http://dx.doi.org/10.

1590/1984-6398201812773

Subtitling, Working Memory, and L2

Learning: A Correlational Study
Legendagem, memória de trabalho
e aprendizagem em L2: um estudo
correlacional
Rafael Matielo*
*Universidade Federal do Ceará (UFC), Fortaleza, Ceará / Brasil
[email protected]
Roberta Pires de Oliveira**
**Universidade Federal de Santa Catarina (UFSC), Florianópolis, Santa Catarina / Brasil
[email protected]
Luciane Baretta***
***Universidade Estadual do Centro-Oeste (UNICENTRO), Guarapuava, Paraná /
Brasil
[email protected]

ABSTRACT: This study investigates the impact of intralingual and interlingual

subtitles on Brazilian English as a Foreign Language (EFL) learners as a
result of their processing of a North-American sitcom. More specifically, it
examines whether subtitling interacts with one’s individual differences, working
memory (WM) as the case in point. Thirty-six intermediate-level EFL learners
were evenly divided into two experimental groups (intralingual subtitles and
interlingual subtitles) and one control group (no subtitles). Participants’
performance was measured based on an L2 video comprehension test and an
L2 vocabulary test. Participants’ performance was correlated with their scores
on two WM tests. The results obtained revealed that both participants’ L2
video comprehension, as well as their L2 vocabulary test performance, did not
significantly interact with their WM capacity under any of the experimental
conditions. These results are discussed in light of the possible processing
mechanisms employed by the participants that may account for the lack of
statistically significant correlations found.
KEYWORDS: subtitled videos; L2 video comprehension; L2 vocabulary
learning; working memory.
RESUMO: Este estudo investiga o impacto de legendas intralinguais e
interlinguais em aprendizes brasileiros de Inglês como Língua Estrangeira
(ILE) resultante do processamento de um sitcom Norte-Americano. Mais

Rev. Bras. Linguíst. Apl., Ahead of Print, 2018 1

 especificamente, ele examina se a legendagem interage com as diferenças
individuais do aprendiz, neste caso, a memória de trabalho (MT). Trinta e seis
aprendizes de ILE foram igualmente divididos em dois grupos experimentais
(legendas intralingual e legendas interlinguais) e um grupo controle (sem
legendas). O desempenho dos participantes foi medido por um teste de
compreensão do vídeo em L2 e um de vocabulário em L2. Ademais, o
desempenho dos participantes foi correlacionado com os resultados de dois
testes de MT. Os resultados obtidos revelaram que a compreensão do vídeo
em L2 pelos participantes e seu desempenho no teste de vocabulário em L2
não interagiram significativamente com sua capacidade de MT em nenhuma
das condições experimentais. Esses resultados são discutidos à luz de diferentes
possíveis mecanismos de processamento empregados pelos participantes que
possam explicar a falta de correlações estatisticamente significativas encontradas.
PALAVRAS-CHAVE: vídeos legendados; compreensão de vídeo em L2;
aprendizagem de vocabulário em L2; memória de trabalho.

1 Introductory Remarks
With the emergence of Karen Price’s seminal work in 1983, a
substantial body of knowledge of the effects of and the effects with subtitles
in language learning has been gathered, and throughout the last thirty years
or so, researchers have moved towards a deeper understanding of how
learners may benefit from subtitling in terms of their language development.
Despite what may seem to be a long period of academic inquiry, much is
still underexplored (VANDERPLANK, 2015).
Investigating the use of intralingual (same language) and interlingual
subtitles (different linguistic pair in the audio/subtitles) in L2 learning has
been gaining prominence in recent years. From the 1980s on, more than
sixty papers have been published in respected journals around the world,
reporting on the results of experiments with different populations and target
languages (MATIELO; D’ELY; BARETTA, 2015). Interestingly, only a
handful of them have addressed Brazilian learners of English (MATIELO;
COLLET; D’ELY, 2013; MATIELO; OLIVEIRA; BARETTA, 2017;
MATIELO; OLIVEIRA; BARETTA, Forthcoming).
In short, subtitles have been found to foster L2 development,
regardless of whether they are interlingual or intralingual (D’YDEWALLE;
VAN DE POEL, 1999; HUANG; ESKEY, 1999; KOOLSTRA; BEENTJES,
1999; MARKHAM, 1999; MARKHAM; PETER, 2003; DANAN, 2004;
STEWART; PERTUSA, 2004; TAYLOR, 2005; CAIMI, 2006; CHANG,

2 Rev. Bras. Linguíst. Apl., Ahead of Print, 2018

2006; VAN LOMMEL; LAENEN; D’YDEWALLE, 2006; SYDORENKO,
2010; WINKE; GASS; SYDORENKO, 2010; ZAREI; RASHVAND,
2011; RAINE, 2013). Nonetheless, some studies have not found significant
differences considering subtitles’ availability and their relationship with
the specific language component being tested, such as general/listening
comprehension (BIANCHI; CIABATTONI, 2008; LATIFI; MOBALEGH;
MOHAMMADI, 2011; MONTERO-PEREZ; PETERS; DESMET, 2013;
SHARIF; EBRAHIMIAN, 2013; MATIELO; COLLET; D’ELY, 2013;
MATIELO; OLIVEIRA; BARETTA, 2017) and L2 vocabulary learning
(MATIELO; COLLET; D’ELY, 2013; ZAREI; GILANIAN, 2013;
MATIELO; OLIVEIRA; BARETTA, Forthcoming).
When adding the simultaneous processing of audio and subtitles
to the equation, the picture becomes blurrier. Since reading subtitles is
considered an automatic process, irrespective of one’s familiarity with them
or knowledge of the target-language available in the soundtrack – evidence
comes from eye-movement recordings (D’YDEWALLE; GIELEN, 1992;
WINKE; GASS; SYDORENKO, 2013) – part of one’s attentional resources
when watching a subtitled film is allocated in the tasks of reading and
processing the subtitles, watching and attending to the story, processing the
motion picture, and arguably attending to the auditory input. Hence, the
ability to successfully attend to the whole set of input – the simultaneous
use of spoken (audio/soundtrack) and written (interlingual and intralingual
subtitles) input modes – could be linked to one’s Working Memory
(WM) capacity, an integrated system in charge of temporary storage and
manipulation of information during the execution of any given cognitive
task (BADDELEY, 1992; BADDELEY, 2011).
Since attending to the whole set of input in a video watching task can
be considered highly cognitively demanding, we posit that the systematic
exploration of the correlation between learners’ WM capacity and their
ability to understand the content of the video and pick up novel words
encountered in subtitled video materials seems to be relevant and necessary.
Insights into this matter may provide L2 researchers and practitioners with
further understanding of the challenges that subtitled video materials may
pose for L2 learners’ simultaneous processing of all the channels at play
while engaged in watching subtitled video materials in and outside their L2
classroom environment.

Rev. Bras. Linguíst. Apl., Ahead of Print, 2018 3

 This article is divided into four sections in addition to this introductory
one. Section 2 presents the main findings related to subtitling availability
and L2 learning from a cognitive oriented perspective. Section 3 centers on
the methodological aspects informing this study. Section 4 focuses on the
descriptive statistics, the results, and the discussion of the main findings.
Finally, section 5 offers a summary of the main findings of the present study
while outlining some of their implications.

2 Subtitling and Cognitive Aspects in L2 Learning

Until the 2000s, intralingual subtitles were mostly found to be more
effective in aiding L2 development (see also Matielo, D’Ely & Baretta,
2015 for a state-of-the-art review), since groups were compared to and
contrasted with controls (with no subtitles available in the experiments).
As for comparative studies that emerged from the 2000s on, the situation
is not quite clear and presents grayer areas. For instance, in two studies
(STEWART; PERTUSA, 2004; HAYATI; MOHMEDI, 2011), better
results were mostly achieved with intralingual subtitles, whereas in one
of the studies (MARKHAM; PETER, 2003), better results were achieved
with interlingual subtitles. Some studies presented better results with
one or the other depending on the language component being tested or
proficiency group (MARKHAM; PETER; MCCARTHY, 2001; BIANCHI;
CIABATTONI, 2008; LATIFI; MOBALEGH; MOHAMMADI, 2011;
ZAREI; RASHVAND, 2011). As far as video comprehension and
vocabulary learning are concerned, there does not seem to be an agreement
as to which type of translational aid can be more beneficial.
The findings originated from eye-tracking studies on the processing
of subtitled materials point to a few poignant aspects. Perego, Del Missier,
Porta, and Mosconi (2010) examined the hypothesis that the processing of
subtitled film is cognitively effective, that is, one should be able to understand
the film content without a significant tradeoff between the processing of
images (visual input) and text (written input). When watching subtitled
material, “attention needs to be flexibly allocated on parallel information
sources during this task” (p. 250). They hypothesized that when attention
is more focused on the subtitles, image processing would be less effective;
hence, the opposite should also be true. They explain that this hypothesis
is generally consistent with attentional theories that postulate the early

4 Rev. Bras. Linguíst. Apl., Ahead of Print, 2018

selection of information channels. Results from the study, however, revealed
no tradeoff effects between subtitle processing and image processing, thus
suggesting that the participants watching the subtitled film did process its
content and subtitles effectively.
Kruger and Steyn (2014) offer novel insights from an experiment
carried out to investigate subtitle reading behavior and performance. In
short, the results obtained with eye-tracking experiments showed that
although no significant differences were found between the performance
of the participants in the intralingual subtitles condition and those in the
control condition, participants in the test group who actually did read the
subtitles performed better on the comprehension test than those who saw
the videos with the subtitles but did not read the subtitles.
Winke, Gass, and Sydorenko (2013) also looked at caption-reading
behavior based on eye-tracking methodologies. Their study’s results revealed
that, in general, participants fixated on the intralingual subtitles 68% of the
time when the captions were shown on the screen. There was also an effect
of video (video familiarity) on caption viewing depending on the L2 being
learned. More specifically, learners of Arabic, Russian, and Spanish spent
similar amount of time reading the captions on both videos. The Chinese
learners, in contrast, spent less time reading the captions when watching the
video with familiar content.
Based on the findings of these studies, it is clear that learners watching
subtitled video materials distribute their attentional resources into processing
both the visual scenes and the subtitles. Bird and Williams (2002) explored a
central prevailing issue regarding the speculations of the effects of subtitling
on L2 learning, that is, whether or not soundtrack is in fact processed when
subtitled videos are watched. Their study examined the effects of single
modality input – either sound or text – and bimodal input – sound and text
– presentation on word learning, with an explicit focus on L2 word learning.
Measures consisted of enhancements in spoken word recognition efficiency
(implicit memory) and recognition memory related to word retention
(explicit memory). Results indicated that auditory lexical decisions in terms
of familiar words were equally primed by prior bimodal and sound-only
presentation modes, though no priming effects for nonwords were found.
Furthermore, the results suggested that the participants were able to attend
to and process both text and sound. Nonetheless, bimodal input failed to
show any significant advantage in relation to sound-only modality. This is

Rev. Bras. Linguíst. Apl., Ahead of Print, 2018 5

an extremely important finding that lends support to the notion that the
addition of text was not conducive to better gains in learning in relation to
single-modality input.
In light of these findings, one’s individual differences – WM, more
specifically – may play a decisive role in one’s successful distribution of
attentional resources while processing the different sources of information
at the same time: soundtrack, motion picture, and subtitles. It could also
be expected that the higher one’s WM capacity is, the better one should do
on subtitled video materials’ processing and language learning tasks. This
is particularly pertinent if we take into account that WM directly relates to
the control of specific cognitive mechanisms, such as attention, processing,
and other regulatory functions, entailing the access to long-term memory
information (BADDELEY, 2000).
According to Baddeley (1992), WM research has mostly developed
under two different, but complementary approaches. The first approach is
the dual-task and neuropsychological approach, focusing on the analysis of
the structure of WM itself, emphasizing its slave subsystems, including the
study of evidence of neuropsychology and the application of dual tasks,
requiring participants to memorize and store digits in a digit-span task while
performing other cognitive tasks, for instance. The second approach, called
the psychometric correlational, refers to the correlation between individual
differences in WM and the performance of cognitive abilities (DANEMAN;
CARPENTER, 1980). The methodology of this approach involves tasks
and correlates performance in these tasks with performance in other high
cognitive tasks. The present study is carried out within the psychometric
approach to the study of WM and its possible relationship with other
cognitive tasks, namely language learning tasks.

3 Method
This mixed design (qualitative and quantitative) study1 examines
whether one’s WM capacity interacts with the processing of intralingual and
interlingual subtitled material by Brazilian English as a Foreign Language
(EFL) learners.

1
For the purposes of this article, only results of correlations among WM tests, the L2
video comprehension test, and the L2 vocabulary test are reported.

6 Rev. Bras. Linguíst. Apl., Ahead of Print, 2018

3.1 Participants
A total of 36 Brazilian intermediate EFL learners, ranging from
18-60 years of age, participated in the data collection (20 female and 16
male) and were chosen based on their proficiency level. Participants were
Brazilian Portuguese native speakers enrolled in level 5 (intermediate) in
the Extracurricular2 (non-credit) Language Courses at Universidade Federal
de Santa Catarina (UFSC), Florianópolis, SC, Brazil. Studies on the effects
of subtitled video materials have typically involved intermediate learners
of the language, and the assumption behind that is that these participants
are usually at a threshold proficiency level that enables them to read the
subtitles on screen in the foreign language, given their short display time
(2-4 seconds, only).
Participants were randomly assigned to one of the treatment groups
or the control group: Intralingual Subtitles Group (n = 12), Interlingual
Subtitles Group (n = 12); and Control Group – no subtitles (n = 12).
Participants received explanations about all the stages in the data collection
in the first meeting, when they were invited to participate in the research;
signed the Informed Consent Form3; and had the opportunity to clarify
doubts related to the goals and the design of the study. However, participants
were not told about specific details that could bias their answers during any
of the stages in the data collection. The participants’ teachers were also given
a questionnaire in Portuguese in order to provide more information about
their education and experience with English language teaching.

2
The Extracurricular (non-credit) Language Courses at UFSC are open to undergraduate
and graduate students enrolled at UFSC or any other higher education institution in the
area, as well as to faculty members and members of the community. Students can enroll
twice a year and can take a placement test to determine their proficiency level in any of
the language courses offered: English, French, German, Italian, Japanese, Portuguese as a
Second Language, and Spanish. In relation to English courses, language instructors consist
of undergraduate and graduate students taking the Letras program, who are supervised
by two coordinators.
3
The research project was submitted to the university’s Ethics Committee and an approval
was obtained, which is logged under code number 36597314.9.0000.0118. Participants’
teachers also signed a consent form.

Rev. Bras. Linguíst. Apl., Ahead of Print, 2018 7

3.2 The TV Series
The participants watched a 20-minute long episode of the American
sitcom The Big Bang Theory, which premiered in 2007. In Brazil, the show
is broadcast with Portuguese subtitles on Warner channel and is a critically
acclaimed show. The sitcom (situational comedy) depicts Leonard Hofstadter
and Sheldon Cooper, two brilliant physicists who are best friends and
roommates. They are also friends with their co-workers Howard Wolowitz,
a mechanical engineer, and Rajesh Koothrappali, an astrophysicist. The gang
spends their time working on their individual University projects, playing
games, watching comic-con related movies or reading comic books. When
Penny, the girl next door moves in the neighboring apartment, the story takes
a turn as Leonard is intent on trying to go out with her.4
The sitcom was selected based on a series of criteria. First of all, it
has been used in previous studies (MATIELO; COLLET; D’ELY, 2013;
MATIELO; OLIVEIRA; BARETTA, 2017). Additionally, a sitcom was
thought to be appealing and appropriate for the target audience. From
the profile questionnaire administered in the very first session in the data
collection, 31 out of the 36 participants reported watching sitcoms, thus
suggesting their familiarity with the genre. The episode used in the study was
“The Grasshopper Experiment”, the eighth episode in the first season. The
episode was chosen, as it contained a complete storyline and did not require
students to be familiar with the series or previous episodes.

3.3 The Video Comprehension Test

Two parts for this test were designed. For the general comprehension
part, three ‘why’ questions were posed and participants were instructed
orally to answer them in Portuguese or in English and were assured that
grammatical errors would be disregarded. To successfully answer the general
comprehension questions, participants needed to understand the story and
how it unfolded throughout the episode. All of the events in the episode
were highly intertwined, which meant that participants had to have some
level of general comprehension as regards the connections among the
goings-on in the episode to be able to answer the questions. An example of
such a question is: “At the end of the episode, why is Sheldon singing and

4
Information retrieved from http://www.imdb.com/title/tt0898266.

8 Rev. Bras. Linguíst. Apl., Ahead of Print, 2018

playing?” In the specific comprehension part, participants had to judge five
statements about the story narrated on screen on whether they were true
or false. The statements essentially provided details about the characters,
their relationships, likes and dislikes, attitudes, and information about the
story. An example of such statement to be judged as true or false is: “( ) Raj
is used to drinking”.

3.4 The L2 Vocabulary Test

The L2 vocabulary test contained three parts: a pre-test, a test, and
a post-test. The pre-test assessed participants’ previous knowledge of the
target vocabulary, in the first session. They received a 20-word list in English
in which 10 of them were distractors and were asked to write their meaning,
a synonym or an explanation in Portuguese or English using their own
words. The target words were chosen taking into account factors influencing
word learnability (LAUFER, 1997); for instance, some words were chosen
because of their facilitated word learning aspect, such as familiar morphemes
(e.g., pointless and membership), whereas others were chosen for their difficulty-
inducing factors, such as the presence of foreign morpheme (e.g., obnoxious),
and some were selected as neutral, such as those related to concreteness
or abstractness (e.g. wrath).5 Moreover, the words chosen are related to
the themes portrayed in the episode, but they are not semantically related
(ERTEN; TEKIN, 2008).
Regarding the number of times the target-words appear in the selected
episode, half of them were uttered and shown in the subtitles – both
intralingual and interlingual – twice (slot, membership, guinea pigs, showdown, and
wrath), whereas half was uttered and shown in the subtitles once (embodiment,
pointless, pushy, obnoxious, and resemblance). The short exposure to the input is
acknowledged, though the video length is short (20 minutes).
Other important criteria considered in this study relate to whether
the words actually appeared in the interlingual and intralingual subtitles and
their relevance to the story, which could facilitate participants’ processing
and future recognition. Distractors, on the other hand, contained likely
familiar vocabulary based on semantic familiarity (e.g., affection and mint) and
unfamiliarity (e.g., award and moisturizer), taking into account their proficiency
level. Word frequency was not controlled.

5
All of the examples have been taken from the actual test.

Rev. Bras. Linguíst. Apl., Ahead of Print, 2018 9

 The L2 vocabulary test required participants to analyze 10 target-
vocabulary words from the pre-test in English (excluding the distractors) and
write their meaning, a synonym, or an explanation in Portuguese or English,
using their own words immediately after watching the video, in the second
session. In the third session, for the post-test, participants were provided
with a test identical to the test to check if they were able to recognize the
words they had encountered by the time the test was administered. The
participants were asked to write their meaning, a synonym, or an explanation
in Portuguese or English using their own words. As in the L2 vocabulary
test, distractors were not included.

3.5 The WM Tests

The Reading Span Test (RST) was originally designed by Daneman
and Carpenter (1980) to explore the relationship between individual
differences in WM capacity of native speakers of English and their
comprehension of sentences in addition to the recall of the last words
of a group of sentences. In the present study, the Brazilian version of the
adopted RST was a modification of the one used in Torres (2003), which
was administered in Portuguese (participants’ L1) to avoid confounds with
participants’ L2 proficiency level and to avoid floor effects (scoring too low)
due to task difficulty. This study’s version of the test comprised 42 unrelated
sentences, ranging from 13 to 17 words in length, presented in sets of 2, 3,
4, and 5 sentences.
Two important modifications in the test were made as a result of
informal piloting. This study’s version of the test did not use sets containing
6 sentences because all of the five participants in the informal piloting could
not remember the words when sets of six sentences were shown on screen,
neither in the correct order nor out of order of presentation. Additionally,
at the end of the test, these participants from the informal piloting reported
finding the test too long and too tiresome. Another important modification
was the color of the computer screen: instead of using a white color, this
study’s version of the test used a black color on the screen, since four out
of the five participants in the informal piloting reported being extremely
uncomfortable with a white screen during the entire practice and testing
sessions, and suggested a black screen instead. The font color was adapted
to the screen color as well, therefore using white instead of black to suit the

10 Rev. Bras. Linguíst. Apl., Ahead of Print, 2018

background of the slides (see also MATIELO, 2016, for more information
on the test administration and scoring).
As for the Operation Word Span Test (OSPAN), the test was originally
designed by Turner and Engle (1989) to investigate the hypothesis that
WM capacity is not language specific and can thus be generalized to any
cognitively complex task, since it has been suggested to be a reliable measure
of WM capacity (CONWAY et al., 2005). The test essentially consists of
asking the participant to solve simple mathematical operations while trying
to recall a set of unrelated words.
The Brazilian Portuguese version of the test used in this study was
designed by Prebianca (2009) based on Turner and Engle’s (1989) test.
The words in the test were disyllabic, unlikely to be unknown by native
speakers of Portuguese in the age range under study (e.g. papel and tinta).
The OSPAN was also administered in Portuguese to avoid confounds with
participants’ L2 proficiency level. As in the RST, the OSPAN consisted of 42
operation strings along with Portuguese words, written in white and placed
directly in the middle of a black computer screen. From the 42 trials, 19
strings presented correct operations, whereas 23 strings displayed incorrect
operations (see also MATIELO, 2016, for more information on the test
administration and scoring).
Both the RST and the OSPAN tests were scored strictly and leniently.
For the strict scoring of RST test, participants’ reading span was calculated
at the level at which s/he was accurate on at least two trials of a given set of
sentences (DANEMAN; CARPENTER, 1980). In accordance with Turner
and Engle (1989), an approximately 85% accuracy rate was required in terms
of participants’ judgment of sentence grammaticality or syntax acceptability
to ensure the processing component of the task, which represented 36 out
of 42 sentences. Half a point was given when the participant passed one
trial at a certain level. For instance, when a participant correctly recalled all
the words in the right order in the three sets of 2 sentences and correctly
recalled just one trial in the three sets of 3 sentences, this participant received
half a point, her/his span being considered 2.5, which is where scoring
would then terminate. In the lenient scoring of the RST, participants were
given points for any set for which s/he recalled all final sentence words,
regardless of the order of recall, as long as s/he recalled all of the words
belonging to two of the trials in the given set, while a half point was given
when the participant passed one trial at a certain level, which is where scoring

Rev. Bras. Linguíst. Apl., Ahead of Print, 2018 11

would then terminate. The approximately 85% accuracy rate was also kept
in the lenient scoring of the RST test in order to maintain the processing
component of the task.
Equally, participants’ answers on the OSPAN test were scored strictly
and leniently, following distinct procedures. In relation to the strict scoring
of the test, one point was credited to each word recalled in the exact order
of presentation, that is, in a test set of three trials, for instance, a participant
who was able to solve at least two math operations correctly and then
recalled their corresponding words, respecting the order of presentation,
was given two points. In this scoring procedure, both operation solving
and word recall were taken into account. By contrast, the lenient scoring
method of the OSPAN test was not as strict in terms of the participants’
processing efficiency. In line with previous studies (TURNER; ENGLE,
1989; KANE; BECKLEY; CONWAY; ENGLE, 2001; KANE; CONWAY;
HAMBRICK; ENGLE, 2007; PREBIANCA, 2009), a criterion of 85%
accuracy in correctly solving all of the mathematical operations of the
entire test was required, since it is useful in ensuring that participants do
not trade-off between processing the mathematical operations of the test
and storing the words that come along with them (UNSWORTH; HEITZ;
SCHROCK; ENGLE, 2005). Hence, all words recalled in the exact order of
presentation and which obeyed the criterion of 85% accuracy were credited
one point, which means that each participant could have gotten up to six
wrong mathematical operations (out of a total of 42 trials). If the participant
responded to a specific operation of a particular set incorrectly but was able
to recall the word following that operation accurately, s/he was credited 1
point as long as s/he had not yet reached six errors.

4 Results and Discussion

4.1 Subtitling, L2 Video Comprehension, and WM Capacity
This section of the paper reports on the results obtained with the
correlation tests that were run in order to examine whether participants’
WM capacity would present any interaction with their performance on the
video comprehension test and/or their performance on the L2 vocabulary
test under any experimental (intralingual or interlingual) condition.
To check whether and the extent to which the performance of the
participants on the general comprehension part of the video comprehension

12 Rev. Bras. Linguíst. Apl., Ahead of Print, 2018

test held a significant relationship with their performance on the WM tests,
different Spearman’s Rank Order Correlation Coefficient tests were run.
The results obtained with these tests are reported in Table 1:
TABLE 1 – General Comprehension and WM Correlations

Correlation GCScore x GCScore x GCScore x GCScore x

Subtitles (n=12)

Test RST Strict RST Lenient OSPAN Strict OSPAN Lenient

Intralingual

Spearman’s
.000 .054 .231 -.336
Coefficient rs

p value 1.000 .867 .470 .286

Correlation GCScore x GCScore x GCScore x GCScore x

Subtitles (n=12)
Interlingual

Test RST Strict RST Lenient OSPAN Strict OSPAN Lenient

Spearman’s
.556 .517 .148 .302
Coefficient rs

p value .061 .085 .645 .340

Correlation GCScore x GCScore x GCScore x GCScore x

Test RST Strict RST Lenient OSPAN Strict OSPAN Lenient
Control
(n=12)

Spearman’s
.006 .247 -.177 .297
Coefficient rs

p value .986 .439 .581 .348

Note. n = sample size; RST = Reading Span test; OSPAN = Operation-Word Span Test;
p = significance level; GCScore = general comprehension score.

Spearman’s Rank Order Correlation Coefficient tests revealed

mostly a positive, though not statistically significant correlation between
the participants’ performance on the general comprehension portion of
the video comprehension test and the WM tests at α = .05. The correlation
tests also revealed two instances of negative correlations, such as in the
OSPAN and the general comprehension scores test with the intralingual
subtitles group (lenient scoring) and the control group (strict scoring),
none of which are statistically significant. The results also indicated a total
absence of correlation between the general comprehension scores and the
OSPAN Strict variables with the intralingual subtitles groups (rs = .000, n
= 12, p > .05).

Rev. Bras. Linguíst. Apl., Ahead of Print, 2018 13

 To investigate whether and the degree to which the performance
of the participants on the specific comprehension part of the video
comprehension test would suggest any relationship with their WM capacity,
Spearman’s Rank Order Correlation Coefficient tests were run. The results
obtained are presented in Table 2:
TABLE 2 – Specific Comprehension and WM Correlations

Correlation SCScore x SCScore x SCScore x SCScore x

Subtitles (n=12)

Test RST Strict RST Lenient OSPAN Strict OSPAN Lenient

Intralingual

Spearman’s -.338 -.271 -.293 -.520

Coefficient rs
p value .283 .295 .355 .083

Correlation SCScore x SCScore x SCScore x SCScore x

Subtitles (n=12)
Interlingual

Test RST Strict RST Lenient OSPAN Strict OSPAN Lenient

Spearman’s .147 .050 .063 .099
Coefficient rs

p value .647 .877 .845 .760

Correlation SCScore x SCScore x SCScore x SCScore x

Test RST Strict RST Lenient OSPAN Strict OSPAN Lenient
Control
(n=12)

Spearman’s -.146 .000 .271 .167

Coefficient rs
p value .651 1.000 .393 .605

Note. n = sample size; RST = Reading Span test; OSPAN = Operation-Word Span Test;
p = significance level; SCScore = specific comprehension score.

Spearman’s Rank Order Correlation Coefficient tests revealed varied

results across the groups. In relation to the intralingual subtitles group, the
correlations found were negative; as for the interlingual subtitles group,
very weak positive correlations were found; finally, regarding the control
group, the correlations obtained were either negative or weak positive.
None of the correlations found are statistically significant considering the
participants’ performance on the specific comprehension portion of the
video comprehension test and the WM tests (p > .05).

14 Rev. Bras. Linguíst. Apl., Ahead of Print, 2018

 As far as the results between the correlation tests performed with
the WM measures are concerned – RST and OSPAN tests – and the video
comprehension test – general and specific parts – it is clear that, in this study,
no statistically significant relationship was found between them in any of
the experimental conditions – intralingual and interlingual subtitles – and
the control condition. We speculate that some factors might help explain the
lack of statistically significant results obtained for those correlational tests.
Firstly, it is important to examine the nature of the WM tests used.
When analyzing what participants actually do on the RST, the common
understanding is that the test taps one’s ability in processing efficiency,
mainly as regards comprehension. Daneman and Carpenter (1980), when
originally designing the test, argued that individual differences in WM
capacity would reflect differences in terms of processing efficiency. They
claim that differences in the processing efficiency would be at the core of
individual differences regarding language comprehension. In other words,
individuals who are not so effective in terms of their processing would
possibly have a smaller storage capacity because they would have to free
more of their attentional resources to meet the processing demands of
a given task. Hence, one’s span is likely to correlate with one’s capacity
to maintain information active in WM long enough to attach it to new
information so as to understand a text, for instance. The scholars also
claimed that the RST can predict comprehension because it involves many
processing aspects that are involved in normal, typical reading.
As for the OSPAN test, Turner and Engle (1989) also intended to
measure WM capacity through a task that could tap one’s ability to store
information for a brief period of time, but that is not language-specific. As
demonstrated by Klein and Fiss (1999), the test has been adopted to measure
WM capacity because of its high reliability and stability scores. The OSPAN
is in tune with Engle and colleagues’ view that WM capacity is related to the
processing efficiency in any given cognitive task in terms of information
processing. Prebianca (2009) explains that “capacity refers to individuals’
ability to bring pieces of information from long-term memory into an active
state and temporarily maintain that information for further processing by
preventing other irrelevant stimuli to enter the focus of attention” (p. 34).
Thus, the lack of a significant relationship among the RST and the OSPAN
measures and the video comprehension test as well as the L2 vocabulary

Rev. Bras. Linguíst. Apl., Ahead of Print, 2018 15

recognition tests in the present investigation may be related to the set of
relatively different types of tasks employed.
Secondly, while watching the TV series, all three groups had either
two or three channels that could compete for their attentional resources.
The intralingual subtitles group was submitted to a watching condition in
which one auditory channel (soundtrack) and two visual channels (subtitles
+ video) were simultaneously provided, all of which shared a common
language (L2); the interlingual subtitles group, however, was performed
in a different condition, as it was provided with both an auditory channel
(L2 soundtrack) and two visual channels (L1 subtitles + video), thus being
presented with two different linguistic sources of input; as for the control
group, two channels were presented to them, that is, one auditory channel
(L2 soundtrack) and one visual channel (video).
Participants did not have their attentional resources directed to any
of the channels (soundtrack, subtitles, and/or video). In other words, no
guiding instructions were provided to them that prompt or influence them
in either prioritizing one of the input sources or even attempting to process
all of them simultaneously. Hence, participants’ attentional resources may
have been allocated in only one verbal channel.
Participants in the intralingual subtitles condition outperformed the
other two groups on both general and specific comprehension portions of
the test (results not included in this paper). Participants performing under
that condition were provided with only one language (L2) in all verbal
input sources (soundtrack + subtitles). The plausible inference is that there
were fewer attention depleting mechanisms involved in comparison to
the interlingual subtitles group. This explanation, however, would not be
applicable to the control group, whose performance was indeed poorer as
compared to the performance of the intralingual subtitles group, but not
statistically different from the interlingual subtitles group’s performance.
Thirdly, although the RST has been found to correlate mostly
with reading comprehension, it may as well be the case that participants’
processing on the video comprehension test involved a set of distinct
mechanisms, especially taking into consideration the nature of the task
employed in this research, involving a multimedia source, and therefore not
regular, typical reading. We suspect that one’s efficiency in reading subtitles
(in L1 or L2) might be more closely linked with one’s experience with and the
amount of exposure to them. Participants in the present research endeavor

16 Rev. Bras. Linguíst. Apl., Ahead of Print, 2018

declared being more used to watching subtitled videos with interlingual
subtitles than with intralingual subtitles or no subtitles whatsoever.
We also suspected that one’s capacity to process the auditory channel
of the video would hold a relationship with one’s WM capacity – or, more
specifically, one’s phonological memory – if, in fact, that channel significantly
drew on one’s attention while processing another channel (subtitles and/
or video) simultaneously. In other words, in the present investigation,
participants might not have consciously attended to the auditory verbal
channel. Given that participants informally stated that they tend to resort
to the subtitles more often than the auditory channel, it could be that no
seemingly major interaction would arise in this scenario. Selective attention
might have played a more prominent role in determining where participants
allocated more of their attentional resources to process the input provided
in a given channel.
Downing (2000) researched the relationship between selective
attention and WM, and contends that selective attention “reduces the load
on limited-capacity cognitive systems by filtering irrelevant information
from the stimulus stream” (p. 467). The author also explains that in a
typical scenario containing many objects, the amount of information
present exceeds the capacity of object representation systems, that is, one’s
capacity to process that (visual) information. Consequently, the objects can
be described as “competing” for attention, and the strongest competitors
– perhaps the more relevant or salient ones – will be likely to become the
focus of selective attention. As a result, they gain access to awareness and
guidance of action.
When watching subtitled audiovisual material, viewers possibly focus
on the most salient channel, which could be their preferred one or the one
that might seem easier for them to process or even one that is more attention
drawing. In other words, visual channels – either verbal or non-verbal –
could constitute their strongest competitors for attention. The problem in
this scenario is that attending to the two channels would require, we assume,
a considerable processing load because of the speed in which they are
processed (short duration on screen as regards the subtitles and the scenes)
and the fact that reading is involved in processing one of the channels. If
true, this would explain why participants informally reported that they tend
to ignore the soundtrack of the video most of the times.

Rev. Bras. Linguíst. Apl., Ahead of Print, 2018 17

 What Downing (2000) has found in his experiments is that “visual
working memory and selective attention share a key functional component:
The contents of working memory guide attention even when there is no
explicit search task” (p. 469). In the present investigation, participants were
not provided with any specific guidance that could direct the allocation of
their attentional resources. It is then plausible to assume that when watching
subtitled videos, selective attention was directed mostly towards the subtitles
and the images of the video.
A somewhat similar perception about this issue is shared by Sydorenko
(2010). Her assumptions relate to learners’ efforts to simultaneously allocate
attention onto all three modalities. She believes that if learners did pay
attention to the three channels simultaneously, this could result in a cognitive
overload, which “occurs even when tasks are performed in the native
language and is attributed to the limits of working memory” (p. 52).
The contradiction surfaces when L1 and L2 studies are contrasted.
Mayer, Heiser, and Lonn (2001), stemming from a cognitive load theory
perspective on multimedia learning, found that L1 speakers of English who
saw an animation and listened to a simultaneous narration in their L1 were
able to gather more information from the narration than those who also
received a third modality/channel to process (intralingual subtitles). Their
view is that subtitle availability is distracting when soundtrack is also available
because they essentially convey the same information, which is thought to
follow a redundancy principle in terms of information processing capacity.
However, the literature on subtitling and L2 development has
suggested quite the opposite: Subtitle availability has been found to be
associated with better L2 comprehension and L2 development. Nonetheless,
the preferred source of input to be attended and processed by viewers has
not yet been extensively researched. To date, there have only been a handful
of studies that have looked at this specific aspect (VANDERPLANK, 1988;
TAYLOR, 2005; SYDORENKO, 2010). In general, what these studies
have found is that, at first, subtitles are considered distracting by the L2
viewers; with time, viewers reported that they get used to them and even
start developing strategies to try to attend to all channels simultaneously.
Fourthly, another possible explanation for the lack of significant
correlation between WM and video comprehension could be due to the

18 Rev. Bras. Linguíst. Apl., Ahead of Print, 2018

sample size and the distribution of high spanners.6 Traditionally, research on
the relationship between WM and language development has been carried
out with much larger sample sizes. In the present investigation, only 12
participants remained until the very end of the data collection in each of the
three groups, and the high spanners were grouped in only one experimental
group, coincidentally. Therefore, it is possible that no statistically significant
correlations were found in this study because of the limited number of
participants and the fact that there were fewer high spanners, who were not
evenly distributed across experimental and control groups.

4.2 Subtitling, L2 Vocabulary, and WM Capacity

The results reported in this section are an attempt to investigate a
possible relationship between the participants’ performance on the L2
vocabulary test (including the pre-test, the test, and the post-test) and their
WM tests (both the RST and the OSPAN tests). The results obtained with
the correlation tests of WM (RST and OSPAN tests) and the L2 vocabulary
pre-test are shown first (Table 3), followed by the L2 vocabulary test (Table
4), and finally the L2 vocabulary post-test (Table 5). The discussion of all
correlation tests is presented altogether, at the end of this section.

6
In this study, as far as the scores on the RST are concerned, three high spanners were
found in the data, all of which belong to the interlingual subtitles group; as regard the
scores on the OSPAN, three high spanners were also found in the data, all of which belong
to the interlingual subtitles group as well.

Rev. Bras. Linguíst. Apl., Ahead of Print, 2018 19

 TABLE 3 – L2 vocabulary pre-test and WM correlations

Pre-Test x Pre-Test x Pre-Test x Pre-Test x

Correlation Test
Subtitles (n=12)

RST Strict RST Lenient OSPAN Strict OSPAN Lenient

Intralingual

Spearman’s
-.419 -.440 -.262 -.102
Coefficient rs

p value .175 .152 .411 .753

Pre-Test x Pre-Test x Pre-Test Pre-Test x

Subtitles (n=12)

Correlation Test
Interlingual

RST Strict RST Lenient OSPAN Strict OSPAN Lenient

Spearman’s
-.194 -.112 .178 .130
Coefficient rs

p value .546 .729 .580 .687

Pre-Test x Pre-Test x Pre-Test x Pre-Test x

Correlation Test
RST Strict RST Lenient OSPAN Strict OSPAN Lenient
Control
(n=12)

Spearman’s
.169 .264 -.184 .286
Coefficient rs

p value .599 .407 .568 .367

Note. n = sample size; RST = Reading Span test; OSPAN = Operation-Word Span Test;
p = significance level.

20 Rev. Bras. Linguíst. Apl., Ahead of Print, 2018

 TABLE 4 – L2 vocabulary test and WM correlations

Test x Test x Test x Test x

Correlation Test
Subtitles (n=12)

RST Strict RST Lenient OSPAN Strict OSPAN Lenient

Intralingual

Spearman’s
-.119 -.254 -.209 -.100
Coefficient rs

p value .713 .426 .515 .756

Test x Test x Test x Test x

Subtitles (n=12)

Correlation Test
Interlingual

RST Strict RST Lenient OSPAN Strict OSPAN Lenient

Spearman’s
-.323 -.281 .360 .230
Coefficient rs

p value .305 .377 .251 .473

Test x Test x Test x Test x

Correlation Test
RST Strict RST Lenient OSPAN Strict OSPAN Lenient
Control
(n=12)

Spearman’s
.203 .318 -.029 .405
Coefficient rs

p value .526 .314 .928 .192

Note. n = sample size; RST = Reading Span test; OSPAN = Operation-Word Span Test;
p = significance level.

Rev. Bras. Linguíst. Apl., Ahead of Print, 2018 21

 TABLE 5 – L2 vocabulary post-test and WM correlations

Post-Test x Post-Test x Post-Test x Post-Test x

Intralingual Subtitles

Correlation Test
RST Strict RST Lenient OSPAN Strict OSPAN Lenient
(n=12)

Spearman’s
.152 -.022 -.283 -.391
Coefficient rs

p value .638 .945 .373 .209

Post-Test x Post-Test x Post-Test x Post-Test x

Subtitles (n=12)

Correlation Test
Interlingual

RST Strict RST Lenient OSPAN Strict OSPAN Lenient

Spearman’s
-.256 -.235 .268 .081
Coefficient rs

p value .421 .462 .399 .803

Post-Test x Post-Test x Post-Test x Post-Test x

Correlation Test
RST Strict RST Lenient OSPAN Strict OSPAN Lenient
Control
(n=12)

Spearman’s
.364 .487 .051 .276
Coefficient rs

p value .245 .109 .876 .385

Note. n = sample size; RST = Reading Span test; OSPAN = Operation-Word Span Test;
p = significance level.

As shown in Table 3, the results obtained with the tests revealed

varied results for the three groups, converging into mostly moderate, weak
negative correlations and a few weak positive correlations, all of which
are not statistically significant (p > .05). In other words, the results of the
correlation tests are suggestive of an absence of significant relationship
between the participants’ performance on the L2 vocabulary pre-test and
the WM tests. Moreover, as indicated in Table 4, the results obtained with
correlation tests revealed varied results across the groups, converging into
mostly weak negative correlations, though there are a few positive weak and
moderate correlations. No statistically significant relationship was found (p
> .05), which can be interpreted as an absence of noteworthy association
between the participants’ performance on the L2 vocabulary test and the
WM tests. Finally, not surprisingly, the results reported in Table 5 also

22 Rev. Bras. Linguíst. Apl., Ahead of Print, 2018

revealed varied results across the groups, which range from an almost perfect
absence of correlation to either a moderate negative or positive correlation.
However, all of correlations obtained are not statistically significant (p >
.05), which are indicative of an absence of substantial relationship between
the participants’ performance on the L2 vocabulary post-test and on the
WM tests in the data analyzed herein.
Firstly, provided that WM measures, especially the OSPAN test, have
mostly been found to correlate with lexical knowledge development, it
would be realistic and possible to find some significant interaction between
participants’ WM measures and their L2 vocabulary test results, mostly in
the immediate L2 vocabulary test. This expectation in part stems from the
fact that no treatment was provided in the pre-test; as for the post-test, a
one-week delay was adopted, so results could possibly interfere with long-
term memory.
One of the reasons underlying the lack of substantial correlations
among the WM measures and the L2 vocabulary recognition tests might be
associated with the nature of the tests regarding time and storage. On both
RST and OSPAN tests, participants had to hold some piece of information
(words) for a very brief period of time in their WM, maintaining it active,
and then had to recall that very same piece of information to verbalize it
(saying it out loud when question marks appeared on screen). Unlike the
WM tests, while watching the video in any of the experimental or control
conditions, participants might not have been consciously trying to hold any
piece of information active for later retrieval. Hence, the nature of the tasks
involved – WM and L2 vocabulary recognition – is apparently dissimilar,
particularly in terms of the processing demands required to accomplish
each of them.
Secondly, as Baddeley (2009) explains, short-term memory (STM) and
the system(s) responsible for it are part of WM. The author also explains
that STM involves the capacity to store small amounts of information for
very brief periods of time. Moreover, Schwartz and Metcalfe (1992) explain
that the STM is deemed an active memory with limited capacity to hold
information for around twenty to sixty seconds. The information received
in the STM is assumed to be stored for a short period of time while being
analyzed and interpreted. Once understood, part of it is transmitted to
the LTM, possibly for permanent storage. The old information that is no
longer needed may fade away from the STM (MAYER; MORENO, 1998).

Rev. Bras. Linguíst. Apl., Ahead of Print, 2018 23

Thus, the operations going on in the STM are thought to be indispensable
for long-lasting storage.
In order for information to be stored in the STM and later transmitted
to the LTM, it must be analyzed and possibly understood, otherwise it will be
lost or discarded. However, information will only be analyzed if such a piece
of information is, we presume, attended to. This view that links the STM to
consciousness is shared by a few scholars (BAARS, 1986; SCHMIDT, 1990).
In sum, their perception is that the STM serves as a type of “broadcasting
station” (BAARS, 1986 apud SCHMIDT, 1990, p. 137) and stresses the role
of consciousness to actuate the learning process. Although consciousness
in learning has been controversially debated in the field of Second Language
Acquisition, many cognitive psychologists, such as Rutherford and Smith
(1985); Schmidt (1990); Koskinen and colleagues (1996); and Hsiao and
Oxford (2002), mostly agree upon the need to raise learners’ consciousness
to raise their awareness of the properties of what to learn.
In the L2 vocabulary test, more specifically the immediate test,
participants watched a 20-minute video and were then given the L2
vocabulary task. It is possible to hypothesize that participants noticed some
of the target lexical items, understood them, and inferred their meaning or
translation counterparts, but by the time they were administered the test –
that is, 20 minutes later – some of those items may have no longer been
active in their STM for some reason, which could be dependent on a larger
storage and processing capacity on their part. Moreover, the simple fact that
participants may have been more concerned with understanding the story
and enjoying the comedy would suggest that their attentional resources were
not consciously allocated towards the lexical items.
Some of these lexical items, unfortunately, may not have even been
integrated into their STM in the first place, because some of them may not
have been processed, understood, and therefore registered. Furthermore, it
is also plausible that lexical items were not held active in participants’ STM
because of a decay in a memory trace due to a lack of frequency/saliency
in the input. Following a similar line of reasoning, it might be the case that,
because very few of the participants were high spanners, in general, their
capacity to hold active pieces of information about the video may have
proven insufficient for later retrieval.

24 Rev. Bras. Linguíst. Apl., Ahead of Print, 2018

 Thirdly, the redundancy produced by subtitled materials may present
some degree of influence upon consciousness and STM capacity. These
theorists argue that visual channels (subtitles and video) need a certain amount
of attention for processing because of their dynamic and graphic nature.
They claim that that the pictures that require attention can be processed with
little effort. Consequently, more attention resources or capacity should be
available to be used (D’YDEWALLE et al., 1991; GRIMES, 1991). If true,
this assumption would mean that if one understands the story narrated
on screen more easily, more attentional resources should be available to
focus on other aspects of the subtitled video. Looking at the results of
the video comprehension test and L2 vocabulary tests, it is clear that the
intralingual subtitles group outperformed the other two groups. Even
though these performances did not always achieve statistical significance
in the comparison across groups, the numbers could suggest that the
availability of subtitles – and mostly English subtitles – enabled better video
comprehension and might have made more attentional resources available to
those participants performing in either experimental conditions to examine
the linguistic content in the input.
It is possible that, from a statistical point of view, no significant
interaction was found between WM measures and L2 vocabulary because
of the low scores obtained on the L2 vocabulary recognition tests (not
reported here due to space constraints), as well as the limited sample size in
each experimental and control group. It is also possible that no statistically
significant interaction between WM measures and L2 vocabulary recognition
tests was found because of the limited number of high spanners. Finally, it is
reasonable that the WM tests and other measures used in the present study
were not able to capture the complexities involved in the actual processing
of subtitled videos, and therefore significant correlations may not have
surfaced due to the nature of the tests adopted in this study.

5 Final Remarks
This mixed design study, focusing on qualitative and quantitative
analyses, set out to investigate whether or not L2 learners’ WM capacity
could show any relationship with their performance on subtitled video
materials’ processing tasks, namely an L2 video comprehension test and an
L2 vocabulary test. Though no statistically significant results were found,

Rev. Bras. Linguíst. Apl., Ahead of Print, 2018 25

we need to bear in mind that correlation analyses cannot be interpreted
as establishing cause-and-effect relationships. In our understanding,
correlation analyses allow, with some degree of caution, to indicate the
level of association among variables. Therefore, the results obtained herein
demonstrate that the variables investigated did not interact significantly.
It may also be the case that WM could possibly be related to participants’
performance on the tests they took, to a lesser or greater degree. Different
results might be obtained with larger sample sizes and modifications in the
instruments or procedures for data collection.
Further research is still necessary to deepen our current understanding
of the impact that subtitled video materials can have on L2 learners.
Furthermore, given the expanding use of subtitled audiovisual materials in
and outside L2 classrooms and the technological advances we are currently
witnessing, we find it crucial for L2 practitioners to help learners develop
strategies to benefit from subtitled video materials, especially in relation to
less strategic L2 learners. Finally, L2 instructors should consider helping
learners to cope with the possible cognitive (over)load that these materials
may present, most of all when any textual aid (subtitles) is used. Directing
learners’ attention to specific features of the video material presented might
be a good start to achieve this end.

Acknowledgment
This study was sponsored by the Conselho Nacional de Desenvolvimento
Científico e Tecnológico (CNPq) – Brazil, granted to the first author of
this paper.

References
BAARS, B. The cognitive revolution in psychology. New York: Guilford, 1986.
BADDELEY, A. D. Working memory. Science, Washington, v. 255, p. 556-559, 1992.
Doi: https://doi.org/10.1126/science.1736359
BADDELEY, A. D. The episodic buffer: A new component of working memory?
Trends in Cognitive Sciences, [s.l.], v. 4, n. 11 p. 417-423, 2000.
BADDELEY, A. D. Short-term Memory. In: BADDELEY, A. D.; EYSENCK, M.;
ANDERSON, M. C. (Org.) Memory. Hove: Psychology Press, 2009.

26 Rev. Bras. Linguíst. Apl., Ahead of Print, 2018

BADDELEY, A. D. Estimating the short-term component in free recall. British
Journal of Psychology, [s.l.], v. 61, n. 1, p. 13-15, 2011.
BIANCHI, F.; CIABATTONI, T. Captions and subtitles in EFL learning an investigative
study in a comprehensive computer environment. Triste: Edizioni Università di Trieste, 2008.
BIRD, S.; WILLIAMS, J. The effect of bimodal input on implicit and explicit
memory: an investigation into the benefits of within-language subtitling. Applied
Psycholinguistics, Cambridge, v. 23, n. 4, p. 509-533, 2002. Doi: https://doi.
org/10.1017/S0142716402004022
CAIMI, A. Audiovisual Translation and Language Learning. The Promotion of
Intralingual Subtitles. Journal of Specialised Translation, London, v. 6, n. 1, p. 85-98,
2006.
CHANG, S. The Interaction between Schemata and Subtitles. Journal of National
Taipei University of Technology, Taipei, v. 39, n 1, p. 209-227, 2006.
CONWAY, A R.; KANE, M. J.; BUNTING, M. F.; HAMBRICK, D. Z.;
WILHELM, O.; ENGLE, R. W. Working memory span tasks: A methodological
review and user’s guide. Psychonomic Bulletin & Review, [s.l.], v. 12, n. 5, p. 769-786, 2005.
DANAN, M. Captioning and subtitling: Undervalued language learning strategies.
Meta, Montreal, v. 49, n. 1, p. 67-77, 2004. Doi: https://doi.org/10.7202/009021ar
DANEMAN, M.; CARPENTER, P. A. Individual differences in working memory
and reading. Journal of Verbal Learning and Verbal Behavior, [s.l.], v. 19, n. 4 p. 450-
466, 1980.
D’YDEWALLE, F.; GIELEN, I. Attention allocation with overlapping sound,
image, and text. In: RAYNER, K. (Org.). Eye movements and visual cognition: Scene
perception and reading. New York: Springer-Verlag, 1992. p. 415-427. Doi: https://
doi.org/10.1007/978-1-4612-2852-3_25
D’YDEWALLE, G.; PRAET, C.; VERFAILLIE, K.; VAN RENSBERGEN,
J. Watching subtitled television: Automatic reading behavior. Communication
Research, Santa Barbara, v. 18, n. 5, p. 650-666, 1991. Doi: https://doi.
org/10.1177/009365091018005005
D’YDEWALLE, G.; VAN DE POEL, M. Incidental foreign-language acquisition
by children watching subtitled television programs. Journal of Psycholinguistic Research,
[s.l.], v. 28, n. 3, p. 227-244, 1999.
DOWNING, P. E. Interactions between visual working memory and selective
attention. Psychological Science, [s.l.], v. 11, n. 6, p. 467-473, 2000.

Rev. Bras. Linguíst. Apl., Ahead of Print, 2018 27

GRIMES, T. Mild auditory-visual dissonance in television news may exceed viewer
attentional capacity. Human Communication Research, Seoul, v. 18, n. 2, p. 268-298,
1991. Doi: https://doi.org/10.1111/j.1468-2958.1991.tb00546.x
HAYATI, A.; MOHMEDI, F. The effect of films with and without subtitles on
listening comprehension of ESL learners. British Journal of Educational Technology,
Cardiff, v. 42, n 1, p.181-192, 2011. Doi: https://doi.org/10.1111/j.1467-
8535.2009.01004.x
HSIAO, T. Y.; OXFORD, R. Comparing Theories of Language Learning Strategies:
A confirmatory Factor Analysis. The Modern Language Journal, Indianopolis, v. 86, n.
3, p. 368-383, 2002. Doi: https://doi.org/10.1111/1540-4781.00155
HUANG, H.; ESKEY, D. E. The effects of closed-captioned television on the
listening comprehension of intermediate English as a second language (ESL)
students. Journal of Educational Technology Systems, New York, v. 28, n. 1, p. 75-96,
1999. Doi: https://doi.org/10.2190/RG06-LYWB-216Y-R27G
KANE, M. J.; BECKLEY, M. K.; CONWAY, A. R.; ENGLE, R. W. A controlled-
attention view of working memory capacity. Journal of Experimental Psychology:
General, [s.l.], v. 130, n. 2, p. 169-183, 2001.
KANE, M. J.; CONWAY, A. R. A.; HAMBRICK, D. Z.; ENGLE, R. W. Variation
in working-memory capacity as variation in executive attention and control. In:
CONWAY, A. R. A.; JARROLD, C.; KANE, M. J. A.; MIYAKE, J. T. (Org).
Variation in Working Memory. New York: Oxford University Press, 2007.
KLEIN, K.; FISS, W. H. The reliability and stability of the Turner and Engle
working memory task. Behavior Research Methods, Instruments, & Computers, [s.l.], v.
31, n. 3, p. 429-432, 1999.
KOOSLTRA, C. M.; BEENTJES, J. W. J. Children vocabulary acquisition in
a foreign language through watching subtitled television programs at home.
Educational Technology Research and Development, [s.l.], v. 47, n. 1, p. 51-60, 1999.
KOSKINEN, P. S.; KNABLE, J. E.; MARKHAM, P. L.; JENSEMA, C. J.; KANE,
K. W. Captioned Television and the Vocabulary Acquisition of Adult Second
Language Correctional Facility Residents. Journal of Educational Technology Systems,
New York, v. 24, n. 4, p. 359-373, 1996. Doi: https://doi.org/10.2190/JKC2-
LE5D-36BC-809F
KRUGER, J.; STEYN, F. Subtitles and Eye Tracking: Reading and Performance.
Reading Research Quarterly, Nashville, v. 49, n. 1, p. 105-120, 2014. Doi: https://doi.
org/10.1002/rrq.59

28 Rev. Bras. Linguíst. Apl., Ahead of Print, 2018

LATIFI, M.; MOBALEGH, A.; MOHAMMADI, E. Movie Subtitles and the
Improvement of Listening Comprehension Ability. Does it help? Journal of Language
Teaching and Learning, [s.l.], v. 1, n. 2, p. 18-29, 2011.
LAUFER, B. What’s in a word that makes it hard or easy? Intralexical factors
affecting the difficulty of vocabulary acquisition. In: MCCARTHY, M.; SCHMITT,
N. (Org.). Vocabulary Description, Acquisition and Pedagogy. New York: Cambridge
University Press, 1997. p. 140-155.
MARKHAM, P. L. Captioned videotapes and second-language listening word
recognition. Foreign Language Annals, Michigan, v. 32, n. 3, p. 321-328, 1999. Doi:
https://doi.org/10.1111/j.1944-9720.1999.tb01344.x
MARKHAM, P. L.; PETER, L. The influence of English language and Spanish
language captions on foreign language listening/reading comprehension. Journal of
Educational Technology Systems, New York, v. 31, n. 3, p. 331-341, 2003. Doi: https://
doi.org/10.2190/BHUH-420B-FE23-ALA0
MARKHAM, P. L.; PETER, L.; MCCARTHY, T. The effects of native language vs.
target language captions on foreign language students’ DVD video comprehension.
Foreign Language Annals, Michigan, v. 34, n. 5, p. 439-445, 2001. Doi: https://doi.
org/10.1111/j.1944-9720.2001.tb02083.x
MATIELO, R. Intralingual Subtitles, Interlingual Subtitles, and L2 Vocabulary
Learning: An Exploratory Study with Brazilian EFL Students. 2016. 272 f.
Dissertation (Ph.D. in English and Applied Linguistics) – Universidade Federal de
Santa Catarina, Florianópolis, 2016.
MATIELO, R.; COLLET, T.; D’ELY, R. C. S. F. The effects of interlingual
and intralingual subtitles on vocabulary learning by Brazilian EFL learners. An
exploratory study. Revista Intercâmbio, Montes Claros, v. 27, n. 1, p. 83-99, 2013.
MATIELO, R.; D’ELY, R. C. S. F.; BARETTA, L. The effects of interlingual and
intralingual subtitles on second language learning/acquisition: a state-of-the-art
review. Trabalhos em Linguística Aplicada, Campinas, v. 54, n. 1, p. 161-182, 2015. Doi:
https://doi.org/10.1590/0103-18134456147091
MATIELO, R.; OLIVEIRA, R. P.; BARETTA, L. Intralingual Subtitles, Interlingual
Subtitles, and Video Comprehension: Insights from an Exploratory Study. Letrônica,
Porto Alegre, v. 10, n. 2, p. 758-774, 2017. Doi: https://doi.org/10.15448/1984-
4301.2017.2.26370

Rev. Bras. Linguíst. Apl., Ahead of Print, 2018 29

MATIELO, R.; OLIVEIRA, R. P.; BARETTA, L. Intralingual Subtitles, Interlingual
Subtitles, and L2 Vocabulary: Developments from an Exploratory Study. Acta
Scientiarum, Maringá. Forthcoming. Doi: https://doi.org/10.4025/actascilangcult.
v40i1.36570
MAYER, R.; MORENO, R. A Split-Attention Effect in Multimedia Learning:
Evidence for Dual Processing Systems in Working Memory. Journal of
Educational Psychology, Washington, v. 90, n. 2, p. 312-320, 1998. Doi: https://doi.
org/10.1037/0022-0663.90.2.312
MAYER, R. E.; HEISER, J.; LONN, S. Cognitive constraints on multimedia
learning: When presenting more material results in less understanding. Journal of
Educational Psychology, Washington, v. 93, n. 1, p. 187-198, 2001. Doi: https://doi.
org/10.1037/0022-0663.93.1.187
MONTERO-PEREZ, M; PETERS, E; DESMET, P. Enhancing Vocabulary
Learning through Captioned Video. An Eye-Tracking Study. Modern Language
Journal, Indianapolis, v. 99, n. 2, p. 308-328, 2015. Doi: https://doi.org/10.1111/
modl.12215
PEREGO, E.; DEL MISSIER, F.; PORTA, M.; MOSCONI, M. The Cognitive
Effectiveness of Subtitle Processing. Media Psychology, [s.l.], v. 13, n. 3, p. 243-272,
2010.
PREBIANCA, G. V. V. Working memory capacity, lexical access and proficiency
level in L2 speech production. Dissertation (Ph.D. in English and Applied
Linguistics) – Post-Graduation in Letters/English and Corresponding Literature,
Universidade Federal de Santa Catarina – Florianópolis, 2009.
PRICE, K. Closed-Captioned TV. An Untapped Resource. MATSOL Newsletter,
[s.l.], v. 12, n. 2, p. 1-8, 1983.
RAINE, P. Incidental Increase in Depth of Vocabulary Knowledge Through the
Viewing of Subtitled, Authentic Videos. In: SONDA, N.; KRAUSE, A. (Org.).
Making a Difference. Tokyo: JALT, 2013. p. 492-505.
RUTHERFORD, W.; SMITH, M. S. Consciousness-raising and universal grammar.
Applied Linguistics, Cambridge, v. 6, n. 3, p. 274-82, 1985. Doi: https://doi.
org/10.1093/applin/6.3.274
SCHMIDT, R. The Role of Consciousness in Second Language Learning. Applied
Linguistics, Cambridge, v. 11, n. 2, p. 129-158, 1990. Doi: https://doi.org/10.1093/
applin/11.2.129
SHARIF, M. R.; EBRAHIMIAN, M. A Comparative Study of ‘Subtitled’ vs.
‘Auditory’ Documentaries Comprehension among Persian Speakers of English.

30 Rev. Bras. Linguíst. Apl., Ahead of Print, 2018

International Journal of Applied Linguistics & English Literature, Footscray, v. 2, n. 2,
p.185-192, 2013. Doi: https://doi.org/10.7575/aiac.ijalel.v.2n.2p.185
STEWART, M. A.; PERTUSA, I. Gains to language learners from viewing target
language closed-captioned films. Foreign Language Annals, Michigan, v. 37, n. 3,
p. 438-447, 2004. Doi: https://doi.org/10.1111/j.1944-9720.2004.tb02701.x
SYDORENKO, T. Modality of Input and Vocabulary Acquisition. Language
Learning & Technology, Mānoa, v. 14, n. 2, p. 50-73, 2010.
TAYLOR, G. Perceived processing strategies of students watching captioned video.
Foreign Language Annals, Michigan, v. 38, n. 3, p. 422-427, 2005. Doi: https://doi.
org/10.1111/j.1944-9720.2005.tb02228.x
TORRES, A. C. G. Working memory capacity and reader’s performance on
main idea construction in L1 and L2. Dissertation (Ph.D. in English and Applied
Linguistics) – Post-graduation in Letters/English and Corresponding Literature,
Universidade Federal de Santa Catarina – Florianópolis, 2003.
TURNER, M. L.; ENGLE, R. W. Is working memory capacity task dependent?
Journal of Memory and Language, New York, v. 28, n. 2, p. 127-154, 1989. Doi: https://
doi.org/10.1016/0749-596X(89)90040-5
UNSWORTH, N.; HEITZ, R. P.; SCHROCK, J. C.; ENGLE, R. W. An automated
version of the operation span task. Behavior Research Methods, [s.l.], v. 37, n. 3,
p. 498-505, 2005.
VAN LOMMEL, S.; LAENEN, A.; D’YDEWALLE, G. Foreign grammar
acquisition while watching subtitled television programs. British Journal of
Educational Psychology, London, v. 76, n. 1, p. 243-258, 2006. Doi: https://doi.
org/10.1348/000709905X38946
VANDERPLANK, R. The value of teletext subtitles in language learning. ELT
Journal, Oxford, v. 42, n. 4, p. 272-281, 1988. Doi: https://doi.org/10.1093/
elt/42.4.272
VANDERPLANK, R. Captioned Media in Foreign Language Learning and Teaching.
Subtitles for the Deaf and Hard-of-Hearing as Tools for Language Learning. London:
Palgrave Macmillan, 2015.
WINKE, P.; GASS, S.; SYDORENKO, T. The effects of captioning videos used
for foreign language listening activities. Language Learning & Technology, Mānoa, v.
14, n. 1, p. 65-86, 2010.
WINKE, P.; GASS, S.; SYDORENKO, T. Factors Influencing the Use of Captions
by Foreign Language Learners. An Eye-Tracking Study. Modern Language Journal,

Rev. Bras. Linguíst. Apl., Ahead of Print, 2018 31

Indianapolis, v. 97, n. 1, p. 254-275, 2013. Doi: https://doi.org/10.1111/j.1540-
4781.2013.01432.x
ZAREI, A. A.; GILANIAN, M. The effect of multimedia modes on L2 vocabulary
learning. International Journal of Management and Humanity Sciences, [s.l.], v. 2, n. 1,
p. 1011-1020, 2013.
ZAREI, A. A.; RASHVAND, Z. The Effect of Interlingual and Intralingual,
Verbatim and Nonverbatim Subtitles on L2 Vocabulary Comprehension and
Production. Journal of Language Teaching and Research, [s.l.], v. 2, n. 3, p. 618-625, 2011.

Data de submissão: 29/11/2018. Data de aprovação: 03/05/2018.

32 Rev. Bras. Linguíst. Apl., Ahead of Print, 2018