The translation process of Brazilian Portuguese-English false cognates

Abstract

The present study is part of the project: “Portuguese-English bilingual lexicon” by the Bilingual and Multilingual Language Processing Laboratory (PLIBIMULT), at the Federal University of Ceará. This study has as its main objective to examine the effect of false cognates in Brazilian Portuguese-English bilinguals, in a language decision task and a multiple-choice translation task. The general objective was divided into two specific objectives: 1) to analyze whether there is any significant difference in the processing cost between false cognates and control words in Brazilian Portuguese (L1) and in English (L2) in a language decision task and in a translation task; 2) to investigate if there are false cognate repetition priming effects in a translation task. The studies carried out by Otwinowska and Szewczyk (2019), Otwinowska, Forys-Nogala, Kobosko and Szewczyk (2020), Brenders, Van Hell and Dijkstra (2011), Lalor and Kirsner (2001), Poort, Warren and Rodd (2015), Gadelha (2021) and Borém (2023) provided the main theoretical support for the present study. For the language decision task (task 1), 25 English false cognates, 25 Brazilian Portuguese false cognates, 100 English control words and 100 Brazilian Portuguese control words were selected. For the translation task (task 2), 50 English words were selected in total to be translated. Out of these 50 words, 25 were English false cognate words and 25 were English control words. 12 English false cognates that were used in task 1 were repeated in task 2. Reaction time and accuracy data for both tasks were collected using the software PsyToolkit (Stoet, 2010, 2017). All lexical items used in this study were controlled regarding frequency and number of characters. A demographic and linguistic questionnaire and an L2 vocabulary knowledge test were used as additional research instruments in this study. For tasks 1 and 2, accuracy was lower for false cognates compared to control words and reaction time was higher for false cognates compared to control words, highlighting that for false cognates the processing cost was higher, Therefore, H1 was confirmed. Regarding the second specific objective of this research, in task 2, results showed that there were no repetition priming effects for false cognate words in a translation task, therefore H2 was confirmed. The findings for this research may contribute to the discussion regarding lexical access and they may also foment translation studies concerning the cognitive processes underlying translation. Furthermore, this research can contribute to studies on false cognates and how differently this type of word is processed by bilinguals, as well as providing effective strategiesfor language learners when accessing lexical items in different languages.