Subtitling words or omitting worlds ? A metafunctionally-oriented analysis

This article looks at the phenomenon of omission of metafuncional meanings in the subtitles of the TV series Heroes drawing on the theoretical framework of Systemic Functional Linguistics, particularly regarding the metafunctional dimension and the methodological framework put forward by Kovačič (1998). On these bases, it aims to investigate the meanings translated out of the subtitles in terms of experiential, interpersonal, and textual realizations. The Revista de Estudos da Linguagem, Belo Horizonte, v. 23, n.2, p. 363-388, 2015 364 analysis reveals a different representation in the Brazilian Portuguese subtitles, which end up preventing the spectator from having access to certain meanings realized in the dialogues between the characters in the series, which also impacts directly upon the representation of the flux of events of the episode under study.

Abstract: This article looks at the phenomenon of omission of metafuncional meanings in the subtitles of the TV series Heroes drawing on the theoretical framework of Systemic Functional Linguistics, particularly regarding the metafunctional dimension and the methodological framework put forward by Kovačič (1998).On these bases, it aims to investigate the meanings translated out of the subtitles in terms of experiential, interpersonal, and textual realizations.The

Introduction
In the context of text analysis and translation (see WILLIAMS, J.; CHESTERMAN, 2002), Systemic functional Linguistics (SFL) has been used by Brazilian and international researchers to investigate 'texts standing in a translational relationship' (HALLIDAY, 1964, p. 124).In the international context, scholars such as Blum-Kulka (1986), Hatim andMason (1990, 1997), Bell (1991), Baker (1992), just to mention a few, have explored the potentials of SFL for the analysis of translations, from different perspectives.In the Brazilian context, Vasconcellos (2009) presents a mapping of the Systemic Functional Translation Studies (SFTS) tradition in the Brazilian environment, from its genesis up to developments in the 2000's.Halliday himself has explored the interface between SFL and Translation Studies (TS) in at least five occasions: 1962, 1964, 1985, 1994, and 2001.
Research in Audiovisual Translation (AVT) has also drawn upon the SFL framework to study subtitles.Mubenga (2009Mubenga ( , 2010) ) uses it to analyze film discourse in AVT and to propose what he calls Multimodal Pragmatic Analysis (MPA), arguing that from a translational point of view, the use of the metafunctional categories enables the construction of profiles of linguistic features in the Source Text (ST) and in the Target Text (TT).Espindola (2010) explores the SFL / TS interface for the integration of the different factors affecting meaning production as realized in subtitles, taking into account the environment in which the text comes into life by giving direct attention to the linguistic manifestations of subtitles (ESpInDOLA, 2010).Kovačič (1998) explores the interface SFL / AVT to propose a methodology for the investigation of the phenomenon of omission in subtitling on the basis of the metafunctional categories.
The present study draws upon Kovačič's methodology to investigate how ideational, interpersonal, and textual components are translated out of the subtitles of the TV series Heroes.Inspired by Kovačič's investigation, and revisiting out our own previous study on subtitles (MATIELO;ESpInDOLA, 2011), we revisit the data previously investigated submitting them to a different theoretical framework.Taking into consideration that what does not get translated also bears significance, we look at ideational, interpersonal and textual meanings omitted in the subtitles.
For the purposes of this study, the working definition of omission is 'dropping a word or words from the ST (oral dialogues) in the production of the TT (subtitles)'.This procedure can be the outcome of the cultural or linguistic clashes that exist between the SL and the TL.In the case of subtitling omission may be also due to technical constraints such as: (i) spotting (the segmentation of the translated dialogues in a subtitle to fit screen spaces and / or to optimize their display); (ii) timing (the duration of the subtitles on the screen, including their in and out time); (iii) change of mode; (iv) (inter)cultural effects; (iv) demands from subtitling laboratories, clients and film distributors (CARVALHO, 2005;CATTRYSSE, 1998;DE LInDE;KAY, 1999;GAMBIER, 2002;GOTTLIEB, 1992GOTTLIEB, , 1994)).
In this context, the research questions put forward to frame the present study are: • Where do omissions occur in the subtitles of the TV series Heroes?• Regarding the three metafunctions, what kind of linguistic items were omitted in the subtitles of the TV Series Heroes?• What meanings do the omitted linguistic items translate out of the subtitles?
This paper is structured as follows.After this introductory section, subsection 2 discusses omission in subtitles; then, subsection 3 presents in more detail the SFL / AVT Interface; subsection 4 presents the research design and analytical procedures for the present study; subsection 5 analyzes and discusses the cases of omission in the data under study; finally, subsection 6 presents some concluding remarks as regards the study.

Omission in Subtitling
One of the most frequent criticisms leveled at subtitles refers to omission or reduction of information which frequently happens, given the constraints to which subtitling is submitted.Subtitling involves the rewriting of a discourse: the written to be spoken original dialogues get rewritten in the form of subtitles appearing on the screen for a different audience.Due to such 'transfer of mode of meaning', subtitles are seldom an absolute verbatim of what was spoken in a given dialogue.One of the motivations for text reduction in the subtitles regards the fact that viewers / listeners can absorb speech more quickly than they can actually read. 1his means that subtitles are rendered in a way that the target audience is able to register, process, and understand what is written at the bottom of the screen (DIAz-CINTAS;REMAEL, 2007) in synchronicity to image.
In the production of subtitles, many aspects account for reductions, one of them being the semiosis of subtitles regarding the mode of meaning.While spoken language tends to be filled with false starts, unfinished sentences, redundant speech and interruptions, written language may present grammatically complex structure with intricate interrelationship between clauses, higher number of nominalization, with higher content words that pack up the information that consequently require economy of language when rendered into subtitles (TVEIT, 2009).Another aspect regards viewers' reading capability.Viewers are expected to have reasonable time to combine reading, watching, and listening to the film so that they can understand the story that unfolds.Finally, the size of subtitles, that is, subtitles are usually no longer than two lines, which constrains the amount of information that the text may contain, depending on the speed that the dialogues were uttered on the spoken version.
According to Diaz-Cintas and Remael (2007), reduction can be partial or total.In practice, what may occur is a combination of both forms, which is usually the typical type of subtitling known today.The typical criterion is to exclude what does not hinder comprehension of the message and to reformulate the remaining message in a concise form.This is not, however, the only force -direct or indirect -behind subtitling production: clients, reviewers, and laboratories may, and typically do have a word on the final decision.
The analysis concerning deletion or compression of information is to be viewed as context-dependent, since text reduction will not only vary from film to film, but also from scene to scene (DIAz-CINTAS; REMAEL, 2007) as most films or programs are produced having a preestablished target audience.Sometimes, losses in lexicogrammatical items can be compensated by other semiotic means, such as pictures, images, and soundtrack.Such a semiotic environment, however, is out of the scope of the present investigation, though it is acknowledged as a necessary background in the study of subtitles and will be referred to when needed in the analysis.

SFL / AVT Interface
According to Vasconcellos (1997, p. 21), the Dictionary of Translation Studies (SHUTTLEWORTH; COWIE, 1997) equates linguistics approaches to translation studies to a monolithic conceptualization of describers of translation as a substitutive operation.To her, and we tend to agree, translation is not a mere transfer or replacement of linguistics units as put forward by formal linguistics.Translation studies adopting linguistics parameters to investigate the translated product are bound to the very definition of what linguistics is being talked about.In this sense, the linguistic approach taken to shed light in the investigation of the translated text of the present studynamely Systemic Functional Linguistics -does not see language as a formal system, but as a system of social semiotics that establishes a 'close connection between the linguistic system and other semiotic systems' (BUTLER, 1988, p. 96).This perspective and the connection of semiotic modes is of utmost importance for the study of subtitles as it allows the researcher to look at the integration existent and the different factors affecting the meaning production construed (at times left out) of subtitles.
The dialogue between SFL and TS has contributed to Systemic Functional Translation Studies (SFTS), as documented by Vasconcellos (2009).SFL has shed light on the study of textual dimension of TS, especially in the 2000's: the author offers a chronicle of SFTS in Brazil and abroad and carries out "a mapping of research (…) in a historically embedded scenario" (p.588).As her chronicle shows, a text-oriented TS in Brazil in the late 90's point to a metafunctionally-oriented research based: the studies focused on the experiential and logical dimensions of language, the interpersonal evaluations of language in translated texts, and the thematic structure and cohesive devices found in texts.
SFL as a conceptual tool has influenced research in TS in a number of ways, providing researchers with a linguistic device to analyze both ST and TT and their linguistic construals.Despite being in an infancy stage in Brazil, Matthiessen (2007, cited in VASCOnCELLOS, 2009) considers it a "feverish phase", to which the present research hopes to contribute.

Kovačič's Experimental Study
Six subtitlers translated a passage from a television drama (Long Day's Journey into Night by Eugene O'neill) and their translations were analyzed based on four categories related to textual parameters.These categories are: (1) number and organization of subtitles; (2) Text reduction in terms of linguistic functions; (3) Language registers; (4) Dramatic and conversational structure.
Regarding (1), Kovačič (1998) observed that five subtitlers used a similar number of subtitles to render the dialogues of a selected passage, resulting in 120 to 134 subtitles each.It is important to point out that only one subtitler deviated considerably from the average number of subtitles, producing 157 subtitles for the same passage.In the experiment, there were only 4% of one-liners, corresponding to 32 subtitles out of 802 (the total amount of subtitles produced) and the one-liners were more frequent in the subtitles produced by subtitlers D and E (8.9% and 6.7%, respectively).
As to (2), the focus was given to linguistic reductions and condensations in the subtitles.The results suggest that the function performed by a linguistic item may determine whether it will be preserved, reduced or discarded.Kovačič states that ideational elements tend to be preserved.The author claims that attributes, circumstances, coordinated and appositional constructions, and information expansions are favorite candidates for omission.As for interpersonal elements, she found they tend to be translated out of the subtitles since they are redundant when combined with the multisemiotic nature of subtitles.
As regards (3), there are, at least, two modes involved in the production of subtitles.Written subtitles play the role of the original spoken dialogue.The dialogues of a film, however, are not merely spoken discourse due to being written -elaborated -to be 'read' / performed as spoken language.In this sense, the two modes encountered in a subtitled film carry, sometimes, distinctive language features.Based on such features, Kovačič raises a theoretical question regarding the nature of subtitles: "are they an independent stratum of language, intermediate between speech and writing, or are they rather a sub-category of either of the two"?(p.80).
In what regards (4), features related to dramatic and conversational structure are explored.The author's findings point to the characters' utterances serving to create the atmosphere in the scenes due to the use of repeated expressions and reuse of adjectives.She adds that a subtitler who is sensitive to the dramatic pattern of the text will try to be consistent in preserving the details that contribute to the overall atmosphere instead of immediately focusing on linguistic expressions and sentence meaning.Subtitlers who are aware of the value of interruptions caused by reductions might choose to sacrifice some ideational content and maintain the interruptions, helping viewers construct the understanding of the characters' personal features.Finally, hesitations, false starts, incomplete sentences, and other similar features of the spoken mode tend to help viewers build the understanding of the characters' personal traits.

The data source
Heroes is a north American science fiction television drama series created by Tim Kring.Heroes premiered on NBC on September 25 th , 2006.The plot tells the stories of disparate and ordinary individuals from around the globe who inexplicably develop superhuman abilities, and it also displays their roles in preventing catastrophes so as to try and save humanity.These disasters are usually foreseen in painted images from precognition painters (<http://www.nbc.com/Heroes>).The series was produced by Universal Media Studios in association with Tailwind Productions, and it was shot primarily in Los Angeles, California.The executive producers for the show are Allan Arkush, Dennis Hammer, Greg Beeman, and Tim Kring.Four complete seasons aired, ending on February 8 th , 2010.(<http://en.wikipedia.org/wiki/Heroes_(TV_series>).The series had a total of 88 episodes, distributed in five volumes / four seasons.
Two main reasons account for the selection of Heroes as the object of study in the present research: (a) the fact that it was analyzed within a different theoretical framework in a previous study (MATIELO;ESpInDOLA, 2011), and was originally looked at from a set of three texts: the ST, containing the original English speeches uttered in the dialogues, and the TT -the subtitles -subdivided into Official Subtitles (OS), rendered by Leonardo neves (Drei Marc Company) and non-official Subtitles (nS), rendered by the internet group 9 th Wonders.Episode Four (Collision) -Season One -was the first episode subtitled by the non-commercial subtitlers, and therefore was used in that analysis; (b) the fact that the same data was then submitted to theoretical constructs from belonging to the interface of AVT and SFL.This paper then focuses on the original spoken discourse and exclusively the official subtitles of Collision.
When dealing with the study of subtitles, defining the source text is not an easy task.As Toury (1995) states, "the identity of the source text(s) will have to be established each case anew" (p.76), since many candidates may be considered as ST.As Carvalho (2005) points out, in order to translate for cable TV, the subtitler may receive a VHS, CD, or DVD of the film to be subtitled.According to her, scripts are not always made available for the subtitlers, suggesting that the film will play "the role of the original" (p.106) in this task.In the case under investigation, Leonardo Neves, Heroes' official subtitler, used the film and the script (containing the speeches) as the ST to translate for the subtitles.This study then takes the English original speeches as the ST and the BP subtitles of the DVD as the TT.

Analytical procedures for data collection and data analysis
From data collection to data analysis, analytical steps were taken to qualitatively interpret the data: 1. Manual transcription of the English spoken dialogues from the DVD Box commercialized in the Brazilian market; 2. Manual transcription of the BP subtitles; 3. Crosschecking of the subtitles with their original counterparts; 4. Identification of occurrences of omissions in the subtitles at clausal level; 5. Segmentation of the clauses; 6. Analysis of segments that presented omissions concerning the metafunctional element translated out of the subtitles; and 7. Metafunctional classification of the analyzed segments according to the type of linguistic element that tended to be omitted.

Metafunctional classification
The occurrences of omission identified in the selected episode are analyzed in the light of the three SFL metafunctions: experiential, interpersonal, and textual, since each of them refers to "distinctive kinds of meaning that are embodied in the structure of a clause" (HALLIDAy; MATTHIESSEn, 2004, p. 59).Also, having a tridimensional view of the occurrences allows for a more holistic view of the meanings left out of the subtitles, for "it is the structure as a whole, the total configuration of functions that construes or realizes the meaning" (p.60).Adapted from Halliday and Matthiessen, Table 1 shows the three metafunctions, definitions, and corresponding status in the clause:  Halliday and Matthiessen (2004).
Omissions were considered to be the lexical items originally present in the ST, but which had no translated equivalent in the TT; an example is presented in tabular form below:2

ST
There 's a man arriving this afternoon from New york.

TT
Um homem chegará hoje de Nova york.
Source: the authors.
Omission is not investigated here in four distinct situations, namely: (i) when occurring below clausal level; (ii) when deriving from differences between the two linguistic systems of the language pair at stake; (iii) when the elements of the visual mode4 could fill in 'the gaps' of the omitted elements or could provide the spectator with complementary information allowing them to overcome the missing written information in the form of subtitles; (iv) when the Subject is retrieved from verbal inflection, which is possible in the linguistic system of the TT, as Table 3 shows.

Analysis and Discussion of Findings
A total of 48 omissions were found to exist in the data set of this study.For the analysis, the omissions were subdivided into three categories, namely: (i) experiential; (ii) interpersonal; and (iii) textual elements, which were subclassified according to their function in the clause.

Omission of experiential elements
Lexical items realizing the experiential metafunction was the most recurrent type of element omitted in the subtitles here investigated.Considering the 48 occurrences of omission found, 26 occurrences of them refer to the experiential metafunction, which means more than half of the occurrences deleted the experiential element from the clause.Table 4 provides an example of such a kind of omission: Source: the authors The above occurrence is a case of an omitted process: in the ST, the realization of two processes, while in the TT the realization of only one process in the clause.The existential process from the ST was rendered into a material process in the TT.According to Halliday and Matthiessen (2004), "while existential clauses are not, overall, very common in discourse" (p.256), they have a central role when "they serve to introduce central participants" (p.257), which is the case of the source text realization as shown above.Moreover, it is possible to notice that there has been a rankshift from a nominalization into a clause to simplify the reading of the subtitle.
Another case of omission of process is the case presented in Table 5: Two occurrences of experiential elements are presented above: (i) omission of process and (ii) omission of Participant.The relational clause assigns the identification to the accidental nature of either the 'rape' or the 'murder'.The accident -the value -realized in the ST can be retrieved by contextual information in the multisemiotic nature of the scene, and also via cohesive ties of reference, therefore saving space in the subtitles.The support of the multimodal environment of subtitling accounts for the omission in this particular case.What is significant, however, is the shift of a single interrogative into two clauses demanding information from Source: the authors.
Source: the authors the listener / interlocutor.In what regards omitting both participant and process, the example below, shows the extreme case of the necessity of omitting an entire dependent hypotactic clause.Two Participants (Carrier and Attribute) and one process (relational) deleted from the subtitles.In fact, it was the entire clause from the ST -and make it a horse race -that was left out in the TT.The omitted clause refers to the competition that Mr. Linderman's money would enable the congressman to engage in.In this case, with Linderman's $2 million, the difference between the candidates would drop from 8 to 3 points, and hopefully a horserace would then be possible, increasing the chances of his winning the elections.Taking this context into account, it is possible to argue that the spectator does not get this particular information from the subtitle.The information stops where the 8 points become 3, and therefore the omission prevents the audience from knowing the entire Participant's thought on the unfolding that the money may cause.However, if translated into BP in its totality, the entire omitted clause would have taken up an extra 20 characters of the screen space and this would have required another subtitle since the information "Com os US$2 milhões dele, os 8 pontos podem virar 3" had already used up a two-liner, not allowing for synchronicity of written and spoken discourses.Not only would space have been a constraint, but also time for the utterance following this one comes immediately after.In turn, this would not allow time for another subtitle to be inserted and properly digested.
The occurrence below presents a case of omission of participant with different functional status: Source: the authors.

ST
you want answers?Go find them.

OM
Part.
In the occurrence above, the ST comprised two clauses, which were rendered into a clause complex.The subtitled clause presents some features that merit closer attention: (i) A relationship of expansion, which, according to Halliday and Matthiessen (2004), occurs when "the secondary clause expands the primary clause, by (a) elaborating it, (b) extending it or (c) enhancing it" (p.377); (ii) The interdependency of the clause is that of hypotaxis because there is a "relation between a dependent element and its dominant, the element on which it is dependent" (p.374); (iii) the characteristic tone is of conjunction, since the subtitled clause makes use of a new element that was not present in the ST -"Se" (Back Translation (BT): If).This conjunction is in fact an addition made in the subtitles so as to guarantee the combination of the two clauses into a clause complex in the TT.There are two participants in this pair of occurrence: One that is elided in the dependent clause -Você (BT: You) -and the other one that belongs to the main clause -the Goal.While in the ST the Goal -answers -is represented in the second clause by a personal pronoun -them -, the same is not true in the TT since such item is retrievable by an anaphoric relation to its lexical element respostas realized in the dependent clause.
It was not only Processes and Participants that were omitted from the subtitles.On the occurrence below, Circumstance is found to be omitted.Source: the authors.
The Circumstance of location translated out of the subtitles realizes in the ST the place being referred to.Circumstances of location tend to "construe the unfolding of the process in space and time" (HALLIDAy; MATTHIESSEN, 2004, p. 263).This circumstantial information of space omitted may be implicit in the chain of events of the episode: minutes before this utterance, the character is leaving his campaign office when Dr. Suresh tries to approach him, but the congressman's bodyguards will not let the scientist come closer.Later on, when the character is telling his brother about this situation, he mentions that Dr. Suresh came by his campaign office that morning.
Most commonly omitted, experiential elements were not the only deletions observed in the TT.Interpersonal elements were also left out and portray a different kind of implication.This is discussed next.

Omission of interpersonal elements
Elements belonging to the interpersonal metafunction were omitted as: mood adjuncts -07 occurrences; interpersonal metaphor -04 occurrences; modal operators -03 occurrences; and comment adjunct -01 occurrence.The pair of clauses displayed on the table below regards an omission of a lexical item which realized an interpersonal metaphor of modality in the ST:

OM Metaphor
Source: the authors.
As Martin et al. (1997) highlight, interpersonal metaphors of modality "can be regarded as comparable with Adjuncts like probably" (p.25) and they "expand the speech functional options, for example in the area of politeness" (p.Furthermore, the character's utterance was significantly altered from an interpersonal standpoint, considering the fact that the interpersonal metaphor of modality was entirely translated out of the subtitle.The assertiveness of the character in the TT is reinforced by the use of the BP adverb "já" (BT: "already").
Continuing with interpersonal metaphors, the tables below present an omission of interpersonal metaphor as well as some other peculiarities worth of discussion.In the first occurrence -Table 10 -there is not only omission of Finite, but also an omission of Subject, that is, the Mood is inexistent as a visual written realization for the target audience, but a realization possible to be retrieved from context, the co-text and also the visual cues available in the scene.The implication of omitting the entire first clause (the one from TABLE 10) implies that the modal operator assigning high obligation -have to -to the interlocutor of the discourse is not available for the Brazilian audience, thus lowering the obligation value on part of the doer of the action as read from the subtitles.By omitting the interpersonal metaphor -I mean -the subtitle ends up by eliminating traces of elaboration that the character makes upon the message that is being exchanged in the ST.
Source: the authors

ST
I mean, you believed it this morning.TT * Acreditava hoje de manhã.

OM Metaphor
Source: the authors Another case of omission of Finite may be seen in the pair of instances next: An omission of the Finite verbal operator may is observed in the TT as given above.The finite element "has the function of making the proposition finite" (HALLIDAy; MATTHIESSEN, 2004, p. 115).Finites make something arguable by means of primary tense or modality.In this particular case, the low modal operator "may" has the role of modalizing the discourse.This modalizer was translated out of the subtitle, which led the TT to bring a new construal with a new "assessment of validity of what is being said" (p.116).The new assessment of the validity contributes to the absence of the expression of indeterminacy in the TT, which means that the new construal erases any trace of doubt or modality in the ST.
Not only modalizers were translated out of the subtitles, but also mood Adjuncts functioning as implicit personal judgments, such case is discussed next:

OM Mood Adjunct
The omission, above, of the mood adjunct "strongly" is compensated by the different choice of process.The strength of the mood adjunct is, to a great extent, expressed or construed in the process used in the TT: odeia (BT: hates).The option of using another process, however, denotes the same process type: mental, which conveys a similar meaning to both audiences.In this sense, such a compensation has not caused 'strangeness' for the audience of the TT who were able to experience the same kind of construal as the one of the ST.

OM Finite
Source: the authors Source: the authors Another case of mood adjunct omission is observed in combination with an omission of a comment adjunct which was realized within two clauses, however, distinctively realized in the TT:

OM Modal Comment Adjunct
Mood Adjunct Martin et al. (1997) define comment adjuncts as adjuncts that "provide an attitude towards, or comment upon, the exchange itself or the information being exchanged" (p.63).Additionally, Halliday and Matthiessen (2004) explain that these adjuncts are more related to the speaker's judgment or attitude towards the message and its content.The mood adjunct is "associated with the meanings construed by the mood system: modality and temporality, and also intensity" (p.126).What may be read from the above realization is that the character's comment or attitude towards the exchange or the information in the message is left out of the meaning realized in the translated text, leaving the target audience with no referent of such construal.Also, the emphasis foreground in the mood adjunct "do" 6 is excluded from the subtitles.Consequently, the meaning that these two interpersonal elements realize as the character's comment or assessment of the message is not made available for the Brazilian spectator.

Omission of textual elements
Omissions of lexical items realizing the textual metafunction in the subtitles of Heroes were the least common type of omission found in the data.Considering the 34 clauses that presented omission of lexical items, only seven occurrences of elements of the textual metafunction were found.The occurrences brought to discussion below are the ones Source: the authors 6 This lexical element is not interpreted here as a finite, such as in the clause "They do not know about this", given the fact that it realizes a relatively similar function of the lexical item "really" in the clause, emphasizing the speaker's attitude toward the exchange.
that are seen to have a higher implication for the comprehension of the subtitles alones.
Conjunctions were the least frequent cases of omission in the data source of the present investigation; however, they bear implications when not realized in the clause: As Halliday and Matthiessen (2004) state that "(…) a conjunction is a word or group of words that either links (paratactic) or binds (hypotactic) the clause in which it occurs structurally to another clause" (p.81).On top of linking and / or binding, conjunctions also establish a relationship of expansion or projection in the clause.In the case presented in table 17, the conjunction translated out of the subtitle had a paratactic function, that is, it belonged to a complex in which the clauses could stand out independently.Analysis of the ST shows that the adversative conjunction establishes a relationship of expansion with the main clause, typical of conjunctions (MARTIN et al., 1997) and it also realizes a relationship of extension, adding meanings to the main clause.However, the adversative meaning realized by the conjunction in the ST is construed in the TT by means of polarity where the opposition is created with the positive on the first clause and the negative on the second clause.Such choice gives the target reader the two extremes of the polarity cline for the two participants involved as Goals in the realization of the material process.
And as the other type of textual element omitted from the TT lines, a continuative is given as an illustration: Source: the authors

OM Continuative
Source: the authors Textual continuatives typically signal a move in the discourse.The omission presented above occurs along with the omission of an entire clause, probably motivated by the change of process, from relational -are -to mental -sabemos -(BT: (we)know).The fact that the 'knowing' is realized through a nominalization in ST cannot be disregarded.However, what is picked up on the TT is a continuation of the dialogues, but one which does not signal the character's development of the idea or the wrapping up of the topic being discussed, since the use of So is nonrealized in the translated text.

Summary of the Findings
Throughout episode four -Collision / Season One, the subtitles of the TV series Heroes presented a total of 48 omissions of lexical items, distributed over 34 clauses.A panoramic view over the findings is provided in Figure 1, which shows the metafunctional components and the number of omission of each component: As Figure 1 shows, the highest number of omissions occurs with lexical items realizing the experiential metafunction, with a total of 12 omissions of participants, 8 omissions of circumstances 6 omissions of processes.The second highest number of omissions occurs with lexical items that realize the interpersonal metafunction, with a total of 7 omissions of mood adjuncts, 4 metaphors, 3 finites, and 1 modal comment adjunct.The lowest number of omissions was found in terms of lexical items realizing the textual metafunction, with a sum of 4 omissions of continuatives and 3 omissions of conjunctions.
The kinds of linguistic items translated out of the subtitles of the TV Series Heroes that were most commonly found in the subtitles were those realizing the experiential metafunction.The omission profile figures a total of 12 omissions of participants -most of them being Goal or phenomenon; 8 omissions of Circumstances -most of them being Circumstances of time and place; and 6 omissions of processes -most of them being relational processes.The second most frequent type of omitted element was that of the interpersonal metafunction, which consisted of 7 omissions of mood adjuncts, 4 interpersonal metaphors, 3 finites, and 1 modal comment adjunct, totalizing 15 omissions.The least frequent type of omission found refers to the textual metafunction: a total of 7 occurrences consisting of 4 continuatives and 3 conjunctions.
These findings suggest that new meanings are construed by force of these omitted linguistic elements, revealing that the TT meanings construed sometimes differ when compared to those of the ST.Considering the omissions of experiential components -Participants, processes, and Circumstances -the goings-on represented in the film are construed by means of an economic language in the subtitles (TVEIT, 2009).Moreover, the flux of events is represented in a somewhat different fashion in the sense that the subtitle does not advance what is about to happen in the scene as the spoken language in the dialogues does, so the information load is left to the next scene, which complements what was subtitled in the previous one.Hence, the spectator who relies on the subtitle alone to follow the unfolding of the scenes in the series is not necessarily provided with the same amount of information at the same time that the dialogues do.
Omission of participants represents 25% of the total number of omissions found in the data (12 out of 48) or around 46% of the experiential omissions (12 out of 26).This may suggest that the TT construal does not place focus on the Participant realizing the process, but on the processes per se.This is an important finding for meaning construal in the subtitles, since processes are the core of any meaningmaking configuration.The new construal of the TT with frequent omissions of Circumstances might be suggestive of the influence of the visual element on the audio and written materials.In the case of Heroes, the multisemiotic elements play an important role in filling in the gaps where omissions were observed.Time and place are typical lexical elements to be inferred by the target spectator through the visual cues of the film, and therefore, when subtitles do not realize them, they do not tend to be an obstacle of meaning processing.However, we are not in a position of making stronger claims, given the fact that the analysis of the multisemiotic environment inherent in this audiovisual subtitled material was out of the scope of the present research.
As for the new meanings construed from the omission of elements from the textual metafunction, the TT construal, at times, does not present links, adversative information, or clear indication of new moves made by the speaker as realized in the ST.Hence, considering the clause as message, these changes in the configuration of the message can affect the way the characters' utterances are structured, as shown in the discussion above.In other words, the linguistic construal of subtitles may achieve coherence through the continuity of the visual material (KOVAČIČ, 1998), for instance, and not by the text (subtitles) alone.This suggests that subtitles stand out as having a language configuration of their own, binding features of spoken and written discourses in one single text.

Final Remarks
This study set out to investigate where omissions occur in the subtitles of the TV series Heroes, what kind of linguistic items are omitted in the subtitles of the TV Series Heroes, regarding the three metafunctions, and what meanings the omitted linguistic items translate out of the subtitles.
When ST and TT are compared, the subtitles show a much more direct and assertive configuration.The TT construal is, frequently, less modalized and less modulated, observed from the omissions of mood adjuncts and finites.This suggests that the character's utterances present a new construal from an interpersonal standpoint, thus presenting a mismatch in terms of directness and modulation, affecting the way a character is perceived in the story being told.The omission of evaluative expressions, which expand on the speech functional options in terms of politeness, impacted on the interpersonal negotiation among the participants, thus generating a new interpersonal configuration on the translated text.
The omissions found in this study confirm Kovačič's findings partially, but refute some of her claims.Concerning the omission of lexical elements in terms of their function, Kovačič (1998) argued that the function performed by a linguistic item determines to a great extent whether it will be preserved, reduced or discarded.The author claims that experiential elements tend to be preserved most, which was not found in the present investigation.In fact, results obtained here suggest the very opposite, since experiential items were omitted the most in translated text here concerned.Therefore, differences in terms of the results found may be accounted for by language specificities inherent to the different language systems at stake, which at this point remains only as a speculation.
Finally, without the intent of disregarding all of the restricting aspects inherent to subtitling, unveiling what could have been construed but was not is also important and should be taken into account in the realm of AVT. As Martin et al. (1997) suggest, a functional view of language, with an emphasis on linguistic choices, gives speakers the right to make their choices among the options they hold.After all, what is derives its meaning from what might have been but was not chosen.In this sense, an additional layer of analysis is brought to the fore: understanding what gets translated and what does not can be useful when it comes to analyzing what translators do and to identifying possible translational patterns across texts types.
58).The subtitle presents a new construal which stands out as much more direct and less modalized discourse -Nós já sabemos o que você é (BT: We already know what you are), instead of what could have been rendered -Eu acho que nós (já) sabemos o que você é (BT: I think we already know what you are).

Figure 1 -
Figure 1 -Overview of the Metafunctional Omissions

Table 1 -
Categories for analysis Source: the authors; adapted from

Table 3 -
Example of omission with elliptical Participant

Table 5 -
Process and Participant

Table 6 -
Process and Participants

Table 7 -
Participant Source: the authors.

Table 14 -
Modal comment adjunct and mood adjunct