Abstract
Previous studies on incidental focus on form (FonF) have mostly focused on its overall effectiveness. This study is an attempt to further examine the effectiveness of incidental FonF with regard to different linguistic categories (vocabulary, grammar, and pronunciation). To this end, eight hours of a free discussion English as a foreign language (EFL) class with 12 upper-intermediate learners were observed, and instances of focus on form episodes (FFEs) were identified. Uptake sheets were also used to identify the learners’ noticing of FFEs and their novelty to them. To measure the effects of incidental FonF on learning different linguistic categories, individualized immediate and delayed post-tests were designed based on learners’ on-the-spot written reports of noticing of novel linguistic forms addressed by FFEs. The results indicated that incidental FonF was almost equally effective in developing learners’ knowledge of different linguistic forms. The findings also revealed that incidental FonF was effective in fostering learners’ knowledge of different linguistic forms at both explicit/implicit and receptive/productive levels.
I Introduction
The observation that second language (L2) learners cannot develop high levels of accuracy through purely meaning-based instruction has led researchers to propose that some degree of attention to form is necessary in order for language acquisition to take place (e.g. R. Ellis, 2001, 2003; Leow, 2015; Loewen, 2018, 2019; Long, 1991, 2015, 2016; Spada & Lightbown, 2008). Attention to form can occur in a number of ways, two of which, focus on form (FonF) and focus on forms (FonFs), have received considerable attention by researchers (e.g. L. Gholami, 2022; L. Gholami & Gholami, 2020; Loewen, 2004, 2005; Nassaji, 2010, 2013; Pawlak, 2007; Pouresmaeil & Gholami, 2019, 2022, 2023; R. Sheen, 2005; Williams, 2001). While in FonFs language is viewed as an object to be mastered, and thus the focus is mainly on developing conscious knowledge of linguistic forms, FonF views language as a tool for communication and integrates any attention to form into meaning-oriented communicative tasks (R. Ellis, 2001; Loewen, 2018). Depending on the number of linguistic forms targeted, FonF is further divided into ‘planned’ FonF (focusing on a limited number of preselected linguistic forms) and ‘incidental’ FonF (focusing on a wide range of linguistic forms arising incidentally in the course of communication) (R. Ellis, 2001; Loewen, 2018).
Previous research has revealed that incidental FonF occurs frequently in communicative classes and addresses various linguistic forms (e.g. L. Gholami, 2022; L. Gholami & Gholami, 2020; Llinares & Lyster, 2014; Loewen, 2004; Nassaji, 2013; Pouresmaeil & Gholami, 2019). Empirical research has also shown that a good proportion of the linguistic forms addressed by FonF episodes (FFEs) is retained by learners (e.g. L. Gholami, 2022; Loewen, 2005; Nassaji, 2010, 2013; Williams, 2001). However, to date, not much research has been carried out to investigate the extent to which FFEs with different linguistic foci lead to L2 development. That is, while previous research has revealed beneficial effects for incidental FonF in general, not much is known about its effects on learning different linguistic forms. Thus, in an attempt to provide a finer-grained picture of the effectiveness of incidental FonF, the present study investigates the effects of incidental FonF on learning different components of language (vocabulary, grammar, and pronunciation).
II Focus on form
In his seminal paper, Long (1991) defined FonF as any attempt that ‘overtly draws students’ attention to linguistic elements as they arise incidentally in lessons whose overriding focus is on meaning or communication’ (pp. 45–46). Closely associated with his interaction hypothesis, Long’s original definition of FonF required it to occur in the form of implicit corrective feedback, such as recasts, when the need arose incidentally in response to communication breakdown. This primary definition of FonF has, however, been expanded by a number of researchers to also include planned attempts (planned FonF) to practice preselected linguistic forms (usually one or two) through focused communicative tasks (R. Ellis, 2001; R. Ellis et al., 2002). FonF has also included incidental attempts to draw learners’ attention to form in the course of communication even when there is no breakdown in comprehension. This is what R. Ellis (2001) and R. Ellis et al. (2002) refer to as ‘incidental FonF’ and categorize it into reactive attempts, as when corrective feedback is provided on learners’ errors, and preemptive attempts, as when the teacher or a learner initiates attention to form without any apparent occurrence of errors. The present study operationalized incidental FonF based on R. Ellis’s (2001) and R. Ellis et al.’s (2002) definition.
Theoretically, FonF is supported by interactionist-cognitive theories of learning (R. Ellis, 2012), such as the Noticing Hypothesis and the Interaction Hypothesis. Schmidt’s (1990, 1993) noticing hypothesis proposes that in order for language acquisition to take place, learners need to notice the input and the linguistic features within it. However, as VanPatten (1990, 1996) argues, it is extremely difficult for L2 learners to simultaneously attend to both meaning and form, and that they opt to attend to either of them depending on the demands of the task. With meaning-focused instruction, it is the meaning which is generally attended to by learners as it is central to the task. FonF, however, provides the learners with some time-out from meaning to attend to form and thus provides them with some opportunities to notice the linguistic features within a meaning-primary instructional context. Attention to form in this way could occur through either ‘negotiation of meaning’ or ‘negotiation of form’. Negotiation of meaning, which is central to Long’s interaction hypothesis, takes place when two or more interlocutors fail to achieve mutual understanding because of failure in comprehension in the course of interaction. The fundamental idea is that through negotiation of meaning learners are provided with both positive and negative evidence, which could result in correction and opportunities for output (Long, 1996). Negotiation of form, on the other hand, occurs when there is no problem in comprehension, but in accuracy in the course of communication (R. Ellis et al., 2002). What is central here is that through negotiation of meaning and form, the problematic areas are highlighted for learners, allowing them to compare their L2 knowledge with the provided input, which in turn makes them notice the gaps in their interlanguage (R. Ellis, 2001, 2016; Long, 2016). According to Long (2016), noticing the gaps in this way is most effective as it occurs in the course of communication when learners are developmentally ready. Similarly, Doughty (2001) argues in favor of FonF on the grounds that it promotes the form-function mapping, which she sees as necessary for L2 development.
1 Controversy over the effectiveness of FonF
The efficacy of FonF, however, is not without controversy. The controversy, as also stated by Storch (2018), basically stems from how L2 acquisition is thought to take place. There are generally three positions on how an L2 is acquired. The first position, advocated by Krashen (1981, 1993), Schwartz (1993), and Truscott (1999), views L2 acquisition to occur in much the same way as L1 acquisition. The advocates of this position argue that all learners need to acquire L2 is positive evidence (comprehensible input) and that negative evidence in any form and explicit instruction of linguistic forms is unnecessary and even harmful. These researchers argue that instruction, at best, may lead to development of some explicit knowledge of simple structures, which cannot be proceduralized for automatic access. Thus, they adhere to the non-interface position, which puts that explicit knowledge does not convert to implicit knowledge. This view, however, is challenged mainly by research contextualized in French Immersion Program (Swain, 1985), which showed that L2 learners’ interlanguage was grammatically flawed even after many years of exposure to L2. Therefore, as stated at the beginning of the article, there is now a general consensus among researchers on the facilitative role of instruction. Thus, today, the controversy is mainly on when and how instruction should take place, rather than whether there should be some form of instruction. There are generally two positions regarding how L2 instruction should be carried out: FonF and FonFs.
While the proponents of FonF (e.g. Doughty, 2001; Long, 1991, 2015, 2016) consider L2 acquisition as a by-product of communication where learners notice linguistic gaps in their interlanguage, the opponents of FonF (e.g. R. Sheen, 2003; Swan, 2005), drawing on Skill Acquisition Theory, argue that in order for L2 acquisition to take place, learners need to receive systematic treatment of linguistic features in the form of explicit instruction. This instruction is followed by some form-based activities and then abundant communicative practice (FonFs), which facilitates the development of procedural knowledge for automatic retrieval. These researchers do not, however, neglect the facilitative role of corrective feedback as long as it is combined with explicit instruction. Thus, the proponents of FonFs stick to the strong interface position, which puts that explicit knowledge has the potential to convert to implicit knowledge. The advocates of FonF, on the other hand, contend that FonF facilitates the development of implicit knowledge as it occurs in the course of interaction while the primary attention is on meaning and communication. The main idea is that, through FonF, the linguistic features of language and the meaning they realize become salient to learners while they are striving for communication. This, in turn, affords optimal opportunities for the occurrence of form-function mapping (Doughty, 2001), which is seen as essential in acquiring implicit L2 knowledge (Doughty, 2001; N. C. Ellis, 2002, 2005, 2015).
Today, there is ample evidence on the beneficial effects of FonF, mostly in the form of corrective feedback provided on a limited number of linguistic structures, without being coupled with any explicit instruction (e.g. Ammar & Spada, 2006; Canals et al., 2021; R. Ellis, 2007; Han, 2002; Kartchava & Ammar, 2014; S. Li, 2014; H. Li & Iwashita, 2021; Yang & Lyster, 2010; Yilmaz, 2012). The efficacy of FonF in these studies could, however, be partly due to the intensive treatment of errors through provision of corrective feedback. Thus, in essence, these studies provide support for the effectiveness of planned FonF. Incidental FonF, however, has not received as much attention as planned FonF as far as L2 development is concerned. Moreover, although a good number of studies on corrective feedback have addressed the acquisition of different linguistic targets (e.g. R. Ellis, 2007; Kartchava & Ammar, 2014; S. Li, 2014; H. Li & Iwashita, 2021; Mackey, 2006; Yang & Lyster, 2010; Yilmaz, 2012), the linguistic forms examined all belonged to one specific linguistic category (e.g. different morphosyntactic features). Not many FonF studies have investigated the acquisition of linguistic forms belonging to different linguistic categories (lexis, grammar, and pronunciation). Thus, in addition to focusing on incidental FonF, the present study also examines its effects on developing learners’ accuracy with regard to grammar, vocabulary, and pronunciation. The limited number of studies on the effects of incidental FonF on L2 development are reviewed in the following section.
2 Studies on incidental FonF
Incidental FonF has received considerable attention by researchers in the area of form-focused instruction (FFI). A good number of studies have been carried out on the frequency of FonF occurrence and the rate of uptake following FFEs (e.g. R. Ellis et al., 2001a, 2001b; L. Gholami, 2021; L. Gholami & Gholami, 2020; Loewen, 2004; Nassaji, 2010, 2013; Pouresmaeil & Gholami, 2019; Wang & Li, 2021). These studies have revealed that incidental FFEs occur frequently in both English as a foreign language (EFL) and English as a second language (ESL) contexts (one FFE per 1.6–2.6 minutes) with a rather high rate of uptake (50%–73%). However, as some researchers have noted, although uptake, which is defined as learners’ subsequent production of the target form or acknowledgement of the provision of corrective feedback, may indicate that learners have noticed the FFEs, it cannot be taken as any direct evidence of L2 development (R. Ellis et al., 2001a; Nassaji, 2011). Thus, in an attempt to investigate whether incidental FonF has any effects on learning, some researchers have made use of individualized posttests designed based on the FFEs related to each individual learner. Using individualized posttests, Williams (2001), for example, tested the retention rate of grammatical and lexical FFEs across different proficiency levels with eight learners. She found the retention rate to range from 50% to 94% for lexical FFEs and 40%–94% for grammatical FFEs, with higher-proficiency learners achieving better scores. In another study, Loewen (2005) investigated the effectiveness of incidental FonF in developing learners’ L2 knowledge and found that learners were able to recall the targeted linguistic forms in the FFEs correctly 60% of the time one day after their occurrence and 50% of the time two weeks later. In a study on reactive incidental FonF, Loewen and Philp (2006) examined the effectiveness of recasts provided in the course of meaning-oriented interactions. The results indicated that recasts were effective at least 50% of the time as measured by immediate and delayed posttests. The study also found that recasts whose corrective force was highlighted in a way or another (for example by declarative intonation, shortened length, or stress) were more effective than the more implicit forms. In another study on reactive incidental FonF, Nassaji (2009) investigated the effectiveness of recasts and elicitations and the role of feedback explicitness. The findings revealed that both corrective feedback types were equally effective, with more explicit forms of each corrective feedback type being more effective. The effects of explicitness were, however, more evident in the case of recasts.
In a more comprehensive study, Nassaji (2010) explored the effectiveness of reactive and preemptive incidental FonF in a university context. The study found that, in total, the learners responded correctly to 58% of the tested FFEs, with preemptive FFEs receiving a higher rate of correct responses than reactive ones. Moreover, similar to William’s (2001) study, Nassaji also found a higher rate of accuracy for more proficient learners on the posttest. In a separate analysis of the data, Nassaji (2013) investigated the effectiveness of incidental FonF based on the type of student-teacher participation structure. The findings indicated that the learners gave correct answers to 69%, 66%, and 48% of the test items designed out of the FFEs occurring in small group, one on one, and whole-class interactions, respectively. In another study, Kim and Nassaji (2018) examined the role of learner extraversion as an individual difference in incidental FonF effectiveness across two proficiency levels (intermediate and advanced). The results revealed the accuracy rate on the posttest to be 27% for the advanced class and 33% for the intermediate class, with no significant correlation between extraversion and incidental FonF effectiveness in either of the classes. The authors discussed the lower accuracy rates compared to similar studies in terms of the type of FFEs, which were mostly reactive and in the form of recasts. As the authors explained, reactive FFEs, particularly when provided in the form of recasts, are less salient than preemptive FFEs and, thus, might go unnoticed by the learners.
In a more recent and innovative study, L. Gholami (2022) explored the effectiveness of incidental FonF in relation to formulaic and non-formulaic forms with a group of advanced EFL learners. Formulaic forms included linguistic features such as idioms, lexical bundles, collocations, and compounds, whereas non-formulaic features encompassed grammar, vocabulary, pronunciation, and spelling. The study showed overall retention rates of 69% and 42% on the immediate and delayed posttests, respectively. Regarding formulaicity, the study found that although formulaic FFEs were followed by uptake more frequently compared to non-formulaic FFEs, learners produced more correct answers to test items designed based on non-formulaic forms than formulaic ones on both the immediate posttest (83% vs. 54%) and delayed posttest (78% vs. 9%). Drawing on Wray’s (2002) needs-only analysis model, L. Gholami (2022) argued that unlike L1 learners who treat combination of words holistically, L2 learners tend to segmentalize formulaic forms and store them as separate bits in their mental lexicon. This, in turn, may pose a challenge to L2 learners to retrieve the whole formulaic form, given that many of these forms are semantically and syntactically arbitrary and, thus, cannot be accounted for by grammar rules.
Studies on incidental FonF have greatly contributed to our understanding of its efficacy. These studies have shown that incidental FonF occurs frequently in meaning-oriented classes, and that it has beneficial effects on L2 development. Overall, the following conclusions could be drawn based on the findings of previous studies:
Incidental FonF occurs frequently in meaning-oriented classes.
Learners notice a significant number of FFEs occurring in the course of communication.
Incidental FonF is facilitative of developing learners’ L2 knowledge.
Preemptive FFEs are generally more beneficial than reactive FFEs due to their higher saliency, and explicit forms of reactive FFEs are more beneficial than implicit forms.
Learners with a higher level of proficiency tend to benefit more from incidental FonF than those with lower levels of proficiency.
Incidental FonF is more effective in relation to non-formulaic forms than formulaic forms.
What, however, remains to be further investigated is the contribution of incidental FonF to developing learners’ L2 knowledge in relation to different linguistic categories. While previous studies have mostly focused on FFEs with different linguistic foci, they have not distinguished between their retention rates (except Williams, 2001). A number of studies on noticing, however, have shown that FonF, in general, and corrective feedback (particularly recasts), in particular, is more noticeable when provided on lexical and phonological forms than when provided on morphosyntactic features (e.g. L. Gholami & Gholami, 2020; Kim & Han, 2007; Lyster, 2001; Mackey et al., 2000; Pouresmaeil & Gholami, 2019; Y. Sheen, 2006). Similarly, some research on planned FonF has revealed that recasts are generally more effective in developing lexical knowledge than grammatical knowledge, although it is not always the case for all morphosyntactic features (Jeon, 2007). These studies suggest that corrective feedback provided on lexical and phonological errors may be generally more effective than feedback on morphosyntactic features. One reason proposed for this suggestion is the learners’ higher perception of the corrective force of feedback provided on these linguistic forms than on morphosyntactic forms. This proposition, however, needs to be further investigated in incidental FonF as it may not necessarily be the case in this type of intervention (for example, see Williams, 2001). This may be mainly due to the fact that incidental FonF is not merely limited to (one type of) corrective feedback. Furthermore, as noted previously, uptake and noticing should not necessarily be taken as evidence of L2 acquisition taking place. Moreover, meta-analysis research on the role of interaction in L2 development has revealed that although learners benefit more from interaction targeting lexical forms in the short term, in the long run, interaction targeting grammatical forms appears to be more beneficial (Mackey & Goo, 2007).
One further limitation of previous studies on incidental FonF concerns their outcome measures. The tests used in these studies to gauge learning tapped only one knowledge aspect (mostly explicit knowledge for grammatical FFEs and receptive knowledge for lexical FFEs). Thus, our knowledge of the effectiveness of incidental FonF is mostly limited to its effects on developing one knowledge aspect. Moreover, as also acknowledged by a number of researchers (e.g. L. Gholami, 2022; Loewen & Philp, 2006), previous studies did not completely account for learners’ prior knowledge of the targeted forms in FFEs. These studies have presumed learners’ queries on linguistic forms (learner-initiated FFEs) or their erroneous utterances (reactive FFEs) as indication of linguistic gaps. While this can be true in the case of learner-initiated preemptive FFEs, not all erroneous utterances can necessarily indicate actual gaps in learners’ interlanguage. Such utterances may indeed be random mistakes (L. Gholami & Gholami, 2020; Nassaji, 2009), or lack of implicit knowledge (R. Ellis, 2004). In an attempt to address these gaps in the literature, the present study investigated whether incidental FonF has differential effects on developing learners’ accuracy with regard to grammar, vocabulary, and pronunciation. As explained later, attempts were also made to design test items based on FFEs addressing actual gaps in learners’ interlanguage. Finally, in order to provide a better picture of the effectiveness of incidental FonF, the study made use of different types of outcome measures. More specifically, the present study sought to find answers to the following questions:
Research question 1: What is the retention rate of grammatical, lexical, and phonological incidental FFEs in the short and long terms?
Research question 2: Does incidental focus on form have a differential effect on developing learners’ knowledge of target forms from different linguistic categories?
III Method
1 Participants
This study was conducted in a free discussion EFL class at a private language school with 12 male and female learners (male = 4 and female = 8), within the age range of 20–36 years, with an average of 25 years. Prior to the study, an announcement was made on the need for participants for a study to be conducted in a free discussion class. Initially, 15 learners volunteered to take part in the study. However, only the learners with an upper-intermediate level of proficiency were selected as the participants of the study. It was assumed that learners with lower levels of proficiency would not be able to contribute much to the class discussions. The proficiency of the learners in English was measured through running the Oxford Placement Test (D. Allan, 2004), and only the learners who scored 135–149 out of 200 were selected as the participants of the study. This score range (135–149) is specified as indicating an upper-intermediate level of proficiency based on the guidelines of the test. However, to further ascertain the proficiency level of the learners in English, the opinion of the teaching staff was also elicited. The teachers also confirmed the results of the placement test as truly indicating the learners’ proficiency level. All learners were L1 speakers of Azeri or Kurdish and were fluent in Persian as their second language. The instructor of the class was a 31-year-old teacher, who held an MA in TESOL and had 14 years of teaching experience, 10 years of which related to running free discussion classes. All participants agreed to take part in the study voluntarily by signing informed consent forms.
2 Instruments
One mini-sized video-recorder was used to record the whole-class interactions occurring in class. In addition, in order to ascertain the novelty of FFEs to learners, uptake sheets, adopted from J. Gholami and Basirian (2011), were distributed among the learners each session. These sheets were in the form of organized charts on which the learners were asked to write down any linguistic forms they noticed during the class discussions and to indicate whether the form was novel to them in the respective section (see Appendix 1). Finally, to measure any learning, immediate and delayed individualized posttests were designed for each learner based on the grammatical, lexical, and phonological FFEs.
3 Data collection procedures
Data for the present study were collected through classroom observation, analysing learners’ uptake sheets, and running individualized posttests. Four two-hour sessions of a free discussion class were observed and video-recorded during four successive days, yielding a total of eight hours of meaning-oriented whole-class interactions. In order to make the class run naturally, the teacher was not informed about the aim of the study. He was merely told that the study was to explore the interactions occurring in class. Thus, he was not required to manipulate the occurrence of any attention to form. Likewise, the learners were not aware that they would be tested on any linguistic forms.
In addition to the video-recording, the first author attended the class each session as an unobtrusive observer and took filed notes of the FFEs occurring in class. These field notes were later used along with the video-recordings to identify the FFEs with more precision. After identifying the FFEs, instances of FFEs the learners did not have prior knowledge about were also identified. This was done through analysing the learners’ verbal reports on their uptake sheets. Finally, individualized posttests were designed based on the FFEs the learners reported to have no previous knowledge of on their uptake sheets and were administered to each learner at two intervals. Below are thorough explanations about how the FFEs were identified and measured in terms of effectiveness.
4 Identification of FFEs
After each session, instances of FFEs were identified based on the transcriptions of the video-recordings and the field notes. An FFE, which was the basic unit of analysis, was defined as the discourse from the point ‘where the attention to linguistic form starts to the point where it ends, due to a change in topic back to message or sometimes another focus on form’ (R. Ellis et al., 2001a, p. 294). Thus, in practice, an FFE started when corrective feedback was provided on a learner’s erroneous utterance or when a learner or the teacher queried a linguistic form and ended by either a learner uptake, shift of attention back to message, or another FFE. FFEs were also categorized based on their linguistic foci. For the purpose of the present study, three linguistic categories were identified: grammar, lexis, and pronunciation.
To check the reliability of coding the FFEs, a random sample of 50% of the FFEs was also coded by a trained research assistant. The Cohen’s kappa coefficient obtained was 97% for coding the linguistic foci of the FFEs (grammatical, lexical, and phonological).
5 Measuring the effectiveness of incidental FonF
Because of the unpredictability of the focused linguistic items, it is not possible to measure the effectiveness of incidental FonF through a pretest–posttest design (R. Ellis et al., 2001a; Nassaji, 2013). This, however, is not to overlook the importance of learners’ previous knowledge of the FFEs. As noted by Nassaji (2016), designing tests based on the FFEs the learners already have good knowledge of does not seem plausible. Thus, researchers have mostly used individualized posttests designed based on the FFEs the learners engaged in. These researchers have assumed the learners’ production errors or request for linguistic information as an indication of their lack of knowledge with regard to the particular linguistic form. While this assumption could be true about some FFEs, the point of caution, as noted previously, is that not all erroneous utterances necessarily reflect actual gaps in learners’ L2 knowledge. In other words, the learners’ erroneous utterance can simply be a mistake, not an error (L. Gholami & Gholami, 2020; Nassaji, 2009). Indeed, previous research has indicated that learners may already have good knowledge of a considerable proportion of the FFEs (Pouresmaeil & Gholami, 2019), indicating that the occurrence of many errors are simply mistakes or at worst, lack of implicit knowledge. Thus, in the present study, attempts were made to design test items based on the FFEs the learners had no or little knowledge about. To this end, the learners were asked to indicate whether any linguistic features (vocabulary, grammar, and pronunciation) they noticed were novel to them in the specified section on the provided uptake sheets. They did so by simply ticking ‘Yes’ (indicating the form was novel) or ‘No’ (indicating the form was not novel).
After identifying the FFEs which were novel to the learners, individualized posttests were developed for each linguistic category, and were administered on two occasions: one immediate (1–4 days after the FFEs) and one delayed (15–18 days after the FFEs). Depending on the linguistic foci of the FFEs, the tests took different formats: Untimed Grammaticality Judgement Test (UGJT) and Oral Elicited Imitation Test (OEIT) for grammatical FFEs, Vocabulary Knowledge Scale (VKS) for lexical FFEs, and Pronunciation Test (PT) for phonological FFEs. Each learner was invited to a quiet room and was given the tests in the following order:
OEIT
PT, sentence section
PT, word section
VKS
UGJT
Below is a thorough description of each testing measure.
a UGJT
This test measured the learners’ explicit knowledge of the grammatical FFEs. UGJTs were designed based on the FFEs with a grammatical focus and were provided in the form of sentences, each having one grammatical error, which was the focus of the FFE that occurred in class. The test also included some distractor items, which were not focused on by any FFEs. The learners were informed in their L1 that each sentence included one grammatical error and were asked to find and correct it. Example 1 illustrates a sample UGJT item, designed based on a grammatical FFE presented in Extract 1.
Example 1: Sample UGJT item Please read the following sentences. Each sentence has one grammatical error. Find the error and correct it. There are not any spelling or punctuation errors. I look forward to meet him. Extract 1: FFE S: I look forward to meet him, and I’m so happy. T: You look forward to meeting him, not meet. S: Meeting? T: Yes. Meeting. Look forward to meeting someone. S: Oh! Look forward to meeting him.
b OEIT
OEIT, developed by Erlam (2006), is a test of implicit knowledge which is based on asking the learners to repeat statements including the target form. Erlam provided arguments and statistical evidence which proved that provided the learners’ primary focus is on meaning, there is a delay between the presentation of the stimuli and repetition, and the test is time-pressured, OEIT can provide evidence of learners’ implicit knowledge of the target structure.
In the present study, the OEIT test items included statements designed based on the FFEs with a grammatical focus as well as some distractor items. The statements were presented to the learners on audiotape only once and in real time. To ensure a primary focus on meaning, following Loewen et al. (2009), the learners were provided with a Beliefs Questionnaire and were asked to first indicate whether they agreed with each statement by ticking one of the three boxes of ‘agree’, ‘disagree’, or ‘not applicable’ and then repeat the statement in correct English. All target items and half of the distractors were grammatically incorrect, but the learners were not informed about this. To ensure that the learners would perform in the test as expected, they practiced the test with four sample items in advance. Example 2 illustrates a sample OEIT test item, designed based on a grammatical FFE presented in Extract 2.
Example 2: Sample OEIT test item Please listen to the following sentences and indicate whether you agree with each one on the questionnaire. You have to repeat the sentence in correct English after checking the box on the questionnaire. You will hear each sentence only once. I only study hard when someone makes me to study. Extract 2: FFE S: I only study hard when someone makes me to study. T: When someone makes you study. No ‘to’ after ‘make’. S: So, when someone makes me study? T: Yes. S: Aha.
c VKS
VKS is a five-level self-report vocabulary knowledge scale developed by Paribakht and Wesche (1993), aiming to measure learners’ different levels of lexical knowledge (ranging from total unfamiliarity to ability to use the word in a sentence). In the present study, the same VKS was used with a minor revision to measure learners’ receptive and productive knowledge of the lexical items the FFEs focused on. The revision concerned the fifth level of the scale, which originally asked the learners to only use the given word in a sentence. However, in this study, the learners were required to first write the definition of the word and then use it in a sentence. This decision was made in order to make sure whether the learners knew the correct meaning of the given words if they chose the fifth level. Example 3 illustrates a sample VKS item, designed based on the lexical FFE presented in Extract 3.
Example 3: Sample VKS item Please choose the most appropriate option based on how well you know each given word.
Extract 3: FFE S: What is دائمی /ˈdɑːemiː/ in English? T: Permanent. S: I think you should have a permanent job when you are 40.
d PT
The retention rate of the phonological FFEs was measured through running the PT, which included two sections: sentence section and word section. In the sentence section, which preceded the word section, the learners were asked to read aloud a list of sentences containing the targeted words in the phonological FFEs, whereas in the word section, the learners were asked to read aloud the same targeted words in isolation. Similar to the other tests, the PT also included a number of distractor items. Examples 4 and 5 illustrate sample PT items, which were designed based on the phonological FFE presented in Extract 4.
Example 4: Sample PT item, sentence section Please read aloud the following sentences. The main obstacle is money. Example 5: Sample PT item, word section Please read aloud the following words. obstacle Extract 4: FFE S: The main obstacle (pronounced as /ˈɔːbstækəl/) is money. T: obstacle (pronounced as /ˈɑːbstəkəl/), obstacle S: Yeah, obstacle (pronounced as /ˈɑːbstəkəl/).
6 Scoring of the tests
Learners’ responses in each outcome measure were marked as either correct or incorrect.
a OEIT
Learners’ responses in this test were marked as correct only if they repeated the statement in correct English. However, if they failed to repeat the statement or did not correct the error, their response was marked as incorrect.
b UGJT
Learners’ responses in this test were considered as correct if they corrected the error in each item. However, if they left an item unanswered, or if they failed to correct the error, their response was marked as incorrect.
c VKS
The VKS included five scales, one of which the learners were required to choose based on how they perceived their knowledge to be regarding the given lexical items. For the purposes of this study, learners’ responses in categories 1 and 2 were considered as evidence of not having either receptive or productive knowledge of the lexical item. In categories 3–5, however, following Paribakht and Wesche (1993), learners’ self-report of the vocabulary knowledge was considered as demonstrated knowledge rather than perceived knowledge. In category 3, if the translation or synonym was correct, the response was scored as category 4, which showed learners’ receptive knowledge of the words. If, however, it was wrong, the response was scored as category 2. Similarly, learners’ responses in category 4 were considered as evidence of having receptive knowledge if the synonym or translation they provided was correct. However, if the definition they provided was wrong, their response was scored as category 2. Learners’ ability to use the lexical items productively was measured via category 5. If the learners provided the correct definition of the word and used it correctly in a sentence, their response was considered as evidence of having productive knowledge, which also presumed having receptive knowledge of the word. If, however, they chose this category, but the definition they provided for the word was wrong, their response was scored as category 2, even if they used the word in a sentence.
d PT
Learners’ responses in this test in either section (words and sentences) were marked as correct only if they pronounced the target words correctly.
To check the reliability of scoring, a second rater also scored a random sample of 50% of learners’ responses to the test items. The Cohen’s kappa coefficient indicated a reliability of 98%.
7 Data analysis
To account for the accuracy rate on both immediate and delayed posttests, raw frequencies were calculated. In addition, to see if there was a significant difference between the scores of the immediate and delayed posttests, Pearson’s Chi-square analysis was performed on the raw frequencies since the data consisted of frequency counts of categorical data. The Alpha level was set at p < .05. All statistical analyses were performed using Statistical Package for the Social Sciences (SPSS), version 18.
IV Results
Overall, 475 FFEs occurred in the observed free discussion EFL class, 322 of which were reported to be novel to the learners. Table 1 shows the frequency of FFEs in each linguistic category. As revealed in Table 1, the highest number of novel FFEs pertained to vocabulary and then to phonological FFEs. The majority of the grammatical FFEs, however, were not novel to the learners.
Frequency of focus on form episodes (FFEs) in different linguistic categories.
This study was, however, mainly concerned with the retention rate of FFEs. For comparison purposes with studies in the literature, first the results with regard to the retention rate of FFEs in total are presented. These results illustrate learners’ performance as measured by the UGJT, category 4 of the VKS, and the word section of the PT. Table 2 presents the overall retention rate of FFEs.
Retention rate of focus on form episodes (FFEs) in total.
As presented in Table 2, the learners could recall the linguistic elements addressed in half of the FFEs on both the immediate and delayed posttests, with no significant difference between the two tests as indicated by the Chi square test, x2 (1, n = 644) = .22, p = .63, phi = .02.
With respect to vocabulary, a total of 245 lexical FFEs were tested on both the immediate and delayed post-tests. Table 3 presents the results in this regard.
Retention rate of lexical focus on form episodes (FFEs).
As revealed in Table 3, the learners could recall the meaning of half of the lexical items addressed in the lexical FFEs in the short term and to a slightly lower extent in the long term. It was also found that the learners had productive knowledge of the vast majority of the learned lexical items in both the short and long terms. The Chi square analysis indicated no significant difference between the retention rates on the immediate and delayed posttests, x2 (1, n = 490) = .21, p = .65, phi = .02.
To measure the retention rate of grammatical FFEs, 26 FFEs with a focus on grammar were tested on the immediate and delayed posttests. The results are presented in Tables 4 and 5 with regard to learners’ explicit and implicit knowledge, respectively.
Retention rate of grammatical focus on form episodes (FFEs) as measured by a Untimed Grammaticality Judgement Test (UGJT) (explicit knowledge).
Retention rate of grammatical focus on form episodes (FFEs) as measured by an Oral Elicited Imitation Test (OEIT) (implicit knowledge).
As shown in Table 4, the learners developed explicit knowledge of half of the grammatical forms addressed through incidental FonF in both the short and long terms, with obviously no significant difference between the immediate and delayed posttest results as revealed by the Chi square test, x2 (1, n = 52) = .000, p = 1.000, phi = .000. Regarding the learners’ implicit knowledge, as Table 5 reveals, the learners developed implicit knowledge of around one-third of the grammatical forms addressed by FFEs in both the short and long terms. The Chi square test revealed no significant difference between the retention rates on the immediate and delayed posttests, x2 (1, n = 52) = .08, p = .77, phi = .08.
With respect to phonological FFEs, a total of 51 FFEs with a focus on pronunciation were tested as separate words as well as inside sentences on both the immediate and delayed posttests. Tables 6 and 7 present the results regarding the retention rate of phonological FFEs as tested in the form of separate words and sentences, respectively.
Retention rate of phonological focus on form episodes (FFEs) at the word level.
Retention rate of phonological focus on form episodes (FFEs) at the sentence level.
As Table 6 reveals, the learners pronounced the majority of the words correctly when presented in isolation in both the short and long terms, with no significant difference between their performance on the immediate and delayed posttests, x2 (1, n = 102) = .000, p = 1.000, phi = .02. The results were almost the same when the words were pronounced inside sentences. The learners pronounced almost the same number of words correctly when presented inside sentences. The Chi square test also revealed no significant difference between the immediate and delayed posttest results, x2 (1, n = 102) = .000, p = 1.000, phi = .02.
V Discussion
This study was primarily concerned with the effectiveness of incidental FonF through measuring the retention rate of FFEs which were novel to the learners. To this end, first, the frequency of FFEs, in general, and the number of novel FFEs, in particular, was calculated in relation to each linguistic category. The results indicated that lexical FFEs occurred much more frequently than grammatical and phonological FFEs. This finding is in line with that of Pouresmaeil and Gholami (2019) and R. Ellis et al. (2001b) and indicates that the teacher and the learners in the investigated context were primarily concerned with lexical forms. One reason for this finding could be the significance of lexical forms in the course of communication. Lexical items generally have a higher communicative value than other linguistic forms and further contribute to the act of communication (Laufer & Goldstein, 2004; Scrivener, 2011). Since the investigated class was meaning-oriented and had a primary focus on communication, lexical forms have received more attention by the teacher and the learners compared to other linguistic forms. Furthermore, as R. Ellis et al. (2001b) state, lexical FFEs typically involve requests for the meaning of words and can easily fit into communicative activities without interrupting the flow of communication unduly. This can also account for why lexical FFEs occurred more frequently than grammatical and phonological FFEs in the investigated meaning-oriented class.
Regarding the novelty of FFEs, the results indicated that while the majority of lexical and phonological FFEs were novel to the learners, only a small number of grammatical FFEs addressed actual linguistic gaps in the learners’ interlanguage. This finding indicates that the majority of the learners’ utterances which were grammatically flawed were indeed mistakes rather than errors. Such erroneous utterances might be the result of learners’ lack of implicit/procedural knowledge of the grammatical structure rather than a complete linguistic gap. This finding further highlights the importance of ascertaining the novelty of FFEs in designing posttests as discussed before. Nevertheless, an interesting area of research could be to investigate whether/the extent to which incidental FonF would contribute to developing learners’ implicit knowledge of grammatical structures about which the learners already have some explicit knowledge.
Regarding the retention rate of linguistic forms addressed by FFEs, the results indicated that the learners could recall the linguistic elements addressed in half of the FFEs in both the short and long terms. This finding points to a rather high retention rate of FFEs, which might be associated with L2 learning. Given that the test items were designed based on the FFEs which the learners presumably had no previous knowledge of further highlights the beneficial effects of this type of instruction. It should, however, be noted that the act of note-taking might have increased the learners’ engagement with the FFEs, possibly allowing them to further process the noticed forms. 1 This, in turn, might have led to a higher retention rate. However, to determine whether the act of note-taking was indeed a contributing factor, further research with and without the incorporation of note-taking activities is required.
The overall retention rates found in this study are roughly in line with those of Williams (2001), Loewen (2005), Nassaji (2010, 2013), and L. Gholami (2022), all of which (except L. Gholami, 2022) occurred in ESL contexts. Some researchers (e.g. Y. Sheen, 2004) argue that learners may benefit more from FonF in EFL contexts where they are more oriented to attending to linguistic forms. The findings of this study, along with those of L. Gholami (2022), however, suggest that incidental FonF is equally effective irrespective of the context in which it occurs. One reason for this finding may be the micro-level context of this study, which was highly communicative and meaning-oriented. Indeed, an informal interview conducted after the second posttest revealed that the learners mostly viewed the main objective of the class to improve their fluency and communicative skills. Thus, it could be argued that what is more important than the macro-level context (e.g. EFL vs. ESL) is the characteristics of the classroom in that context. This postulation is also echoed by Pouresmaeil and Gholami (2019) and Wang and Li (2021), who found no significant difference in the overall uptake and repair rate following corrective feedback between EFL and ESL contexts, arguing that similar classroom characteristics and objectives, such as improving communicative competence, in EFL and ESL contexts may result in similar uptake and repair rates.
What, however, distinguishes this study from those in the literature is its focus on the retention rate of FFEs with different linguistic foci. As far as vocabulary is concerned, the findings indicated that the learners could recall the meaning of half of the lexical items addressed by FFEs. They could also use the vast majority of the learned lexical items in sentences. The findings were even more encouraging with regard to pronunciation. The learners were able to pronounce over 60% of the target words correctly on the immediate posttest and to a slightly lower extent on the delayed posttest. The results were quite similar for grammatical FFEs. The learners developed explicit knowledge of 50% of the grammatical forms addressed by FFEs. This finding points to a rather equal effect of incidental FonF on developing learners’ knowledge with regard to different linguistic categories in this study. It, however, runs contrary to what Jeon (2007) states about the learnability of linguistic forms. Jean argues that lexical items generally have a higher communicative value and perceptual salience compared to morphosyntactic features, which in turn enhance their learnability. Similarly, studies on noticing have mostly revealed a higher rate of noticing and uptake following lexical and phonological forms. However, this study found FFEs to be almost equally effective irrespective of the linguistic foci. While not neglecting the higher communicative value of lexical forms and their importance in the act of communication compared to morphosyntactic features as explained before, it seems that factors other than merely communicative value may also play a role in the effectiveness of focusing on different linguistic forms. One reason for this rather contradictory finding could be the explicitness of FFEs. In Jeon’s (2007) study, as well as many comparative studies on noticing different linguistic features (e.g. Kim & Han, 2007; Lyster, 2001; Mackey et al., 2000; Pouresmaeil & Gholami, 2019; Y. Sheen, 2006), attention to (grammatical) form, occurred in the form of implicit corrective feedback, such as recasts. In such cases, the further saliency of lexical forms compared to morphosyntactic features might contribute to learners’ further noticing of lexical items (Harley, 1995; Nassaji, 2010). However, saliency may not be considered as an advantage as far as attention to form occurs explicitly. As Nassaji (2010) argues, ‘when FonF is more direct and explicit, students are more likely to notice the linguistic forms’ (p. 926). Previous studies on explicitness of corrective feedback have also revealed that explicit forms of a specific type of corrective feedback are generally more beneficial than its implicit forms (Loewen & Philp, 2006; Nassaji, 2009). Similarly, in the present study, attention to grammatical and phonological forms occurred mostly through explicit forms of reactive FonF. Further analysis of the data revealed that although the majority of the grammatical and phonological FFEs were reactive (as opposed to lexical FFEs which were mostly preemptive and explicit), they were almost entirely of the explicit type of corrective feedback, with the teacher writing the FFEs on the board (and thus further enhancing the noticeability of the linguistic forms). Therefore, in essence, there was not a significant difference in the explicitness of the reactive (grammatical and phonological) and preemptive (lexical) FFEs. This might explain why the learners benefited from FonF on different linguistic categories to almost the same extent and promises some pedagogical and theoretical implications. Based on this finding, it might be suggested that, in terms of noticing and learning, the explicitness of FFEs compensates for the lower communicative value of the linguistic forms, which has been proposed as a factor impacting learning. This, however, should not be taken as suggesting no difference in the noticeability of different linguistic items in general nor should it be assumed that there is no difference in the importance of different linguistic features in the act of communication. Obviously, different linguistic forms vary in saliency and communicative value. Lexical items, for instance, are generally more salient to notice than morphosyntactic features and have a higher communicative value compared to other linguistic forms. The higher contribution of lexical items to the act of communication may prioritize treatment of lexical errors compared to morphosyntactic errors. L2 teachers, thus, should consider the significance of different linguistic items in a communicative act when treating different errors in a meaning-oriented activity. Our findings, however, point to the role of explicitness of FFEs in the extent to which incidental FonF may be effective when treating errors which may possibly go unnoticed due to their lower communicative value. We are, however, aware that due to the lack of a comparative group, which was inevitable based on the nature of the study, this suggestion might be premature and warrants further research.
Another finding of the study concerns the contribution of grammatical FFEs to developing learners’ explicit and implicit knowledge. The finding that incidental FonF contributed to development of explicit knowledge more than to implicit knowledge may be simply due to the ease in developing explicit knowledge compared to implicit knowledge. However, that the learners further developed implicit knowledge over time, although not statistically significant, may be due to the extra time required for internal processing, which is considered essential for the conversion of intake to implicit knowledge (Gass, 1997; Nassaji & Fotos, 2004; VanPatten, 1996). This finding is in line with that of Mackey and Goo (2007), who also found further delayed effects of interaction on the acquisition of grammatical features in their meta-analytical study. Some interaction researchers have suggested that interaction targeting grammatical forms may show further effects in the long run (e.g. Mackey & Goo, 2007; Mackey et al., 2000). Our findings provide some support to this suggestion.
VI Conclusions and limitations
The present study was an attempt to further our knowledge of the effectiveness of incidental FonF. While further corroborating the findings of previous studies in the literature, the study revealed that incidental FonF had an almost equal effect on developing learners’ knowledge of different linguistic categories. The reason for this finding was proposed to be the explicitness of the FFEs. However, further research is warranted to ascertain whether the explicitness of the FFEs was the sole contributing factor. The study also found that incidental FonF was more effective in developing learners’ explicit knowledge than their implicit knowledge, although there was nonsignificant evidence of further improvement of implicit knowledge in the long run. Overall, given that attention to form was transient with a focus on a rather wide range of linguistic features with minimum interruption in the course of communication, drawing attention to form through incidental FonF could be regarded as a beneficial technique to be implemented by L2 teachers in communicative classes.
This study, however, has some limitations, which provide some avenues for further research. Firstly, the study was conducted with a rather small number of participants. Thus, any claims based on the findings of this study are supposed to be suggestive. More studies with a larger number of participants are required before we can safely generalize the results. Another limitation of the study concerns its relatively short observation period. Longer observation hours would possibly provide more instances of different types of FFEs with different linguistic foci, which would in turn provide more items to test. One more limitation of the study concerns the lack of novel test items in the outcome measures. Future studies may incorporate test items which do not repeat the exact words occurring in FFEs. Such novel test items would possibly provide us with a better picture of the ability of learners to use the target forms in novel situations, which might be more indicative of L2 development compared to retention rates. Future research may also investigate the effects of different categories of incidental FonF (i.e. reactive and preemptive) on developing learners’ knowledge of different linguistic features. Finally, it is suggested that interested researchers explore the role of individual differences, such as language analytic ability, working memory capacity, anxiety, motivation, etc., in learners’ ability to benefit from incidental FonF provided on different linguistic targets including vocabulary, pronunciation, and grammar.
