20.500.12556/RUL-188675 Slovenščine tujejezičnih govorcev 3.0 Slovene languages of foreign speakers 3.0 V prispevku nadgrajujem raziskavo Pirih Svetina in Ferbežar (2005) o sprejemljivosti slovenščine kot neprvega jezika med domačimi govorci. Raziskavo, izvedeno leta 2025, sem prilagodila značilnostim digitalne dobe in izvedla anketo, v kateri je sodelovalo 99 domačih govorcev. Ocenjevali so pet besedil na isto temo – štiri besedila študentov, govorcev slovenščine kot neprvega jezika, od katerih je bilo eno prevedeno v slovenščino s pomočjo strojnega prevajalnika, in eno besedilo, ki ga je tvorila generativna umetna inteligenca. Rezultati kažejo, da so domači govorci do vseh besedil kritični, a tolerantni. Med jezikovno motečimi elementi so izpostavili neustrezni besedni red, končnice in besedišče. V splošnem so kot bolj razumljiva, zanimiva, osebna in pravilna ocenili besedila z manj napakami, zato sta najvišje ocene v vseh kategorijah dobili besedilo umetne inteligence in strojni prevod. Drugačen odziv na besedilo umetne inteligence je mogoče zaznati le v občutno manjšem številu neobveznih splošnih komentarjev, ki jih je dobilo to besedilo v primerjavi z ostalimi. Poleg tega so komentarji na ostala besedila napisani s precejšnjo stopnjo empatije do tvorcev, medtem ko pri besedilu umetne inteligence anketiranci takšne empatije niso pokazali. This article builds on the research by Pirih Svetina and Ferbežar (2005) on the acceptability of Slovene as a non-native language among native speakers. The present study, conducted in 2025, adapted the aforementioned study to the characteristics of the digital age. A survey was conducted in which 99 native speakers assessed five texts on the same topic. Four of these were written by students for whom Slovene is a non-native language, one of which was translated into Slovene using machine translationsoftware. The fifth text was generated by artificial intelligence. The findings indicate that native speakers adopt a critical stance while also exhibiting a high degree of tolerance towards all texts. The linguistically disturbing elements that were highlighted by the respondents included inappropriate word order, endings, and vocabulary. Texts with fewer errors were generally perceived as more understandable, interesting, personal, and correct. Consequently, the highest ratings in all categories were attributed to the texts generated by artificial intelligence and the machine translation. The sole discernible discrepancy in the response to the AI text was the conspicuously lower number of optional general comments it received in comparison to the other texts. The comments on the other texts also demonstrate a significant degree of empathy towards the producers, while the respondents exhibited a notable absence of such empathy towards the AI text. slovenščina kot neprvi jezik sprejemljivost besedil umetna inteligenca strojno prevajanje jezikovne napake Slovene as non-native language text acceptibility artificial inteligence machine translation language errors true false true Slovenski jezik Angleški jezik Licenca Creative Commons je navedena na pristajalni strani članka (glej izvorni URL): "To delo je licencirano pod Creative Commons Priznanje avtorstva-Deljenje pod enakimi pogoji 4.0 mednarodno licenco." (Datum opombe: 25. 9. 2026 ) Članek v reviji 2026-09-25 09:39:30 2026-09-25 09:39:36 2026-09-26 04:31:33 0000-00-00 00:00:00 2026 0 0 Str. 121-148 št. 1/2 Letn. 71 2026 0000-00-00 Zaloznikova Objavljeno NiDoloceno 0000-00-00 0000-00-00 2026-07-23 811.163.6 0021-6933 10.4312/jis.71.1-2.121-148 289272579 Stritar+Kucuk.pdf Stritar+Kucuk.pdf 1 3CBE8EDE4C942FC9211143B4A4FF7121 67460c704910e3aee52de889663c5085af5212e0b170fdb765b8c3ce5348e672 432c462b-b8b5-11f1-8bc5-0050569b8976 https://repozitorij.uni-lj.si/Dokument.php?lang=slv&id=248980 https://journals.uni-lj.si/jezikinslovstvo/article/view/24195 1 3040bf39-b8b4-11f1-8bc5-0050569b8976 https://repozitorij.uni-lj.si/Dokument.php?lang=slv&id=248979 Filozofska fakulteta 0 0 0