Censorship and Taboo Maintenance in L1 and LX Swearing

Abstract

In this paper, we consider the censorship of public swear word usage as a function of, and continued maintenance of, taboo with a focus on L1 and LX swearing and its management. In research with multilingual speakers, first-language swear words are consistently perceived as more taboo, and thus more emotional/powerful than equivalent words from a second or third language. While the public use of English-language swear words may be subject to censorship in Anglophone contexts, it is not censored to the same extent in LX contexts. On the other hand, L1 swear words are censored. Such perceptions of differences in strength between one’s L1 and LX languages also seem to affect the work of language professionals: translators’ tendency to self-censor may at least in part be explained by this bias. The existence of a two-tier system of swearing and censorship serves to reinvigorate L1 swear words, while diminishing the power of English swear words. We thus examine how censorship works as a means of maintaining and/or attenuating taboo, potentially moderating the power of swearing itself in cross-linguistic and multilingual contexts.

1. Introduction

Swearing may be defined as the use of potentially offensive lexical items that denote a taboo. As a direct result of its taboo status, swearing holds power for speakers and hearers in that it produces specific rhetorical, interpersonal, emotional, and psychological effects (Beers Fägersten and Stapleton 2022; Stapleton et al. 2022; Stapleton and Beers Fägersten 2023). The potential for swearing to cause offence means that it is subject to surveillance, sanctions, and censorship. Censorship acknowledges the taboo of swear words in public space: this serves to reflect the power of swearing, but also to further imbue swear words with power. That is, the act of censorship may be seen as a means of highlighting and thus strengthening the notion that these particular words are unacceptable, while others dealing with the same semantic matter do not carry the same force.

The tradition of censoring swear words in Anglophone contexts is well known. For example, English-language swear words are subject to censorship in the media, such that there is a recognizable sfet of words that may not be used on network television or commercial radio stations or must be censored to obfuscate their appearance or sound (Beers Fägersten and Bednarek 2022). However, English swear words, analogous to the English language, have been spread worldwide and appropriated by speakers of other languages. The result is that English swear words are used by LX speakers of English in ways that are at odds with norms of swearing in public in both Anglophone cultures and their own L1 cultures (Beers Fägersten and Stapleton 2017).

In this paper, we begin from the premise that the (self-)censorship of public swear word usage is not only a function of taboo, but also a means by which taboo is maintained and potentially strengthened. We consider this issue via a focus on censorship practices in multilingual settings and a parallel consideration of the respective emotional force of first- and second-language swearing for multilingual speakers.

We take the following set of concepts as our starting point:

(1)

Swearing is both emotionally arousing and emotionally forceful because it is taboo and potentially offensive;

(2)

Censorship works as a means of reflecting and maintaining existing taboos;

(3)

Research with multilingual speakers shows that the emotional force of swearing is stronger for one’s first language (L1) than for other languages that one might speak or subsequently encounter (LX).

Taking these findings together, we suggest that censorship practices in multilingual settings reinforce and strengthen the power/taboo of swearing in the L1 while simultaneously reflecting, and arguably, further attenuating, the lesser power/taboo of swearing in other languages.

In advancing this idea, our paper is primarily conceptual in nature: i.e., it does not aim to specifically test or prove a hypothesis via structured data and analysis, but rather to explore the validity of the proposal through two case studies of specific forms of swear word censorship in multilingual settings (here, Swedish and Finnish). Empirical examples are used as the basis for discussion in each setting. Through these case studies, we observe and discuss the management of L1 and LX swearing, as well as the borrowing of English-language swear words for use in other linguistic settings.

In the remainder of this paper, we first review the research on swearing and emotion among multilingual speakers to establish the dynamic between L1 and LX swearing. We then consider in detail the two case studies of L1 and LX swearing: In Sweden, we analyze the use of Swedish and English swear words in comparable contexts, and in Finland, we examine the translation of swear words, with the main focus on English-to-Finnish translation. Against the background of these case studies, and in light of the emotional power of swearing and its derivation, we then discuss how censorship works as a means of maintaining and/or attenuating taboo, potentially moderating the power of swearing itself in cross-linguistic and multilingual contexts.

2. Swearing, Taboo, and Emotion: Multilingual Studies

The taboo, and thus potentially offensive nature of swearing, is strongly linked with emotional arousal, as shown in self-reports/speaker perceptions, and in empirical studies of physiological (autonomic) changes (see Stapleton et al. 2022 for an overview of this research). For the purposes of the present paper, it is useful to further explore these findings within the framework of multilingualism and emotion, and, specifically, multilingual swearing. In this section, we will briefly outline the research on multilingualism and emotion before discussing, in more detail, the findings for multilingual swearing.

Studies of emotion and multilingualism have consistently found that emotionality is stronger in one’s first language (L1) than in languages learned subsequently (LX) (Pavlenko 2005, 2012; Colbeck and Bowers 2012; Caldwell-Harris 2015). Affective processing is faster and more automatic in the L1, and speakers perceive greater emotional force when using their first language (Dewaele 2018). A number of explanations have been offered for this. For example, the context of acquisition is usually different for L1 and LX (Harris et al. 2006). L1s are learned in naturalistic settings, involving interactions and relationships with family and caregivers, while LXs are often acquired in situations with fewer personal attachments or associations, for example, in more formal settings, such as school or work, or through popular culture. Thus, associative memories can be seen to underpin the heightened emotionality of L1 (Caldwell-Harris 2015). Pavlenko (2005) further proposes language embodiment and autobiographical memory as factors in establishing L1 emotionality.

These factors are highly relevant to a consideration of the effects of multilingual/LX swearing. Swearing in the L1 is strongly linked to emotional expression and emotional arousal, and according to a considerable body of self-report evidence, swearing is felt to have more emotional force in the L1 (Dewaele 2013, 2018; Vélez-Uribe and Rosselli 2019). However, as shown by Stephens and Robertson (2020), the surface properties of swear words do not in themselves produce emotional arousal, suggesting that the emotional power of swear words is a learned association (see Jay et al. 2006; Stapleton et al. 2022).

Two related bodies of research have explored the emotional force of L1 vs. LX swearing. The first examines self-report studies of LX swearing in terms of emotional force and usage frequency (see Dewaele 2010a). Two main comparative designs have been used: comparisons of L1 vs. LX for bi/multilinguals, and comparisons of English-language swearing by native English speakers (English L1) vs. by non-native English speakers (English LX). A key finding is that the perceived emotional force of swear words is higher for L1 than for LX, and speakers generally prefer the L1 for swearing (Dewaele 2004a, 2004b, 2004c, 2005, 2010b, 2011, 2013, 2017; Shakiba and Dewaele 2022; Shakiba and Stapleton 2022). However, the emotional force of swearing in the L2 is moderated by a number of factors (Pavlenko 2012; Dewaele 2018). A key factor is socialization and acculturation into the L2 culture, indicated in part by L2 proficiency and usage (see Pavlenko 2012; Dewaele 2018; Shakiba 2019). Both higher proficiency in the LX and higher frequency of LX usage are associated with stronger emotional force of LX swear words and stronger preference for LX for swearing (Dewaele 2004a, 2004b, 2011, 2013). Moreover, participants no longer dominant in their L1 judged L1 swear words to be less powerful (Dewaele 2004a, 2004c). Thus, it would appear that “intense affective socialisation in the LX can drain the emotional power of L1 swearwords” (Dewaele 2018, p. 223). In addition, acquisition factors play a significant role in the perceived emotional force of LX swear words and the choice of LX for swearing. Higher emotional force and/or more frequent use of the LX for swearing is correlated with participants’ having learned the LX in a naturalistic context, rather than in an instructional setting, and also with an earlier age of acquisition (Dewaele 2004a, 2005, 2011, 2013).

The work of Dewaele and colleagues has produced consistent and large-scale findings on the perceived emotional force of L1 vs. LX swearing, based on self-reports by multilingual speakers. The second, complementary, body of research has examined responses to L1 vs. LX swear words under laboratory-based experimental conditions. These studies compare autonomic and cognitive changes to provide evidence of relative levels of emotional arousal in each language (see Pavlenko 2012). The main measures used, sometimes in combination, have been attention/interference, memory/recall, and physiological (autonomic) effects.

Caldwell-Harris and colleagues conducted a series of experiments exploring autonomic responses to L1 and LX verbal stimuli, with a particular focus on “negative emotion words”, including swear words (Harris et al. 2003, 2006; Ayçiçegi and Harris 2004; Harris 2004; Ayçiçegi-Dinn and Caldwell-Harris 2009; Caldwell-Harris and Ayçiçegi-Dinn 2009; Caldwell-Harris et al. 2011). These and other studies show some variability in outcome depending on the specific languages, stimuli, measures, and comparisons used. However, some broad tendencies and findings can be extrapolated. In general, negative emotion words are seen to produce more arousal in the L1 than in the LX, as measured in higher skin conductance (Harris et al. 2003, 2006; Harris 2004; Caldwell-Harris and Ayçiçegi-Dinn 2009). Moreover, analysis of responses to L1/LX swear words has shown consistently that swear words elicit greater arousal than neutral words on the various measures deployed, including recall (Ayçiçegi and Harris 2004; Ayçiçegi-Dinn and Caldwell-Harris 2009), skin conductance rates (Harris et al. 2003; Harris 2004; Eilola and Havelka 2011), and Stroop task interference (Eilola et al. 2007; Eilola and Havelka 2011). However, direct comparisons of emotion responses to L1 vs. LX swear words have shown mixed findings (Ayçiçegi-Dinn and Caldwell-Harris 2009). There is nonetheless some evidence for greater emotionality for L1 swearing. Eilola and Havelka (2011) showed higher skin conductance for L1 swear words compared to LX swear words. Colbeck and Bowers (2012) also found less interference from LX (compared to L1) swear words in a Stroop task with Chinese/English bilinguals, thereby indicating more arousal from L1 swear words.

To summarize this section: Taboo language, specifically, the set of verbal items commonly identified as swear words (see Stapleton and Beers Fägersten 2023) is strongly associated with emotion and can produce emotional arousal. When viewed in the context of multilingualism and emotion (wherein L1 is typically more emotional than LX), L1 swearing may be seen to have stronger emotional associations, to produce stronger emotional effects, and to be more fruitful as a means of emotional expression than LX swearing. By contrast, LX swearing is perceived as less offensive, less taboo, and thus less potent as a verbal and/or psychological resource. Arguably, it is the act of censorship, stronger in the L1, that maintains the taboo on L1 swearing. LX swearing is more widely permitted and less emotionally forceful: both in its use by multilingual speakers, and also in its borrowed forms, whereby swear words from one language (often English) are used in another linguistic setting, either in their original forms or via translation. It is from this basis that we consider the two case studies which form the focus of this paper.

3. Divergent Censorship Practices in L1 and LX Swearing: A Case Study of Sweden

In this section, we explore the censorship of L1 swear words juxtaposed with non-censored LX swearing practices in written texts in the public forum. The focus on written texts allows for an examination of swearing in a context from which it tends to be absent, as all language use in a public forum is normally subject to censorship so as to comply with standard language practices. Unlike spoken language, which may occur in spontaneous, unmonitored form, written language—especially in a publicized context—is both planned and often available to concurrent or post-hoc editing and can thus be held to more stringent standards of propriety. Examining swearing in the context of freely available, public written texts allows us to avoid any “accidental” occurrences and to focus on cases where there was a deliberate decision to feature swearing, either censored or uncensored.

The context of Sweden is particularly fruitful for this investigation, as it has already been established that swearing in English is a recurring feature across many forms of Swedish media and public texts. For example, in a series of studies, Beers Fägersten analyzed the Swedish use of English-language swear words in book, television, and film titles; in national newspapers; in live television broadcasts; in advertising campaigns; in comic strips; and by social media personalities (Beers Fägersten 2012, 2014, 2017a, 2017b, 2017c, 2018). Such examples provide a foundation to the argument that English-language swearing is “not only in contrast to Anglophone norms for media language, but conspicuously so, such that the use of English swear words in Swedish media can be understood as exploiting the opportunity to swear in English with impunity and serves to both ratify and standardize the general use of English swear words in Sweden” (Beers Fägersten 2014, p. 70).

English-language popular media productions (such as television series, films, and music) are regularly imported to Sweden, but any constraints on their broadcast due to offensive language are not. In Sweden, such imported media are neither regulated for consumption nor subject to censorship on the basis of offensive language. Consequently, exposure to non-standard language such as swear word usage in English-language popular media is divorced from Anglophone broadcasting practices that foster the development of a cultural sensitivity to swear word usage. English swear words are thus introduced into both the Swedish collective consciousness and the vernacular, and subsequently appropriated in ways that often conflict with Anglophone norms of usage (Beers Fägersten 2012, 2014, 2017b, 2018). Beers Fägersten has previously argued, however, that “the social complexities of swearing are nevertheless made salient to [Swedes] by American media coverage of public swear word usage” (Beers Fägersten 2014, p. 83), such that, while they are aware of Anglophone inhibitions and restrictions regarding swear word usage, Swedes are less likely to experience any similar levels of reluctance, offence, or impropriety toward LX swear words, in accordance with other foreign- or second-language speakers (Dewaele 2004b, 2010b; Jay and Janschewitz 2008).

In Sweden, English swear words have been appropriated and are widely used in private interactions but notably in very public contexts, too. Such occurrences of swearing are not entirely without scrutiny (Beers Fägersten 2017b), but ultimately, the decision to use or incorporate English swear words in public media in particular has a legitimizing effect and serves as evidence that swearing is deemed not only appropriate but perhaps even advantageous by those directly involved in the media productions in question. Indeed, as a marked, non-standard use of language, swearing attracts attention and thus can entail an added value from a marketing perspective (Beers Fägersten and Pereira 2021).

The practice of English-language swearing in Sweden suggests a distinct and recognized value, exemplified by its occurrence in cross-media titles, advertising campaigns, newspaper quotes, or comic strips (as listed above). The examples presented in previous studies serve to establish the use of English-language swear words as widespread across Swedish-language media productions, which ultimately has the function of standardizing and legitimizing English-language swearing. But the great extent to which English-language swearing has been appropriated as a practice in the Swedish context raises the question of what effect this has on Swedish-language swearing. Whether English-language swearing has had any subtractive effect on Swedish-language swearing is not immediately evident—what is potentially absent or replaced cannot be observed. For this reason, the focus in the present analysis is on English- and Swedish-language swearing in public written texts as observable and thus comparable instances of swearing. Beers Fägersten (2020) previously analyzed the use of non-censored English-language swearing and censored Swedish-language swearing in Swedish-language comic strips. In this paper, the two case studies presented below provide additional evidence of swearing in English and Swedish in similar text types but exhibiting different practices of censorship.

The first example set is extracted from a local Stockholm newspaper, Mitt i Södermalm (“Södermalm central”). Page 2 of the weekly issues always includes a brief column and an inset by that issue’s acting editor, the latter of which is titled Veckans känsla, “Emotion of the week”. The insets in Example 1 and Example 2 both feature swearing. Example 1 (22 February 2020) includes uncensored swearing in both English and Swedish; Example 2 (26 August 2023) includes censored swearing in Swedish. Both texts were written by Swedish, white, male journalists, 47 and 46 years of age, respectively.

ÖVERPEPPAD. Lite drygt en vecka kvar till kort semester i London. Jag vill inte ta i för mycket, men here fucking jävlar så skönt det ska bli.

[Over-excited. A little more than a week left until a short vacation in London. I don’t want to exaggerate too much, but holy fucking shit so nice it’s going to be.]

LEDSEN. Minst ett år borta från Stockholm är planen. Hur f*n hamnade jag här?

[Sad. At least a year away from Stockholm is the plan. How the d*vil did I get here?]

As evident in the translations, the “Emotion of the week” texts are characterized by a personal, informal, and conversational tone, one that is in line with the use of swear words (Beers Fägersten 2012). Example 1 includes both an English (“fucking”) and Swedish (“jävlar”) swear word, but neither has been censored. This example thus illustrates the interplay between the two swearing systems, with each swear word functioning equally as an expletive (McEnery 2006) to express positive emotion, as framed by the featured emotion, “over-excited.” In Example 2, on the other hand, there is only one swear word, the Swedish “fan”, which has been censored. Both jävlar, which is derived from djävul, and fan refer to the devil, and thus both are swear words from the category of religion. Both have also been suggested as among the “worst” swear words in Swedish, with “herre [d]jävlar” in particular cited as a particularly “bad word” expression (Stroh-Wollin 2010, p. 9). As equally taboo terms, both would be candidates for censorship; the fact that only “fan” was censored in Example 2 may be due to the framing provided by the emotion, “sad”, contributing to a negative, more powerful interpretation of “fan”, thus prompting censorship. Alternatively, it could be argued that, by censoring “fan”, the writer imbues the word with power, which in turn strengthens the distraught state expressed in “How the d*vil did I get here?” Finally, the cases may simply illustrate a difference in sensibilities between the two writers. In the next set of examples, we consider additional evidence of censorship serving to strengthen Swedish swear words.

Example 3 and Example 4 represent advertising campaigns that included print adverts placed on the buses and trains of Stockholm’s public transit system.

Example 3 features a tagline (“Luxury that doesn’t fuck up our planet”) for Residus (https://residusofficial.com/, accessed 20 March 2024), “an uncompromising clothing brand based in Sweden.” As pictured, the tagline appeared on placards that could be found on public transport buses, but as shown in Example 3, the placard also faced outward, such that the bus served to mobilize the advertising campaign, exposing it not only to the bus’s passengers but also to anyone else on its route. On the receiving end of the campaign was a British tourist, who lodged a complaint with the Advertising Ombudsman’s Council, stating that the use of the word “fuck” was a violation of the International Chamber of Commerce’s rules for advertising and marketing communication and that it should not be normalized for young children as it is “coarse.” According to the advertiser’s defense, “a little provocation is needed to change and transition to a more sustainable fashion industry”, and that the phrase “fuck up” has a “completely different meaning” than “fuck”’ on its own. Finally, the advertiser claimed that the complaint is due to the British tourist being more sensitive to the phrase than an American would be. The Council ruled in favor of the advertiser, agreeing with their semantic distinction and addressing the tourist’s concern for children by noting that the poster does not depict children or young people and that it cannot be considered to be aimed at children or young people either through the products advertised or because of its design. The fact that the poster has been displayed in the subway, where children and young people are staying [sic], does not change the assessment (https://reklamombudsmannen.org/en/decisions/enskilt-beslut/?caseid=1908-170#, accessed on 20 March 2024).

The complaint against the swearing in Example 3 invokes age-old arguments against swearing, namely, that censorship (or avoidance) is necessary to protect children who may be unintentionally exposed (Hutton 2019, p. 276), and that swearing in the public context can serve as inappropriate behavior modelling (Hutton 2019, p. 283), in effect encouraging or legitimizing similar usage of swear words. Notably, this example illustrates a divergence between L1 and foreign-language speakers in the assessment of English-language swearing as well as a dismissal of the sensitivity of the former toward the use of swear words by the latter.

Example 4 illustrates another advertising campaign deployed in the context of the public transportation system. Like Residus, the advertiser featured in Example 4, Jobbland, is also a Swedish company, but their advertising campaign is exclusively in Swedish. Jobbland maintains a job search site, using the tagline “Ny dag, samma j@%la jobb” (“New day, same damn/fucking job”) to appeal to those potentially experiencing dissatisfaction with their career. Because of the censorship achieved through the use of typographical symbols, it is not possible to determine if the swear word is jävla or its euphemistic alternative, jäkla, conventionally translated as “fucking” and “damn”, respectively (Stroh-Wollin 2008), and comparable to their respective English equivalents in force (Höder 2023, p. 88). However, we would argue that the censorship encourages the reader to assume it is the former, as the stronger term would more likely be subject to censorship. Similar to Example 2, the censored swearing in the tagline is framed in such a way as to encourage its interpretation as annoyance swearing (Ross 1961; Montagu 1967), which is also generally experienced as more offensive than social swearing (Beers Fägersten 2012). Once again, it is the censorship that not only suggests an offensive term that should be mitigated by obfuscation in print form, but also serves to imbue the Swedish swear word with force: were the word not inherently capable of offence, it would not need to be censored. Thus, we argue that the act of censorship serves to highlight, maintain, and potentially strengthen the taboo power of the Swedish swear word in contrast to the offensive English swear word examples, which are not similarly sanctioned.

It is not the case that Swedish swear words are regularly censored, as Example 1 illustrates. However, the method of incidental observation that is often employed in swear word research (see Beers Fägersten 2017b), as it is in these case studies, suggests that English-language swearing is more prevalent than Swedish-language swearing in Swedish print media, and that the latter, though infrequent, is exclusively subject to censorship. Together, the examples presented in these cases not only detail different strategies of censorship but also show how censorship functions to highlight, not hide, swear word usage. These particular instances of censorship thus illustrate the status of English swear words in Sweden as not only generally inoffensive, but also culturally less problematic. Swedes are not beholden to the cultural norms or social constraints associated with English swear word usage in L1 contexts, and as such they are free to swear in English with abandon. The result is a swearing system that is two-tiered, with English swear words being used in ways that diverge from Anglophone (L1) norms yet do not cause offence among English L2 speakers. Indeed, it is the general lack of offence experienced that makes this divergence possible. The censorship of Swedish swear words used in similar contexts establishes their inherent value as offensive and thereby also maintains this value. Thus, the force and potential offensiveness of Swedish swear words are conveyed and preserved by virtue of their being subjected to censorship that contrasts with the lack of censorship applied to English swear words, which in turn conveys and maintains their status as inherently less offensive. It is unclear whether the advent of English swear words has boosted the taboo status of Swedish swear words, but observable asymmetry in censorship makes salient a two-tier system—ideally confirmed by additional evidence.

4. Caution and Self-Censorship in Using Swear Words in Translation: A Case Study of Norm Statements by Finnish AVT and Literary Translators

In this section, we examine the act of translating swear words in fictional works, which has been suggested to commonly involve a reduction in the number and emotional force of the words used (see below), and we consider whether the above-discussed differences in perceptions of L1 and LX swear words could partly explain the caution exercised in using strong L1 swear words as translations of strong LX swear words.

Below, we first briefly examine a large body of research on audiovisual and literary translation to investigate whether such reduction patterns exist widely. Secondly, we focus on translations into one specific target language, Finnish, and revisit the datasets of two earlier survey studies with professionals of literary and audiovisual translation (reported in, e.g., Hjort 2009, 2014) from a previously unexplored point of view: to investigate norm statements reflecting L1/L2 perceptions.

A number of authors have suggested that translators have a general tendency to use milder and fewer taboo words/swear words in translations. Such claims have been made with regard to both literary and audiovisual translation and within and across language pairs. For this paper, we surveyed 75 empirical studies in the fields of literary translation (LT) (24) and audiovisual translation (AVT) (51) on the translation of swear words, swearing, and/or taboo words1. The AVT studies examined subtitling (31), dubbing (18), or both (2). Most were analyses of professionally made translations but there were also studies on fan/amateur-produced translations. In total, 39 different language pairs were addressed in the studies. Particularly in AVT but also in LT, English was the most common source language (SL) of the translations investigated. This mostly reflects the significant share Anglo-American cultural products often occupy of the television, film, and literature markets (cf., e.g., De Bens and de Smaele 2001; Sevänen 2007; Crane 2014). However, the language pairs also included combinations such as Italian into Chinese, Swedish into Finnish, Norwegian into French, and German into Basque. The datasets of the studies varied from individual case studies (e.g., Pujol 2006; Klungervik Greenall 2011) to analyses of mid-sized or large corpora (e.g., Bucaria 2009; Alsharhan 2020; Valdéon 2020).

To mention a few examples, De Meo (2014, p. 246), in her analysis of the subtitling and dubbing of two of Ken Loach’s films into Italian, states that “both using swearwords and censoring its usage are common human phenomena” and demonstrates how in her data, the English taboo words cunt and fuck/fucking in particular have been omitted or toned down. Looking at only dubbing, Bucaria (2009) found similar trends in the same language pair in a corpus comprising material from twelve TV series. Trupej (2019), who studied a corpus of 50 English-language films subtitled into Slovenian, writes that “throughout the centuries, translators have frequently decided to avoid elements of the original which could potentially clash with reader sensitivity”, and concludes that in his data, offensive language was avoided more often than transferred (Trupej 2019, p. 68). Similarly, Alsharhan (2020, p. 7) argues that there is a “long history” of euphemizing, ’sanitizing’ and omitting swear words in Arabic translation, and that her study shows how this tendency prevails even in the English-to-Arabic subtitling of modern Netflix shows, despite Netflix’s policy of encouraging translators to refrain from omitting or toning down taboo language. Moreover, looking at translations of Roddy Doyle’s work, Horton (1998) and Ghassempur (2011) found that some of the translators they studied applied a “consistent process of toning down and omissions” (Horton 1998, p. 428).

Finally, we want to highlight a study focusing specifically on colloquialisms in Finnish novels and novels translated into Finnish. Nevalainen (2004) analyzes a large monolingual comparative corpus consisting of Finnish novels and novels translated into Finnish from multiple languages and finds that there are significant differences between the use of swear words and obscene language. Firstly, there are generally fewer swear or obscene words in the translations. Secondly, the translations into Finnish contain much fewer strong swear words such as vittu (lit. “cunt”, similar in usage as “fuck”, see below) and saatana (“satan”) than the novels originally written in Finnish. Instead, mild (such as hitto, lit. “devil”, similar to “damn”) or euphemistic interjections (voi luoja, lit. “oh creator”, similar to “oh god”) are more common in the Finnish translations.

When comparing the major strategies preferred in the data of the reviewed studies, a pattern does emerge: as many as around 90% of the studies identify the attenuation and/or omission of taboo words as major translation strategies. This was particularly true in the case of omissions in subtitling, as expected: time and space restrictions and the additional information provided by the visual and audio content understandably result in a reduction in the number of swear words. Attenuation can partly be explained by a common perception by translators according to which swear words that appear in writing (the subtitles) have more emotional force than in speech (the audio track), an effect which is not fully supported by research (see, e.g., Hjort 2009; Briechle and Eppler 2019). However, these trends also seemed to appear in subtitling studies in which the technical restrictions are taken into consideration (see, e.g., Ávila Cabrera 2016; Trupej 2019), and they also appeared in dubbing, which involves less time constraints and the medium of the original and translated texts is the same (audio). Furthermore, examples of the phenomenon could also be found in literature studies despite unlimited space and identical medium. While none of the translators studied completely erased swearing, these studies almost never report full swear word volume or strength retention, and findings of increase/intensification are extremely rare. A notable exception is Valdéon (2020), who argues that contemporary English-to-Spanish dubbing has seen an increase in swearing. An interesting discussion (for which we do not have space here), is whether omitting swearing sometimes means equivalent swearing because of linguistic and/or cultural differences in swearing habits; for discussions, see Ghassempur 2011 and Klungervik Greenall 2011. However, these arguments mainly seem to work in one direction: to justify reductions rather than increase/intensification of swearing in order to accommodate to the target culture.

It should be noted that the above body of research is diverse: a challenge for a literary review on this topic is, firstly, that the object of study can vary even when the category name remains the same. Secondly, similar objects may be studied under a variety of terms because of a lack of established terminology. Perhaps because of these challenges and other reasons of convenience, many have chosen to examine a single lexeme or a small selection of lexemes (e.g., Ghassempur 2011; De Meo 2014; Diáz-Pérez 2020) rather than the full range of taboo/swear words in a dataset. There are also no standard categories of translation strategies for taboo expressions: the names and number of categories vary. With these caveats in mind, however, these studies can be deemed similar enough to warrant a comparison: they all address an act of interlingual transfer where the object of transfer is subject to restrictions because it is considered somehow inappropriate, offensive, or sacred (i.e., taboo), and all of them record, at the minimum, whether swear words are replaced with equivalent terms or somehow manipulated (omitted, attenuated, euphemized, rendered with a non-taboo word).

What other explanations could there be, then, for these patterns? Santaemilia (2008) suggests that such tendencies arise from translators having a general tendency towards self-censorship when working with taboo language. While Brownlie (2007, p. 2006) defines translational self-censorship as gatekeeping exercised by a translator who voluntarily removes or edits something in a translation to conform to prevailing societal norms or to seek approval from an authority, and Merkle (2010) argues that the term should be reserved for such societally motivated manipulation, Santaemilia (2008, pp. 221–22) applies a more encompassing definition. For him, self-censorship is “an ethical struggle between self and context”, in which translators “produce rewritings that are ‘acceptable’ from both social and personal perspectives”.

Are translators generally more conservative or cautious than the creators of the original work or could other factors be at play? While the task of translators arguably is to recreate the voices of others (writers and characters), it has long been accepted in translation studies that translators have a voice (Hermans 1996) and a habitus (Simeoni 1998) of their own, and their personal backgrounds can impact the choices they make professionally. Santaemilia (2008, p. 227) suggests that when translators attempt to produce socially and personally acceptable translations of tabooed vocabulary and topics, they aim to “safeguard their professional status or their socio-personal environment”. Thus, there can be an element of risk avoidance (Pym 2008; Hjort 2017).

There is not a lot of research on translator attitudes towards swearing but, for example, a previous analysis of the below survey of Finnish literary translators (Hjort 2006, 2017) points to the existence of some conscious caution, for example in a statement according to which a translator was not personally comfortable with elements in the original text because of their personal religious beliefs. However, when the same translators were asked about their main strategies regarding the translation of swearing, they overwhelmingly said that they aim at equal offensiveness and avoiding prudishness. Statements to the contrary were rare.

Could there therefore be further factors that increase the likelihood of the reduction in swear word frequency and strength in translations, even when translators make a conscious effort to avoid it? We have previously suggested that the impact of the discrepancy between the perceived emotional force of L1 and LX swear words should be investigated in more detail (Hjort 2014, 2017, p. 165), and we reiterate that suggestion here. Below is a first attempt to do so: we revisit two datasets from previous surveys to see whether there is any evidence of L1/L2 perceptions that might point to the impact of such perceptions.

We have not found other discussions on this question, apart from a mention by Valdéon (2008, p. 369) regarding novice translators. He suggests that a discrepancy between swear words used by student translators (milder than in the original) as compared to those used by professional translators (equal to or even stronger than in the original) could be due to the findings that language learners” find offensive words in the L1 more emotionally charged than L2 swear words” (ibid.). Below, we consider whether these perceptions can be so significant that they also impact professional translators.

In a survey of 44 Finnish AVT translators (from Hjort 2009; “Dataset 1”) and another of 46 Finnish literary translators (Hjort 2006, 2017, “Dataset 2”), both with multiple source languages, the respondents were requested to explain, in their own words, their principles for translating swearing, and to report on any related instructions or feedback they have received. The issue of interlanguage swear word strength was not explicitly mentioned in the survey questions, but it did arise spontaneously.

The analysis reveals 15 comments (11 in Dataset 1 and 4 in Dataset 2) in which the issue of the strength of the L1 target language (TL) (Finnish) as compared to that of the LX source language (SL) is raised in some form. Many involve comparisons between Finnish and a specific SL or SL word. In addition, some respondents describe the Finnish swear word vocabulary as “exceptionally” or “comparatively” rich, such as the respondent in the first quote (the Finnish quotations are the original, the English quotations are translations):

Example 5: Suomessahan on esim. englantiin verrattuna erittäin rikas kirosanaperinne

Finland in fact has a very rich swearing tradition compared to for example English.

(AVT, woman, 18–30 years)

In this second quote, a literary translator comments on the emotional force of Finnish swear words at large and describes it as generally stronger than those of the language(s) he translates from (not known):

Example 6: Suomennoksessa käytän vähemmän kirosanoja, koska suomen sanat [ovat] yleensä voimakkaampia kuin lähtökielen.

When I translate into Finnish, I use fewer swear words because Finnish swear words [are] usually stronger than those of the source language.

(LT, man, 41–50 years)

To align with the first analysis in this article, what follows will focus on references to English. Multiple responses in the two datasets raise the question of the strength of one English word in particular, fuck, and how this word compare to its Finnish equivalents. In Dataset 1 (AVT), there are a total of six comments on the emotional force of fuck. In the first quote, the respondent views fuck as a weak swear word when used by a certain demographic:

Example 7: On myös huomattava, että kirosanojen painoarvot vaihtelevat voimakkaasti eri kielissä, vaikka sanojen merkityssisältö olisi sama (esim. englannin fuck on melko heppoinen täytesana, jos puhuja kuuluu sanan luontaiseen käyttäjäryhmään), eivätkä merkityssisällöltään samoilla kirosanoilla ole aina sama funktio eri kielissä.

It must also be noted that the weight of swear words varies strongly between languages, even when they have the same meaning (e.g., the English word fuck can be a relatively weak filler if the speaker belongs to the word’s natural user group), and swear words with a similar meaning do not always have the same function in different languages.

(AVT, woman, 31–40 years)

In the second quote, the respondent refers to the conventional Finnish equivalent of fuck, namely, vittu, and argues against it being used as frequently as fuck:

Example 8: Tämä riippuu tietysti elokuvan tyylistä, mutta monissa elokuvissa viljellään esim. "fuckia" niin arkipäiväisesti ja tiheään, että sen kääntäminen jatkuvasti "vituksi" häiritsisi ainakin minua katsojana.

This depends, of course, on the style of the movie, but in many movies the word fuck is thrown around in such a mundane manner and so often, that translating it always as vittu would be annoying; at least I would be annoyed as a viewer.

(AVT, woman, 31–40 years)

In the following quote, the respondent refers to an assumed equivalence between fuck and vittu but argues that using a milder translation is not necessarily producing a milder effect, because of the longer history of fuck:

Example 9: Vaikka fuckin kääntäisi muuksi kuin vituksi, kyse ei ole välttämättä lieventämisestä, vaan siitä että se on ollut pidempään yleinen kuin vittu ja siten menettänyt tehoaan. Jos siis joka fuckin kääntää vituksi, itse asiassa tulee voimistaneeksi kirosanaa. (Otin nyt tuon fuckin esimerkiksi, kun se ehkä yleisimmin herättää keskustelua).

Even if fuck was to be translated as something else than vittu, it’s not necessarily about making it milder, but about it having been common longer than vittu, and therefore it has lost some of its force. Therefore, if every fuck is rendered as vittu, you will actually intensify the swear word. (I chose fuck as the example because it is most often being debated).

(AVT, woman, 31–40 years)

One translator reported that fuck had been specifically referenced in instructions she had been given:

Example 10: Fuckia viljellään kuin mitä tahansa sanaa ja sitä vastaavat suomeksi monet sanat. Fuck on korostusta eikä sitä todellakaan käännetä aina.

Fuck is being cultivated as if it were any old word and there are many Finnish words that correspond to it. Fuck is emphasis and should definitely not always be translated.

(AVT, woman, 41–50 years)

Fuck is specifically mentioned by several literary translators, too. In the quote below, emotional force is not addressed directly but the response reveals a perception of fuck as repetitive, and a general view of Finnish swear words being more diverse than those of English:

Example 11: Koska esim. ”fuck” ja ”fucking” jatkuvasti hoettuna ovat toivottoman yksitoikkoisia, pyrin usein hakemaan vaihtelua suomen kielen monipuolisemmasta kiro- ja voimasanavalikoimasta.

Because for example ”fuck” and ”fucking”—when constantly repeated—are helplessly boring, I often have to look for variation from the more versatile swear word selection of the Finnish language.

(LT, man, 61+ years)

The respondent below, on the other hand, lists fuck and hell as examples of words that have, in her view, lost force as compared to their Finnish equivalents:

Example 12: Yleensä käytän suomessa lievempää ilmaisua kuin englannissa käytetään. Ilmeisesti englannin hell, fuck ovat virttyneempiä kuin suomen vastaavat, jotka kyllä tekstissä pistävät pahasti silmään.

I usually use a milder expression in Finnish than in English. Apparently, the English hell, fuck are more worn than similar words in Finnish, which do really stand out in a text.

When analyzing the statements of Finnish AVT translators, it should be borne in mind that the context is subtitling, and thus space is limited and the modalities of the original and the translation (audio vs. subtitles) are different (the audience is often able to gain swearing-related information from the audial and visual clues). The above-mentioned common perception that swear words have a stronger impact when they are used in the subtitles than in the audio track does not appear to be the main concern here: the respondents are comparing languages and seem to share a perception that the swear word fuck is milder, older, or is used in a more frequent and less offensive manner than vittu. However, in many respects, the words are arguably very similar. While testing and comparing the emotional force of these two words is difficult, both are also words that are still described as (amongst) the strongest non-slur swear words in their respective languages (e.g., Hjort 2015; Ofcom 2021), while also being the most or amongst the most frequent (see swearing paradox, Beers Fägersten 2012). Furthermore, while fuck might overall have longer roots, the periods marking the frequent use of both fuck and vittu in popular culture are not that far apart: for example, in US literary fiction, fuck started to appear around the 1950s and increased in frequency from the 1960s to 1970s (Sheidlower 1999, pp. xiv–xv), while in Finland, vittu started to appear in the 1960s and increased significantly around the 1970s and 1980s (Hjort 2015, p. 321). In television, their rise took a decade or two more to begin (e.g., Sheidlower 1999 for fuck). Although attitudes have loosened (e.g., Ofcom 2021, p. 3), both are still subject to certain restrictions and resistance in the public sphere (with regard to TV, see, for example, Ofcom (2021) for the concept of watershed, Ofcom (2021) and BSA (2010) on attitudes towards the use of fuck in TV). The excessive use of both is regularly criticized, particularly with regard to young people (for vittu, see, e.g., Priiki 2020).

These responses illustrate the thinking process of two groups of L2 to L1 translators who process swearing in their work. Translators are language professionals and arguably more sensitive to the nuances of language than an average language user. However, it is still possible that not even professionals are immune to the effects of the L1. We argue that the above views on language differences and the exceptional nature of the translators’ L1 could be indications of this. The general tendency of translators to exercise caution (through omission and/or attenuation) in using the strongest words in their L1 shown in the literature review is unlikely to be caused by a single cause, but combined with other causes—such as the above-mentioned technical issues and self-censorship stemming from conservative attitudes and professional risk avoidance—such safeguarding of one’s embodied L1 swear words will result in bodies of translated works that are very different from their originals in terms of swearing.

5. Discussion

Native or first-language (L1) swear words are consistently perceived as more taboo, and thus more emotional and powerful than equivalent words from a second or third language (LX) (Dewaele 2004a, 2010b; Pavlenko 2008). In this article, we focused on taboo in the form of L1 and LX swearing, providing examples from Swedish and Finnish contexts as both visualizations and verbalizations of this perception, respectively. In similar contexts of usage, English-language swearing in written texts is not censored while Swedish-language swearing is; English-language swearing in fictional works is not reliably recognized as warranting similar frequency or force when translated into Finnish. We suggest that these practices serve as taboo maintenance and strengthening of L1 swear words. In the Swedish context, contrasting censorship practices position Swedish swear words as more powerful in their potential to offend; in the Finnish context, examples of translations and norm statements of English-to-Finnish translators reveal the belief that the true Finnish counterparts of English swear words would stand out as both unusual and overly offensive.

Overt censorship of written texts (e.g., by means of asterisks) both highlights and obscures swearing: the propositional content is nevertheless conveyed, but the observable censorship also communicates an inherent taboo and an ideological stance that may be transferred to and adopted by the observer as a pragmatic effect. Similarly, the omission or weakening of source-material swearing in literary or audiovisual translations reduces exposure to (strong) L1 swearing, sustaining a lesser frequency that may serve to preserve the force of (strong) L1 swear words. Such subjective practices are certainly a function of the acquired emotional force of L1 swearing, but they also feed into and perpetuate this force.

Importantly, the examples featured in our analyses illustrate L1 swearing in relation to LX swearing, particularly when English is the LX. While L1 swearing is framed as potent, English swear words are invoked as non-offensive tools of verbal stylization, edgy entertainment, or advertising gimmicks. And while L1 swearing is perceived as warranting concern and conservative practices, English swear words pale in comparison as “overused” and “diluted”. Evidence of both the parallel use of English and L1 swear words and of the resistance to attributing to English swear words the same force as L1 swear words serves to establish a two-tier system of swearing in multilingual contexts, whereby the taboo of English swear words is lessened or even neutralized, while the taboo of L1 swear words is strengthened.

The case studies presented in this article provide indications of the relationship between taboo and L1/LX swearing, but the scope is, admittedly, limited. Further research is recommended to investigate L1 and LX swearing in a variety of interactional settings and with a variety of communicative goals. The use of English LX swear words prompts particular attention. The global reach of Anglophone popular culture contributes to the spread of English, generally, and English-language swearing, specifically (Beers Fägersten 2023). If widespread English LX swearing jeopardizes its potency in the LX, what is the effect, if any, on L1 English swearing? The proposed two-tier (L1 and LX English) system of swearing does not similarly apply to L1 English speakers, prompting the question of whether (and if so, how) the emotional force of L1 English-language swear words can be sustained despite their global spread. We welcome further and broader investigation of the taboo nature of L1 vs. LX swearing.