in a determined spot at a determined time. In our corpus, these questions are often introduced by formulae such as “Where were you when...” or “When was the last time...”.


ST: [B] So, then where were you the day that Sara got killed? [Castle 1x02]




TT: ¿Y dónde estabas el día en que mataron a Sara? [And where were you the day they killed Sara?]




ST: [B] So, the last time you spoke to Sara was when? [Castle 1x02]




TT: ¿Cuándo la vieron por última vez? [When did you see her for the last time?]




ST: [B] When was the last time you saw your son alive? [Castle 1x04]




TT: [...]-¿Cuándo fue la última vez que vieron a su hijo con vida? [When was the last time that you saw your son alive?]




ST: [B] Do you know, maybe, where he went after? [Castle 1x04]




TT: ¿Saben adónde fue después? [Do you know where he went afterwards?]

Notice that (30–32) provide us with samples of fictional orality in TV dialogue. The adverbs then and maybe in (30) and (33), the marker so or the non-neuter syntactic order in (31) try to evoke colloquial register in the characters’ speech. Interestingly, these elements are not rendered in the Spanish version, which therefore becomes more formal.
c) Examples (34–37) depict questions trying to pinpoint the emotional state or the victim’s ideas in the last days prior to their death:


ST: [B] Do you know of any reason he’d be in the park? [Castle 1x04]




TT: ¿Sabéis por qué estaba en el parque? [Do you know why he was in the park?]




ST: [B] Do you know what he would have been doing in the park? [Castle 1x04]




TT: ¿Saben qué podía estar haciendo en el parque? [Do you know what he could have been doing in the park?]




ST: [C] Did she say anything to you? [Castle 1x02]




TT: ¿Te contó algo? [Did he tell you anything?]




ST: [C] Did she seem upset? [Castle 1x02]




TT: ¿Parecía triste? [Did she seem sad?]

The subtitles of (34–37) also resort to stereotyped questions within the target culture. This kind of question is so well known by audiences of both the ST and the TT cultures that they can be subverted in order to introduce humour. It is quite typical for the character being interrogated to anticipate the officer’s questions. An example of this can be found in Ben Aaronovitch’s novel, Whispers Underground: “This is where, as police, you have to make a decision—do you ask for an alibi or not? Fifty years of detective dramas that even the densest member of the public knows what it means when you ask where they were at a certain time or date” (2012: 125). This subversion of stereotypical blocks links back to our earlier point about innovative situations continuing to make use of fixed frames.

Other stereotypical elements of interrogation are the interrogator’s strategies for gathering information from their interlocutors. The approach of the police in interrogation is to divide up roles between good cop or bad cop. We have found that, in their pursuit of information and confessions, they resort to lexical and syntactic repetition in order to emphasise their hypotheses and cause the interrogated character to speak, maybe against their own best interests:


ST: [B] You followed her down to the basement, didn’t you?


You followed her in the basement and then you killed her. [Castle 1x02]




TT: Usted la siguió hasta el sótano, ¿verdad? [You followed her to the basement, right?]


La siguió hasta el sótano y la mató. [You followed her to the basement and you killed her.]

Again, the Spanish TT in (38) imitates the telegraphic style of the interrogators being much shorter than the ST. The viewer may fill in the omitted adverb then by the logical order of the reported actions, i.e. first the chase and then the murder.

4.2.2. Interrogatees’ reactions

Let us now turn our attention to the reactions of the characters who are asked the above described questions. These reactions often involve a defensive attitude. In this section we discuss the use of the turn-taking discourse marker look and focus on three examples where scriptwriters used genre stereotyped answers as a device for introducing humour.

Previous studies have argued the low presence of certain discourse markers in subtitled crime fiction (cf. Mattsson 2009: 264). In our corpus, the high frequency of the discourse marker look (cf. Biber et al. 1999: 549) in interrogation sequences called our attention. The marker is typically employed by interrogatees before providing an alibi whereby they try to convince the police of their innocence. Consider the examples below:


ST: [BR] Look, we have video cameras there and a sign-in sheet. [Castle 1x02]




TT: Hay cámaras de seguridad y firmamos [There are security cameras and we sign]


|la hoja de control al entrar y salir. [the control sheet when entering and leaving.]




ST: [P] Look, you ask Chloe. She’ll tell you, I was there. [Castle 1x02]




TT: Pregúntenselo a Chloe, les dirá que estaba allí. [Ask Chloe, she will tell you I was there.]




ST: [BR] Look, I didn’t have anything to do with this. [Castle 1x02]




TT: Yo no tuve nada que ver. [I did not have anything to do with it.]




ST: [P] Look, every night that I was supposed to be home having dinner? I wasn’t. [Castle 1x02]




TT: Verá, todas las noches que debía estar [See, all nights that I had to be]


|en casa cenando, no estaba. [home having dinner, I was not.]




ST: [H] Look, we had sex, okay? [Castle 1x02]




TT: Mire, nos acostamos, ¿vale? [Look, we slept together, okay?]

While a quantitative analysis of look in a larger corpus may shed light on this matter, the above examples do not show a clearly inclusive or exclusive tendency in the subtitling of this marker. As can be seen, the subtitles in examples (39–41) omit the marker, while (42) and (43) provide the forms “Verá, ...” and “Mire, ...”, respectively. Both turn-taking markers consist of verbs in the formal second person in Spanish (usted), and thus contribute to elevating the register of the dialogue.

In the previous section we alluded to the role played by humour in the questions of interrogators. But these humoristic effects are not exclusive of this kind of character. Our corpus has provided several samples of interrogated characters responding with such stereotypical answers that they allow a humoristic interpretation by the audience. Some defensive declarations are so stereotyped and therefore empty in terms of alibi that they might make viewers smile. Let us close this section with three examples:


ST: [PE] Well, it was difficult to tell. It was dark. [Dexter 1x05]




TT: Es difícil de decir. Estaba oscuro. [It is difficult to tell. It was dark.]




ST: [P] It’s not what you think. [Castle 1x02]




TT: -Primero, no es lo que creen. […] [First, it's not what you think.]




ST: [BR] No, no, no way, it wasn’t like that. [Castle 1x02]




TT: No, no. No fue así, en absoluto. [No, no. It was not like that, not at all.]

5. Final considerations

The present paper has given evidence that recurrent sequences pave the way for the appearance of recurrent lexical units and syntactic structures. In spite of the limitations of our corpus — more series and episodes would obviously enrich the findings made — we have demonstrated how police procedural scriptwriters play with genre conventions and repeatedly resort to similar linguistic structures. It is our hypothesis that, by the same token, this is also the case for other audiovisual genres. Developers of automatic and computer-aided translation tools could take advantage from further quantitative research on this matter.

Likewise, we have posited the idea that professionals from the field of subtitling could benefit from gaining awareness of the text blocks present in their STs. As we have seen, in practice, subtitlers already fall into step with scriptwriters in that they also tend to provide genre stereotyped structures and lexicon in their translation. It has been argued that the language of subtitles is even more conventionalised than the language of STs (Deckert 2013: 61–63), which leads us to believe that the recurrent text blocks of crime fiction are well set in the subtitling context as well, and thus do not pose a special challenge for the professional. The translations in our examples seem to support this idea: slight changes in the syntactic structure of the translation are observable in just a few instances. We may therefore argue that we have not found examples of content omission caused by the time-and-space constraints of subtitling, nor significant condensation of clauses, which are two frequent strategies for text reduction in subtitling.

6. Acknowledgements

This work was co-funded by the research programme of the Spanish Ministerio de Economía y Competitividad, within the framework of the project La traducción del diálogo ficcional. Textos literarios y textos multimodales (FFI2010-16783), and the UPF-IULA PhD mobility grant program (COFRE).


Dexter (2006). Developed for television by James Manos Jr. Primera temporada. English original version; Castilian Spanish subtitles. DVD. Madrid: Showtime.

Castle (2009). Various directors. La primera temporada completa. English original version; Castilian Spanish subtitles. DVD. [s. l.]: ABC Studios. Distributed by Walt Disney Studios Home Entertainment.

  • Aaronovitch, Ben (2012). Whispers Under Ground. London: Gollancz.
  • Arias Badia, Blanca (2014). “Fictional Orality in the English and Spanish subtitles of Fontane Effi Briest, by R. W. Fassbinder.” Jenny Brumme and Sandra Falbe (eds) (2014). The Spoken Language in a Multimodal Context: Description, Teaching, Translation. Berlin: Frank and Timme, 197–211.
  • Balló, Jordi and Xavier Pérez (2005). Yo ya he estado aquí. Ficciones de la repetición. Barcelona: Anagrama.
  • Baroni, Raphaël (2007). La tension narrative. Suspense, curiosité et surprise. Paris: Éditions du Seuil.
  • Bartoll, Eduard (2010). “Marcas de oralidad en los subtítulos en catalán de Das Leben der Anderen.” Gemma Andújar and Jenny Brumme (eds.) (2010). Construir, deconstruir y reconstruir. Mímesis y traducción de la oralidad y la afectividad. Berlin: Frank and Timme, 187–217.
  • Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad and Edward Finegan (1999). Longman Grammar of Spoken and Written English. London/New York: Longman.
  • Bönnemark, Margit (1997). The Mimetic Mystery. A Linguistic Study of the Genre of British and American detective fiction including a comparison with suspense fiction. PhD thesis. Stockholms Universitet.
  • Brumme, Jenny and Anna Espunya (2012). The Translation of Fictive Dialogue. Amsterdam/New York: Rodopi.
  • Brumme, Jenny (2012). Traducir la voz ficticia. Berlin/Boston: De Gruyter.
  • Cattrysse, Patrick and Yves Gambier (2008). “Screenwriting and translating screenplays.” Jorge Díaz Cintas (ed.) (2008). The Didactics of Audiovisual Translation. Amsterdam/Philadelphia: John Benjamins, 39–55.
  • Chaume, Frederic (2003). Doblatge i subtitulació per a la TV. Vic: Eumo.
  • — (2004). Cine y traducción. Madrid: Cátedra.
  • Deckert, Mikołaj (2013). Meaning in Subtitling. Toward a Contrastive Cognitive Semantic Model. Frankfurt am Main: Peter Lang.
  • de Linde, Zoé and Neil Kay (1999). The Semiotics of Subtitling. Manchester: St. Jerome.
  • Díaz Cintas, Jorge (2003). Teoría y práctica de la subtitulación inglés-español. Barcelona: Ariel Cine.
  • Genette, Gérard (1979). Introduction à l'architexte. Paris: Éditions du Seuil.
  • — (2002). “Des genres et des œuvres.” Gérard Genette (2002). Figures V. Paris: Éditions du Seuil, 39–133.
  • Gläser, Rosemarie (1979). Fachstile des Englischen. Leipzig: VEB Verlag Enzyklopädie.
  • González, Montserrat (2012). “Pragmatic markers in translation.” Jenny Brumme, and Anna Espunya (eds) (2012). The Translation of Fictive Dialogue. Amsterdam/New York: Rodopi, 217–232.
  • Göpferich, Susanne (1995). Textsorten in Naturwissenschaft und Technik. Pragmatische Typologie — Kontrastierung — Translation. Tübingen: Gunter Narr.
  • Inigo, Marta and Debra Westall (1997). “Los realia lingüísticos y la traducción cinematográfica.” Antonio Briz, Maria Josep Cuenca and Enric Serra (eds) (1997). Sobre l’oral i l’escrit. Quaderns de Filologia. Estudis Lingüístics II. Valencia: Universitat de València.
  • Kovačič, Irena (1994). “Relevance as a factor in subtitling reductions.” Cay Dollerup and Annette Lindegaard (eds.) (1994) Teaching Translation and Interpreting 2. Amsterdam/Philadelphia: John Benjamins, 245–251.
  • Kuyumcuyan, Annie (2012). “La polyphonie dans l’interrogatoire policier de fiction: les fonctions textuelles de l’intertextualité locale dans le «Corpus Maigret» de G. Simenon.” Lorda Mur and Clara Ubaldina (eds.) (2012). Polifonía e intertextualidad en el diálogo. ORALIA. Análisis del discurso oral. Anejos, 6. Madrid: Arco/Libros, 63–82.
  • Mattsson, Jenny (2009). The Subtitling of Discourse Particles. A corpus-based study of well, you know, I mean, and like, and their Swedish translations in ten American films. PhD thesis. University of Gothenburg. (consulted 06.02.2014).
  • McCarthy, Michael and Ronald Carter (1994). Language as Discourse: Perspectives for Language Teaching. Essex: Longman.
  • Nusser, Peter (2003). Der Kriminalroman. Dritte Auflage. Stuttgart, Weimar: Metzler.
  • Pfister, Manfred (2001[1977]). Das Drama. Theorie und Analyse. 11th edition. München: Wilhelm Fink.
  • Propp, Vladimir (1968[1928]). Morphology of the Folktale. Second Edition. Revised and edited with preface by Louis A. Wagner. Introduction by Alan Dundes. Austin: University of Texas Press.
  • van Dijk, Teun Adrianus (2001). “Algunos principios de una teoría del contexto.” ALED, Revista latinoamericana de estudios del discurso 1(1), 69–81.
  • Wellek, René and Austin Warren (1949). Theory of Literature. New York: Harcourt, Brace and Company.
  • Wilss, Wolfram (1997). “Textbausteine.” Eberhard Fleischmann (ed.) (1997). Translationsdidaktik. Grundfragen der Übersetzungswissenschaft. Tübingen: Gunter Narr, 36–45.
  • Wolf, Mauro (1984). “Géneros y televisión.” Anàlisi. Quaderns de Comunicació i Cultura 9, 189–198.
  • Zabalbeascoa, Patrick (2008). “The Nature of the Audiovisual Text and its Parameters.” Jorge Díaz Cintas (ed.) (2008). The Didactics of Audiovisual Translation. Amsterdam/Philadelphia: John Benjamins, 21–37.

Badia 4Blanca Arias Badia is a predoctoral researcher, member of the InfoLex research group of the University Institute for Applied Linguistics (IULA) at Universitat Pompeu Fabra. She holds a MA in Comparative Literature and Literary Translation and a BA in Translation and Interpreting. She is currently developing a thesis on audiovisual translation, teaching Spanish and Translation in the Department of Translation at UPF, and working as a literary translator.


Jenny BrummeJenny Brumme is a professor of translation and applied linguistics at Universitat Pompeu Fabra. She leads a research project on the translation of fictional dialogues in literary and multimodal texts (FFI2010-16783). She has presented publications on: neology and lexicology; language policy and standardisation; historical linguistics and historiography of linguistics; language for special purposes and specialised communication; and contrastive linguistics and translation.



Note 1:
The concept text block was introduced by Gläser (“syntaktische Fertigstücke”), who defined the notion as “phraseological units readily available for the author of a text” (1979: 90; our translation). Later, Göpferich consolidated the concept in her classification of text genres in specialised communication (“Textversatzstücke”; 1995: 217). Wilss coined the term text block (“Textbaustein”) and described its main features such as recurrence, suppleness, easy to recognise and to remember (Wilss 1997: 36). The term has always been connected to specialised discourse and contrastive analysis. Therefore, it seems to us that its application to our study is appropriate and useful.
Return to this point in the text

Note 2:
On the examples in this paper: The letter(s) between square brackets provide the initial of the speaking character. We have preserved the form of the subtitles — the symbol [|] portrays a change of line in the subtitle. At the end of each ST, we provide information about the episode from which the caption was extracted, but not the exact timing, in order to favour legibility and because it is not especially relevant for the purposes of the present paper, since we do not deal with technical aspects of subtitling here. In order to help readers understand the Spanish version, a literal back translation is provided for each example.
Return to this point in the text

Note 3:
For a thorough description of the components of the audiovisual text, cf. Zabalbeascoa (2008).
Return to this point in the text

Note 4:
Note the nuance present in the ST “relieved of her fingertips” which is semantically neutralised in the translation (“le cortaron los dedos” vs., for example, “la despojaron de sus dedos”).
Return to this point in the text

Note 5:
Defined as “any attempt to recreate the language of communicative immediacy in fictional texts, including both narrative and theatrical texts as well as audiovisual or multimodal texts” in Brumme and Espunya (2012: 13). For a state of the art on the matter, cf. also Brumme (2012).
Return to this point in the text