928577

research-article2020
LTR0010.1177/1362168820928577Language Teaching ResearchBenati

LANGUAGE
TEACHING
Article RESEARCH

Language Teaching Research

The effects of structured input

1­–21
© The Author(s) 2020
Article reuse guidelines:
and traditional instruction on sagepub.com/journals-permissions
DOI: 10.1177/1362168820928577
https://doi.org/10.1177/1362168820928577
the acquisition of the English journals.sagepub.com/home/ltr

causative passive forms: An

eye-tracking study measuring
accuracy in responses and
processing patterns

Alessandro Benati 
The University of Hong Kong, Hong Kong

Abstract
The present study explores the effects of structured input and traditional instruction on the
acquisition of English causative passive forms using online measurements (eye-tracking). Previous
empirical research investigating the effects of processing instruction through offline measurements
(sentence and discourse) has overall shown positive results for this pedagogical intervention.
Research investigating the main factor responsible for the effectiveness of processing instruction
has confirmed that it is the structured input component that is the causative factor for the positive
effects of processing instruction. The main questions of this study are: (1) what are the effects of
structured input and traditional instruction on accuracy when measured by an eye-tracking picture
selection task? (2) would possible difference in accuracy between structured input and traditional
instruction be accompanied by changes in eye-movement patterns? To provide answers to the
two questions formulated in this study, one eye-tracking study was carried out. Fifty-two adult
learners (aged 19–21 years) participated and were assigned to one of two groups: structured
input (n = 26) or traditional instruction (n = 26). Neither instructional groups received explicit
information. A pre and post-training design was adopted and the two groups received two different
instructional treatments (structured input vs. traditional instruction). Participants were assessed
through a picture selection eye-tracking task to measure accuracy and eye-movement patterns
while they were processing auditory sentences. Results of the eye-tracking task indicated that
the structured input group achieved significantly higher accuracy scores compared to the group
receiving traditional instruction. The main findings from the present study reveal that structured
input training causes a change in learners’ eye-movement patterns.

Corresponding author:
Alessandro Benati, Centre for Applied English Studies, University of Hong Kong, 6th Floor, Run Run Shaw
Tower, Centennial Campus, Pokfulam, Hong Kong.
Email: [email protected]
2 Language Teaching Research 00(0)

Keywords
Accuracy, English causative passive forms, eye movement patterns, input processing, processing
instruction, structured input

I Background
1 Introduction
The input processing theory (VanPatten, 1996, 2004, 2015a) is concerned with how sec-
ond language (L2) learners initially perceive and process input. Some of the key questions
addressed by VanPatten in his theory are: (1) why is the input processed by L2 learners
reduced (intake)? And (2) why would L2 learners process some elements in the input and
not others? This theory consists of two main principles addressing different aspects of
processing: The Primacy of Meaning Principle and The First Noun Principle. According
to The Primacy of Meaning Principle, L2 learners are driven to get meaning first while
processing language input. In the case of verbal inflection, for example, the English verbal
inflection -ed encodes past as in talked. The same semantic notion is also expressed in
English by words such as yesterday or last year. Given that L2 learners are driven to pro-
cess content words before anything else in a sentence, they would attend to lexical tempo-
ral references of ‘pastness’ before verbal inflections of past tense in English and fail to
make an appropriate form–meaning connection. Form–meaning connections are the rela-
tionship L2 learners make between referential meaning and the way it is encoded linguis-
tically. For example, when L2 learners hear the sentence I talked to my teacher and
understand that talked means the action is in the past, a form–meaning connection is
made. In the sentence Yesterday, I played tennis with John in the park, in the attempt to
make moment- by-moment connection of surface form with meaning, L2 learners need to
tag played with the fact that it is a verb (<+V>, <−N>), that its meaning refers to play-
ing a sport, that it is past tense and not present (<+present> <−past>), and so on.
According to the First Noun Principle, L2 learners, in the attempt to make moment-
by-moment computation of sentence structure during comprehension, would process the
first element they encounter in the sentence as the subject of the sentence. In the case of
a sentence such as Alessandro was kissed by Bernadette, L2 learners would misinterpret
the sentence as if it were ‘Alessandro who kissed Bernadette’. The meaning of the sen-
tence is actually that ‘Bernadette kissed Alessandro’. In figuring out who did what to
whom, L2 learners rely on word order. By assigning the role of subject to the first ele-
ment they encounter in the sentence, L2 learners would (in this case) misinterpret the
meaning of this sentence causing delay in acquisition.
Processing instruction is a pedagogical intervention derived from the input processing
theory, and it aims at facilitating the cognitive processes by which L2 learners make
form–meaning connections and/or compute sentence structure during comprehension.
For example, L2 learners prefer to process first lexical items before grammatical items
when both items encode the same semantic information.
Processing instruction has been described in detail in previous work (Benati, 2013b,
2017, forthcoming; Farley, 2005; Lee & VanPatten, 2003; VanPatten, 2015b; Wong,
Benati 3

2004). The main characteristic of processing instruction is that it makes use of a par-
ticular type of input practice (structured input) to push learners away from non-optimal
processing principles so that they are more likely to process input correctly and appro-
priately. Processing instruction is mainly concerned with the processing of morpho-
phonological units in input strings and the development of underlying linguistic
representation. Most simply put, a main objective of processing instruction is to ensure
that L2 learners process forms and structures (one at a time) accurately and efficiently
in the input they receive.
Processing instruction consists of two main components (Benati & Lee, 2008): (1) L2
learners are given explicit information about a linguistic structure or form and the par-
ticular processing principle that may negatively affect their picking up of the form or
structure during language processing; and (2) L2 learners are pushed to process (not
produce) the target form or structure during structured input activities. In structured input
activities, the input is structured so that L2 learners can process the grammatical markers
that otherwise would not be processed and at the same time make correct form–meaning
connections.

2 Research measuring processing instruction versus traditional instruction

VanPatten and Cadierno (1993) published the first investigation on processing instruc-
tion and set the stage for carrying out a processing instruction study. They wanted to
demonstrate that processing instruction was effective type by comparing its effects to
those of traditional instruction on the acquisition of Spanish direct object pronouns.
Spanish direct object pronouns occur in Objectpro–Verb–Subject (OproVS) sentence
patterns.
The processing problem for L2 learners is that they tend to identify the object pronoun
as the agent and the subject in post-verbal position as the patient (The First Noun
Principle). The processing group received explicit information about the target feature
(info about the processing problem) which was followed by structured input practice. In
the traditional instruction treatment, L2 learners were required to produce the target form
in each practice item. VanPatten and Cadierno (1993) designed two types of assessments:
one to measure interpretation of the target form; and the other to measure form produc-
tion. Only the processing instruction group showed significant improvement on the inter-
pretation task. Both the processing instruction and traditional groups showed significant
improvement on the form production task. VanPatten and Cadierno (1993) found no
change in post-test scores one month after instruction. This double interpretation/produc-
tion effect for processing instruction has been supported by other research comparing
processing instruction and traditional instruction (Benati, 2001, 2013a; Cadierno, 1995;
Cheng, 2004; Lee & Benati, 2007b; Marsden, 2006; VanPatten & Wong, 2004) on the
acquisition of forms/structures affected by processing principles such as the First Noun
Principle and The Lexical Preference Principle (Learners process lexical items before
grammatical items when they both encode the same semantic information). Overall, the
main research findings, have shown that processing instruction is better at improving
learners’ rate of processing and at increasing learners’ accuracy in production than tradi-
tional instruction.
4 Language Teaching Research 00(0)

3 Research measuring the effects of structured input practice

The effects of structured input practice through structured input activities have been
examined following two main lines of enquiries: (1) measuring its effects as the possible
causative factor of processing instruction in changing language processing behaviors;
and (2) comparing structured input practice versus structured-output practice.
This line of research, within the processing instruction research framework, is the one
which has investigated what the effects are of the components of processing instruction.
The main purpose of this classroom-based research was to establish which factor is the
most effective in processing instruction. Is it the explicit information component or the
structured input practice component? Or perhaps a combination of the two?
The original study, which attempted to address these questions, was conducted by
VanPatten and Oikkenon (1996). The item investigated was the object pronouns in
Spanish and the processing principle under investigation was the First Noun Principle.
Three groups were compared: one receiving only explicit information about the target
feature; the other structured input activities; and the third both components (full process-
ing instruction). The performance of the three groups was measured on the interpretation
and production tasks. The outcome of this study was that structured input activities were
found as being responsible for L2 learners gains. A significant finding was that the struc-
tured input activities group performed as well as the processing instruction group. These
findings strongly suggested that it was the structured input activities themselves and the
form–meaning connections being made during input processing that were responsible
for the relative effects observed in this and previous studies. Research within this frame-
work has repeatedly demonstrated (Benati & Lee, 2015; Lee & Benati, 2009) that the
structured input component is the causative variable responsible for the way L2 learners
process input and can eventually access for production.
These positive effects for structured input have been confirmed and observed in differ-
ent processing principles, languages, linguistic items and assessment tasks (Benati, 2004a,
2004b; Farley, 2004; Lee & Benati, 2007a; Sanz, 2004; Wong, 2004). Lee and Benati
(2007b) compared the performance of L2 learners who received enhanced structured
input, and unenhanced structured input on the acquisition of Italian future tense. All in all,
enhanced and unenhanced structured input were equally effective in helping L2 learners
process verbal morphology. The results of this study further indicated that structured input
practice represents the most significant variable within processing instruction.
Research comparing structured input versus structured output practice (structured
output and traditional instruction) have demonstrated the superiority of structured input
training. Benati and Batziou (2017, 2019) have explored the discourse and long-term
effects of structured input and structured output practice when delivered in isolation or
in combination on the acquisition of the English causative forms. The results of both
studies were similar and showed that L2 learners who received the structured input
instructional training benefitted more than L2 learners receiving structured output
instructional training. In both studies, structured input practice alone was sufficient to
improve learner’s performance on both interpretation and production discourse tests
containing English causative forms. One possible explanation of these results might be
found in the nature and purpose of structured input activities. As outlined by Wong
Benati 5

(2004, p. 35), structured input practice ‘push[es] learners to abandon their inefficient
processing strategies for more optimal ones’.
The main finding of both lines of research, teasing out the main factors responsible for
the effectiveness of processing instruction and measuring the effects of structured input
activities compared to structured output, confirmed that it is the structured input compo-
nent that alone is responsible for learners’ improved performance. The positive effects of
structured input practice can be observed in different processing principles, languages,
and linguistic items and can be measured by offline assessment tasks (sentence and dis-
course-level tasks). No empirical research has been conducted so far to investigate online
effects of structured input practice.

4 Research measuring the effects of processing instruction/structured

input practice using eye-tracking tasks
The bulk of the evidence in support of the beneficial effects of processing instruction has
come largely from offline tasks. The use of online measurements to investigate the
effects of processing might offer us the possibility of more fine-grained information and
analysis about moment-by-moment sentence comprehension. More recently, eye track-
ing has been used in processing instruction research to measure the location and duration
of eye gazes while speech unfolds.
Studies within this research branch have focused on the First Noun Principle and
investigated (1) the allocation of attention (passive vs. active sentences) before and after
receiving processing instruction and (2) whether processing instruction reduce learners’
reliance on the First Noun Principle during online comprehension of passive sentences.
Wong and Ito (2018) compared changes in processing patterns between L2 learners
receiving processing instruction and traditional instruction on the acquisition of the
French causative. In this pre-and post-test experimental study, a dichotomous scene
selection eye-tracking task was used to measure eye movement patterns and accuracy in
picture selection while learners were processing auditory sentences. The results from this
study indicated that the processing group gained higher scores for accuracy than the
traditional instruction group. A change in eye movement was also observed in learners
after the processing instruction training, but not after the traditional instruction training.
In a pair of follow-up experiments, Ito and Wong (2019) tested whether consistency
in input modality (i.e. auditory PI training prior to the visual world task) might lead to an
earlier shift in looks to the correct picture. They also manipulated voice familiarity (same
vs. different) across training and online task. Surprisingly, the auditory processing
instruction training led to a longer delay in the use of online cues to sentence structure
during spoken language comprehension in that looks to the correct picture did not
increase until after sentence offset, regardless of voice familiarity.
Lee and Doherty (2019) compared native and nonnative processing of active and pas-
sive sentences in Spanish. Accuracy and response time were measured in a paired picture
matching task. After receiving processing instruction, the nonnative speakers showed no
significant difference from the native speakers in accuracy and response time. The
behavior of the nonnative participants became more native-like after exposure to pro-
cessing instruction.
6 Language Teaching Research 00(0)

Issa and Morgan-Short (2019) examined online effects of textual enhancement and
structured input practice on the acquisition of Spanish direct object pronouns. Results for
both treatments indicated the following: (1) learner attentional allocation to the form was
affected; (2) L2 gains were evidenced, although only the internal manipulation led to
above-chance performance; and (3) L2 gains were related to attention allocated to the
form under the external manipulation and to a lesser extent the internal manipulation.

5 Eye-tracking and language processing

Eye-tracking is an online measure used in psycholinguistic research to monitor, examine
and record learners’ visual attention, visual search and language processing during spo-
ken or reading language processing. Eye-tracking is a methodological procedure used to
inform researchers with regard to eye movement behaviors. For example, it records
where and for how long a participant looks at an element in a sentence and/or where his/
her eyes move next. Eye-tracking gives researchers the opportunity to examine the
moment-by-moment comprehension processes in a more natural way than, for instance,
self-paced reading. Furthermore, it provides and records a more fine-grained reading
profile of the different processing stages in reading: the so-called ‘first fixations’ which
is the first time the eyes fixate on the region of interest (e.g. a particular word, sentence
segments, etc.); the ‘first-pass’ times which sums up the time spent reading the region of
interest from the first fixation until the eyes exit to the right or to the left; and how often
the word was returned to for re-reading (regressions).
The present study adopts a type of eye-tracking technique called visual world para-
digm (Trueswell and Tanenhaus, 2005). This technique records the fixations that a reader
makes on objects or scenes displayed on a computer while listening to a recorded sen-
tence (speech input) about them. The images on display usually consist of black-and-
white or colored line drawings of objects or semi-realistic scenes. The presentation of the
images and the auditory input is carefully synchronized. The visual word paradigm has
been adopted by scholars such as Altmann and Kamide (1999) to measure incremental
online syntactic processing. This work has demonstrated that L2 learners integrate the
grammatical role and meaning of an incoming word as soon as sufficient sound informa-
tion of the word unfolds in speech.
The same paradigm used by Wong and Ito (2018) was used in the present study. It
consists of two nearly identical scenes which differ only in the thematical roles of two
people. In the present study, participants received visual targets (two pictures, one show-
ing Jack pouring his coffee and the other showing Jenny pouring the coffee for Jack)
while they listened to speech input. The two pictures differed only in who was doing the
action. Participants were simply asked to choose the picture that matched the sentence
they heard. L2 learners of English viewed the two pictures while they listened to the
causative structure (‘Jack had his coffee poured by Jenny’), which tends to be incorrectly
interpreted with the First Noun Principle. L2 learners have the universal tendency to
assign the role of agent to the first element they encounter in a sentence. This strategy is
effective when learners are processing SVO order sentences but it will lead to incorrect
interpretations when the first noun is not the agent like in the case of English causative
forms. Previous research (Stromswold et al., 2002) using this picture-match paradigm,
Benati 7

has demonstrated that adults swiftly shift their gaze to the correct picture as soon as they
hear the main verb of the passive form (e.g. The boy was kicked) while children do not
look at the correct picture until they hear the prepositional phrase that specified the agent
(by the boy).
In the present study, L2 learners of English viewed two scenes while they listened to
the causative structure which tends to be incorrectly interpreted with the First Noun
Principle (VanPatten and Wong, 2004). Considering that there is robust evidence
(VanPatten, 2015a) that has demonstrated the use of The First Noun Principle by L2
learners when processing syntactic structures such as passive constructions and causa-
tive forms, it is predicted initial incorrect looks by participants in this study to the picture
where the first noun plays the agent role (‘Jack . . .’ will send listeners’ eyes to the pic-
ture where the man is pouring the coffee) before they receive any instructional treatment.
The main question of this study is whether different pedagogical interventions (struc-
tured input vs. drill-based traditional instruction) lead to differences in online processing
of English causative forms. Previous research (Benati and Batziou, 2017, 2019) have
demonstrated that processing instruction training (and in particular structured input prac-
tice) is effective in ensuring correct interpretations of sentences and discourse containing
causative forms.

II Motivation and research questions of the present study

A large body of research has compared the effects of structured input practice. This
work has affirmed the importance of structured input activities (Lee, 2004, 2015; Benati,
2017) as the main component of processing instruction. Previous research also suggests
that participants who are trained with processing instruction are more successful on
sentence interpretation tasks than participants who are trained with traditional instruc-
tion. Recent eye-tracking research has also demonstrated the effectiveness of process-
ing instruction (Lee & Doherty, 2019; Wong & Ito, 2018) through online measurements.
Collectively, this small body of work yields promising findings, however more research
is needed within the processing instruction research paradigm to measure whether
structured input practice leads to qualitative changes in online processing strategies
between pre- and post-test. More specifically, whether L2 learners receiving structured
input practice improve their accuracy in language processing and whether these effects
are accompanied by changes in eye-movement patterns. The use of online measure-
ments in this field of enquiry could provide, in addition to the large database of research
available using offline pre and post-test designs to measure the relative effects of pro-
cessing instruction, more refined information about moment-by-moment language pro-
cessing and comprehension.
The aim of the present study is twofold: (1) to compare the effects of structured input
and drill-based traditional instruction by measuring how these instructional treatments
can affect accuracy in sentence processing; and (2) to measure possible changes in eye-
movement patterns caused by the instructional training treatments. Based on the main
aims of this study and the specific intent to compare structured input and traditional
instruction on the acquisition of English causative passive forms, two specific questions
were formulated:
8 Language Teaching Research 00(0)

•• Question 1: What are the effects of structured input and traditional instruction on
accuracy when measured by an eye-tracking picture selection task?
•• Question 2: Would possible difference in accuracy between structured input and
traditional instruction be accompanied by changes in eye-movement patterns?

Based on the results of previous research (offline) measuring the effects of structured
input and recent eye-tracking studies measuring the effects of processing instruction
and structured input practice, it is predicted that those who receive structured input
activities will show reductions in initial looks to the incorrect picture (due to a faster
switch to the correct picture) to a larger degree than those who receive drill-based tradi-
tional instruction activities. As a result of these predictions, the following two hypoth-
eses are formulated:

•• Hypothesis 1: L2 learners receiving the structured input training will achieve sig-
nificantly higher accuracy scores compared to the group receiving traditional
instruction.
•• Hypothesis 2: The structured input training will cause a change in eye-movement
patterns.

III Design
1 Participants
The population of this study consisted of 52 adult native speakers of Chinese (aged 19–21
years) who were enrolled in an early intermediate English course in the United Kingdom
during the summer period. By drawing names out of a box, participants were randomly
assigned to two groups: structured input (n = 26) and traditional instruction (n = 26).The
criteria for inclusion of subjects in this study were (1) no previous knowledge of the target
form, (2) no history of hearing or visual impairments, (3) and no language disability. All
the voluntary participants completed an informed consent form and were give a small
financial incentive. The background questionnaire revealed that the population was quite
a homogeneous one in terms of their experience and knowledge of English.

2 Target grammar feature

Because it is affected by the First Noun Principle, the target grammar feature selected
for this study was the passive English causative forms (VanPatten, 1996, 2004, 2015a).
According to this principle, L2 learners tend to assign agent status to the first noun or
pronoun they encounter in a sentence. For example, in the sentence Lina had her dress
mended, L2 learners would process Lina as the person who actually mended the dress.
The use of this default processing principle by L2 learners would cause misinterpreta-
tion and delay in the acquisition of the target feature (word order pattern). The main
goal of the structured input practice in this study is to aid L2 learners in processing
English causative correctly and appropriately. Previous research within the processing
instruction framework has provided evidence for the positive effects of structured
Benati 9

Pre-training Pre-training Task = Eye-tracking selection picture

task (36 items)

Instructional training period Structured input (SI) via computer

(75 minutes) Traditional instruction (TI) via computer
NO Explicit info for both groups
Minimal amount of feedback

Post-training Post-training Task = Eye-tracking selection picture

task (36 items)

Figure 1.  Overview of the study.

input activities in circumventing the First Noun Principle and helping learners to cor-
rectly interpret and produce sentences containing the target feature (VanPatten &
Cadierno,1993; VanPatten & Oikkenon, 1996; VanPatten & Wong, 2004; Morgan-
Short & Bowden, 2006).

3 Procedure
A Tobii Pro X2-120 Eye tracker was used to capture accuracy in comprehension and
eye tracking movements. The device has a temporal resolution of 120 Hz and binocu-
lar accuracy of 0.4 degree. To address the two main questions of this study, a pre and
post-training design was adopted (see Figure 1). Before the pre and post-training ses-
sions, participant’s eyes were calibrated. In this eye-tracking experiment participants
received an eye-tracking picture selection task in the pre-training phase. Participants
responded to 36 items consisting of 12 causative forms (target items), 12 non-causa-
tive forms and 12 fillers. The two pictures for each item in the picture selection task
were equidistant from the center on the screen. Tobii Pro Studio was used to create
and display all the stimuli. For each two pictures, participants had two seconds to
respond and choose the accurate picture after one sentence was played through a
headphone set. In each trial, a pair of scenes appeared on the screen. After 1.5 sec-
onds, a sentence describing one of the two scenes was played through a headphone
set. Participants were told to listen to the entire sentence before clicking on the picture
that best matched the sentence they heard and move to the next pair. After the pre-
training, participants moved to a different laboratory where they worked on the com-
puter to complete their instructional treatment (ninety target items) through an online
program. After the two 60-minute instructional training sessions (structured input and
traditional instruction), all participants went back to the eye-tracking computer for the
post-training session which included another 36 items. There was no time pressure
during any phase of the study.
10 Language Teaching Research 00(0)

Jane had her dress mended last Monday (stimuli)

Who mended the dress?

A) Jane B) someone else

Correct answer: Jane

Figure 2.  Sample of structured input training.

4 Instructional training
Two online instructional training treatments (structured input and traditional instruction)
were developed for this eye-tracking experiment. The material in both trainings was bal-
anced in terms of number of activities and number of target features. High frequency
vocabulary items and familiar words were used. The only difference between the two
instructional treatments was that one group received simultaneous-interpretation (SI)
practice in interpreting sentence structure while the other group created causative sen-
tences by following a model (as is typical with traditional drill practice).

a  The structured input training.  As previously said, structured input is the practice com-
ponent of processing instruction. During structured input activities, L2 learners are
pushed to process the targeted form through activities in which the input is manipulated
in particular ways to push learners to become dependent on the form to process meaning.
Structured input activities created for this study were strictly referential. The main role
of referential activities is to push L2 learners to pay attention to form and make decisions
(for example like or wrong, present or past, likely or unlikely).
The material for the experiment (i.e. the structured input treatment) was developed
according to the following guidelines provided by Lee and VanPatten (2003) and Farley
(2005) for developing structured input activities: (1) present one target feature at a time;
(2) keep meaning in focus; (3) move from sentence to discourse; (4) use both written and
oral input; (5) have learners do something with the input; and (6) keep learners’ process-
ing strategies in mind. The six referential activities (aural and written input) were struc-
tured in a way so that L2 learners relied on the causative form to understand correctly
meaning in the input. They were developed to aid L2 learners to circumvent the First
Noun Principle by manipulating word order and contrasting passive (English) causative
forms to SVO order active forms where the first noun was the causer/agent of the action.
All the activities were communicative and meaningful and L2 learners were asked to
interpret input correctly. No activities were included where learners had to produce the
target form. The structured input training developed for this experiment contained a total
of 42 target items (see sample below). L2 learners were exposed to target items and were
asked to respond (interpret sentences containing the target form) to the stimuli by press-
ing a button on a button box. Once either A or B was chosen, the display showed the
correct answer in blue to provide feedback. No other information was provided and
subjects moved to the next sentence by pressing the rightmost button (see Figure 2).
Benati 11

Model
Jane had the car washed

Paul/his dress/mended

Paul had his dress mended

Correct answer: Paul had his dress mended

Figure 3.  Sample of traditional instruction training.

b  The traditional instruction training. The traditional instruction training consisted of

drill-type mechanical activities (42 target items in total). Each prompt appear in the
screen individually. Participants were asked to create sentences with English causative
forms. After typing the sentence with the form in a box, they could press the space bar to
see if the answer was correct. No explicit information or additional information was
provided. They could then press the space bar to move to the next item. The only differ-
ence between the structured input and traditional instruction practice was that the struc-
tured input group received training in interpreting causative forms while the traditional
instructional group produced sentences containing the target forms as in traditional drill
practice. All the activities used in this treatment were constructed to make learners pro-
duce the causative passive forms at sentence level (see Figure 3).

5 Assessment task
To test online interpretation of English causative forms, an eye-tracking picture selection
task was carried out. Thirty-six paired pictures were presented in the causative and non-
causative conditions, 12 as target items in the causative passive forms (The manager had
the report written by the assistant) and 12 non-causative sentences (The manager wrote
the report). Twelve fillers were also used (six passive and six active sentences). The
items were randomly arranged so that no causative passive forms appeared contiguously.
The passive causative forms consisted of five words (see Section IV). L2 learners of
English viewed two pictures while they listened to the causative forms which tends to be
incorrectly interpreted with the First Noun Principle (VanPatten & Wong, 2004). The two
pictures were displayed on the screen and after a few seconds a sentence describing one
of the two pictures was played through headphones. Participants were simply asked to
click once on the picture that matched the sentence they heard and move to the next set
of pictures. This task was developed to ascertain whether different types of pedagogical
interventions (structured input vs. traditional instruction) lead to differences in online
processing of the English causative forms. A native speaker recorded all the experimen-
tal sentences. Previous research on processing instruction suggests that participants who
are trained with this pedagogical intervention are more successful on sentence interpreta-
tion tasks than participants who are trained with traditional instruction (Benati, 2005,
2013b; Lee & Benati, 2007a; VanPatten, Farmer & Clardy, 2009). Figure 4. provides an
12 Language Teaching Research 00(0)

Figure 4.  Picture selection.

example. Learners would hear a sentence once, simultaneously see two different charac-
ters doing the same action, and then decide who is performing the action.

IV Results
Data from all participants were included in the analysis. The use of a visual world para-
digm was because of its ecological validity and because it allows for the examination of
auditory comprehension. A mixed-effects logistic regression model (using a SPSS pack-
age) was used to measure the effects of the two instructional training types (structured
input vs. traditional instruction/drill practice), the time (before training and after training
effects) and their interaction as predictor factors. This model would test possible effects
instruction and time on participant’s click responses.

1 Accuracy
The analysis of the effects of time and training type on mouse click responses indicated
that the accuracy of picture selection for both the structured input (8%) and the traditional
Benati 13

Figure 5.  Traditional instruction: Fixation patterns while listening (after training).

instruction (7%) groups was very low before the beginning of the instructional training
with no significant statistical difference between the two groups (z = .34, p = .76). After
receiving the instructional training, the structured input group improved to 76% in accu-
racy picture selection while the traditional group did not improve at all. The results of the
statistical regression model clearly indicated that there was an effect for time (z = 7.20,
p < .000), instructional training (z = 535, p < .000) and a significant interaction between
the two (z = 4.50, p < .000). The effect of Time (pre and post training) was only signifi-
cant for the structured input training (z = 7.85, p < .000) and not for the traditional
instruction group (z = 4.65, p < .82).

2 Eye-movement patterns
The fact that the structured input training had positive effects on the accuracy of L2
learners to select the correct picture might signify that this instructional training had an
effect on learners’ eye-movements (before they click the mouse to choose the right pic-
ture). It is expected that in the pre-test, the First Noun Principle would directly affect
eye-movements behavior (fixation towards the incorrect picture). Post-training should
show a change in eye-movement patterns (between correct and incorrect picture) at least
in the case of the structured input instructional training. The main question here is
whether the fixations patterns would capture the initial bias towards the picture congru-
ent with the First Noun Principle (agent interpretation) and whether there would be a
reduction after the training.
Participants in this study viewed either the correct or the incorrect picture for each
item at any point in time. The changes in gazes were expressed in logit, or the log of the
ratio between the numbers of fixations on the two areas of interest. The logit functions
were plotted with the baseline of zero in the middle of y-axis such that the relatively
higher numbers of fixations on the correct picture would be shown above the midline
whereas the relatively higher number of incorrect fixations would be shown below the
midline (see Figure 5 and Figure 6). The expectation is that, the stronger the First Noun
Principle bias is, the larger downward deviation of the function from the midline we
should observe. The fixation logits were aligned at the utterance onset, and at the onsets
of critical windows according to the duration information of each auditory stimulus. The
structured input and traditional instruction groups’ fixation patterns were captured while
14 Language Teaching Research 00(0)

Figure 6.  Structured input: Fixation patterns while listening (after training).

they listened to the target causative structure in the before-training session and in the
after-training session. In the pre-training session, both groups’ fixation patterns indicated
a gradual increase in incorrect fixations (while participants were listening to the target
forms) to the first noun as the agent throughout the causative sentence. The after-training
session, however, showed a contrast between the two instructional training groups (for
fixation patterns, see Figure 5 and Figure 6). The traditional instruction group showed a
similar look to the incorrect pictures to the one showed in the before-training session.
The structured input treatment group instead, showed a decrease in the incorrect fixa-
tions towards the end of the sentences heard. Participants in the structured input treat-
ment group started to shift their gazes to process the sentence correctly. The turning point
in the fixation logits appeared at the beginning of the prepositional phrase.
In the statistical analysis, the empirical logit was calculated for each time window
(W1: Jack; W2: had; W3: his coffee; W4: poured; W5: by Jenny) of each trial. A set of
mixed-effects models tested the effects of training type, pre-/post-test/time, and the inter-
action between the two on the elogits for each window (the training type and pre-/post-
test/time were both sum coded). The p-value of each effect was tested through comparison
of the models that were minimally different with the presence or absence of the factor in
question. The results of this analysis are presented in Table 1.
The constantly significant intercept across the five windows indicates an overall bias
toward the incorrect picture throughout the sentences regardless of the training type and
time. For the last window (by Jenny), a significant effect of training and an interaction
between training type and time were confirmed, indicating that the structured input
group made more fixations to the correct pictures than the traditional instruction group.
Overall the structured input group made a statistically significant increase in their gaze
direction toward the correct pictures across time (pre and post-test) compared to the tra-
ditional instruction group. The effect of time (pre-/post-test) was found in the structured
input group (z = 1.25, p = .02) but not in the traditional instruction group who showed
no effects (z = 1.28, p = .31).

3 Summary of findings
The first research question was: What are the effects of structured input and traditional
instruction on accuracy when measured by an eye-tracking picture selection task? The
 Benati

Table 1.  Summary of mixed effect regression on the fixation logits.

Predictor W1 W2 W3 W4 W5

  Est. SE t Est. SE t Est. SE T Est. SE t Est. SE t

Intercept 0.32 0.57 −1.98** −0.56 0.12 −4.76** −0.36 0.06 −4.32** −0.34 0.05 −4.54** −0.23 0.06 −564***
Training 0.35 0.12 1.48 0.24 0.23 0.76 0.02 0.18 0.18 −0.07 0.13 −0.54 0.30 0.12 2.3*
Time 0.003 0.13 0.04 −0.25 0.18 −1.04 −0.32 0.17 −1.13 0.04 0.12 0.41 0.01 0.12 0.04
Training: Time 0.09 0.26 0.26 −0.29 0.36 −0.78 0.35 0.34 0.56 0.15 0.23 0.78 0.45 0.25 1.78

Notes. ***p < .002, **p < .01, *p < .04.

15
16 Language Teaching Research 00(0)

results from this eye-tracking study confirmed previous research findings using off-line
measurements that structured input training improves the processing and interpretation
of sentences containing English causative forms. The structured input group made almost
70% improvement from pre to post-test. The first hypothesis in this study is fully con-
firmed. L2 learners receiving the structured input training do achieve significantly higher
accuracy scores compared to the group receiving traditional instruction.
The second research question was: Would possible difference in accuracy between
structured input and traditional instruction be accompanied by changes in eye-movement
patterns? The main findings from this eye-tracking study indicate a change in eye-move-
ment patterns. Structured input practice changed the way that this group processed the
target structure but the traditional instruction training did not lead to any processing
changes for this group. The structured input group made a statistically significant increase
in their gaze direction toward the correct pictures across time (pre and post-test) com-
pared to the traditional instruction group. The first hypothesis in this study is fully con-
firmed. The structured input training does cause a change in eye-movement patterns.
The First Noun Principle bias in eye movement was reduced after training for the
structured input group participants only. A closer examination of the gaze patterns sug-
gests that while the structured input training may have changed participants’ processing
mechanism for the target structure, the traditional/drill practice activities did not lead to
any changes in processing. The structured input group show great improvement in accu-
racy of picture selection and change in their eye-fixation patterns after training. Structured
input group started to shift their gaze to the correct picture.

V Discussion and conclusions

1 Discussion of the findings
The use of the First Noun Principle by all participants in the two groups is confirmed
both in terms of eye-movement patterns and picture selection. Before training, partici-
pants in both instructional groups increasingly gazed at the incorrect picture throughout
the causative sentence, leading to only around 7% to 8% accurate picture selection. As
stated in the First Noun Principle, L2 learners tend to assign the role of agent to the first
animate noun in dichotomous picture selections.
The results of this eye-tracking empirical study confirm the previously reported
effects of structured input practice (see a summary of this research in Benati, 2017; Lee,
2015). The structured input activities in this training directed participants’ attention to
‘who did what’ in each sentence significantly improving the processing and comprehen-
sion of the target causative forms as compared to the traditional instruction training that
forced mechanical production of causative sentences. The effect of structured input on
the accuracy of picture selection was significantly greater than the traditional instruction
training and it was effective without any explicit instruction.
The results of this study also indicated that the reliance to the First Noun Principle in
eye movement was reduced after instruction. However, this is only the case for the struc-
tured input group. A closer examination of the gaze patterns further suggests that while
the structured input activities may have changed participants’ processing mechanism for
Benati 17

the target structure, the traditional instruction activities did not lead to any changes in
processing for this group. There was no visible change in the eye-fixation patterns after
training for the traditional instruction group. In contrast, the structured input group
showed the turning point in the fixation logit function at the onset of the final preposi-
tional phrase (‘by Jenny’).
The overall findings from this study support the view that structured input practice is
successful in altering the way learners process input and provides evidence of the effec-
tiveness of this treatment through online measures. Structured input practice changed the
way that L2 learners processed the target form (altering the reliance on the First Noun
Principle) but the traditional instruction training did not lead to any processing changes.
Online data revealed that eye-movement patterns are affected by structured input train-
ing and this effect can lead to an immediate change in incremental sentence comprehen-
sion patterns. Once again, the possible explanation of these results might be found in the
nature and purpose of structured input activities. As outlined by Wong (2004, p. 35)
structured input practice ‘push learners to abandon their inefficient processing strategies
for more optimal ones’. In the case of this study for example, structured input signifi-
cantly improves learners’ interpretation and processing of sentences containing English
causative forms. The accuracy scores in this study for the structured input group are
higher than the ones obtained by a processing instruction group in one of the studies
conducted by Wong and Ito’s (2018). One of the possible explanations is that in the pre-
sent study, participants in the structured input group were exposed to an higher number
of target items through the structured input training they received.
To conclude, prior to instruction, both groups gazed at the incorrect picture through-
out causative sentences. After instruction, the structured input group made signifi-
cantly more looks to the correct picture, whereas the traditional/drill-based group did
not. The fact that a difference in looks emerged prior to sentence offset strongly sug-
gests that the structured input practice affects the strategies that learners use to process
input in real time.

2 Implications for theory and pedagogy

Overall, the main findings from this study make a number of theoretical and pedagogical
contributions. First, the main results from this study contribute directly to the discussion
on the crucial role that input processing plays in second language acquisition. The eye-
movement patterns and the picture selection observed in participants in this study sup-
port the use of the First Noun Principle. The eye-tracking analysis in this study further
demonstrates that L2 learners rely on the First Noun Principle to process forms such as
English passive causative.
Second, this study provides further online evidence (Ito & Wong, 2019; Lee &
Doherty, 2019; Wong & Ito, 2018) of the effectiveness of structured input practice com-
pared to traditional instruction. Structured input training is an effective pedagogical
intervention for grammar instruction designed to alter a number of processing princi-
ples. Traditional instruction is unsuccessful in bringing about similar effects as those
brought about by structured input. Structured input training led to more accurate pro-
cessing of the target linguistic item. This is the first eye-tracking study measuring the
18 Language Teaching Research 00(0)

effects of structured input training without explicit information. Structured input train-
ing does change eye-movement patterns in real-time input processing.
Third, the main findings from this study reaffirm the importance of input-based prac-
tice as a key pedagogical tool. Structured input practice provides an effective option for
teaching grammar when linguistic features are affected by processing problems.
Grammar instruction should be not be viewed as the explanation and practice of gram-
matical forms but as a pedagogical intervention to facilitate the processes and strategies
used by L2 learners during input processing. Structured input practice does not intend to
‘pour knowledge’ into L2 learners heads, but rather to assist certain processes which can
aid the growth of the internal language system by making appropriate for meaning con-
nections or like in the case of this study processing English causative forms correctly.
In this respect, a subsequent pedagogical implication of this study is that structured
input should precede output-based practice. Acquiring L2 grammar requires learners to
process meaning-bearing input first. Such input should be structured to highlight a par-
ticular grammatical feature to aid its acquisition. A coherent grammar lesson is one that
takes L2 learners from processing a grammatical feature in the input to accessing the
feature from the internal language system to create output.

3 Limitations and further research

Despite the positive outcomes of the present study, there are some limitations and sug-
gestions for future research. First, the lack of a post-test battery (due to practical reasons)
is a methodological limitation. Further research should measure durability of the online
effects of the structured input training. Second, although the present study is the first one
of its kind measuring online effects of structured input practice on the acquisition of
English passive forms, future research should also investigate the effects of this peda-
gogical intervention on other linguistic features in other languages affected by similar or
different processing principles.
Third, future research should consider a longer instructional period to see whether
structured input practice might have further effects in further improving accuracy and
processing of forms/structures. More online research within this framework would
need to be carried out in different languages, processing principles and among differ-
ent populations (e.g. first languages, language backgrounds) in order to generalize this
initial finding.
Online data revealed that eye-movement patterns are affected by structured input
training and this effect can lead to an immediate change in incremental sentence interpre-
tation patterns. However, the results of this study do not clearly show change in incre-
mental sentence interpretation patterns. Future research should make use of time windows
rather than using a growth curve analysis which takes full advantage of the time- course
of auditory processing and its effects on eye-movements.
The use of online methods are an important advancement for processing instruction
research and a necessary step given the focus on moment-by-moment processing. This
research is very much in its infancy and needs to be expanded to include a wider variety
of input processing principles, L2s, target forms and structures, and online methods.
Benati 19

Acknowledgements
I would like to thank all the students and teachers who participated to the study. I also express my
gratitude to a number of colleagues who read this article and provided me with valuable comments
and suggestions.

Funding
The author received no financial support for the research, authorship, and/or publication of this
article.

ORCID iD
Alessandro Benati   https://orcid.org/0000-0002-5999-560X

References
Altmann, G.T.M., & Kamide, Y. (1999). Incremental interpretation at verbs: Restricting the
domain of subsequent reference. Cognition, 73, 247–264.
Benati, A. (2001). A comparative study of the effects of processing instruction and output based
instruction on the acquisition of the Italian future tense. Language Teaching Research, 5,
95–127.
Benati, A. (2004a). The effects of structured input and explicit information on the acquisition of
Italian future tense. In VanPatten, B. (Ed.), Processing instruction: Theory, research, and
commentary (pp. 207–255). Mahwah, NJ: Erlbaum.
Benati, A. (2004b). The effects of processing instruction and its components on the acquisition of
gender agreement in Italian. Language Awareness, 13, 67–80.
Benati, A. (2005). The effects of PI, TI and MOI in the acquisition of English simple past tense.
Language Teaching Research, 9, 67–113.
Benati, A. (2013a). The input processing theory. In Garcia Mayo, P. (Ed.), Contemporary
approaches to second language acquisition (pp. 93–110). Amsterdam: John Benjamins.
Benati, A. (2013b). Age and the effects of processing instruction on the acquisition of English pas-
sive constructions among school children and adult native speakers of Turkish. In Lee, J.F.,
& A. Benati (Eds.), Individual differences and processing instruction (pp. 83–104). Sheffield:
Equinox.
Benati, A. (2017). Classroom-oriented research: Processing instruction. Language Teaching, 52,
343–359.
Benati, A. (forthcoming). Input processing and processing instruction: The acquisition of Italian
and Modern Standard Arabic. Amsterdam: John Benjamins.
Benati, A., & Batziou, M. (2017). The effects of structured-input and structured-output tasks on
the acquisition of English causative. International Review of Applied Linguistics in Language
Teaching, 57, 265–287.
Benati, A., & Batziou, M. (2019). Discourse and long-term effects of structured-input and struc-
tured-output tasks in combination and isolation on the acquisition of passive English causa-
tive forms. Language Awareness, 28, 77–96.
Benati, A., & Lee, J. (2008). Grammar acquisition and processing instruction: Secondary and
cumulative effects. Bristol: Multilingual Matters.
Benati, A., & Lee, J. (2015). Processing instruction: New insights after twenty years of theory,
research and application. Special Issue in International Review of Applied Linguistics in
Language Teaching, 53, 87–90.
20 Language Teaching Research 00(0)

Cadierno, T. (1995). Formal instruction from a processing perspective: An investigation into the
Spanish past tense. The Modern Language Journal, 79, 179–193.
Cheng, A.C. (2004). Processing instruction and Spanish ser and estar: Forms with semantic-aspec-
tual value. In VanPatten, B. (Ed.), Processing instruction: Theory, research, and commentary
(pp. 119–141). Mahwah, NJ: Erlbaum.
Farley, A. (2004). Processing instruction and the Spanish subjunctive: Is explicit information
needed? In VanPatten, B. (Ed.), Processing instruction: Theory, research, and commentary
(pp. 227–239). Mahwah, NJ: Erlbaum.
Farley, A. (2005). Structured input: Grammar instruction for the acquisition-oriented classroom.
New York: McGraw-Hill.
Issa, B., & Morgan-Short, K. (2019). Effects of internal and external attentional manipulations on
second language grammar development: An eye-tracking study. Studies in Second Language
Acquisition, 42, 389–417.
Ito, K., & Wong, W. (2019). Processing instruction and the effects of input modality and voice
familiarity on the acquisition of the French causative construction. Studies in Second
Language Acquisition, 41, 443–468.
Lee, J. (2004). On the generalizability, limits, and potential future directions of processing instruc-
tion research. In B. VanPatten (Ed.), Processing instruction: Theory, research, and commen-
tary (pp. 311–323). Mahwah, NJ: Erlbaum.
Lee, J. (2015). The milestones in twenty years of processing instruction research. International
Review of Applied Linguistics in Language Teaching, 53, 111–126.
Lee, J., & Benati, A. (2007a). Delivering processing instruction in classrooms and virtual con-
texts: Research and practice. London: Equinox.
Lee, J., & Benati, A. (2007b). Second language processing: An analysis of theory, problems and
possible solutions. London: Continuum.
Lee, J., & Benati, A. (2009). Research and perspectives on processing instruction. Berlin: Mouton
de Gruyter.
Lee, J., & Doherty, S. (2019). Native and nonnative processing of active and passive sentence:
The effects of processing instruction on the allocation of visual attention. Studies in Second
Language Acquisition, 41, 853–879.
Lee, J., & VanPatten, B. (2003). Making communicative language teaching happen. New York:
McGraw-Hill.
Marsden, E. (2006). Exploring input processing in the classroom: An experimental comparison of
processing instruction and enriched input. Language Learning, 56, 507–566.
Morgan-Short, K., & Bowden, H.W. (2006). Processing instruction and meaningful output-
based instruction: Effects on second language development. Studies in Second Language
Acquisition, 28, 31–65.
Sanz, C. (2004). Computer delivered implicit versus explicit feedback in processing instruction.
In VanPatten, B. (Ed.), Processing instruction: Theory, research, and commentary (pp. 241–
255). Mahwah, NJ: Erlbaum.
Stromswold, K., Eisenband, J., Norland, E., & Ratzan, J. (2002). Tracking the acquisition and
processing of English passives: Using acoustic cues to disambiguate actives and passives.
Unpublished paper presented at the CUNY Conference on Sentence Processing, New York,
USA.
Trueswell, J.C., & Tanenhaus, M. (2005). Approaches to studying world-situated language use:
Bridging the language-as-product and the language-as-action traditions. Cambridge, MA:
MIT Press.
VanPatten, B. (1996). Input processing and grammar instruction: Theory and research. Norwood,
NJ: Ablex.
Benati 21

VanPatten, B. (Ed.) (2004). Processing instruction: Theory, research, and commentary. Mahwah,
NJ: Lawrence Erlbaum.
VanPatten, B. (2015a). Input processing in adult SLA. In VanPatten, B., & J. Williams (Eds.),
Theories in second language acquisition (pp. 113–135). New York: Routledge.
VanPatten, B. (2015b). Foundations of processing instruction. IRAL, 53, 91–109.
VanPatten, B., & Cadierno, T. (1993). Explicit instruction and input processing. Studies in Second
Language Acquisition, 15, 225–243.
VanPatten, B., & Oikkenon, S. (1996). Explanation vs. structured input in processing instruction.
Studies in Second Language Acquisition, 18, 495–510.
VanPatten, B., & Wong, W. (2004). Processing instruction and the French causative: Another rep-
lication. In VanPatten, B. (Ed.), Processing instruction: Theory, research, and commentary
(pp. 97–118). Mahwah, NJ: Erlbaum.
VanPatten, B., Farmer, J., & Clardy, C. (2009). Processing instruction and meaning-based output
instruction: A response to Keating and Farley (2008). Hispania, 92, 116–126.
Wong, W. (2004). The nature of processing instruction. In VanPatten, B. (Ed.), Processing instruc-
tion: Theory, research, and commentary (pp. 33–65). Mahwah, NJ: Erlbaum.
Wong, W., & Ito, K. (2018). The effects of processing instruction and traditional instruction on L2
online processing of the causative construction in French: An eye-tracking study. Studies in
Second Language Acquisition, 40, 241–268.