Respectus Philologicus eISSN 2335-2388
2025, no. 48 (53), pp. 109–123 DOI: http://dx.doi.org/10.15388/RESPECTUS.2025.48.9
Maryna Bielova
University of Birmingham
Edgbaston, Birmingham B15 2TT, United Kingdom
H. S. Skovoroda Kharkiv National Pedagogical University
29 Alchevskykh St, Kharkiv, Ukraine
Email: m.bielova@bham.ac.uk
ORCID iD: https://orcid.org/0000-0002-1306-9395
Scientific interests: Cognitive linguistics, Cognitive stylistics, Multimodal studies, Advertising discourse, Translation of tropes, Translation of neologisms, Audiovisual translation, Machine translation
Abstract. Stylistic neologisms (SN), new words crafted to achieve pragmatic effects, pose significant challenges for translators, particularly between typologically distant languages like English and Ukrainian. With advances in machine translation (MT), evaluating its handling of SNs – especially vs. human translation (HT) – is crucial. Despite growing research on HT vs MT of neologisms, studies focusing on English > Ukrainian translation remain absent, leaving a critical research gap. This study addresses this gap by analysing how HT and MT rendered SNs formed through various morphological patterns in English language chick (ELCL) into Ukrainian. The findings reveal that while HT demonstrated exceptional abilities to creatively render English language SNs into Ukrainian, MT lacked creative and cultural sensitivity. MT exhibited an error rate of over 80%, significantly higher than HT’s 14%, with the most frequent errors occurring with loss of connotative meaning and incorrect word formation. The lack of innovation and contextual awareness in MT outputs underscores the necessity for algorithmic advancements and post-editing strategies to improve the rendering of SNs in literary translation.
Keywords: stylistic neologism; chick lit; linguistic creativity; cognitive metaphor and metonymy; machine vs human translation strategies and techniques.
Submitted 7 February 2025 / Accepted 20 May 2025
Įteikta 2025 02 07 / Priimta 2025 05 20
Copyright © 2025 Maryna Bielova. Published by Vilnius University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution License CC BY 4.0, which permits unrestricted use, distribution, and reproduction in any medium provided the original author and source are credited.
In ELCL, a genre marked by verbal inventiveness, cultural satire, and emotional immediacy (Harzewski, 2011), SNs are genre-bound lexical innovations that blend morphological extravagance with deliberate stylistic intent. The term is supported in scholarly literature (e.g., Cañete-González, Llopart-Saumell, 2023) alongside the related terms “extravagant neologisms” (Lensch, 2022) and “extragrammatical neologisms” (Mattiello, 2017). These neologisms challenge translators for their high degree of creativity and contextual sensitivity – especially between typologically distant languages like English and Ukrainian, where morphological structures and cultural references may not align.
The relevance of this research stems from ongoing technological advances that encourage the integration of MT tools into literary translation workflows. While previous studies have dealt with the challenging domain of MT creativity in literary discourse and explored the translation of metaphor (Dorst et al., 2024), creative word forms including unconventional orthography, derivation, compounding, and wordplay (Kenny, 2001), manipulated multiword expressions (Corpas Pastor, Noriega-Santiáñez, 2024), formal neologisms in classical and dramatic texts (Kolb et al., 2023), fantasy, mystery, and children’s literature (Noriega-Santiáñez, Corpas Pastor, 2023) between high-resourced languages like English, German, and Spanish, research has yet to be done on the HT vs. MT of SNs in ELCL into Ukrainian. This pilot study, which seeks to address this notable gap, is the first to evaluate the comparative performance of HT vs. MT specifically in the domain of SNs in ELCL by addressing the following objectives:
1. To assess the effectiveness of HT in rendering SNs from ELCL into Ukrainian.
2. To evaluate how MT performs in reproducing SNs from ELCL into Ukrainian.
3. To compare Google Translate and DeepL in their ability to translate SNs.
4. To determine whether certain SNs are more easily processed by the two MT engines than others.
5. To determine whether MTs offer viable solutions or are sufficiently intriguing to warrant consideration by HTs from a pragmatic perspective, in alignment with the presumed intent of the original author.
The structure of the argument is as follows: following the introduction, the data and the applied method are described (Section 2). This is followed by an analysis of the HT of the sampled neologisms (Section 3), a comparison with MT (Section 4), and a conclusion (Section 5).
This study adopts a mixed-method approach. It draws on 89 SNs sampled from 13 ELCL novels by using a close reading methodology and following a morphosemantic approach, which yielded a wide variety of SNs, including derivatives (49%), compounds (40%), clippings (3%), and conversions (8%). The collected SNs meet the following criteria: (1) They are novel, absent from standard English dictionaries. However, some appear in the historical Oxford English Dictionary (OED), dictionaries of unconventional English, or electronic corpora, suggesting early usage. (2) Their novelty is influenced by origin and spread. (3) They show phonological, morphological, or semantic deviation. (4) Their meanings are highly context-dependent. (5) Most lack broad semantic applicability, remaining genre-specific. (6) They are often explicitly marked – through punctuation, meta-comments, or typography – to flag their creative, stylistic nature.
The original SNs are treated as source texts (ST) and their Ukrainian HTs and MTs as target texts (TT). The HT equivalents were sampled from 13 printed Ukrainian translations of the original texts. For machine-generated translation, two widely used general-purpose MT engines were selected, Google Translate (GT) and DeepL (DL), as they both operate on artificial neural networks (Specia, Wilkes, 2022) and are commonly employed for the English>Ukrainian language pair. The neologisms were translated within their broader textual contexts rather than as standalone units in November 2024.
This study approaches translation of neologisms with an evaluative focus, concentrating specifically on the assessment of translation quality. Quality is assessed through a set of pre-defined criteria grounded in translation theory (House, 1997), alongside insights from Newmark (2008, p. 143), who emphasises that in fiction, neologisms should be creatively reimagined using equivalent morphemes and sound effects while preserving naturalness in morphology and phonology. Translation quality is explicitly defined as the extent to which the target text (TT) preserves the semantic fidelity, stylistic function, morphological creativity, and pragmatic appropriateness of the original. Evaluation transparency is ensured through the use of annotated examples.
Drawing on prior research into neologisms translation (Delabastita, 2004; Kenny, 2001; Kolb et al., 2023; Salich, 2020), this study categorises strategies to render the meaning and effect of new words as follows: occasionalistic, non-occasionalistic translation, and omission.
Occasionalistic translation generates new TL expressions to retain source language (SL) neologisms’ creativity and style. In cases where an SN was translated using this strategy, The Dictionary of the Ukrainian Language in 11 volumes (Rusanovskyi, 2017) was consulted to determine whether the equivalents could be perceived as new words by the reader and the Corpus of Ukrainian Language Texts (CULT). Occasionalistic strategy includes: literal translation (direct rendering of the neologism); calque (literal translation of its components); naturalisation (adaptation to TL phonology or morphology); compensation (recreating stylistic effect elsewhere in the text); and transference (borrowing the SL term directly). Non-occasionalistic translation relies on existing TL terms, often at the cost of creativity, through replacement (using a pre-existing equivalent), explanation (adding clarifying information), or description (paraphrasing the neologism’s meaning). Omission removes neologisms entirely, eliminating their meaning or stylistic effect.
Applying these strategies to the 89 sampled neologisms, Table 1 summarises their translation into Ukrainian, including an additional category for major errors such as semantic distortion, loss of nuance, or nonsensical renderings:
Table 1. Strategies and errors for Ukrainian translation of SN in ELCL, %
|
Translation type |
Occasionalistic |
Non-occasionalistic |
Omission |
Error |
|---|---|---|---|---|
|
HT |
43 |
43 |
0 |
14 |
|
GT |
2 |
13 |
0 |
85 |
|
DL |
7 |
7 |
0 |
86 |
This interdisciplinary methodology merges insights and tools from stylistics, morphology, sociolinguistics, and translation studies to examine both the creative construction of neologisms and their cross-linguistic transfer and effectively addresses the objectives set.
To address the objectives set, the analysis begins with an examination of how HTs handle SNs, starting with the most prevalent type in the sample – derivatives – in Section 2.1. This will be followed by the analysis of compounds in Section 2.2 and the least common types in the sample, clippings and conversion, in Section 2.3.
Only 22% of the sampled derivatives were translated into Ukrainian by creating a new word and achieving a comparable effect on target readers. One such instance is the literal translation of the suffixed derivative mentionitis in (1):
(1) “It’s Mentionitis,” Jude was saying. … “Oh, you know, when someone’s name keeps coming up all the time, when it’s not strictly relevant … (Fielding, 2014, p. 41).
The SL mentionitis is a humorous blend of mention and the disease-mimicking suffix -itis-, typically used to denote inflammation or illness (LDOCE). This playful formation frames compulsive name-dropping in romantic contexts as a mock-medical condition. The Ukrainian equivalent згадоніт/zgadonit (Fielding, 2020, p. 53) mirrors the original’s structure, combining згад/zgad “mention” with the suffix -іт-/-it-, which, like -itis, commonly signals acute inflammation in medical terms (Seligej, 2014, p. 146) and is metaphorically extended to imply obsessive behaviour, echoing the original’s stylistic effect.
One other technique employed to create new words in Ukrainian while rendering English language SNs formed by affixation is replacement through prefixation, like, for instance, in (2):
(2) Ended up walking along with Rebecca’s nephew: sub-Leonardo DiCaprio lookalike … whom everyone referred to as ‘Johnny’s boy’ (Fielding, 2014, p. 103).
The author-produced sub-Leonardo DiCaprio reflects women’s tendency to assess male attractiveness when seeking potential partners. Fielding (2014) creatively uses the actor’s name metonymically substituting CATEGORY for SALIENT PROPERTY (see Littlemore, Tagg, 2018). Additionally, by prefixing it with sub-, she draws on the orientational metaphor GOOD LOOKS ARE UP (see Lakoff, Johnson, 1980), thereby implying that someone is attractive, but not to the level of Leonardo DiCaprio. The HT equivalent is недо-Леонардо Дікапріо/nedo-Leonado DiCaprio (Fielding, 2020, p. 128). In Ukrainian, the prefix недо-/nedo- can be used to convey the idea of something being incomplete (like in недоспати/nedospaty “not have enough sleep”) which fits the usage of sub-, yet in the sense of “lesser” or “not fully”. This is an example of replacement, where the stylistic nuance of the SL neologism is preserved but expressed using a different prefix that conveys a similar meaning in the TL.
In 55% of cases of SNs created by affixation in ELCL, HTs opt for a non-occasionalistic approach. One example is Ricki Lake-ism in (3):
(3) “Becky?” says my father gravely … . “Are you in some kind of trouble we should know about? Only tell us if you want to,” he adds hastily. “And I want you to know – we’re there for you.” That’s another bloody Ricki Lake-ism, too (Kinsella, 2001).
In this context, the suffix -ism- conveys a behaviour or attitude associated with Ricki Lake’s persona, known for promoting open, empathetic dialogue as a talk show host. Ricki Lake-ism humorously critiques the father’s attempt to mimic her emotionally supportive tone. The Ukrainian translation – дурнувате слово від Рікі Лейк/durnuvate slovo vid Riki Leik (Kinsella, 2009, p. 223) – uses a descriptive phrase. While this version narrows the original meaning by reducing a broader mannerism to a word, it retains the speaker’s sarcastic tone through the use of дурнувате/durnuvate “stupid”, thereby preserving the humour and implied criticism of the father’s emotional imitation.
In ELCL, a wealth of SNs is characterised by using the prefix un- with atelic verbs to convey a reversative action. Rendering these neologisms in Ukrainian requires descriptive techniques as the Ukrainian equivalent of un- (не/ne-) only denotes the absence of an action, not its reversal. This is evident in the translation of un-fire in (4):
(4) “You are fired.” Alex’s voice is like a bullet. “End of.” “You can’t do that!” I cry, outraged. “Un-fire her!” (Kinsella 2017b, p. 348).
The Ukrainian translation, скасуй це звільнення/skasuj ce zvil’nennja (Kinsella, 2017a, p. 429) “cancel this dismissal”, is a descriptive rendering that attempts to capture the idea of reversing an action.
What came as a surprise is that in several cases HTs erroneously rendered prefixed derivatives with un-. One example is (5):
(5) Okaaaay. I’ve seen the whole lot. Can’t unsee that (Kinsella, 2020, p. 230).
By adding the prefix un- to see, the author humorously conveys being stuck with an unwanted visual memory. The Ukrainian HT renders can’t unsee that using double negation as я не можу цього не бачити/ja ne možu cjogo ne bachyty (Kinsella, 2023, p. 243) “I can’t help seeing this”, which completely distorts the original meaning.
In contrast to derivatives, compounds are more often rendered into Ukrainian following the stylistic choices made by the authors (61%). One example is the translation of matchy-matchy in (6):
(6) It’s irrelevant. Matt and I have a different kind of relationship. Not so matchy-matchy. More…Well. More un-matchy-matchy (Kinsella, 2020, pp. 286–287).
In this example, matchy-matchy and its negative counterpart un-matchy-matchy are borrowed from fashion discourse, where they describe overly coordinated outfits, and creatively repurposed in ELCL to depict emotional or behavioural compatibility (or its absence) in relationships. Morphologically, matchy-matchy is a reduplicative compound, while un-matchy-matchy adds a prefix to indicate negation. The Ukrainian translation сумісні-сумісні/sumisni-sumisni (Kinsella, 2023, p. 302) employs an occasionalistic strategy, calquing the structure of the original. This preserves both the semantic content and the playful stylistic effect of the SL.
HTs resort to normalisation in rendering compounds in 27% of cases. To illustrate the point, consider the translation of clothes-berg in (7):
(7) This cannot be happening. I’m on holiday with Luke, in a smart hotel – and I have no clothes. What am I going to do? I can’t tell him the truth. I just can’t admit that my dinky suitcase was only the tip of the clothes-berg (Kinsella, 2004, p. 48).
In this scene, the narrator arrives at a hotel with only a small makeup bag, having secretly shipped her larger suitcases after promising to pack lightly. When the suitcases fail to arrive, she cannot admit the truth to her boyfriend. To capture her distress, the author coins the neologism clothes-berg. It alludes to the image of an iceberg, indirectly referring to the well-known idea of an iceberg’s hidden mass, while simultaneously surfacing a visual metaphor that humorously portrays the narrator’s hidden excess of luggage beneath the surface of minimalism. The translation of clothes-berg as одяговий айсберг/odjagovyj ajsberg (Kinsella, 2019, p. 63) “clothes iceberg” into Ukrainian is a successful attempt to preserve the core metaphor of the original by decomposing the compound into a free (Andj + N) phrase. This adapts the translation to the TL norms, as in Ukrainian N + N compounds, where the left-hand noun cannot be used in an attributive function.
HTs generated errors with compounds with a heavy cultural load. One example is striped-shirted in (8):
(8) Every striped-shirted Hugo, Johnny or Jerrers who had ever idly wondered what was under Jude’s suit was catapulted into a state of priapic frenzy, and she seems to have a different one on the phone every night (Fielding, 2014, p. 12).
Morphologically, striped-shirted is a compound adjective formed by adding -ed to striped shirt. It refers to men wearing striped shirts, but contextually it signals affiliation with professional or upper-middle-class circles. Paired with the posh diminutives Hugo, Johnny, and Jerrers, striped-shirted reinforces the stereotype of privileged, conformist men condensing rich social and cultural meaning into a single word. Its translation into Ukrainian as у смугастих сорочках/u smugastyx sorochkax (Fielding, 2020, p. 21) “in striped shirts” shifts from a neologistic compound adjective to a prepositional phrase, which lacks stylistic creativity and completely misses the humour and cultural nuance of the original.
My corpus contains only 3% of clippings and 8% of conversions. All clippings are translated with new words in the TL. For example, “witchies” and “Ho” in (9):
(9) They do strange things, too. … Like they call sandwiches ‘witchies.’ Moreover, a drink is a ‘titchy’ (except if it’s water, which is ‘Ho’) (Kinsella, 2001).
The clipped witchies, in this context, refer to a family-specific way of referring to sandwiches and have a magical association with the word witch. Its translation as броди/brody (Kinsella, 2009, p. 84) (shortened for бутерброди/buterbrody “sandwiches”) into Ukrainian imitates the word formation pattern of the original. Although броди/brody evokes a rural or less formal tone, which may fit the whimsical nature of the family’s unique language, it slightly shifts the playful connotation of witchies away from its magical association. Ho could have been chosen because of its phonetic similarity to H2O, a common way to refer to water. This might explain why Ho stands for water, not other drinks. Its translation equivalent in Ukrainian – буль/bul’ (Kinsella, 2009, p. 84) – captures the onomatopoeic sound of a water droplet. The HT’s choice retains the original intent of a quirky family language, adding a subtle layer of sound-based humour specific to Ukrainian, which may enhance the whimsical family language.
SNs formed by conversion are translated through literal and non-literal procedures. One example is be Demetered in (10):
(10) “Bloody Demeter.” “Listen, Hannah,” Rosa continues, … . “You’ve been Demetered” (Kinsella, 2017b, p. 27).
The verb is coined from the proper noun Demeter (a female name) by conversion and implies “being mistreated or subjected to something negative due to Demeter’s management style and decisions”. This stylised language captures the speaker’s disdain and disbelief about the situation, adding a significant emotional intensity. The Ukrainian деметризували/demetryzuvaly” (Kinsella, 2017a, p. 36) “have been Demetered” is formed through verbal conversion, adding the suffix -ува-/-uva-, commonly used to turn nouns into verbs in Ukrainian. This mirrors the English use of -ed- in forming past participles but aligns with the verb formation rules of Ukrainian. The translation is a faithful naturalisation successfully preserving the essence of the English neologism, retaining its humour, sarcasm, and critical nuances.
The non-literal translations of SNs formed through conversion are approximate denotational equivalents. One case is (11):
(11) I altered my path to pass his table, at which he immersed himself deep in conversation with trollop, glancing up as I walked past and giving me a firm, confident smile as if to say “business meeting.” I gave him a look which said, “Don’t you business meeting me,” and shuttered on (Fielding, 1996).
In (11), business meeting functions syntactically as a verb, derived via conversion from the compound noun. While N-to-V conversion is conventional in English (Booij, 2005, p. 67), the verb business meeting is defective – it resists tense formation. Hohenhaus (2007, pp. 28–29) refers to such formations as “delocutive conversion” – speech-act-based neologisms not tied to direct quotation and typically found in communicative situations of conflict, where a speaker rejects the form of something said to him/her. Although Ukrainian also allows delocutive conversion (e.g., не мамкай/ne mamkai “don’t mother me”), the Ukrainian HT version, мене не обдуриш/mene ne obdurysh (Fielding, 2019, pp. 135–136) “you won’t fool me” opts for an approximate rendering. It conveys the intended scepticism but loses the satirical allusion to the performative façade of a professional meeting.
Some of the SNs formed by conversion were translated erroneously, as in example (12):
(12) “I have a message from Demetra,” I announce. “She’s just a bit … tied up.” Dyed up crosses my mind, and I almost give a little snort of laughter (Kinsella, 2017b, p. 41).
In this passage, Demetra dyes her hair at work with her subordinate’s help. When a man calls to say he’s coming, she, embarrassed by her appearance, sends her subordinate to intercept him. Annoyed by the task, the subordinate humorously coins dyed up, a playful twist on tied up (“busy”). The Ukrainian HT, трохи пофарбована/trohy pofarbovana (Kinsella, 2017a, p. 54) “a bit dyed”, completely shifts away from the linguistic twist that made the original phrase playful and also fails to convey the sense of urgency.
The results reveal MT’s limitations with neologisms in ELCL, as both GT and DL often erred, relying on minimal occasionalistic strategies and producing few acceptable translations, especially for compounds and derivatives.
The occasionalistic MT solutions for compounds are mainly literal translations.
(13) People talk about the generation gap. Generation chasm, more like (Kinsella, 2017b, p. 13).
The neologism generation chasm is formed analogously to generation gap, emphasising a deeper separation. The latter has an established Ukrainian equivalent розрив поколінь/rozryv pokolin’ (Rusanovskyi, 2017). DL successfully recognises the language model and produces an occasional output прірва між поколіннями/prirva miž pokolinnjamy, decomposing the original compound into a syntactic phrase and adapting it to the rules of Ukrainian grammar. GT’s output – безодня поколінь/bezodnja pokolin’ “a chasm that contains generations” – is, however, erroneous. As mentioned, in this case, the focus should be on the relationship between the generations rather than them being contained within the chasm.
My corpus contains 4 suffixations for which one of the MT systems offered occasionalistic translation. One example is London-ness in (14):
(14) It doesn’t have that London-ness that makes my heart soar (Kinsella 2017b, p. 15).
This is an example of a successful neologistic adaptation performed by DL. It effectively utilises a Ukrainian suffixation pattern to preserve the stylistic and semantic nuances of the original. In the MT, лондонськість/londons’kist’, the equivalent suffix -ість-/-ist’- is used, a technique often employed in Ukrainian to denote an abstract quality, as in українськість/ukrajins’kist’ “ukraineness”. The MT effectively mirrors the inventive style of ELCL, where such playful or evocative terms are common. GT, unlike DL, fails to do a similar job. The engine omits the suffix -ness- and translates London-ness as a root-noun London, undermining the speaker’s attempt to convey an emotional connection to “London vibe” and introducing a significant grammatical error.
One other example is MT’s non-occasionalist outputs for showy-offy in (15) referring to a display of unnecessary or excessive behaviour:
(15) Everyone is chattering about their mornings, but Demeter’s voice rings out above the hubbub, a constant, showy-offy, clarion sound (Kinsella, 2017b, p. 232).
DL’s translation хвалькуватий/xval’kuvatyj “boastful” – a codified adjective in Ukrainian (Rusanovskyi, 2017) – captures some of the meaning but loses part of the stylistic and tonal nuances of the original, where the double suffix -y- creates a whimsical, exaggerated effect, emphasising a less serious critique. The TL adjective fits the sentence structure and does not disrupt the flow of the translation. GT’s translation ефектний, яскравий звук/efektnyj, jaskravyj zvuk “effective, bright sound” diverges from the original in tone and meaning. It entirely lacks the idea of boastfulness or self-importance and suggests a positive or neutral impression rather than the critical, playful tone of the original.
Both engines handle stylistic neologisms with significant errors (85% and 86%, respectively), which have a profound impact on the translation quality. One peculiar error with compounds relates to orthographic similarity of their elements with other words, like in (16):
(16) I tried it, and you know, it’s OK. It’s protein. It’s fine. I think I could be … semitarian, maybe (Kinsella, 2020, p. 365).
The original semitarian humorously describes someone who could be “half vegetarian”, implying they are not vegetarian but are willing to make sacrifices to align with their partner’s lifestyle. DL’s output напівлікар/napivlikar “half-doctor” is incorrect. The machine engine appears to have confused -terian with a similar-sounding word related to a profession. This error might have happened because of the algorithm’s reliance on word patterns, where the suffix -ian- is sometimes associated with professions or roles in certain contexts. GT’s output напівмітарій/napivmitarij (non-existent in Ukrainian) fails to convey the meaning completely. The MT system apparently treated the original -tarian as an unknown token, copied it into the TT and rendered the initial element semi, which led to a meaningless translation.
The MT engines demonstrate a high error rate in translating compounds, with 89% of cases being rendered incorrectly, which is a stark contrast with HTs, who accurately translated 83% of the compounds. One example is (17):
(17) “Pashminas,” I slurred, slurping on my Chardonnay. “That’s it!” said Tom excitedly. “It’s fin-de-millennium pashmina-ism, Shazzer is Simon’s “pashmina” because she wants to shag him most so he diminishes her and Simon is Shazzer’s pashmaster.”
… we could come up with a list of further definitions, as follows:
Pashmincer: A friend who you really fancy who’s actually gay. …
Pashmarried: A friend who you used to go out with and is now married with children who likes having you around as memory of old life but makes you feel like mad barren pod-womb imagining vicar is in love with self.
Ex-pashspurt: An ex-partner who wants to get back with you but pretends just to want to be friends, then keeps making passes and getting cross.
“What about ‘pash-hurts’?” said Shaz sulkily. “Friends who turn your own private emotional disaster into a sociological study at the expense of your feelings” (Fielding, 2014, pp. 191–192).
In this context, pashmina is humorously repurposed from its original meaning – a type of luxurious scarf – to symbolise someone (like Shazzer) who is romantically invested in another person (Simon) but is not receiving the same emotional attention. This word sense sets the stage for further playful language games, exploring social and emotional relationship dynamics in a light-hearted, tongue-in-cheek manner.
The HT’s choice of мохнатки/moxnatky (Fielding, 2020, p. 234) “something fuzzy or fluffy” for pashminas is a deliberate visual metaphor for pubic hair. It cleverly embeds an extra layer of meaning, enhancing the sexualised subtext while maintaining humour and informality. HT follows the stylistic choice made by the author for pashmina-ism and provides the suffixed derivative мохнатизм/moxnatysm (Ibid.), retaining the metaphoric connotation of the root noun. Combining мох/mox (a clipping from мохнатки/moxnatky) with велитель/velytel’ (a clipping from повелитель/povelytel’ “lord”), HT produces a compound noun мохвелитель/moxvelytel’ (Ibid.) to render pashmaster, capturing the dominance implied by master and maintaining humour. The translation equivalent мохпедик/moxpedyk (Ibid.) for pashmincer uses мох/moh combined with педик/pedyk (slang for a “gay man”), aligning with the humorous and slightly edgy tone of pashmincer (a friend one fancies who is gay). The blend of мох/mox with жонатик/žonatyk (a colloquialism of “married man”) (Ibid.) suggested as a translation for pashmarried preserves the idea of an ex-partner who is now married but keeps the speaker as a nostalgic memory. The translation of ex-pashspurt as ексмох’якулятор/eksmox’jakyljator (Ibid., p. 235) uses екс-/eks- “ex-” and мох’якулятор/mox’jakyljator, a playful creation of мох/mox + якулятор/jakyljator (еякулят/ejakuljat “ejaculate” + ор-/-or “or”) that adds a sexual undertone to the TT reflecting the ex-partner’s advances while pretending to want friendship. Combining мох/mox with кривдник/kryvdnyk “offender” (Ibid., p. 234) to reproduce pash-hurt, the HT keeps the sense of someone turning emotional pain into an impersonal case study, aligning with the original meaning.
Unlike the HT’s creative and contextually rich rendering of these SNs, the MT engines fail to capture their nuanced meaning, humour, and inventiveness. The errors mainly stem from the machines’ inability to recognise the semantically surprising combinations of the original. Firstly, both engines transliterate pashminas as пашміни/pashminy, failing to capture the metaphorical humour of the original. This literal approach strips the term of its playful nuance and leads to other meaningless transliterations, like пашмінаїзм/pashminajizm (DL) and пашмінер/pashminer (GT), which retain some phonetic resemblance but lose the playful blend of the constituents. In some cases, GT resorts to omission, for instance, removing the element pash- from pashmarried and reproducing it as одружений/odryženyi “married”. DL reproduces pashmarried as паш-одружений/pash-odryženyi, preserving the clipped “pashmina” but failing to integrate it creatively. Overall, the machine engines primarily transliterated/transferred or partially translated original SNs into Ukrainian, losing the humour and layered meaning. They failed to adapt suffixes, prefixes, and wordplay, which are vital to preserving the tone.
One difficulty that MT faced with compounds is the reproduction of those formed by reduplication. One example is eensy-beansy in (18):
(18) …kids, who would invariably appear in the alumni magazine clad in eensy-beansy black-and-orange outfits … (Weiner, 2002, p. 34).
The TT by GT is nonsensical. The engine mistranslates eensy-beansy outfits as вбрання з квасолі/vbrannja z kvasoli “bean dress,” failing to recognise that eensy-beansy is a playful expression for something tiny formed in performance by analogy with teensy weensy. This mistranslation distorts the meaning and loses the whimsical tone, rendering the phrase flat and conventional. Compounds formed by reduplication are not rare in Ukrainian (Nikolashyna, 2017), yet machine engines algorithms fail to recognise them.
The translation of prefixed derivatives by machine engines highlighted a profound misinterpretation of the SL. A notable example is (19):
(19) Shazzer suddenly detranced (Fielding, 2014, p. 160).
GT’s translation зійшла з розуму/zijshla z rozumu “went mad” is a significant error, conflating detrance (“to come out of a trance”) with mental instability. DL’s output відсторонилася/vidstoronylasja “withdrew” is somewhat closer but still misses the subtlety of the character emerging from a trance-like state.
Suffixed derivatives are equally problematic. One example is tentery in (20):
(20) I have this secret little vocabulary for my husband. Words I’ve invented, just to describe him. … Like … Tentery: that taut, defensive way he behaves whenever the subject of my father comes up in conversation (Kinsella, 2018, p. 1).
In Kinsella (2018), the protagonist is raised in privilege thanks to her father, while her husband comes from a modest background. Unable to match his father-in-law’s support, he develops a strained relationship with him, prompting his wife to describe his emotional state as “tentery” whenever her father is mentioned. GT’s тентері/tenteri is a transference that lacks interpretation or contextual adaptation. DL’s натягнутість/natjagnutyst’ “tenseness” more accurately conveys the sense of being taut and defensive. Still, it fails to retain the grammatical form of the original – rendering the adjective as a noun.
Many MTs struggle with playful neologisms and are unable to replicate the humour or stylistic ingenuity evident in the ST.
To the best of my knowledge, this article is among the first to explore the issue of creativity in MT when applied to stylistic neologisms in ELCL. As a pilot study, it supports the view that MT systems, notably DL and GT, still struggle with the creative demands of literary texts, particularly when tasked with coining new, morphologically unconventional words for stylistic effect. In fact, the findings suggest that creativity remains a deeply human quality that is difficult for MT systems to replicate, with error rates exceeding 80%. Notably, this study also reveals that HTs working with the English–Ukrainian translation of SNs employ a wider range of techniques than those reported for high-resourced language pairs, such as English–German or English–Spanish (imitation of clippings, use of cognate metaphors, capturing of onomatopoeic effect absent in the source). Although MT handled suffixations reasonably well in some cases, HT consistently outperformed MT, especially when neologisms involved cultural or metaphorical nuances. While DL and GT demonstrated potential in recognising some morphological patterns, they generally fell short of reproducing the intended creativity, indicating that literary translation still requires significant human input for post-editing or retranslation. Furthermore, this study reflects the author’s dual perspective as both a researcher and practitioner, warning that overreliance on MT could restrict the creative autonomy of literary translators. However, given the exploratory nature of this pilot study and its several limitations, including a small sample size and some degree of subjectivity in assessing translation quality, the findings should not be readily generalised. To overcome this limitation, future research should involve broader datasets that incorporate more problematic cases and include human evaluation by multiple professional translators. Overall, this study contributes to ongoing discussions about the intersection of technology and creativity in translation, highlighting the need to carefully integrate MT tools into professional practice to avoid potential misuse or ethical concerns.
Fielding, H., 1996. Bridget Jones’ Diary. Available at: <http://www.kkoworld.com/kitablar/Helen%20Fielding%20-%20Bridget%20Jones%27s%20Diary.pdf>. [Accessed 2 September 2024].
Fielding, H., 2014. Bridget Jones. The Edge of Reason. London: Picador Books.
Fielding, H., 2020. Bridžyt Džons: Na Meži Zdorovogo Gluzdu [Bridget Jones: The Edge of Reason]. Translated by O. Mel’nyk. Lviv: Staryj Lev publishing. [In Ukrainian].
Fielding, H., 2019. Shchëdennyk Bridžyt Džons [Bridget Jones’ Diary]. Translated by L. Lucan. Lviv: Staryj Lev publishing. [In Ukrainian].
Kinsella, S., 2001. Confessions of a Shopaholic. [online] Available at: <https://allnovel.net/confessions-of-a-shopaholic-shopaholic-1/>. [Accessed 2 September 2024].
Kinsella, S., 2020. Love your Life. London: Penguin.
Kinsella, S., 2023. Lyuby Svoje Žyuttya [Love your Life]. Translated by E. Baskakova. Kiyv: KM Buks PG. [In Ukrainian].
Kinsella, S. 2017a. Moe ne Nadto Doskonale Žyttja [My not so Perfect Life]. Translated by K. Bulkina. Kiyv: KM Buks PG. [In Ukrainian].
Kinsella, S., 2017b. My not so Perfect Life. London: Black Swan.
Kinsella, S., 2004. Shopaholic Takes Manhattan. London: Dell.
Kinsella, S., 2009. Shopogolik [The Secret Dreamworld of a Shopaholic]. Translated by N. Tretjakova. Kharkiv: Klub simejnogo dozvillja. [In Ukrainian].
Kinsella, S., 2019. Šopogolik na Mangetteni [Shopaholic Abroad]. Translated by N. Tretjakova. Kharkiv: Klub simejnogo dozvillja. [In Ukrainian].
Kinsella, S., 2018. Surprise me. London: Bantam Press.
Weiner, J., 2002. Good in Bed. Washington: Washington Square Press.
Booij, G., 2005. Compounding and Derivation. Evidence for Construction Morphology. In: Morphology and its demarcations. Amsterdam / Philadelphia: John Benjamins, pp. 109–132. https://doi.org/10.1075/cilt.264.08boo.
Cañete-González, P., Llopart-Saumell, E., 2023. Are Stylistic Neologisms More Neological? A Corpus-Based Study of Lexical Innovations of Women and Men. Languages, 8 (3), pp. 175–187. https://doi.org/10.3390/languages8030175. Available at: <https://www.mdpi.com/2226-471X/8/3/175>. [Accessed 2 September 2024].
Corpas Pastor, G., Noriega-Santiáñez, L., 2024. Human vs. Neural Machine Translation Creativity: A study on Manipulated MWEs in Literature. Information, 15, pp. 530–548. https://doi.org/10.3390/info15090530. Available at: <https://www.mdpi.com/2078-2489/15/9/530>. [Accessed 2 September 2024].
CULT – Corpus of Ukrainian Language Texts. Available at: <http://www.mova.info/corpus.aspx?l1=209>. [Accessed 2 September 2024].
Delabastita, D., 2004. Literary Style in Translation: Archaisms and Neologisms. In: Translation: An International Encyclopedia of Translation Studies, 1. Berlin: De Gruyter, pp. 883–888. Available at: <https://doi.org/10.1515/9783110137088.1.12.883. https://scholarlypublications.universiteitleiden.nl/handle/1887/3704727>. [Accessed 2 September 2024].
Dorst, A. G., Rothwell, A., Way, A., Youdale, R., 2024. Metaphor in Literary Machine translation: Style, Creativity and Literariness. In: Computer-Assisted Literary Translation. N.Y. / London: Routledge, pp. 173–186. https://doi.org/10.4324/9781003357391-12. Available at: <https://scholarlypublications.universiteitleiden.nl/access/item:3704728/view>. [Accessed 2 September 2024].
Harzewski, S., 2011. Chick Lit and Postfeminism. Charlottesville: UVP.
Hohenhaus, P., 2007. How to Do (Even More) Things with Nonce Words (Other than Naming). In: Lexical Creativity, Texts and Contexts. Amsterdam, Philadelphia, PA: John Benjamins, pp. 15–38. https://doi.org/10.1075/sfsl.58.08hoh.
House, J., 1997. Translation Quality Assessment: A Model Revisited. Tübingen: Gunter Narr.
Kenny, D., 2001. Lexis and Creativity in Translation: A Corpus-Based Study. Manchester: St Jerome.
Kolb, W., Dressler, W. U., Mattiello, E., 2023. Human and machine translation of occasionalism in literary texts. Johann Nestroy’s Der Talisman and its English translations. Target, 35 (4), pp. 540–572. https://doi.org/10.1075/target.21147.kol.
Lakoff, G., Johnson, M., 1980. The Metaphorical Structure of the Human Conceptual System. Cognitive Science, 4, pp. 195–208. Available at: <https://opessoa.fflch.usp.br/sites/opessoa.fflch.usp.br/files/Lakoff-Johnson-Metaphorical-Structure.pdf>. [Accessed 2 September 2024].
LDOCE – Longman Dictionary of Contemporary English. Available at: <https://www.ldoceonline.com/dictionary>. [Accessed 2 September 2024].
Lensch, A., 2022. Diggers-out, leaf clearer-uppers and stayer-onner-for-nowers. On creativity and extravagance in English -er nominalisations. In: Extravagant Morphology: Studies in Rule-bending, Pattern-extending and Theory-challenging Morphology. Amsterdam / Philadelphia: John Benjamins, pp. 73–100.
Littlemore, J., Tagg, C., 2018. Metonymy and text messaging: A framework for understanding creative uses of metonymy. Applied Linguistics, 39 (4), pp. 481–507. https://doi.org/10.1093/applin/amw018. Available at: <https://pure-oai.bham.ac.uk/ws/portalfiles/portal/27178399/Littlemore_Metonymy_text_messaging_Applied_Linguistics_Accepted.pdf>. [Accessed 2 September 2024].
Mattiello, E., 2017. Analogy in Word-formation: A Study of English Neologisms and Occasionalisms. Berlin / Boston: De Gruyter.
Newmark, P., 2008. A Textbook of Translation. New York: Prentice Hall.
Nikolashyna, T., 2017. Reduplikacija v suchasnij ukrajinskij movi: struktura, semantika, funkciji [Reduplication in the modern Ukrainian: Structure, Semantics, Functions]. Studia Ucrainica Varsoviensia, 5, pp. 87–99. [In Ukrainian].
Noriega-Santiáñez, L., Corpas Pastor, G., 2023. Machine vs Human Translation of Formal Neologisms in Literature: Exploring E-tools and Creativity in Students. Revista Tradumàtica. Tecnologies de la Traducció, 21, pp. 233–264. https://doi.org/10.5565/rev/tradumatica.338. Available at: <https://revistes.uab.cat/tradumatica/article/view/n21-noriega-santianez-corpas-pastor/338-pdf-en>. [Accessed 2 September 2024].
OED – Oxford English Dictionary. Available at: <Oxford English Dictionary>. [Accessed 2 September 2024].
Rusanovskyi, V. M., 2017. Slovnyk Ukrayinskoyi Movy v 11 tomakh [Dictionary of the Ukrainian Language in 11 Volumes]. Kyiv: Naukova Dumka. [In Ukrainian].
Salich, H., 2020. Authorial Neologisms in Translation: Is the Translator a Smuggler? Między Oryginałem a Przeklam, 47, pp. 59–78. https://doi.org/10.12797/MOaP.26.2020.47.03. Available at: <https://journals.akademicka.pl/moap/article/view/1204/1081>. [Accessed 2 September 2024].
Seligej, P., 2014. Etymologichnyj Slovnyk Zapozychenyx Syfiksiv i Sufiksoijdiv v Ukraijnskij Movi [Etymological Dictionary of Borrowed Suffixes and Suffixoids in the Ukrainian Language]. Kyiv: Akademperiodika. [In Ukrainian].
Specia, L., Wilkes, Y., 2022. Machine Translation. In: The Oxford Handbook of Computational Linguistics. Oxford: Oxford University Press, pp. 817–870.