Information & Media eISSN 2783-6207
2026, vol. 105, pp. 49–61 DOI: https://doi.org/10.15388/Im.2026.105.3
Loreta Huber
Vilnius University, Lithuania
loreta.huber@knf.vu.lt
https://orcid.org/0000-0001-5453-7995
https://ror.org/03nadee84
Aušra Pokvietytė
Vilnius University, Lithuania
ausra.pokvietyte@knf.stud.vu.lt
https://ror.org/03nadee84
Abstract. Environmental documentaries play an important role in communicating climate change, yet the translations through which they reach international audiences remain underexplored. This study examines how the representation of truth in Before the Flood (2016) is reshaped in two Lithuanian voice-over translations: a professional translation (TT1) and an amateur translation (TT2). A comparison of professional and amateur translations of the same documentary within the same language pair provides a controlled basis for examining how different translation practices may shape the mediation of documentary truth claims. Representation of truth is operationalised through four dimensions: factual content, specialised climate-related terminology, emotional register, and markers of expert authority. A comparative, product-oriented mixed-methods analysis of 176 source-text segments combines qualitative analysis of translational shifts with quantitative comparison across the four categories. The findings show that TT1 preserves the factual content and specialised terminology more consistently than TT2 and retains more markers of expert authority. The emotional register presents a different pattern: both translations modify the same proportion of emotionally charged instances, predominantly reducing their intensity.
The findings demonstrate that the professional translation’s advantage is dimension-specific rather than absolute. Translation thus emerges as an active site of environmental discourse mediation through which the factual, terminological, affective, and epistemic resources constructing documentary truth claims may be reshaped.
Keywords: audiovisual translation; voice-over; climate change communication; documentary translation.
Santrauka. Aplinkosaugos dokumentiniai filmai atlieka svarbų vaidmenį komunikuojant klimato kaitos problemas, tačiau vertimai, kuriais jie pasiekia tarptautinę auditoriją, iki šiol tirti nepakankamai. Šiame tyrime nagrinėjama, kaip tiesos reprezentacija dokumentiniame filme Prieš potvynį (2016) perkuriama dviejose lietuviškose užklotinio vertimo versijose: profesionaliame vertime (TT1) ir mėgėjiškame vertime (TT2). To paties dokumentinio filmo profesionalaus ir mėgėjiško vertimų palyginimas, išlaikant tą pačią kalbų porą, suteikia pagrindą tirti, kaip vertimo praktikos veikia dokumentinio filmo tiesos teiginių perteikimą. Tiesos reprezentacija analizuojama pagal keturis matmenis: faktinį turinį, specializuotą su klimato kaita susijusią terminiją, emocinį registrą ir eksperto autoriteto žymenis. Atliekant lyginamąją, į vertimo produktą orientuotą mišrių metodų analizę, nagrinėjamos 176 šaltinio teksto atkarpos, derinant kokybinį ir kiekybinį keturių kategorijų palyginimą. Rezultatai rodo, kad TT1 nuosekliau nei TT2 išsaugo faktinį turinį ir specializuotą terminiją, taip pat išlaiko daugiau eksperto autoriteto žymenų. Emocinio registro atveju išryškėja kitokia tendencija: abiejuose vertimuose pakeičiama tokia pati emociškai paveikių atvejų dalis, daugiausia sumažinant jų intensyvumą.
Tyrimo rezultatai atskleidžia, kad profesionalaus vertimo pranašumas priklauso nuo konkretaus analizuojamo matmens ir nėra absoliutus. Taigi vertimas atsiskleidžia kaip aktyvi aplinkosaugos diskurso mediacijos erdvė, kurioje gali būti keičiami dokumentinio filmo tiesos teiginius konstruojantys faktiniai, terminologiniai, afektiniai ir episteminiai ištekliai.
Pagrindiniai žodžiai: audiovizualinis vertimas; užklotinis vertimas; klimato kaitos komunikacija; dokumentinių filmų vertimas
Received: 2025-08-06. Accepted: 2026-09-10.
Copyright © 2026 Loreta Huber, Aušra Pokvietytė. Published by Vilnius University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Climate-change communication is shaped by framing choices across linguistic, media, and cultural contexts. Research has shown that climate meanings are constructed through language and through interactions among verbal, visual, and other semiotic resources (Guenther et al., 2023; Kramar, 2023; Dancygier, 2023). Within this broader communicative landscape, environmental documentaries are particularly relevant because they combine scientific information with narrative and affective framing, presenting environmental realities as both factual and consequential.
When such documentaries circulate across linguistic boundaries, translation introduces a further layer of mediation. In audiovisual translation, lexical choice, terminological precision, evaluative language, and the representation of expert voices may preserve or modify source-text framing. This is especially relevant to documentary voice-over, where translated speech coexists with visual material, and often with an audible trace of the original voice. Translation can therefore be examined not only as linguistic transfer but also as a form of mediation through which environmental discourse may be reshaped for another linguistic community. This cross-linguistic mediation takes place within the broader linguistic conditions of international climate communication. English occupies a prominent position in international climate research and media production, and many globally circulated environmental documentaries originate in anglophone contexts. Their terminology and rhetorical framing may therefore already reflect the English-language climate discourse even before interlingual translation takes place. Translating such documentaries consequently entails mediating an already framed representation of climate change for a new linguistic and cultural environment. This linguistic asymmetry is particularly relevant in smaller-language contexts, where climate concepts may enter national public discourse through English-language scientific, institutional, and media channels.
The present study does not treat anglophone dominance as an empirical variable but, rather, recognises it as part of the communicative setting in which English–Lithuanian translation operates. Lithuanian voice-over thus provides a relevant context for examining how the globally circulating climate vocabulary and rhetorical framing are rearticulated within a language with its own institutional mechanisms of terminology standardisation. Despite extensive research on climate communication and audiovisual translation, the intersection of these fields remains comparatively underexplored.
Spinzi (2020) examined the Italian dubbing of Before the Flood, while Vyzas (2022) investigated specialised terminology in science-documentary translation. However, comparisons of professional and non-professional translations of the same audiovisual source within the same language pair remain rare, not least because parallel versions of this kind seldom coexist. Such a comparison offers a particularly useful basis for investigating whether different translation practices produce different patterns in the mediation of the same source material.
The research problem addressed here concerns how translation mediates the verbal resources through which a climate documentary constructs its truth claims. To investigate this question, the study compares a professional Lithuanian voice-over translation (TT1) and an amateur version (TT2) of Before the Flood (Stevens, 2016). A comparison of two translations of the same English source into the same target language limits variation in the source material and language pair, thus allowing differences between the translation products to be examined more directly.
Rather than assessing the objective truth of the documentary’s environmental claims, the study examines their representation in translated discourse through four analytical dimensions: factual content, specialised climate-related terminology, emotional register, and markers of expert authority. The analysis identifies how these dimensions are preserved or modified in TT1 and TT2, and whether the two versions display different patterns.
Documentary discourse is conventionally associated with factuality, yet documentary meaning is necessarily mediated through selection, framing, narration, image, sound, and editing. In environmental documentary, scientific evidence, quantitative data, specialised terminology, and expert testimony coexist with rhetorical and affective strategies that communicate scale, significance, and urgency. Cognitive and affective appeals can operate together in climate communication rather than functioning as opposing modes (Bieniek-Tobasco et al., 2019). Documentary truth is therefore understood here not as reality reproduced intact on screen, but as a discursively and audiovisually constructed truth claim whose credibility and force depend on how information, evaluation, and authority are represented.
For analytical purposes, this representation of truth is divided into four interconnected dimensions. Factual content concerns preservation of propositional information; specialised terminology concerns conceptual and scientific precision; markers of expert authority concern the epistemic status attributed to speakers; and emotional register concerns the evaluative and affective framing through which factual claims acquire urgency and significance. The emotional register is not treated as a measure of factual accuracy, but as part of the rhetorical environment in which facts are communicated.
The four dimensions are analytically distinct but frequently co-occur. A statement may contain a verifiable proposition, employ a specialised climate term, be delivered by an institutionally identified expert, and simultaneously use evaluative language that frames the proposition as urgent. Their separation is therefore methodological rather than ontological: it enables the analysis to specify what has changed in translation instead of treating every departure from the source as an undifferentiated loss of ‘truth’. A factual proposition can remain intact while its terminology becomes less precise; similarly, an expert’s statement can be translated accurately while the credential establishing that speaker’s authority is omitted.
The multimodal character of documentary is central to this process. Images, graphics, editing, music, vocal delivery, and verbal language may reinforce or qualify one another, and so the epistemic and affective force of a statement cannot always be inferred from words alone (Dancygier, 2023). For this reason, the present study treats the verbal translation as one component within a larger audiovisual configuration and refers to the multimodal context where it materially changes interpretation.
This framework also avoids treating emotional language as opposed to factuality. Environmental documentaries often place evidence within narratives of risk, responsibility, and anticipated consequences. Affective wording may increase or reduce the salience of a factual proposition without changing its basic truth conditions. The analytical question is therefore whether translation preserves the relationship among factual information, scientific precision, emotional framing, and attributed expertise.
Translation can modify each of these resources. Documentary voice-over is especially relevant because it is strongly associated with factual audiovisual genres and typically maintains a perceptual relationship between source speech and translated speech (Franco et al., 2010). A target text may preserve a proposition while changing terminology, emotional intensity, certainty, or attribution of expertise. Translation is therefore an additional site of documentary mediation.
Voice-over is especially suitable for this investigation because the translated speech does not wholly replace the source voice. The continued audibility of the original creates a distinctive relation between source and target discourse and reinforces expectations associated with factual audiovisual genres (Franco et al., 2010). At the same time, voice-over remains a constrained form of translation: condensation, timing, syntactic restructuring, and the coordination of spoken translation with on-screen information may all affect what can be retained. The analysis therefore distinguishes modifications that can be described linguistically from claims about why a translator made them.
Translator subjectivity is used here in a restricted textual sense. Drawing on Chesterman (1997), the analysis distinguishes obligatory changes arising from structural differences between English and Lithuanian from non-obligatory choices among plausible target-language alternatives. Such choices do not in themselves imply error or reveal translator intention. Selected strategies – particularly paraphrase, omission, addition, generalisation, and specification – are used descriptively to identify how observable Source Text (ST) ↔ Target Text (TT) relationships are produced.
The professional-amateur distinction is likewise treated descriptively rather than normatively. Non-professional translation may operate under production conditions and norms different from commissioned professional practice (Pérez-González & Susam-Saraeva, 2012). TT1 is therefore not used as the standard against which TT2 is judged; both versions are compared independently with the same source text. This distinction is important because the professional status is not treated as a synonym for quality. Professional and non-professional audiovisual translation may differ in access to briefs, terminology resources, revision, technical support, and production conventions, but these conditions cannot be reconstructed from the finished texts alone. The comparison therefore evaluates TT1 and TT2 against the same source text and according to the same analytical criteria. Any professional advantage is identified only where the coded source features are demonstrably preserved more consistently in TT1 than in TT2.
The study adopts a comparative, product-oriented mixed-methods case-study design, combining qualitative analysis of translational shifts with descriptive quantitative comparison. Before the Flood was selected because it offers an unusually controlled case: two Lithuanian voice-over versions circulate for the same English source, an officially commissioned professional version (TT1) along with an independently produced amateur version (TT2). Their coexistence permits comparison while holding the source film and target language constant. The internationally circulated National Geographic documentary also combines scientific explanation, interviews with recognised experts and public figures, specialised environmental vocabulary, and overtly evaluative climate rhetoric.
Three analytical claims guided the study. C1 proposes that documentary translation is shaped not only by linguistic constraints but also by translator subjectivity, understood as non-obligatory choices among plausible target-language alternatives. C2 proposes that such choices may reshape environmental communication by modifying factual, terminological, affective, or epistemic resources. C3 predicts that TT1 will preserve the source documentary’s representation of truth more consistently than TT2.
The corpus comprises 176 unique source-text segments selected through criterion-based purposive sampling from the documentary’s spoken English text. A segment was included if it contained at least one of four features: factual information, specialised climate-related terminology, emotionally charged language, or a marker establishing expert authority. Due to the fact that categories overlap, the 176 segments yielded 209 category occurrences: 74 factual-content instances, 75 climate-related terms, 42 emotionally charged instances, and 18 markers of expert authority. The source segment is the unit of corpus selection; the relevant feature within it is the unit of category-specific analysis.
Two researchers, namely, both authors of the study, coded the corpus by using the same category definitions. Relevant segments were aligned with TT1 and TT2, and translational shifts were classified; interpretatively difficult cases were discussed and resolved through compromise and agreement. Both target texts were compared directly with the ST according to identical criteria.
As C3 predicts professional advantage, particular care was taken not to build that expectation into the procedure. TT2 was never evaluated against TT1; each target segment was aligned separately with the same ST. A modification was recorded only in relation to the relevant source feature. This allows professional-amateur differences to emerge from the data rather than from treating the professional version as a normative reference.
The four dimensions were operationalised separately even when they co-occurred. For example, a sentence spoken by an identified scientist could simultaneously count as factual content, contain a specialised term, and include a marker of expert authority. This separation makes it possible to distinguish a preserved proposition from a weakened expert attribution, or terminological imprecision from factual alteration.
Analysis operated at two levels. First, each ST-TT relationship was assessed for its effect on the relevant dimension: preservation or modification of the factual content, conceptual specificity of terminology, preservation/intensification/reduction of the emotional register, and preservation or omission of expert-authority markers. Second, relevant operations were described with selected Chestermanian strategies. Quantitative results report frequencies and distributions descriptively rather than inferentially. Multimodal discourse analysis was used selectively where the visual or auditory context materially affected interpretation (O’Halloran, 2011). The study does not infer translator intention or measure audience reception.
The factual content was analysed in terms of whether source propositions and causal relations were preserved. Among 74 coded factual instances, TT1 modified 29 (39%) and TT2 modified 51 (69%) cases. Example 1 illustrates a consequential difference in the explanation of oil-sands extraction.
The ST presents a causal sequence: steam is pumped into the ground, the ground warms, and oil is released. TT1 retains both human agency and the intervention but renders steam as ‘karštą srovę’ [a hot stream]. The mechanism is therefore largely preserved, although terminological precision is reduced through paraphrase. Whereas, TT2, ‘Kai žemė šyla, nafta tiesiog kyla į viršų’ [When the ground warms, oil simply rises to the surface], omits steam injection and human agency. This omission changes the propositional structure by presenting warming as the initial condition rather than the consequence of an extraction procedure. The example shows that a shift may extend beyond wording to modification of factual explanation.
Table 1. Translation of a fact-based segment describing oil-sands extraction
|
Timecode |
ST |
TT1 |
TT2 |
Translation |
|---|---|---|---|---|
|
00:11:11–00:11:19 |
We just pump steam into the ground and then as the ground warms… literally the oil drops. |
Pumpuojame karštą srovę į žemę ir jai sušilus iš jos ima varvėti nafta. [We pump a hot stream into the ground, and once it warms, oil starts dripping from it.] |
Kai žemė šyla, nafta tiesiog kyla į viršų. [When the ground warms, oil simply rises to the surface.] |
TT1: paraphrase; TT2: omission of steam injection and human agency |
Source: authors’ analysis
The segment is spoken by Marc Mageau, Vice President of Oil Sands Operations at Suncor. His institutional role contributes to expert authority, while his explanation of steam injection constitutes factual content; the rendering of steam also intersects with specialised terminology. These overlapping features were coded separately. In the factual category, the decisive issue is whether the extraction mechanism remains recoverable from the translated proposition.
The audiovisual context does not compensate fully for TT2’s omission. The industrial landscape signals an extraction process, but the verbal source specifies how warming is produced. TT1 remains broadly compatible with that explanation because the action of pumping is retained. TT2 removes the intervention that links industrial activity to the resulting release of oil. The example therefore illustrates why omission can be analytically more consequential than a lexical paraphrase even when both target versions remain superficially coherent.
The corpus contains 42 emotionally charged instances, coded according to whether their affective force was preserved, intensified, or reduced. Both target texts modified 28 instances (67%): TT1 reduced the emotional force in 23 and intensified it in five cases, while TT2 reduced it in 25 and intensified it in three instances. Meanwhile, fourteen instances remained unchanged in each version. Thus, unlike the factual content, the emotional register shows no difference in the overall rate of modification; reduction predominates in both cases.
Table 2. Translation of an emotionally charged segment on increased coal use
|
Timecode |
ST |
TT1 |
TT2 |
Translation |
|---|---|---|---|---|
|
00:35:37–00:35:42 |
If those households move to coal, you have that much more use of fossil fuels, then the entire world is fried. |
Jeigu jie visi ims deginti anglis, jų poreikis išaugs ir nuo to kentės visas pasaulis. [If they all start burning coal, their demand will grow, and the whole world will suffer from it.] |
Jei tie namai naudotų anglį, jie naudotų iškasenų kurą, tuomet visas pasaulis būtų užnuodytas. [If those households used coal, they would be using fossil fuel, and the whole world would be poisoned.] |
TT1: paraphrase, reduced force; TT2: paraphrase, imagery redirected |
Source: authors’ analysis
Sunita Narain’s warning about an increased coal use provides a representative case. The expression “the entire world is fried” combines catastrophic evaluation with heat-related imagery. TT1 renders it as ‘nuo to kentės visas pasaulis’ [the whole world will suffer from it], thus preserving negative evaluation while reducing figurative vividness and emotional intensity. TT2 uses ‘visas pasaulis būtų užnuodytas’ [the whole world would be poisoned], retaining strong negative force while redirecting the metaphor from heat to toxicity. Both versions employ paraphrase, but their rhetorical effects differ. This supports treating the translation strategy and the discursive effect as separate analytical levels.
The multimodal context reinforces the source utterance: Narain is foregrounded visually, he speaks with a serious expression, and gestures emphatically. These visual resources remain available in both target versions while the verbal expression changes. The emotional register therefore contributes to the documentary’s representation of truth not by determining factual accuracy, but by framing the significance and urgency of factual and projected consequences.
The contrast is methodologically important because the same Chestermanian strategy – paraphrase – produces different effects. Strategy identifies the operation, whereas the four-dimensional framework identifies its consequence for the documentary’s truth claim. TT1 neutralises the source metaphor; TT2 maintains intensity but changes the conceptual image. A strategy-based analysis alone would therefore miss the difference between a reduced emotional force and redirected emotional framing.
Terminological analysis distinguishes lexical variation from conceptual modification. Of 75 coded climate-related terms, TT1 modified 24 (32%) and TT2 42 (56%). TT1 is more consistent for recurrent concepts such as climate change, fossil fuels, and global warming, whereas TT2 uses a wider range of paraphrases and broader or narrower expressions.
Variation, however, is not inaccuracy by default. For climate change, TT1 predominantly uses ‘klimato kaita’, while TT2 also uses ‘klimato pokyčiai’ [climate variations], ‘klimatas’ [climate], and expressions referring to climate warming or effects on climate. For fossil fuels, TT2 ranges from near-equivalent forms to the broader ‘kuras’ [fuel] and paraphrases such as ‘senųjų išteklių’ [old/ancient resources], which reduce or modify conceptual specificity. Sustainable energy is a useful counterexample: TT1 uses ‘švari energija’ [clean energy] and ‘atsinaujinanti energija’ [renewable energy], while TT2 uses ‘neišsenkanti energija’ [inexhaustible energy] and ‘atsinaujinanti energija’. Neither version consistently preserves the source concept, showing that professional status does not guarantee terminological equivalence.
Table 3. Comparative rendering of five recurrent climate-related terms
|
Source term |
TT1 rendering(s) |
TT2 rendering(s) |
Comparative assessment |
|
climate change |
klimato kaita [climate change], rare variation |
klimato kaita; klimato pokyčiai [climate variations]; klimatas [climate]; related reformulations |
TT1 more consistent |
|
fossil fuels |
iškastinis kuras [fossil fuel], few exceptions |
iškasamas kuras; iškasenų kuras; kuras [fuel]; senųjų išteklių [old/ancient resources] |
TT1 more consistent; some TT2 forms reduce specificity |
|
sustainable energy |
švari energija [clean energy]; atsinaujinanti energija [renewable energy] |
neišsenkanti energija [inexhaustible energy]; atsinaujinanti energija |
Both versions use related but non-equivalent concepts |
|
global |
visuotinis atšilimas [global warming]; one case of pasaulinis atšilimas |
klimato kaita; klimato problemos [climate problems]; globalinis atšilimas |
TT1 more consistent; TT2 sometimes broadens the concept |
|
environmentalist |
paraphrase: aktyviai kovojate už aplinkosaugą [<you> actively fight for environmental protection] |
paraphrase: nesėdi sudėjęs rankų klimato klausimais [do not sit idle on climate issues] |
Both paraphrase; contextual meaning largely retained |
Source: authors’ analysis
In the context of Lithuania’s institutional terminology infrastructure, including the Lietuvos Respublikos terminų bankas [The Term Bank of the Republic of Lithuania], coordinated by the State Commission of the Lithuanian Language (VLKK, 2026), the terminology observed in this corpus may be characterised as relatively “standardised at the core but dynamic at the periphery”. Central recurrent concepts show greater lexical stability, while other expressions attract semantically related but non-equivalent renderings. This is a description of the corpus pattern rather than a claim that Lithuanian climate vocabulary has fixed stable and unstable zones; the analytical question remains whether a particular rendering preserves the boundaries of the source concept.
The same distinction is observed in the field of global warming. TT1 predominantly uses ‘visuotinis atšilimas’ [global warming], whereas TT2 sometimes substitutes ‘klimato kaita’ [climate change] or ‘klimato problemos’ [climate issues]. These alternatives broaden the source concept rather than merely varying its wording. By contrast, both translations render environmentalism through paraphrase rather than a compact lexical equivalent, and yet they largely preserve the contextual meaning of active environmental engagement. The examples confirm that consistency and conceptual precision must be evaluated separately.
Eighteen coded instances identify or contextualise speakers as experts. TT1 omits ten (56%) but partially retains eight (44%), typically through on-screen identification; whereas, TT2 omits all 18. These omissions do not necessarily change the propositions spoken by experts, but they remove verbal or textual cues that position those statements as professionally or institutionally grounded. Expert identification is therefore analytically distinct from factual accuracy: the information may remain while its epistemic framing changes.
This category also demonstrates the interaction between verbal and visual channels. Some expert names or affiliations may remain visible on screen even when they are not voiced, but their accessibility depends on whether the information is translated and on how long it remains available. TT1 partially preserves this layer in eight cases; TT2 provides no equivalent identification in the coded instances. The result is therefore not a claim that expert authority disappears from the film altogether, but that the translated verbal/textual resources supporting it are reduced.
This distinction matters because expertise is not identical with factual content. Removing a title or affiliation does not necessarily make the expert’s proposition less accurate, but it changes the textual basis on which viewers are invited to recognise the speaker as a qualified source. Expert authority is therefore coded separately from the information communicated by that expert. In a documentary that repeatedly moves between narration, political figures, scientists, activists, and industry representatives, such identification forms part of the epistemic organisation of the discourse.
The category totals reveal a dimension-specific professional-amateur contrast. TT1 modifies fewer factual and terminological instances than TT2 and retains more markers of expert authority. Emotional register differs: both translations modify the same proportion of coded instances, predominantly by reducing the emotional force. TT1 therefore shows its clearest advantage in preserving factual, terminological, and epistemic resources rather than affective framing.
The quantitative pattern therefore does not support a single scale on which one translation is uniformly closer to the source. TT1’s advantage is clearest in factual, terminological, and epistemic resources, while emotional framing is vulnerable in both versions. This dimension-specific result is important for interpreting the professional status: the data support greater source preservation in TT1 in several categories, but they do not justify treating every TT1 choice as preferable or every TT2 variation as loss.
Table 4. Summary of coded instances and translational modifications by analytical category
|
Analytical category |
ST instances |
TT1 |
TT2 |
|
Factual content |
74 |
29 modified (39%) |
51 modified (69%) |
|
Climate-related terminology |
75 |
24 modified (32%) |
42 modified (56%) |
|
Emotional register |
42 |
28 modified (67%): 23 reduced, 5 intensified |
28 modified (67%): 25 reduced, 3 intensified |
|
Markers of expert authority |
18 |
10 omitted (56%); 8 partially retained (44%) |
18 omitted (100%) |
Source: authors’ analysis
Taken together, the findings provide textual support for C1 and C2 and qualified support for C3. Recurrent non-obligatory shifts support C1’s restricted conception of translator subjectivity as observable choice among target-language alternatives. Their effects on factual specificity, terminology, emotional force, and expert positioning support C2’s claim that translation may reshape environmental discourse. C3 is supported for factual content, terminology, and expert authority, which TT1 preserves more consistently, but not for the overall frequency of emotional-register modification. Professional advantage is therefore dimension-specific rather than absolute.
The findings can also be considered alongside Spinzi’s (2020) study of the professionally dubbed Italian version of Before the Flood, which identifies substantial creative intervention in a professional target text. The difference should be interpreted cautiously because the studies involve different languages, production contexts, analytical frameworks, and translation modes. Dubbing and voice-over impose different formal conditions, making mode one possible factor, but the present data cannot establish causality. Together, the studies nevertheless show that professional audiovisual translation is not inherently source-proximate, and that the form of translational mediation cannot be predicted from the professional status alone.
The distinction between preservation and modification also has practical implications. Recurrent technical terms can be prepared through a short project glossary based on authoritative Lithuanian resources, while professional titles and affiliations should be treated as information contributing to epistemic framing rather than as automatically expendable detail. The shared reduction of emotional force suggests that translators should also consider the rhetorical function of affectively charged expressions alongside factual accuracy.
This study shows that translation functions as an additional layer of mediation in the construction of environmental documentary truth claims. C1 and C2 receive textual support: both Lithuanian versions contain non-obligatory choices that modify factual, terminological, affective, or epistemic resources. C3 receives qualified support. TT1 preserves the factual content, specialised terminology, and markers of expert authority more consistently than TT2, whereas both translations modify the emotional register at the same overall rate. The professional advantage identified here is therefore dimension-specific rather than absolute.
The findings also distinguish terminological variation from conceptual modification and suggest a Lithuanian climate vocabulary that is relatively standardised at the core but dynamic at the periphery. Professional translation cannot be equated with complete preservation, nor can amateur translation be related with conceptual loss; the relevant question is how a particular choice affects the resources through which a documentary establishes factuality, precision, urgency, and authority.
The study is limited to one documentary, one language pair, and two voice-over versions. It is product-oriented and does not undertake to establish translator intentions or audience effects. Future research should test the four-dimensional framework across other documentaries, languages, translation modes, and professional conditions; reception studies could examine whether the textual differences identified here correspond to differences in comprehension, perceived expert credibility, or emotional engagement. A broader English-Lithuanian terminology corpus could also test whether the ‘standardised core/dynamic periphery’ pattern extends beyond audiovisual translation. Overall, the findings indicate that environmental documentary translation involves not merely the transfer of information but also the mediation of verbal resources through which documentary truth claims are constructed.
Loreta Huber: conceptualization, methodology, validation, writing – original draft, writing – review and editing.
Aušra Pokvietytė: formal analysis, writing – original draft, visualization, investigation, methodology.
The authors declare that they have no financial interests or benefits arising from the direct applications of this research and no conflicts of interest to disclose.
Bieniek-Tobasco, A., McCormick, S., Rimal, R. N., Harrington, C. B., Shafer, M., & Shaikh, H. (2019). Communicating climate change through documentary film: Imagery, emotion, and efficacy. Climatic Change, 154(1–2), 1–18. https://doi.org/10.1007/s10584-019-02408-7
Chesterman, A. (1997). Memes of translation: The spread of ideas in translation theory. John Benjamins.
Dancygier, B. (2023). Multimodal media: Framing climate change. Discourse Studies, 25(2), 220–236. https://doi.org/10.1177/14614456231154724
Franco, E., Matamala, A., & Orero, P. (2010). Voice-over translation: An overview. Peter Lang.
Guenther, L., Jörges, S., Mahl, D., & Brüggemann, M. (2023). Framing as a bridging concept for climate change communication: A systematic review based on 25 years of literature. Communication Research, 51(4). https://doi.org/10.1177/00936502221137165
Kramar, N. (2023). Construction of agency within climate change framing in media discourse: A corpus-based study. Respectus Philologicus, (43(48)), 36–48. https://doi.org/10.15388/RESPECTUS.2023.43.48.106
O’Halloran, K. L. (2011). Multimodal discourse analysis. In K. Hyland & B. Paltridge (Eds.), Continuum companion to discourse analysis (pp. 120–137). Continuum.
Pérez-González, L., & Susam-Saraeva, Ş. (2012). Non-professionals translating and interpreting: Participatory and engaged perspectives. The Translator, 18(2), 149–165. https://doi.org/10.1080/13556509.2012.10799506
Spinzi, C. (2020). (Re)creating actuality in environmental documentaries: The case of Before the Flood. In A. Rizzo (Ed.), Translation and accessibility for all in the creative industries: Digital spaces and cultural contexts (Vol. 1, No. 2, pp. 107–131). Department of Translation Studies, Faculty of Arts Constantine the Philosopher University in Nitra.
Valstybinė lietuvių kalbos komisija [State Commission of the Lithuanian Language]. (2026). Lietuvos Respublikos terminų bankas [Term Bank of the Republic of Lithuania]. https://terminai.vlkk.lt/
Vyzas, T. (2022). Specialised vocabulary in subtitling science documentaries. Journal of Audiovisual Translation, 5(1), 48–72. https://doi.org/10.47476/jat.v5i1.2022.152
DiCaprio, L. (Producer), & Stevens, F. (Director). (2016). Before the Flood [Motion picture]. National Geographic Documentary Films.
DiCaprio, L. (Producer), & Stevens, F. (Director). (2016). Prieš potvynį [Motion picture]. National Geographic Documentary Films. Amateur translation.
DiCaprio, L. (Producer), & Stevens, F. (Director). (2016). Prieš potvynį [Motion picture]. National Geographic Documentary Films. Professional translation by Mindaugas Lapinskas.