The long and the short of it: the translation of non-finite adverbial clauses and ly-adverbials

(1)

The long and the short of it

– the translation of non-finite adverbial clauses and ly-adverbials

Author: Anna Görman Supervisor: Magnus Levin Examiner: Jenny Ström Herold Term: Spring 2020

(2)

Abstract

This study investigates the translation of non-finite supplementive clauses and one-word adverbials with a suffix of -ly in an English non-fiction text of academic prose and its Swedish target text. The results show that the non- finite supplementive clauses often are translated into either a new main clause, a coordinated clause or a subordinate clause, where the latter in a majority of cases involves the use of explicitation. The main clause strategy proved the most frequently used, indicating a possible connection between choice of translation strategy and source text sentence length. The ly- adverbials show a clear tendency for translation into one-word open-class adverbials in Swedish, most frequently with a suffix of -t.

Clear differences were found between the investigated adverbial structures regarding movement; the supplementive clauses retain their source text positions in a vast majority of cases whereas the ly-adverbials show a higher frequency of movement, most commonly from their original source text position to sentence-final position in the target text.

Other factors proven to impact the choice of translation strategies are compliance with Swedish preference regarding adverbial placement (in turn dependent on the type, grammatical structure and length of the adverbial as well as the register of the source text), the clarity and readability of the target text as well as style and level of formality.

Key words

Adverbial placement, explicitation, ly-adverbials, non-finite supplementive clauses, translation studies, transposition.

(3)

1 Introduction

From a translation perspective, adverbials are rewarding structures to study. They occur in an abundance of text types, come in a variety of shapes and sizes and can take a range of positions in a sentence, making them grammatically versatile and flexible. However, they also frequently pose a challenge for the translator as they often are closely linked to the information structure of the text, at the same time as they can be rather semantically ambiguous and sometimes lack a correspondent structure in Swedish.

As a small contribution to the rich, and still relatively new, research field that is Translation Studies, the present study investigates adverbial translations from English to Swedish in a non-fiction material. This study is limited to focusing on two specific adverbial structures. Firstly, adverbial clauses are investigated, limited to non-finite clauses that contain a verb with the ing- or ed-participle, as in (1) and (2), and secondly one-word adverbials ending in -ly, as in (3) and (4).

(1) The revolution in taking the kind of approach that I have just done, recognising vampires as culturally constructed […] (Whisker, 2016:166)

Den revolution det innebär att närma sig saken på det sätt jag just gjort – att se vampyrer som en kulturell konstruktion […]

(2) Brite reclaims the vampire’s engagement with an expression of contemporary issues, creating a place of lost hopes, emptied out values […]

(ibid., 2016:170-171)

Med nutida problem som uttrycksmedel ger Brite vampyren sin aktiva roll tillbaka och skapar en plats av förlorat hopp och uttömda värderingar […]

(3) The female vampire is conventionally represented as abject because she disrupts identity and order. (ibid.:161)

Den kvinnliga vampyren skildras traditionellt som en abjekt varelse eftersom hon stör känslan för ordning och identitet.

(4) The historical Elizabeth Baathory was finally found guilty […] (ibid.:162)

Den historiska Elizabeth Baathory befanns till slut skyldig […]

As shown above, different strategies can be used when translating these structures. An adverbial ing-clause can be translated into an infinitive clause (as in (1)) or a coordinated clause (as in (2)), but also into, for example, a new main clause, a subordinated clause or a noun phrase. Ly-adverbials can be translated into, for example, other adverbials (as in (3)), adverbial phrases or prepositional phrases (as in (4)), but also be restructured into other word classes. As the choice between the available translation strategies often is less than straightforward, this study hopes to provide some insight as to the how different strategies can be used, as well as the possible factors behind them.

(5)

1.1 Aim and scope

The aim of this essay is to investigate a non-fiction translation from English to Swedish, focusing on the translating of supplementive ing- and ed-clauses and ly-adverbials.

This study addresses the following research questions:

1) What Swedish correspondences of the investigated grammatical constructions are used in the target text and to what extent?

2) What sentence positions do the investigated grammatical constructions occupy in the English source text and Swedish target text and to what extent?

3) What factors affect the translation choices of the investigated grammatical constructions?

1.2 Material and method

1.2.1 Source text

The source text (henceforth ST) in the present study consists of pages 160–173 of “Vampire Bites”, chapter 7 in Contemporary Women’s Gothic Fiction by G. Whisker (2016), a non- fiction work that summarises and critically explores previous literary research on the topic of Gothic fiction written by women. The text introduces and discusses different aspects of the vampire character in Gothic writing, using psychoanalytic theories as framework.

Using Reiss’ (1971/2000, in Munday 2016:115–118) function-based text classification categories, the ST is mainly informative, as it aims to convey information, knowledge and opinions from the author to the reader regarding a given topic. According to Reiss, this means that the main focus when translating this type of text is to produce a target text (henceforth TT) that carries over the factual contents of the ST, using explicitation when required to carry the message and without stylistics being a major factor (ibid.). However, subordinate functions of the present ST can be identified, which in turn have a bearing on translation choices. For example, there is an expressive element to the text, where the style and preference of the author (both in syntax and lexis) shines through. Even if the main function of the ST is informative, which means that this will be the highest priority in the TT, the form of the text has also been taken into account, in the hope of constructing a TT that produces a reader response equivalent to that of the ST.

The style of the text is quite formal, as can be expected from academic writing. The syntax is characterised by long, complex sentences and compressed information structures, in which the investigated ing- and ed-clauses are often found. On a vocabulary level the ST

(6)

showcases a variety of advanced terminology within its research field, giving an impression of the intended reader being someone with previous knowledge in the fields of literature, psychoanalysis and gender studies.

1.2.2 Method

As a first step, the translation of the ST was carried out. After having examined the aim, purpose and style of the text a rough translation was done. The software Memsource was used for this initial part of the translation process, facilitating consistency in terms and vocabulary by enabling the use of a term bank and translation memory. To assist the translation several online dictionaries were used, as well as the Corpus of Contemporary American English, COCA (english-corpora.org/coca/), and Språkbanken (spraakbanken.gu.se). For field-specific vocabulary and expressions parallel texts within psychoanalysis, feminist gender studies and Gothic literary criticism were consulted.

Once this first translation draft was complete, the translation segments were copied to a Word document in order to facilitate a better overview when finalising the translation via a revision procedure, facilitating the achievement of ‘naturalness’ in the TT (Newmark 1988:24ff.). Newmark claims that a sufficient level of naturalness is achieved only by

“temporarily disengaging yourself from the SL text, by reading your own translation as though no original existed” (ibid.:24).

The second step, isolating the target structures of the present study, was largely done by manually identifying the ing-/ed- or ly-structures of the ST and their correspondences in the target text (henceforth TT), with some help from the search function in Word 2018. As the material used is a text of no more than 13 pages, this was not particularly time consuming. An alternative solution would have been to use a concordance software to isolate the structures, which would had been more appropriate, not to say necessary, if the data set had been more extensive, but in this case a manual process of identification and counting proved sufficient.

Finally, an analysis of the identified ST structures and their correspondences was carried out, focusing on what structural shifts and translation strategies are used and to what extent, as well as what factors influence the translation choices. This step also involved the categorisation of the TT structures in focus, applying similar categories to those used by Ström Herold & Levin (2018) for the supplementive clauses and those used by Lindquist (1989) for the ly-adverbials (see section 3.1.2 for more detail).

(7)

2 Theoretical background

This part of the paper presents background information on different theoretical aspects relevant to the investigation at hand, starting with a general overview of adverbials in section 2.1, followed by information on the adverbial structures in focus (sections 2.1.2 and 2.1.3) and a summary of some of the most relevant previous research carried out in this field in section 2.2. The third part, section 2.3, is dedicated to a presentation of some of the concepts within translations theory that will be of importance and referred to in the analysis of the present paper.

2.1 Adverbials

Adverbials are a type of clause elements that add information about the clauses they belong to. There are different grammatical frameworks with slightly different approaches to the grammar of adverbials, but Biber et al. (1999:763–764) divide adverbials into three different types based on their functions: circumstance adverbials, stance adverbials and linking adverbials.

Circumstance adverbials add information to the clause regarding for example time, place, process, contingency and degree. Stance adverbials add information about how the speaker’s opinion on what is said in the clause or on how it is said, whereas the third adverbial class, linking adverbials, serve a connective function and give information on how one part of the discourse connects to another.

Adverbials are syntactically realised in a number of different forms (adverbs, adverb phrases, noun phrases, prepositional phrases, finite clauses and non-finite clauses) where adverbs and prepositional phrases are by far the most common (Biber et al. 1999:768).

Because of the variety of forms an adverbial can take, they also vary in length; an adverbial realised as an adverb, as in (5), will consist of a single word, whereas an adverbial in the form of a clause, as in (6), will be considerably longer. This in turn affects what positions the adverbial can occupy.

(5) Proudly, Mr Ahmed accepted the award. (Quirk et al.1985:488) (6) She telephoned to ask for an interview. (ibid.:489)

Adverbials occur in three different positions in a clause: initial (I), medial (M) or final (F), with the final position, where the adverbial is positioned after all obligatory clause elements (as in (6)), being the most common in English (Biber et al.1999:772). The initial position is

(8)

found in one of the possible positions between obligatory initial and final elements in the clause, for example:

The present study focuses on two types specific types of adverbial structures: supplementive clauses (a type of non-finite clause) and adverbials with a suffix of -ly, both of which will be explained further in the following sections.

2.1.1 Adverbials in Swedish and English

Adverbial use in Swedish and English largely follow the same rules, but there are some important differences that may impact my results and therefore are worth mentioning.

Svartvik & Sager (1996:401–402) list more differences than the ones below, but as not all of them will be relevant in the present study the list has been limited to the following ones:

1. Certain adverbial types are as a rule placed directly before the main verb (see (8)) or (if there are auxiliary verbs) after the first auxiliary, whereas Swedish dictates that the adverbial is placed after the main verb (Svartvik & Sager 1996:401)

(8) She never returned. Hon återvände aldrig. (ibid.)

2. Swedish uses different rules for adverbial placement in main clauses (see (8) above) and subclauses (9), whereas English does not (ibid.).

(9) They said that she never returned. De sade att hon aldrig återvände.

(ibid.)

For the purpose of this study, two major grammatical works have been used as sources for any reference to Swedish adverbial use and differences between the languages: Svenska Akademiens Grammatik (Teleman et al. 1999) and Engelsk Universitetsgrammatik (Svartvik

& Sager 1996).

2.1.2 Supplementive ing- and ed-clauses

Supplementive ing- and ed-clauses, in this study defined as non-finite, subjectless adverbial clauses without a subordinator (Quirk et al. 1985:1123–1125; Biber et al. 1999:820), are often challenging structures to translate. This is mainly due to the fact that they often lack clear indicators of the semantic relationship to the surrounding clauses, which also makes this type (7) Carrie had often dreamed about coming back. (Biber et al. 1999:771)

(9)

of clause useful; their ambiguity makes them versatile. Consider the following example, taken from the ST of the present study, illustrating the semantic versatility of these types of clauses:

(10) [D]riven by her lust for blood, she does not respect the dictates of the laws […]

(Whisker 2016:162)

As there are several possible semantic readings of the above sentence, there are also several possible Swedish translations, for example:

(10a) (Temporal reading) När hon drivs av sin blodtörst, […] [‘when’]

(10b) (Causal reading) Eftersom hon drivs av sin blodtörst , […] [‘because’]

In some cases, the ambiguity can remain intact in the translation, but as seen in (10) there are also cases where the translator needs to make a choice as to how the semantic relationship is to be interpreted, which in turn affects the structures used in the TT.

2.1.3 Ly-adverbials

The second object of this study, ly-adverbials, fits under the semantic category of adverbials syntactically realised as an adverb (Biber et al. 1999:767). In the present study they are defined as adverbials in the form of a single adverb with a suffix of -ly, thereby excluding adverbials realised as adverb phrases of two words or more, even if these are headed by an ly- adverb. The basis for this limitation is to enable a comparison between the longest adverbial structure available, the adverbial clause, with the shortest adverbial structure, i.e. a single word. This also facilitates a comparison between the results of this study and the findings of Biber et al. regarding placement of these kinds of adverbials.

One-word adverbials ending in -ly are relatively common, as a common way of creating an adverb is to add this suffix to a base of an open-class word, such as a noun or adjective (Svartvik & Sager 1996:305–306; Biber et al. 1999:64). For example, adding -ly to the noun careful will produce the adverb carefully. The fact that these types of adverbs also have proven to be common in the text genre of academic writing (Biber et al. 1999: 549) makes them interesting to investigate.

Adverbials with the -ly suffix can occur in all three semantic categories (circumstance, stance and linking adverbials), as illustrated in (11), (12) and (13):

(11) It was just flowing along smoothly. (Biber et al.: 803)

(12) That sort of gossip certainly should be condemned. (ibid.:854)

(13) Similarly, natural systems would probably evolve to avoid chaos (ibid.:876)

(10)

Ly-adverbials also occur in all different positions (as shown in the examples above), even if certain positions are preferred to others, depending on the adverbial’s function and relation to the clause, or part of the clause, it modifies.

2.2 Previous Research

Several studies have been carried out on the topic of English adverbials from a semantic perspective, both mono-lingual (cf. Kortmann (1991) and Biber et.al. (1999)) and contrastive (cf. Behrens (1998) and Behrens & Fabricus-Hansen (2005) who investigate translations of ing-participle adverbials in Norwegian and German). The present study, however, focuses on adverbial renderings primarily from a structural perspective between Swedish and English.

With this in mind, two of the most important studies within this area are presented below.

Lindquist (1989) investigates how professional translators of fiction solve the problem of translating adverbials, focusing on the grammatical form of the Swedish renderings and their positions in the target text. The ly-adverbials in Lindquist’s data (called AO, defined as

“adverb phrase, with open class adverb as head” (Lindquist 1989:35)) are found most commonly (40% of the material) realised as an adverb phrase with an open class adverb in Swedish, e.g. loudly > högt. The second largest group: AC – adverb phrase with a closed class adverb, e.g. finally > slutligen – represents 19% of the material and the third largest group (PP – prepositional phrase) represents 16%. A total of 81% of the English ly-adverbials were given an adverbial translation in Swedish as well.

In the case of adverbial ing-clauses in his material (1989:120–130), Lindquist finds that the most common TT rendering is into a finite clause (45%). In this category he includes three subcategories: verb phrase coordination, adverbial subclause introduced by a conjunction and a new full finite clause. In regard to the adverbial ed-clauses of his study he finds that they are in clear minority in his material (8 ed-clauses compared to 93 ing-clauses (1989:128)) and that a majority of them (38%) are, as in the case of the adverbial ing-clauses, translated into a finite clause.

When investigating the position of the different adverbial structures he notes that the majority of the adverbials in I or F position in the ST are transferred into the same position in the target text (I: 82%, F: 88%), whereas adverbials in M position appear less stable (Lindquist 1989:141). He finds that the ly-adverbials overall are moved more frequently than other types and that most of those are moved from their original position into F position. He also notes that the supplementive ing- and ed-clauses in his data set usually keep their original position and show a clear preference for F position (ibid.: 144–155, 162).

(11)

In a more recent study by Ström Herold & Levin (2018), non-fiction data sets from the Linneaus English-German-Swedish corpus (LEGS) are used to investigate the English supplementive ing-clause and its Swedish and German correspondences. It should be noted that Ström Herold & Levin use a slightly different categorisation scheme from that of Lindquist (1989), as they separate what Lindquist refers to as the ‘finite clause’ category into 3 separate categories: coordination, subordination and main clause.

Their results show that the most frequent Swedish correspondence of supplementive ing-clauses is that of coordination (40.5%), which they link to coordination being “compact and semantically rather indeterminate” (ibid.:136); a trait shared with the supplementive ing- clause. The other main correspondence categories are subordination (27.8%), main clause (12.7%) and prepositional phrase.

Ström Herold & Levin (2018:125–127) also investigate the positions of the supplementive ing-clauses and find that F position is the most common one in the English material and that the vast majority (90%) of these are translated into the same position in the Swedish TTs. Of the total 69 sentence-initial clauses in the English material 78% keep their position in the Swedish translations. The sentence-medial clauses prove to be the most uncommon and also the position with the lowest rate of correspondences keeping the same position in the Swedish TTs.

Furthermore, Ström Herold & Levin’s results (2018:135–136) show the presence of the translation procedure of ‘explicitation’ (further explained in section 2.3.2) in their material, supporting theories saying that translations TTs typically are more explicit than their STs (cf.

Blum-Kulka 1986, Séguinot 1988, Chesterman 2004, Klaudy 2008). In Ström Herold & Levin the most frequently used explicitation devices are temporal and causal conjunctions in Swedish, such as när [‘when’], eftersom [‘because’], så att [‘so that’] and därmed [‘thereby’]

(Ström Herold & Levin 2018:135–136).

2.3 Translation theory

This section introduces concepts within translation theory relevant for the discussion of the results of the present study: transposition (2.3.1) and explicitation (2.3.2). These are examples of what Vinay & Darbelnet identify as “translation procedures” belonging to the strategy of

“oblique translation”, i.e. a type of free translation to be used in cases where a literal translation is deemed impossible for grammatical, syntactic or pragmatic reasons (Vinay &

Darbelnet 1995, in Munday 2016:88–89).

(12)

2.3.1 Transposition

Vinay & Darbelnet define transposition as a method that involves changing one part of speech into another, keeping the meaning of the message intact (1995:36), as in (14):

(14) They have pioneered They have been the first (Munday 2006:90)

Extending this concept and applying it to larger units than single parts of speech, Eriksson (1997:20–21) introduces the idea of a type of transposition (by Eriksson referred to as

“strukturomvandlingar” – “structural shifts”) on clause and phrase level. Eriksson argues that in the process of translating, the translation units are larger than a single word, which means that it is more adequate to consider transposition as a structural shift occurring on the levels of phrase and clause (ibid.:21)

Eriksson distinguishes between 4 main categories of structural shifts (ibid.), described and exemplified below (C = clause, P = phrase, Ca/Cb = different clause types, Pa/Pb = different phrase types):

Type Example

1. Ca > Cb ST subordinate clause > TT main clause

2. C > P ST relative clause > TT postmodifying adjective phrase 3. P > C ST noun phrase > TT subordinate clause

4. Pa > Pb ST adverb phrase > TT adjective phrase

It should be noted, however, that structural shifts also can be viewed as occurring on a higher level than that of a phrase or a clause (cf. Frankenberg-Garcia (2019) and Fabricius-Hansen (1999)), as sentences being split or joined in translation also exemplify a type of structural shifts.

Closely related to the concept of transposition is Catford’s model of “category shifts”

(Catford 1965:75–82, 2000:143–147) which are defined as a type of “translation shift” on a structural level. A translation shift, in turn, is said to occur when there is a divergence between two system-based concepts (one in the SL and on in the TL), provoking a shift when translating between the two languages (ibid. 1965:73, 2000:141). Category shifts are subdivided into 4 groups (ibid. 1965:75–82, 2000:143–147). To facilitate application of these in the analysis section of this paper, a code system similar to that of Eriksson (1997:20) has been applied to the below definition and examples where appropriate (C = clause, P = phrase, W = word, Wa/Wb = different word classes).

(13)

1. Structure shifts: shifts in grammatical structure, e.g. adverbial placement in Swedish subclauses compared to that of English subclauses.

2. Class shifts: shifts between different word classes (Wa > Wb), e.g. an adjective in the SL is translated into an adverb in the TL.

3. Unit/rank shifts: where the formal correspondence is on another hierarchical level in the SL than in the TL, e.g. a phrase or a clause in the SL is translated into a single word in the TL (or the reverse) (P > W, W > P, C >W, W > C).

This type of shift overlaps to an extent with those described by Eriksson (1997:20) (see above), but as Eriksson does not specify shifts between single words and larger grammatical units there is a need for the additional tools.

4. Intra-system shifts: shifts occurring between SL and TL systems that are similar, but where the translator is obliged to use an equivalent in the TL system that does not entirely correspond to that of the SL, e.g. translating numbers between English and French (as these language systems both use numbers, but differ in how the numbers are structured).

For the purpose of this essay, Vinay & Darbelnet’s (1995:36) term transposition will be used as an umbrella term when referring to the application of any of the shifts above (either as described by Eriksson (1997) or Catford (1965, 2000)). I will, when applicable, be using these terms to describe the changes occurring in the translation process of the present material. As the transposition categories of Eriksson (1997:20–21) and Catford (1965:75–82, 2000:143–

147) complement each other (despite some overlap) and both assist in sharpening the somewhat vague distinctions provided by Vinay & Darbelnet (1995:36), a combination of these systems will be used in this study.

2.3.2 Explicitation

The term “explicitation” was coined by Vinay & Darbelnet and defined as ”[a] stylistic translation technique which consists of making explicit in the target language what remains implicit in the source language because it is apparent from either the context or the situation”

(1995:342). This could mean, as in (15), that the gender of the French possessive pronoun son (which is implicit in the ST because of the context) is made explicit in the TT:

(15) Il prit son livre He took his/her book (ibid.)

(14)

The concept of explicitation has been discussed from several different angles by different researchers (cf. Blum-Kulka 1986, Séguinot 1988, Chesterman 2004, Klaudy 2008).

In her influential work from 1986, Blum-Kulka formulated what since then has been referred to as her “explicitation hypothesis”, stating that even if some instances of explicitation can be attributed to differences in grammatical systems between the SL and the TL or stylistic differences between languages, a major part of the explicitation introduced by translators is due to the translation process itself, resulting in a type of redundancy in the TT:

The process of interpretation performed by the translator on the source text might lead to a TL text, which is more redundant than SL text. This redundancy can be expressed by a rise in the level of cohesive explicitness in the TL text. This argument may be stated as ‘the explicitation hypothesis’, which postulates an observed cohesive explicitness from SL to TL texts regardless of the increase traceable to differences between the two linguistic and textual systems involved. It follows that explicitation is reviewed here as inherent in the process of translation.

(Blum-Kulka 1986:19)

Blum-Kulka’s claim that explicitation produces a redundant text is, however, questioned by Séguinot (1988:108) who states that explicitation does not necessarily mean redundancy, as a greater number of words in a TT compared to a ST can be explained by for example differences in stylistics (in her case between English and French). She also claims that explicitation may occur not only when some sort of addition of information has taken place, but also in instances where “something which was implied or understood through presupposition in the source text is overtly expressed in the translation, or an element in the source text is given a greater importance in the translation through focus, emphasis, or lexical choice” (ibid.).

Building on the work of Blum-Kulka, Chesterman (2004) includes explicitation in his list of so called “S-universals”, which he defines as “universal differences between translations and their source texts” (Chesterman 2004:39). He claims that translated texts tend to be longer than their STs and that they show a higher level of cohesion (ibid.:40), where especially the latter introduces a more positive angle of the effects of explicitiation than that of Blum-Kulka’s ‘redundancy’.

Another scholar building on Blum-Kulka’s ideas is Klaudy (2008:107) who divides explicitation into 4 different categories: obligatory, optional, pragmatic and translation- inherent. Obligatory explicitation, which is the type most relevant to the present study, occurs because of syntactic and semantic differences in structure between languages and is viewed as obligatory as the TT sentences would be ungrammatical without it. Klaudy states that one of the most obvious examples of obligatory explicitation can be found in the translation of so

(15)

called “missed categories” (ibid.). Klaudy here gives the example of translating from Russian into English (Russian lacks the definite article and English uses it frequently), but this would also be highly relevant when translating supplementive ing- and ed-clauses from English to Swedish. In many cases the translator is forced to make the corresponding TT structure more explicit (as illustrated in (16)), as no formal equivalent to these clauses exists in Swedish.

(16) Walking past the lineup of tables set up by the Heartland conference’s sponsors, it’s not terribly hard to see what’s going on.

När man går förbi de bord som ställts upp av Heartlandkonferensens sponsorer är det inte alls svårt att inse vad som är på gång. (Ström Herold &

Levin 2018:121)

The example above shows an instance where the translator is obliged to add a subordinator (när), making the implicit relationship between the matrix clause and the subordinate clause in the ST explicit in the TT.

However, not all translators choose the same strategies when posed with situations where no formal equivalence exists in the TL. Dirdal (2014) investigates differences between translators regarding “clause building” (which involves e.g. translating a non-finite clause into a finite clause, a coordinated clause into an independent clause or one clause to two or more clauses (ibid.:122)). She notes that there is considerable variation and that this can be connected to the translators’ individual styles, but also that syntactic changes often are an effect of lexical choices (ibid.:138).

Considering its status as an S-universal (Chesterman 2004:39), it seems likely that the concept of explicitation will be applicable when analysing the material of this paper, especially the ing- and ed-clauses and their translations into Swedish. Building on the base definition provided by Vinay & Darbelnet (1995:342) and applying the more detailed definitions by Séguinot (1988:108) and Klaudy (2008:107), the following analysis will refer to instances of explicitation and relate these to the translation choices made.

3 Analysis

This section examines and analyses the findings of my investigation, supported by examples from the data set at hand. Each sub-section is introduced by a presentation of the correspondence categories used before moving on to a quantitative overview and analysis and finishing with a section focusing on adverbial positions and changes therein.

(16)

3.1 Supplementive ing- and ed- clauses

3.1.1 Correspondence categories

To facilitate a comparison between this study and previous research, similar correspondence categories to those found in Ström Herold & Levin (2018:120–123) were applied. Even if Ström Herold & Levin do not include ed-clauses in their study, the same categories apply because of the similarities between these two clause types. In the present material, 3 major correspondence categories were identified (Main Clause, Subordination, Coordination) as well as four minor categories (Infinitive Clause, Omission, Rephrased and Participle). These are defined and explained below using examples drawn from the ST and TT.

• Main Clause

The Main Clause category includes instances where the ST adverbial is translated into a new main clause in the TT, as in (17). All instances are full finite clauses with their own subjects.

(17) [p.163]

[…] the tale is set at the advent of the First World War, working with familiar settings and scenarios of nineteenth- century romantic fiction […]

[…] äger berättelsen rum i tiden före första världskriget. Den arbetar med välbekanta miljöer och scenarier för 1800-talets romantiska litteratur […]

• Coordination

This category includes instances where the ST supplementive clause ist restructured into a coordinated clause introduced by the conjunction och, with the subject (and sometimes auxiliary verb) omitted in the second clause (conjunct).

(18) [p.163]

This tale utilises romantic fiction tropes, rewarding the vampire hero’s chosen partner […]

Den här berättelsen använder sig av uttryck från den romantiska litteraturen och belönar vampyrhjältens valda partner […]

• Subordination

The category of Subordination is comprised of ST clauses translated into subordinate clauses, either in the form of an adverbial clause (see (19)) or a post-modifying relative clause.

(19) [p.162]

[D]riven by her lust for blood, she does not respect the dictates of the laws […]

Eftersom hon drivs av sin blodtörst följer hon inte lagen som dikterar reglerna […]

• Infinitive Clause

This category includes instances where the ST adverbial clause is translated into an infinitive clause, as in the following 2 examples:

(17)

(20a) (b) [p.166]

The revolution in taking the kind of approach that I have just done, recognising vampires as culturally constructed, exploring those gendered cultural and social conventions […], could well have begun […]

Den revolution det innebär att närma sig saken på det sätt jag just gjort – att se vampyrer som en kulturell konstruktion, att utforska dem som genusbetingade och sociala konventioner […] – kunde mycket väl ha börjat […]

• Participle

The Participle category contains examples of supplementive clauses rendered as Swedish present or past participles:

(21) [p.172]

[…] Poppy Z. Brite’s androgynous vampire trio, swagger into town like a bunch of outlaws, boozing, […]

[…] Poppy Z. Brite’s androgyna vampyrtrio, svassar supande in till stan som en hop laglösa […]

• Noun Phrase

This category consists of instances where the ST supplementive clause is translated into a noun phrase:

(22) [p.165]

Rusty, who lost her partner, Gregor Bathory (referencing Countess Elizabeth Baathory, and Gregor from Kafka’s The Metamorphosis (1915), is a night-time photographer […]

Rusty, som förlorat sin livspartner Gregor Bathory (en referens till grevinnan Elizabeth Baathory och Gregor i Kafkas Förvandlingen (1915)) är en nattfotograf […]

• Omission

In this category the ST adverbial clause is omitted in the TT:

(23) [p.163]

[…] women who fall for the promise of eternal romantic love, […] and women who foolishly seek eternal beauty (and/ or eternal life), having bought into the beauty myth.

[…] De är kvinnor som faller för löftet om evig romantisk kärlek, […]

samt kvinnor som naivt nog söker evig skönhet (och/eller evigt liv) Ø.

The section below gives a quantitative overview of the findings of ing- and ed- correspondences, which are then discussed and compared to the findings in previous studies.

3.1.2 Ing- and ed-clause correspondences: quantitative overview and analysis

The material of the present study contains a total of 35 instances of supplementive clauses. An overwhelming majority of these (30 tokens) are ing-clauses and the remaining are ed-clauses.

The ing-clauses proved considerably more common, which mirrors the findings of Lindquist (1989:128) and may indicate a general tendency when comparing the frequency of use

(18)

between these two clause types. This is also supported by Biber et al. (1999:826), who describe ed-clauses as rare by comparison.

Table 1: Correspondences of supplementive ing- and ed- clauses

Presented in Table 1 are the combined instances of ing- and ed-clauses divided into their respective categories. The findings here are in some ways similar to those of previous studies. Considering the top 3 correspondence categories in Table 1, the same categories proved to be the largest both in Ström Herold and Levin (2018) and Lindquist (1989), even if their internal order differs slightly. In Ström Herold & Levin (2018:128) Coordination, followed by Subordination and Main Clause, are the largest categories and together make up 81% of the total number of ing-clauses found, which is close to the results in the present study where the same categories make up 71% of the total. In Lindquist’s study (1989) however, the same 3 categories account for only 44.6% (ing- and ed- clauses combined) (1989:121, 128). Ström Herold & Levin argue (2018:128) that this may be linked to the difference in material. It could be that the non-fiction data in their study shows less translator creativity than Lindquist’s fiction data, which would mean that there would be less variety in the correspondence categories when investigating non-fiction translations.

One of the differences between the above-mentioned previous studies and the present one is that the Main Clause category in my data proved to be the largest, whereas Coordination proved the most frequent category in both Ström Herold & Levin (2018:128) and Lindquist (1989:126), even if Lindquist does not state exact numbers.

Coordination has a distinct advantage to the main clause strategy, which, according to Ström Herold & Levin (2018:129), could be one of the reasons behind coordination being the most frequently used strategy in their study: it allows for the omission of the subject in the coordinated clause, as in (24):

(24) [p.170–

171]

Brite reclaims the vampire’s engagement with an expression of contemporary issues, creating a place of lost hopes, emptied out values, […]

Med nutida problem som uttrycksmedel ger Brite vampyren sin aktiva roll tillbaka och skapar en plats av förlorat hopp och uttömda värderingar; […]

TT Correspondence N

Main Clause 11

Coordination 7

Subordination 7

Infinitive Clause 5

Participle 3

Noun Phrase 1

Omission 1

TOTAL 35

(19)

As the subject (Brite) is omitted in the second conjunct of the coordination, the structure is more compact than if one had opted for the main clause strategy, and now mirrors the structural compactness found in the TT clause. It could also be added that using compact structures in many cases could be an advantage in translations, where you often are limited by space constraints.

There are also clear similarities between the main clause and coordination strategies.

They both fit under Eriksson’s (1997:20–21) transposition category Ca > Cb as the subordinate ing- and ed-clauses are translated into a new type of clause. Both strategies have the effect that the matrix clause and the ing-/ed-clause rendering are on the same clause level in the TT sentence, which in turn gives the translator increased flexibility of choice. Coordinated clauses, whether the subject is omitted or not, are easier to move and/or adapt to the context than less forgiving structures, such as a main clause with dependent clauses attached.

Both strategies have also proven popular choices for translators (as seen in their frequency in my material, as well as that of Lindquist (1989:126) and Ström Herold & Levin (2018:128)), which probably is partly explained by the lack of explicitness in the TT renderings. The semantic ambiguity in the relationship between the ST supplementive clause and its main clause can remain intact in the TT sentence, as no explicitation takes place.

As both of these categories have similar advantages, I would argue that the main reason behind the difference in most popular correspondence categories between my material and that of Lindquist (1989) and Ström Herold & Levin (2018) lies in the characteristics of the ST at hand. As mentioned in section 1.2.1, the ST sentences are overall quite long and complex, and one of the aspects to take into account when translating is the standard sentence lengths of the SL and TL. According to Ingo (2007:74), this is connected to e.g. text type and style, but in general, English allows for more complex sentences than Swedish. This in turn means that the translation process between these two languages often includes dividing a ST sentence in the TT. This is supported by the findings of Fabricius-Hansen (1999:211) who states that sentence-splitting to some extent can be attributed to structural preferences in languages.

Consider (25), where the ing-clause has been rendered as a main clause.

(25) [p.

169]

Many contemporary women’s vampire fiction writers align vampires themselves with a new feminist carnivalesque, infusing the age-old figure with new life and new potential to comment on what it means to be […] human.

Många nutida författare inom den kvinnliga vampyrlitteraturen ansluter vampyrerna till den nya feministiska karnevalismen. De blåser nytt liv i den urgamla karaktären, som får nya möjligheter att berätta vad det innebär att vara […] mänsklig.

(20)

In the example above, the supplementive ing-clause provides an opportunity for a natural division point in an otherwise rather long sentence, which is why creating a new main clause has been preferred over coordination. Using coordination would have resulted in a longer sentence and would further have impacted negatively on readability, as the subject would have ended up quite far away from the second VP.

The suggestion that the high number of tokens in the Main Clause category could be due to ST sentence length is supported by the fact that 9 out of the total 10 instances of main clause renderings in this study are cases where the main clause is separated from the matrix clause by a full stop. Had the ST contained less long and complex sentences, chances are the frequencies would have been more similar to those of Lindquist (1989) and Ström Herold &

Levin (2018).

The second largest category in my material is Subordination, which differs from the other major categories in several ways. No transposition takes place as the ST subordinate ing- or ed-clause is rendered as another subordinate clause in the TT (Eriksson, 1997:20–21).

However, explicitation was present in all 7 cases, all of which are instances of what Klaudy (2008:107) categorises as obligatory explicitation (see section 2.3.2).

The subordination strategy provides a higher degree of semantic clarity than the ST structure as the conjunction introducing the clause gives specific information about the semantic relationship between the supplementive clause and its matrix clause, which, as pointed out by Ström Herold & Levin (2018:127), provides the reader with a frame in which the matrix clause information is to be understood. Consider (26) and (27):

(26) [p.162]

[D]riven by her lust for blood, she does not respect the dictates of the laws which set down the rules […]

Eftersom hon drivs av sin blodtörst följer hon inte lagen som dikterar reglerna […]

In (26) the reader is informed that it is because the vampire is driven by her bloodlust that she does not follow the rules, spelling out the causal relationship between the subclause and matrix clause. This was deemed the most likely scenario in the given context; the passage describes a general tendency in this female vampire. If the vampire’s bloodlust had been described as a type of temporary insanity, när would have been more appropriate introduction to the subclause, to indicate a temporal relationship. This situation points to the importance of considering context when translating these types of clauses (cf. Behrens 1998:265).

(27) [p.164]

[…] Count Ferenc Zohary, who, spotting her copy of Dracula, seduces her […]

[…] Greve Ferenc Zohary, som när han ser hennes utgåva av Dracula förför henne […]

(21)

In (27), the subordinate clause is introduced by när, which in this case leaves some of the ST ambiguity intact, as när in this context also can be interpreted as indicating a level of causality (SAOB.se). The ST sentence highlights that several semantic interpretations can be possible at the same time (as noted by Kortmann (1991:112)) which, if possible, should be transferred to the translation as well. The TT clause is not necessarily saying that the count seduces his victim straight after having spotted her copy of Dracula, but it is certain that him spotting it triggered his will to do so. Another alternative would have been to use då, as it provides the same causal/temporal ambiguity in Swedish as när (SAOB.se), even if it would have given a slightly more formal impression.

A vast majority of the instances of explicitation found in the present study (8 out of 9 tokens) involve temporal or causal conjunctions (such as därmed, eftersom, när etc.) which is the case in both (26) and (27) above. This result mirrors that of Ström Herold & Levin (2018:135–136). None of the explicitations found can be said to be unnecessary and contribute to any redundancy in the TT, which contradicts Blum-Kulka’s definition of explicitation (1986:19) in support of Séguinot’s (1988:108) (see 2.3.2). However, the TT sentences have still been extended as an effect of explicitation, which gives some support to Chesterman’s claim of explicitation being an S-universal (2004:39–40).

A somewhat surprising find in the present material is the relatively high percentage of infinitive clause renderings (14%), compared to Ström Herold & Levin (2018:128) where the same category only accounts for 6.2% of the total number of correspondences. However, it should be noted that 3 of the 5 renderings in the present data were found in the same sentence, (28), which is likely to impact the result. In the below examples we see 3 instances where Eriksson’s transposition type Ca > Cb has been used (Eriksson 1997:20–21).

(28a) (b) (c) [p.166]

The revolution in taking the kind of approach that I have just done, recognising vampires as culturally constructed, exploring those gendered cultural and social conventions, recognising that the figure of the vampire was a product of their time, place and cultural mores, and which could be unpacked, re-scrutinised as a vehicle for cultural critique, could well have […]

Den revolution det innebär att närma sig saken på det sätt jag just gjort – att se vampyrer som en kulturell konstruktion, att utforska dem som genusbetingade och sociala konventioner och erkänna att vampyrkaraktären var ett barn av sin tid och plats, och av sina kulturella sedvänjor, som kunde plockas isär och omprövas som en bärare av kulturell kritik – kunde mycket väl ha […]

In (28) the ST ing-clauses act as inserted clarification to the matrix clause, meant to provide

(22)

clauses in this sentence, and as they together form what gives the impression of a list of examples, the same needs to be reflected in the TT sentence. This could be achieved either by coordination or by using an infinitive structure (commonly found in lists). The option of coordinating all the clauses would have rendered a very repetitive translation, which would have gone against Swedish preferences (Teleman et al. 1999, Vol. 4:880–881) and probably been perceived as stylistically inferior (especially as the second ing-clause contains an instance of and to begin with). I therefore opted for the infinitive clause solution, which keeps the TT structure and stylistic level as close as possible to that of the ST.

This section has provided a quantitative overview and qualitative analysis of the results regarding ing- and ed- correspondences. In summary, the results are in several respects similar to those of previous studies, even if main clause was proven to be the most commonly used strategy in this study; a tendency that may be explained by factors such as sentence length. In the following section we turn our attention to the positions and position changes of the supplementive clauses at hand.

3.1.3 Position

The results in this section are based on the position classification system proposed by Quirk et al. (1985:490–501), which, as this is the system used in both Ström Herold & Levin (2018) and Lindquist (1989), facilitates comparison of results. However, it should be noted that some of the categories used by Quirk et al. (1985:490–501) have been combined for the purpose of this study, as Quirk et al. distinguish between different internal positions within I, M and F.

Table 2: Positions of supplementive ing- and ed- clauses

As shown in Table 2, the most common position for the supplementive ing- and ed-clauses in the present study is F. This is similar to Ström Herold

& Levin (2018:125), where ing-clauses in F position are found to be even more common than in the present study. Lindquist’s data (1989:162) shows the same tendency; 86% of the ing-clauses and 63% of the ed-clauses in his material are found in F position.

That F position is found to be the most common is not a major surprise, as Biber et al.

(1999:830–833) state that this is the unmarked choice for non-finite clauses of this kind.

However, my data shows a slightly lower percentage of clauses in F position than Ström Herold & Levin (2018) and Lindquist (1989). This does not seem to be related to the fact that

Position ST TT

N N

I 8 8

M 7 6

F 20 20

Omitted - 1

TOTAL 35 35

(23)

the ing- and ed-clauses have been combined into one supplementive clause category; the percentages, if counting these clauses as separate categories, would still be very similar. I would rather argue that this difference may be connected to the comparatively high counts of M position clauses in the present material. According to Biber et al. (1999:830–833) this position is considered very rare for clauses of this type, but in my data they are almost as common as those found in I position. This could in part be due to the fact that 3 of the M tokens all derive from the same sentence (see (25)), which combined with the restricted size of the present data would produce a relatively high result in the M category and thereby a lower percentage in the F category.

The second most common position in the present material is I. A vast majority of the ST supplementive clauses in this position were translated into subclauses in the TT, as in (29):

(29) [p. 163]

Then, turning the tables, lesbian vampires appear as celebrated figures […]

När rollerna sedan byts framträder de lesbiska vampyrerna som hyllade karaktärer […]

This result is similar to that of Ström Herold & Levin (2018:126–127), who found that the sentence-initial position favours translation into subclauses. They also found a large majority of these realised as adverbial clauses, which is the case in the present study as well; 5 of the total 6 subclauses had adverbial translations in the TT, as exemplified in (29).

Few position changes were found in the present material. One of the two available examples is given in (30), where the sentence-medial ing-clause in the ST is realised as an infinitive clause in I position in the TT:

(30) [p. 167]

[…] including faeces, menstrual blood, vomit and other abject substances, and then that first Other, succouring not the infant self, the figure of the mother.

[…] Detta inkluderar avföring, mensblod, spyor och andra abjekta substanser. Utan att hjälpa spädbarns- jaget avvisas sedan även den där första Andra: modersgestalten.

This very low number of position changes is supported by previous research. Lindquist shows that 94.6% (ing- and ed-clauses combined) of all positions remained unchanged (1989:147, 155) which is similar to Ström Herold & Levin’s figure of 86.7% (2018:126).

This section has provided insight into the positions of the ing- and ed-clauses in the present material. The results show F position to be the most common and position movements to be very rare. The next two sections are focused on the correspondences and positions of ly- adverbials.

(24)

3.2 Ly-adverbials

3.2.1 Correspondence categories

To facilitate a comparison between the present data and that of previous research, similar categories to those in Lindquist (1989:35) were applied. In the present material, 4 of Lindquist’s nine correspondence categories were identified: AO (adverb phrase with open class adverb as head), AC (adverb phrase with closed class adverb as head), PP (prepositional phrase) and Restructured. In the section below, these categories are defined and explained using examples drawn from the ST and TT.

• AO: adverb phrase with open class adverb as head

The AO category includes instances where the SL adverbial is translated into a single open class adverb (as in (31)) or in some cases a multiple-word adverb phrase with an open class adverb as head in the TT. Lindquist defines open class adverbs as adverbs formed with a productive adverbial suffix but does not specify which Swedish suffixes this includes (ibid.:36). According to Teleman et al. (1999, Vol. 1:632) only the adverbial suffixes -vis (as in naturligtvis) and -ledes (as in således) are considered productive. However, as Lindquist (1989:108), also includes the suffix of -t (as in kärleksfullt) as an open-class suffix in his analysis, the t-ending will be included in the AO category for the purpose of this study.

(31) [p.164]

He also offers the erotic opportunity to actually achieve her own sexual desires […]

Han erbjuder henne också den erotiska möjligheten att faktiskt kunna uppnå sina egna sexuella begär […]

• AC: adverb phrase with closed class adverb as head

This category includes instances where the SL adverbial is translated into a closed class adverb in the TT. These are defined as adverbs not formed with an adverbial suffix:

(32) [p.161]

[…] although there seems to be less interest in the figure currently.

[…] även om intresset för karaktären numera verkar minskat.

• PP: prepositional phrase

The category PP is comprised of ST adverbials translated into prepositional phrases:

(33) [p.162]

The historical Elizabeth Baathory was finally found guilty and sentenced to immurement […]

Den historiska Elizabeth Baathory befanns till slut skyldig och dömdes till följd av sina grymma dåd till att bli inmurad i en vägg […]

(25)

• Restructured

This category is comprised of ST adverbials restructured into adjectives:

(34) [p.165]

Not surprisingly then, rescuing friends and lovers from fatal diseases into vampiric being is a common humanitarian act.

Då är det föga förvånande att det ses som en vanlig humanitär insats att rädda vänner och älskare från dödliga sjukdomar och in i ett vampyrväsen.

The section below provides a quantitative overview of the findings of ly-adverbials the present study, which are then discussed and compared to the findings of previous studies.

3.2.2 Ly-adverbial correspondences: quantitative overview and analysis

The ST material of the present study contains a total of 32 instances of ly-adverbials. As shown in Table 3, a majority of these are adverbials of circumstance and the remaining adverbials of stance. This distribution is not surprising, as circumstance adverbials is the most frequent and varied class of adverbials (Biber et al. 2007:763). There were no instances of linking adverbials ending in -ly in the ST, which could be explained by the fact that the -ly ending is not particularly common in linking adverbials (ibid.: 887) combined with these being less frequently used than the other two adverbial types over all.

The TT findings regarding ly-adverbials are presented in Table 4, which shows instances of ly-adverbials divided into their respective correspondence categories.

Table 3: ly-adverbials divided by adverbial type Table 4: Correspondences of ly-adverbials

The reason that only 4 of Lindquist’s (1989) categories were found in the present material could be explained by the difference in data quantities. In Lindquist’s material (1989:94) 129 instances of ly-adverbials were found, which is about 4 times as many as in the material used for this study. It is not unrealistic to believe that more categories would have been found if my data set had been of a similar size.

The correspondence rates in the present study are in many ways similar to Lindquist’s (1989:94), where the same 3 categories (AO: 56% of the total, PP: 16% and AC: 16%) showed the highest percentage of correspondence. In Lindquist (ibid.), AO is in unrivalled

TT Correspondence N

AO 18

AC 5

PP 5

Restructured 4

TOTAL 32

Adverbial type N Circumstance 19

Stance 13

TOTAL 32

(26)

first place with 40% of the total. The AC and PP categories in Lindquist’s material are of very similar sizes (AC: 19%; PP: 16%) (ibid.), which is the case in my material as well.

The relatively high frequency of ly-adverbials being translated into open-class adverbs rather than any other category, both in present study and in Lindquist (1989), is probably to some extent connected to their level of use in the Swedish language as a whole. All instances of AO in the TT (except two) are realised as adverbs ending in -t (as in (35)), a seemingly common adverbial suffix in Swedish, whereas the AC category mainly consists of adverbs with the -ligen ending (see (36)), which are considerably less common. A search in Språkbanken (sprakbanken.gu.se) gives an indication in support of this; the adverbial ending of -t is more than 10 times as common as -ligen in the corpus texts.

(35) [p.161]

The female vampire is conventionally represented as abject because she disrupts identity and order.

Den kvinnliga vampyren skildras traditionellt som en abjekt varelse eftersom hon stör känslan för ordning och identitet.

In (35), the stance adverbial conventionally modifies the rest of the clause (represented as abject […]) which should be reflected in the translation as well. As there is no auxiliary verb present in the TT sentence, the adverbial needs to move either to the position after the finite verb or to the beginning of the clause in the TT in order to comply with Swedish preferences regarding placement of these types of adverbials (Svartvik & Sager 1996:401; Ingo 2007:178). The alternative would be to let the adverbial introduce the sentence as a whole (Traditionellt skildras den kvinnliga vampyren […]). This would have been a grammatically acceptable solution but would have emphasised traditionellt as the foundation of the sentence and thereby acted as the frame through which the following information would have been viewed (Teleman et. al. 1999:431–433). As this is not the information structure of the ST, the choice was made to keep the adverbial in closer proximity to the main verb. There is in this case no specific reason to change the order in which the information is given, and therefore the aim should be to keep the ST information structure as intact as possible in the TT (Lindquist 1989:139).

In (36) a slight shift of focus has taken place. The circumstance adverbial finally, which in the ST adds information as to when the subject is rejected, has been transferred into a new role: to modify the verb ersätts, adding information regarding when the subject is replaced.

(36) [p.163]

[…] or finally rejected en route to a more dependable and rewarding (and conven- tional) love of one’s life.

[…] eller så förkastas de under berättelsens gång och ersätts slutligen med en mer pålitlig, givande (och konventionell) livspartner.

(27)

As the construction of rejected […] to (here meaning ‘reject in favour of another alternative’) does not have a readily available equivalent in Swedish of the same compact structure, restructuring the sentence was deemed necessary to reflect the full meaning. An alternative solution would be to translate this passage into förkastas slutligen under berättelsens gång till förmån för, as förkastas […] till förmån för gave 9 more hits in Språkbanken (spraakbanken.gu.se) than förkastas […] ersätts. However, the combination of slutligen and under berättelsens gång comes across as somewhat illogical, as it indicates that this action takes place simultaneously, despite the different connotations of the two adverbials.

According to Svenska Akademiens Ordbok (saob.se), slutligen typically denotes that an action occurs at the end of a process, whereas (under […] gång) denotes an action occurring during a process. Therefore, the choice was made to specify that characters are replaced after having been rejected, as this renders this sentence more logical in Swedish and fulfils the aim of naturalness (Newmark 1988:24).

In both (35) and (36), the adverbial, realised as an adverb, has been translated into another adverb, meaning that no transposition has taken place. There are however several instances in my material where transposition has occurred. The most obvious examples of transposition are found in the categories of PP (see (37a)) and Restructured (see (38)), but there are also instances of transposition in the AO category (see 37b).

In (37a), a unit/rank shift (Catford 1965:79) of the type W > P has occurred, where the adverb effectively has been rendered as a prepositional phrase: på ett effektivt sätt. The formula used in this particular PP (preposition + adjective + noun, expressing ‘in an x manner’) is by Lindquist described as quite typical (Lindquist 1989:114).

In the same sentence, (37b) we see a shift of the same type, but here the adverb (physically) has been translated into the adverb phrase rent fysiskt. Both these adverbials are adverbials of circumstance and express the manner in which the persuasion takes place.

(37a) (b) [p.164]

It also emphasises the erotic charge of the (venomed) vampire embrace, which effectively physically persuades the victim that [..]

Den betonar den erotiska laddningen i vampyrens giftiga omfamning, som på ett effektivt sätt övertalar offret rent fysiskt att […]

The choice was made to separate these two adverbials in the TT and move the second one to F position after the verb and object. Both these positions are common in Swedish for adverbials of this type (Ingo 2007:178). The main factor behind this translation is to prioritise clarity and readability. Firstly, to keep the two adverbials in their original positions would not be a viable

The long and the short of it: the translation of non-finite adverbial clauses and ly-adverbials