The Linguist

28 The Linguist Vol/64 No/2 ciol.org.uk/thelinguist REVIEWS pretending to be someone else online) was translated as γατόψαρο, as in the species of fish. The software is even less consistent when it comes to the translation of acronyms. In most cases they are left in English. Instead of being rendered correctly as ΚΑΡΠΑ, CPR (cardiopulmonary resuscitation) is left in English or translated as KPP, an acronym that has no meaning in Greek. Similarly, the AI tools do not consistently transliterate the names of people and places. The same tool will leave the name Cheryl in English at times and transliterate it as Σέρυ at others. I hypothesise that other non-Latin languages, such as Arabic and Macedonian, could face similar issues. Do the tools help? Among my tentative conclusions from this testing is that, despite the noted issues, AI- assisted technologies can greatly facilitate the translation of subtitles, especially when dealing with large volumes of media content. I found that when it comes to synchronising the subtitles with the video, the results are astonishingly accurate. I can already see how AI could aid the laborious task of manual synchronisation, significantly reducing the amount of time it takes. Which tool to choose will depend on your field of work. For TV and film, tools like Limecraft, Happy Scribe and Stellar, which are designed for this type of output, may be the best option. At the same time, when it comes to content, the intervention of professional linguists is paramount. This is especially true when it comes to cultural references and slang, which require cultural knowledge and a sensitive approach to language. The sharp increase of content in the media industry, combined with the extremely demanding timeframes for subtitle production, requires us to re-view, re-visit, re-think and re-design the current localisation production models with the help of AI. However, to achieve efficiency, consistency and quality, it is essential that human expertise remains at the heart of the process of subtitle creation. Panayiota Vatikioti MCIL is a certified translator, localisation consultant, blogger and transcreator; www.pvtranslations.com. Readers of The Linguist do not need to be persuaded of the importance and benefits of learning new languages – it is our passion and our livelihood. In recent years, the field of language learning has seen a seismic shift fuelled by digital technology, with mobile apps emerging as a popular tool. For professional linguists, language teachers and translators, this phenomenon is a source of both opportunities and questions: How effective are these tools beyond the beginner stage? What pedagogical models do they adopt? And where do they succeed or fall short compared to traditional instruction? As the global demand for multilingualism continues to grow – driven by international mobility, migration, digital communication and the multilingual workplace – language learning apps are increasingly being used not only by casual learners but also by students, teachers and professionals seeking supplementary or flexible practice. While no app can replace classroom-based learning or linguistic immersion in a country where the language is spoken, many offer structured, adaptive and increasingly sophisticated systems based on principles from second language acquisition (SLA) research. As a language teacher (German, French and Russian), my students often ask me if there is a language learning app that I would recommend, so I decided to test some of the most popular platforms, starting with Duolingo, Memrise and Babbel. I focused on their pedagogical structure, linguistic scope, pricing models and distinctive features. Duolingo Duolingo was founded in 2011 by Luis von Ahn and Severin Hacker at Carnegie Mellon University. The purpose of the project was to provide an accessible means to learn languages. Initially, the founders considered a non-profit organisation model, but it was not sustainable. Originally conceived as a dual-purpose platform for language learning and a digital text translation service, the US- based app quickly pivoted to focus solely on education. It has since grown into one of the most downloaded education apps worldwide, and in the first quarter of 2025, it had 130 million active users. Duolingo uses short, interactive lessons that blend translation, matching games, quizzes, and listening and speaking exercises. These are presented in a colourful, gamified Battle of the apps gxv{ÇÉÄÉzç Anna Rioland tries out three of the most Duolingo 43 languages, Free (with ads); premium £4.99 pcm/£48 pa Memrise 23+ languages, Free (limited content); premium from £4.50 pcm/ £79.99 pa; lifetime subscription £254.99 Babbel 14 languages, From £48 for 3 months/£96 pa

• Cover