Human Translation Research Articles

An Assessment of the Quality of Post-Edited Text From CAT Tools Compared to Conventional Human Translation: An Error Analysis Study

An Assessment of the Quality of Post-Edited Text From CAT Tools Compared to Conventional Human Translation: An Error Analysis Study

POETIC TRANSLATION AND STUDENTS’ APPRECIATION THROUGH HUMAN TRANSLATOR AND MACHINE SYSTEM

The engagement with poetry is a personalized journey that transcends standardized methodologies. Key to this process is complete immersion in the poetic experience, alongside evaluators’ openness to both human and machine-generated translations from Indonesian to English. The overarching goal is to enhance students’ discernment and appreciation of these translated works. The study’s specific objectives involved comparing poetic translation assessments by evaluators for both human translators and machine systems. It was to assess students’ appreciation of poetry through the lens of both translation methods across three different institutions. The research employed a mixed-methods approach, combining quantitative and qualitative descriptive analyses, including descriptive statistics. The dataset comprised two Indonesian poems by Taufik Ismail, rated using a score as proposed by Nababan. Findings indicate that human translators outperformed machine systems in terms of accuracy, acceptance, and readability. While students from Institutions 1 and 2 preferred human-translated poetry, students from Institution 3 favoured machine-translated versions. This suggests that human translation quality remains superior.

A Study Comparing the Advertising Language Ability of ChatGPT-4o and Kimi in the Translation Beauty Industry

This paper explores the ability of ChatGPT-4o and Kimi to translate advertising slogans for different beauty brands. Both AI tools are translated with the ICIO framework and the same background knowledge supplement. Through the comparison of the results of translation, research find out the similarities and differences between the two types of AI translation. Translation can be judged in terms of coherence, the communication of its emotion, and whether it reflects the deep. And the combination of the advantages of human translation and machine translation will provide more accurate and insightful translations in the future. By setting up questionnaires and interviews, the research can explore the similarities and differences between the two AI tools to translate specific text with more “humanized” characteristics in the future and improve the translation level of text translation. At the same time, this study still has some limitations in the selection and classification of advertising languages, and future studies can further explore the methods of translating domain-specific texts and promote the improvement of translation quality.

Exploring challenges in audiovisual translation: A comparative analysis of human- and AI-generated Arabic subtitles in Birdman.

Movies often use allusions to add depth, create connections, and enrich the storytelling. However, translators may face challenges when subtitling movie allusions, as they must render both meaning and culture accurately despite existing language and cultural barriers. These challenges could be further complicated by the use of available AI tools attempting to subtitle movie allusions, while probably unaware of existing cultural complexities. This research investigates these challenges using qualitative and descriptive quantitative approaches by analyzing the movie Birdman or (The Unexpected Virtue of Ignorance), comprising13.014 words, to identify the types of allusions used and compare the human- vs. AI (ChatGPT)-generated Arabic subtitles in terms of the subtitling strategies, their frequency, and quality. The results revealed that the movie used 52 Noun Phrase (NP) allusions, where the writer intertextually employed a proper name to convey meaning, and 8 Key-Phrase (KP) allusions, where the writer used phrases that convey implicit meaning easily perceived by members of the source culture (by referring to religious, literary, or entertainment texts). For NP allusions, both the human translator and AI opted for retentive strategies; however, the human translator's preference to add guidance/parentheses to mark NP allusions was distinct. Additionally, it was observed that AI used neologism to render technology-related allusions, which could be a suggested strategy for NP subtitling into Arabic. For KP allusions, while the human translator seemed to be cognizant of the idea that KP allusions typically require a change in wording, AI fell short. Specifically, the human translator employed reduction in 5 out of 8 KPs, opting for minimum change/literal translation only three times. Conversely, AI utilized literal translation in all 8 examples, despite its awareness of the allusion and its intricate meaning/reference. As for the FAR assessment, for NP allusions, it revealed minor semantic errors in AI's subtitles that did not affect the plot. Regarding KP allusions, AI's subtitles were penalized in 5 out of its 8 Arabic renditions, in contrast to the human translator. Most of the errors were serious semantic errors that likely disrupted the flow of reading the subtitles due to conveying irrelevant meanings in the movie's/scene's context. Despite its functionality, this study suggests adding an extra parameter to the FAR model: consistency, as it plays a role in enhancing audience involvement and understanding. Its absence, as observed in some AI instances, can be misleading.

Preparing Translators for the Age of Generative AI: Revising Educational Curricula to Prioritize Essential Skills in an AI-Dominated Landscape

The recent ability of machines to generate text and images that are fluent, coherent, and culturally nuanced is already causing major upheavals in the translation sector. Machine translation tools have exploded in number, sophistication, and quality, and the profession of human translator needs to adapt to work within this new environment and avoid being replaced by it. This paper examines the major changes that have taken place, and which could take place in the near future, and suggests ways of revising translator-training curricula to adapt to these challenges. This is essential at a time when the profession is undergoing profound changes and when all translators and future translators need to be prepared for new skills in the field of post-editing, machine-reading, machine-cultures and multidisciplinarity. After outlining the challenges that AI is causing to the profession and to the norms and values of the translation sector in its current state, the author suggests that translator training should adapt to the new reality of the profession and not to what it used to be. The profession is undergoing profound changes, and all translators and future translators need to find new skills and knowledge in order to continue to work with AI. The article concludes with recommendations on how to design AI-friendly translation programmes that can train students for the post-AI era.

« Vous pouvez entendre les mâchoires de l’arme de l’ange ». Le traducteur humain face à la traduction automatique dans le contexte littéraire : Les Livres de Jakób d’Olga Tokarczuk

The dynamic development of automatic translators raises questions about the future of human translators. The aim of this article is to evaluate the French versions of an excerpt from The Books of Jacob by Olga Tokarczuk, made by Systran, Google Translate and DeepL, and to compare them with the reference translation made by a human translator. The translations were carried out in 2019 as part of the ATLAS association project ‘L’Observatoire de la traduction automatique’. The analyses show that the inability to interpret the original in a complex way leads to weaknesses in machine translators. This proves that the position of the human translator remains unthreatened in the field of translating high-quality literary texts.

Das Übersetzten zwischen Mensch und Maschine

Abstract Artificial intelligence is changing our world and will change it ever more profoundly in the future. As a result, numerous areas of work are also changing, professions and jobs are disappearing, while others must change and adapt to survive. These transformations also affect the translation market and the work of translators. Machine translation (MT) transfers entire texts from one language to another in a matter of seconds, enabling communication on a scale and at a speed that could never have been dreamed of in the past. However, these MT systems have been around for over 70 years and are still not able to deliver the perfect translation or make human translators superfluous. So how should aspiring translators deal with these technologies, how should they use them to make their work more efficient and therefore more profitable? These are questions that we, as trainers of these future translators, should address in our teaching in order to prepare our students for the labour market of the 21st century. This article wishes to discuss some theoretical aspects of MT to address some concrete elements that I believe could and should be introduced into translation courses and that are intended to stimulate discussion and reflection.

Translating classical Arabic verse: human translation vs. AI large language models (Gemini and ChatGPT)

Translating classical Arabic verse: human translation vs. AI large language models (Gemini and ChatGPT)

What Can We Learn from the 21st Century Translator Pen? A Case Analysis on Comparing English-to-Chinese Rendition Differences between Human and Machine Translation

This study investigated and compared differences in rendering from the same English source text into two versions of target text in Chinese produced respectively by human and machine translator represented by translator pen. Using news report as case analysis, this study found that improper segmentation of punctuation marks appeared most frequently in machine translation, followed by lexical vacancy, and by inconsistency of terms. This study also identified rendition differences between human and machine translation in the handling of terms as well as in the treatment of punctuation marks. Overall, the human translator showed more flexibility in the selection of words to match target text expression than the machine translator.

The Future of the Translation Profession in the Era of Artificial Intelligence. Survey Results from Polish Translators, Translation Trainers, and Students of Translation

Technological advancements in computer science have led to the progressive change of the translation profession. In particular, this concerns machine translation, which has fundamentally changed the role of the translator. AI-driven tools, such as neural machine translation systems, have significantly improved translation speed and accuracy, with the translator role resembling that of an editor or proofreader. While translators have viewed the technological progress as a threat to their jobs, machine translation comes with some persistent issues, such as bias or inaccuracies stemming from insufficient context understanding and cultural nuances. Accordingly, the question of the future of human translation is more complex than it may initially appear. This paper presents the results of a pilot study on the perception of the future of the translation profession in the era of artificial intelligence.

Translating Four-Character Structures in Chinese Literary Works on Traditional Architecture—A Case Analysis of Canal Towns South of the Yangtze and Folk Houses South of the Yangtze

Traditional Chinese architecture boasts a unique style and profound cultural connotations. Accordingly, the related literary works attain value for international dissemination. The four-character structures in these works serve as concise expressions for describing architectural details and design concepts. However, due to their linguistic implicitness, grammatical structure diversity, and rich cultural images, their translation into English requires multidimensional shift, redundancy elimination, and the avoidance of cultural defaults. This research has developed an effective translation workflow that leverages the synergistic strengths of Large Language Models (LLMs) and the human translator. By strategically employing tailored methods and techniques, it addresses the challenges in the translation process. Analysis suggests: It is recommended that the technique of shift be employed to address the implicit content and complex grammatical structures in translating four-character structures in Chinese literary works on traditional architecture into English. Besides, the selective application of combination and omission can help deal with information redundancy. Furthermore, it is advisable to apply foreignization and a combined method of literal translation and paraphrasing in conveying cultural images. These approaches will, to a certain extent, convert the untranslatability of four-character structures into translatability, enhancing the dissemination of the essence of traditional Chinese architecture.

Cultural Discourse in African Poetry: Output by Human Translators and Machine Translation Systems

Cultural Discourse in African Poetry: Output by Human Translators and Machine Translation Systems

Placental Nanoparticle-mediated IGF1 Gene Therapy Corrects Fetal Growth Restriction in a Guinea Pig Model.

Fetal growth restriction (FGR) caused by placental insufficiency is a major contributor to neonatal morbidity and mortality. There is currently no in utero treatment for placental insufficiency or FGR. The placenta serves as the vital communication, supply, exchange, and defense organ for the developing fetus and offers an excellent opportunity for therapeutic interventions. Here we show efficacy of repeated treatments of trophoblast-specific human insulin-like 1 growth factor (IGF1) gene therapy delivered in a non-viral, polymer nanoparticle to the placenta for the treatment of FGR. Using a guinea pig maternal nutrient restriction model (70% food intake) of FGR, nanoparticle-mediated IGF1 treatment was delivered to the placenta via ultrasound guidance across the second half of pregnancy, after establishment of FGR. This treatment resulted in correction of fetal weight in MNR + IGF1 animals compared to sham treated controls on an ad libitum diet, increased fetal blood glucose and decreased fetal blood cortisol levels compared to sham treated MNR, and showed no negative maternal side-effects. Overall, we show a therapy capable of positively impacting the entire pregnancy environment: maternal, placental, and fetal. This combined with our previous studies using this therapy at mid pregnancy in the guinea pig and in two different mouse model and three different human in vitro/ex vivo models, demonstrate the plausibility of this therapy for future human translation. Our overall goal is to improve health outcomes of neonates and decrease numerous morbidities associated with the developmental origins of disease.

Foreign Language Student Linguistic Personality as a Vector for the Professional Bilingualism

The relevance of the study manifests itself through the use of a polyparadigmatic approach to the consideration of the linguistic personality of a translator, including cognitive, cultural, synergetic, and competence aspects. The cognitive aspect is based on the study of the linguistic consciousness of the individual, the synergetic aspect is based on achieving a harmonious translation worldview, the cultural aspect is based on the need for an interconnected study of languages and cultures, and the competence aspect is based on the formation of personal intercultural competence. The purpose of the study is to substantiate a polyparadigmatic approach to the study of the linguistic personality of a foreign student – a future translator. As a result of the study, it was established that in the digital age, with artificial intelligence constantly evolving, new trends arise in the professional bilingualism formation of a human translator interacting with artificial intelligence. This study is based on the results of accumulated pedagogical experience and theoretical provisions in this area, which unites the educational space of two universities (PNRPU, Russia and KRU named after A. Baitursynuly, Kazakhstan). The results obtained can of interest for teachers engaged in training bilingual students, linguistic program students, for the tertiary education practitioners as well as those involved into decision making and curriculum development.

Comparative Analysis: Machine vs Human Translation of Educational Scientific Texts. Challenges and Solutions

Introduction. The purpose of this article is to analyze the features of identity formation and identification processes in postmodern society and in digital society. The article explores various approaches to the concept of «identity» within the framework of socio-humanitarian theories. The interrelation of real and virtual identity, the emergence of new forms and models of identity are considered. Two approaches to the analysis of identity in a digital society are analyzed and compared: digital optimism and digital pessimism.Methodology and sources. The article uses a comparative analysis of various theoretical approaches to identity formation. An interdisciplinary approach is used to understand the features of identity in the era of digitalization. As the main sources for the study of the phenomenon of identity there were used works of classics of sociology and psychology E. Erickson, J.G. Meade, E. Durkheim, as well as studies of representatives of modern philosophical and sociological theories Z. Bauman, A. Giddens, S. Turkle, A.A. Lisenkova, T.G. Leshkevich.Results and discussion. It is shown that in postmodern society, identity is characterized by the processes of individualization, fragmentation and the absence of a single cultural center. In a digital society, along with the processes of individualization, there are processes of standardization and unification of person. It is concluded that in a digital society, identity has positive and negative sides. On the one hand, there is a positive vector for the use of online communications, network interaction of people, which destroys spatial and temporal boundaries, as well as social barriers. On the other hand, new types of virtual identity are being formed, challenging a person's own identity.Conclusion. It is concluded that the digital society changes human identity, identification practices and strategies. The consequence of these processes is the formation of a person fused with digital technologies, which leads to the transformation of the social order, social norms and society as a whole.

Cross-Lingual Short-Text Semantic Similarity for Kannada–English Language Pair

Analyzing the semantic similarity of cross-lingual texts is a crucial part of natural language processing (NLP). The computation of semantic similarity is essential for a variety of tasks such as evaluating machine translation systems, quality checking human translation, information retrieval, plagiarism checks, etc. In this paper, we propose a method for measuring the semantic similarity of Kannada–English sentence pairs that uses embedding space alignment, lexical decomposition, word order, and a convolutional neural network. The proposed method achieves a maximum correlation of 83% with human annotations. Experiments on semantic matching and retrieval tasks resulted in promising results in terms of precision and recall.

What are the differences? A comparative study of generative artificial intelligence translation and human translation of scientific texts

Generative artificial intelligence translation (GenAIT) brings convenience yet also imposes severe challenges on the dissemination of knowledge. The respective (dis)advantages of GenAIT and human translation (HT), and the ways to promote their effective interaction have not been sufficiently explored yet. This study investigates the linguistic features of GenAIT and HT of scientific texts rendered from English to Chinese from lexical and syntactic levels. The GenAIT is generated by ChatGPT 3.5, a representative GenAI platform, while HTs are done by 19 Master-of-Translation-and-Interpreting students in China. Data shows that GenAIT and HTs present distinguished linguistic features in both levels. At the lexical level, HT exhibits lengthier texts with a lower average word diversity; GenAIT presents higher accuracy in translating terminology. At the syntactic level, the average sentence count in HT is greater, whereas its average sentence length measured in tokens is shorter. Moreover, human translators tend to transform sentences from passive voice into active voice more frequently than ChatGPT 3.5 does. Furthermore, human translators exhibit superior skills in deconstructing lengthy and complex sentences into shorter, more comprehensible clauses. This study reveals that ChatGPT 3.5 and human translators exhibit complementary capabilities in the translation of scientific texts, thereby suggesting an optimisation of the strengths inherent to both parties. Implications are provided for future translator training, language service providers and the ongoing development of GenAIT and HT.

Universal Basic Income as a Response to Automation? Attitudes of Human Translators Facing Neural Machine Translation

How do professionals who may be at risk of job loss due to AI-driven automation view a universal basic income (UBI) as a policy response to technological unemployment? This article examines the attitudes of translators from Turkey, a country with the highest perceived technological risks across the OECD. Based on interviews, the article reveals varying views on UBI, the framings of which reflect participants’ political views and understandings of capitalist societies. While some participants depict a perspective of economic individualism, emphasizing individual responsibility and a work ethic in opposing UBI, others advocate an anti-capitalist outlook, promoting collective action yet appearing skeptical of UBI. However, a majority supports UBI as a social right. Furthermore, those concerned about job or wage loss are primarily found among those who favor UBI and those who specialize as freelancers in areas perceived as more susceptible to automation, suggesting that job insecurity shapes views of UBI.

Translating Arabic Free Poetry Texts into English by ChatGPT: Success and Failure

The purpose of this study is to examine the success and failure of ChatGPT in translating some Arabic free poetry texts selected from Wahib's 'Hulm 'ala Nasiat Aljabin 'A Dream on the Corners of Nostalgia.’ In this study, we examined ChatGPT’s ability in the translation field. In particular, the focus is on translating in Arabic context. The free poetry texts were first translated by the authors and then by ChatGPT. Both translations were then analyzed and compared to find out to what extent ChatGPT translation deviates from human translation. A qualitative approach is adopted with the help of comparative and analytical methods. The results show that ChatGPT is, to some extent, successful in producing a natural English translation of the Arabic poems in this study. In contrast, humans adopt translation strategies like paraphrasing, adapting, recreating, and transliteration. ChatGPT adopts translation strategies like paraphrasing addition, word-for-word, and literal translation. However, there is no care for the aesthetic values in ChatGPT’s translation, and many types of translation errors result from applying these strategies.

Investigation in yeast of novel variants in mitochondrial aminoacyl-tRNA synthetases WARS2, NARS2, and RARS2 genes associated with mitochondrial diseases.

Aminoacyl-transfer RiboNucleic Acid synthetases (ARSs) are essential enzymes that catalyze the attachment of each amino acid to their cognate tRNAs. Mitochondrial ARSs (mtARSs), which ensure protein synthesis within the mitochondria, are encoded by nuclear genes and imported into the organelle after translation in the cytosol. The extensive use of next generation sequencing (NGS) has resulted in an increasing number of variants in mtARS genes being identified and associated with mitochondrial diseases. The similarities between yeast and human mitochondrial translation machineries make yeast a good model to quickly and efficiently evaluate the effect of variants in mtARS genes. Genetic screening of patients with a clinical suspicion of mitochondrial disorders through a customized gene panel of known disease-genes, including all genes encoding mtARSs, led to the identification of missense variants in WARS2, NARS2 and RARS2. Most of them were classified as Variant of Uncertain Significance. We exploited yeast models to assess the functional consequences of the variants found in these genes encoding mitochondrial tryptophanyl-tRNA, asparaginyl-tRNA, and arginyl-tRNA synthetases, respectively. Mitochondrial phenotypes such as oxidative growth, oxygen consumption rate, Cox2 steady-state level and mitochondrial protein synthesis were analyzed in yeast strains deleted in MSW1, SLM5, and MSR1 (the yeast orthologues of WARS2, NARS2 and RARS2, respectively), and expressing the wild type or the mutant alleles. Pathogenicity was confirmed for most variants, leading to their reclassification as Likely Pathogenic. Moreover, the beneficial effects observed after asparagine and arginine supplementation in the growth medium suggest them as a potential therapeutic approach.