Sequential Use of <scp>L1</scp> and <scp>L2</scp> Captions: Exploring the Benefits for Vocabulary Acquisition

There has been a proliferation in recent years of studies on second language (L2) vocabulary acquisition from watching audiovisual materials, such as TV programs and movies. A recurring theme this line inquiry the extent to which captioning facilitates (Montero Perez, 2022; Teng, 2021). Studies have examined compared benefits captions learners' first (L1 captions, often referred literature subtitles) that is used video's audio-track (L2 captions) (e.g., Peters, Heynen, & Puimège, 2016; Pujadas Muñoz, 2019). It stands reason L1 are especially helpful for learners comprehend content video, while L2 segment speech into words match spoken their written representations. thus complement one another, providing both simultaneously recently investigated well (Wang Pellicer-Sánchez, 2022). concern with simultaneous provision find it hard allocate attention sources input. scenario that, our knowledge, not given consideration research sequential use where watch video assisted by then again captions. This brief report explores scenario. growing body demonstrates audio-visual input made more likely when on-screen text available Majuddin, Siyanova-Chanturia, Boers, 2021; Montero Clarebout, Desmet, 2014; 2019; Both found be beneficial regard (Peters et al., 2016). help material (Baranowska, 2020). On downside, may feel little need process audio-recording because suffice understand video. If they do try attend also speech, connect specific lexical items word-order differences between two languages far easy discern word boundaries connected speech. other hand, what hear how target language, so foster facet knowledge (written form) fail (Bird Williams, 2002). While give direct access meaning items, stream distinct words, whose can guess (Vanderplank, 1993). some degree comprehension, who better at decoding discourse than discourse. Because benefit complementary ways, combined particularly beneficial. One way doing provide types together. study Wang Pellicer-Sánchez (2022) confirmed leads greater gains no onscreen text. Moreover, brought about results only least level acquiring words. When came form was watched fared best. must challenging take lines L2) real time. Eye-movement data collected showed attracted most attention, expense tend receive available. An alternative reaping rewards combination captioning, explored present study, sequentially. entails same twice. Several materials involved repeated viewing Peters Winke, Gass, Syodorenko, 2010). because, become familiar after viewing, will easier them express content. The twice instead once were demonstrated Majuddin al. (2021). Despite remain elusive if insufficient contextual support crucial part rationale sequence starts helps establish meanings. Viewing but learning connection forms provides an incentive recall meanings supplied through subsequently view without any case carefully retrieve based forms. effortful retrieval consolidation recognized (Karpicke, Lehman, Aue, Roediger III Butler, 2011), holds true Barcroft, 2015; Candry, Decloedt, Eyckmans, In exploratory reported below, another three times, consistently 5-minute TED Talk entitled brie(f) History Cheese selected (Paul S. Kindstedt: history cheese | Talk), consisting 688 running Knowledge 4000 frequent families English close 95% coverage text, according Cobb's (n.d.) profiler https://www.lextutor.ca/vp/. suggests demand concerned, suitable participants van Zeeland Schmitt, 2013). At time, there yet could serve targets learning. Eleven (see Table 1). test adopted pretest, posttest, delayed posttest. presented 20 11 9 fillers (higher-frequency known). asked write down either or L1. Half prompts writing, half aurally Jelani 2018, Sydorenko, 2010, similar approach). Although primarily meant assess participants' ability supply done examine whether exposed repeatedly (in L2–L2–L2 condition) reproduced these accurately those saw during sequence. comprehension administered announced beforehand focus consisted six questions: short-answer questions, true/false multiple-choice question. None included test. No tasks tests beforehand. After invited complete online anonymous questionnaire inquiring habitual ways using outside classroom. questions interest once, type textual (if any) prefer. conducted over 3-week period. session, gave informed consent completed pretest. following week, times treatment condition had randomly assigned to. would followed viewings opportunity refine rectify understanding. Shortly third (unannounced) took all collected, They questionnaire. Responses scored dichotomously (either 1 0). twice: strict criteria proposed needed exact, leniently credit response approximated word. statistical trends scoring systems, here sake conciseness. responses assessed raters unaware participants. Any cases disagreement (mostly lenient scoring) resolved discussion. We groups' performance viewing. considered useful ascertain provided assistance associated gains. Given format above), coded correct incorrect (no partial given). Mixed-effects logistic regression compare participant groups, GAMLj module jamovi (The Jamovi Project, 660 observations. fixed effects (the sequences), time (pre, immediate, test), modality versus prompts), scores (covariates). interested interaction effect (pre vs. posttests) condition, indicate other. random factors Alpha significance set .05. According descriptive statistics (Table 2), group students finally slightly poorer prior peers comparison condition. However, posttests noticeably group's, showing pretest large. (Neither attrition L1–L2–none even marginally test, reflect re-taking test.) 2.09 (1.61) [1.80, 2.38] 4.91 (1.78) [4.59, 5.23] 5.18 (1.59) [4.90, 5.47] 2.56 [2.20, 2.91] 3.89 (2.44) [3.41, 4.37] (2.39) [3.42, 4.36] Recall aurally, before supplying seemed lead accuracy production evidence effect, however. conditions score increase median 3 4 immediate needs acknowledged, however, number (n = 5) very small, relatively guessable phonological form. Part captions–L2 captions–no free up attentional resources deal unfamiliar vocabulary. therefore worth examining content-comprehension As expected, obtained (M 4.64, SD 1.21, Mdn 3.56, 1.81, 3). mixed model 3) shows significant predictor [χ2 (2) 30.944, p < .001], fell short despite considerable odds ratios (OR 2.33 OR 2.71) favor gain did reach significance. course bearing mind small groups. reveal (1) 5.646, .017]. higher chance meanings, increasing 31.93% participant's increased unit. recalling tended detected interactions. total amount variance explained 47.6% (conditional R2 .476), 13.4% (marginal .134) attributed leaving 34.2% items. question might realistic engage material. suggested participated Eighteen responded Seventeen indicated habitually purpose understanding improving knowledge. Fourteen respondents additionally usually seek although necessarily here. For example, wrote like themselves seeking promoted retention significantly recall. promising enhancing difference significance, gained large observed × finding hypothesis helped interpret subsequent viewings. Logistic factor learning, indeed conclusions drawn tentative, considering sample size, obvious limitation study. Having control additional limitations. interesting tried (2022). exploring sequences, including ones mentioned (for instance, captions). preliminary first—to knowledge—to explore promising, further research, participants, facets degrees addition recall, clearly necessary confirm effectiveness procedure. somewhat surprising researchers questionnaire, something appear resort spontaneously classroom acknowledging findings preliminary, we believe nonetheless prompt designers, teachers, consider diverse strategic materials. project received funding. Neither authors declares conflict interest. Ethics approval Research Board University Western Ontario: REB# 116673. Injung Wi doctoral student Applied Linguistics University. She MA TESOL university. experience teaching EFL learners. Her interests include grammar Frank Boers professor His early publications field lexicology, focused figurative he published mostly foreign education, phraseology.