Introduction

Oral narratives have been considered rigorous tasks which demand various aspects of language form, meaning and use, and provide rich and reliable data on oral language proficiency (McFarland, 2013; Zanchi et al., 2020). Therefore, narrating requires several linguistic, communicative, and cognitive skills such as oral discourse coherence and fluency. Oral narrative performance is referred to as “the ability to tell a rich, coherent and well-structured story” (Zanchi et al., 2019, p. 3). Furthermore, oral narrative performance has been associated with high academic achievement and literacy skills (Pinto et al., 2016). In the past five decades, numerous studies (e.g., Appose & Karuppali, 2018; Karusoo-Musumeci et al., 2021; Kunnari et al., 2016; Mills et al., 2013; Zanchi et al., 2020) have addressed oral narratives as an effective instrument to assess first language (L1) and/or second language (L2) oral proficiency in children with or without language impairment such as stuttering, and to predict academic achievement, especially in reading and writing skills.

However, little research has been conducted on oral narratives as potential tasks to longitudinally improve some of the inherent features of oral proficiency in young adult English as a Foreign Language (EFL) learners, such as oral discourse coherence and fluency. Therefore, this study aimed at exploring and describing the development of Chilean EFL preservice teachers’ oral discourse coherence and fluency, through the implementation of oral fictional and personal experience narrative tasks over four months. The research questions that guided this study were: 1) Is it possible to establish a relation between oral narrative tasks and oral discourse coherence and fluency improvement over the course of four months? and 2) How do incoherence and dysfluency indicators affect oral narrative performance in participants’ oral narratives?

Literature review

Oral narratives

Oral narratives are defined as the accounts of connected events that can be expressed through personal experiences or fictional stories (Boudreau & Chapman, 2000; Hessling & Brimo, 2019; Salamanca González, 2015). Oral narrative production includes a variety of communicative skills and mastery of pragmatics, syntax, and semantics (Suggate et al., 2018). Oral narrative tasks promote naturalistic language skills, providing researchers with better chances to describe strengths and weaknesses on the use of language components (Granados Ramos et al., 2019; Karusoo-Musumeci et al., 2021). Thus, oral narratives are commonly used at schools and significant associations have been found between oral narrative performance and academic achievements (Lucero, 2016; Zanchi et al., 2020).

Oral narrative samples can be analyzed from two different perspectives: macrostructural and microstructural (Appose & Karuppali, 2018; Kunnari et al., 2016). The macrostructure of narratives “includes statements regarding time, place, and characters and temporal sequencing of narrative events” (Mills et al., 2013, p. 2). Consistently, classic investigators of narratives such as Labov and Waletzky (1997) argued that the following elements fully structure oral narratives:

• orientation (time, place, situation, and participants)
• complication (conflict arisen from events and happenings)
• evaluation (meaning and significance of some events)
• resolution (result of the happenings)
• coda (verbal perspective returned to the present)

These constituents of narrative macrostructure are also defined in terms of story grammar (Janssen et al., 2020; Karusoo-Musumeci et al., 2021, which was first proposed by Stein and Glenn (1979). Thus, macrostructure analysis examines higher-order organization and the overall quality and coherence of narratives (Heilmann et al. 2010b; Heilmann et al., 2010c; Janssen et al., 2020; Kahveci & Güneyli, 2020; Karusoo-Musumeci et al., 2021; Lucero, 2016), all of which is understood as discourse coherence. Consequently, the narrative macrostructure perspective is commonly referred to as story coherence; story coherence (or macrostructure) denotes the way the events are connected to each other, and to the global topic of the narrative (Karlsen et al., 2021).

While narrative macrostructure is referred to as story coherence, microstructure refers to linguistic cohesion, depicting the semantic interactions between sentences, often with an emphasis on word use and linguistic complexity of the narratives (Karlsen et al., 2021). The microstructure perspective implies “the way that words and sentences work together to build a cohesive story. It is determined based on semantic and syntactic productivity, complexity, and accuracy” (Appose & Karuppali, 2018, p. 1). Thus, an oral narrative demands not only the learners’ discourse coherence, but also knowledge of structural language throughout the domains of semantics, syntax, morphology, and phonology (Hughes et al., 1997; Westerveld & Roberts, 2017). In recent years, oral fluency measures have increased in use with Spanish-English oral narrative microstructural analysis (Hessling & Brimo, 2019; Miller et al., 2015; Wood et al., 2018). Although narrative macrostructure and microstructure can be conceptually distinctive, past research has stressed the relevance of examining oral narratives at both macrostructure and microstructure levels while assessing oral narrative performance (Westerveld & Roberts, 2017).

Oral narrative samplings, in accordance with Janssen et al., (2020), provide a “reliable assessment of whether narrative ability will transfer to everyday language use” (p.183). So far, many studies have proved the benefits of oral narrative performance on literacy skills (Karlsen et al., 2021; Schaughency et al., 2017; Spencer & Petersen, 2018; Zanchi et al., 2020). Since the effect of narrative performance on literacy and higher order skills has been largely investigated, longitudinal studies on narrative performance influencing other language skills or components have been suggested as future research areas (Norbury et al., 2014). Therefore, the current study sought to study the role of oral narratives in improving oral discourse coherence and oral fluency in EFL young adult learners.

Discourse coherence

Discourse coherence has gained relevance after decades of being ignored by linguists (Arévalo Balboa & Briesmaster, 2018), and appears as “a key concept, perhaps even the key concept, in discourse…analysis” (Bublitz, 1999, p. 1). Coherence constitutes a significant support to the whole discourse logic and rationality, and it is the most important factor in putting a casual group of utterances into an intelligible discourse; coherence is essential to efficacious discourse (Chatterjee & Chakraborty, 2019; Freed & Cain, 2021). Coherence, regarding narrative discourse, constitutes the temporal and causal structure of the narrative (Karmiloff-Smith, 1985; Pinto et al., 2019). In this respect, discourse coherence denotes the way the events are connected to each other, and to the global meaning of the narrative (Karlsen et al., 2021).

Numerous assessment models have been used for narrative discourse coherence analysis, such as narrative scoring scheme, story grammar analysis, and the narrative assessment profile (Heilmann et al., 2010a; Wood et al., 2018). Yet, the Narrative Assessment Profile (NAP) is a more culturally sensitive measure of narrative discourse in comparison with others; it avoids bias in favor of typically western European narratives (McCabe & Bliss, 2003). Moreover, the NAP was developed by Bliss et al. (1998) to specifically assess discourse coherence and it is pertinent to evaluating the narratives of both children and adults with impaired language or not (McFarland, 2013). That’s why NAP is the most suitable narrative assessment model for the purposes of the present study.

The NAP consists of six dimensions, namely topic maintenance, event sequencing, informativeness, referencing, conjunctive cohesion, and fluency (McFarland, 2013). McCabe and Bliss (2003) suggested that weak event sequencing and informativeness, besides topic maintenance, compromise discourse coherence. McFarland (2013) describes topic maintenance, event sequencing, and informativeness dimensions of the NAP as follows:

• Topic maintenance is concerned with how well all the different utterances in a narrative connect to a main topic, thus avoiding deviations such as insignificant, tangential, or ambiguous utterances that detract attention from the focused theme and disrupt discourse coherence.
• Event sequencing entails the order of presentation of events in chronological or logical order, which corresponds to the order of events in real life.
• Informativeness implies the elaboration necessary to tell a story lacking nothing, which involves information of the type and level of detail needed to for the listener to understand the gist of the story, Informativeness: Police officer (IPO), according to McFarland (2013. p. 139). and the embellishment required to tell a narrative appealing to listeners, as it would be demanded by a teacher. Informativeness: Teacher (TCH) (p. 139).

Conjunctive cohesion, and the fluency dimensions of the NAP, were omitted from the current study. This decision was made as they were considered to be overlapping with other macro and micro structural dimensions, such as fluency.

Fluency

Up to the present date, L2 oral fluency has been defined, classified, and investigated from different perspectives (Huensch & Tracy-Ventura, 2017; Nergis, 2021; Peltonen, 2018). In the 1990s, L2 fluency was distinguished as overall oral language proficiency from a broad sense, and as a single dimension of L2 oral proficiency from a narrow sense (Lennon,1990). More recently, the latter perspective, which is the pertinent one to the current study, was classified by Segalowitz (2010) into three categories of L2 fluency: cognitive fluency, utterance fluency, and perceived fluency.

Cognitive fluency is related to oral performance (Velásquez-Hoyos, 2021), and involves cognitive processes such as lexical retrieval, information access, and articulation, which allow the speaker to deliver the intended message by tailoring the utterances (Nergis, 2021). Utterance fluency refers to the acoustic and temporal measures of L2 oral fluency, namely pauses, hesitations, repetitions, and repairs (Segalowitz, 2016). Lastly, perceived fluency consists of subjective judgments and impressions about L2 oral fluency (Nergis, 2021; Segalowitz, 2016; Velásquez-Hoyos, 2021). 

Cognitive fluency is considered to underlie utterance fluency (Peltonen, 2017; 2018; Segalowitz, 2010) because speakers cannot properly perform oral discourse without the linguistic and cognitive functioning to do so, regardless of the listeners’ perceptions or the speakers’ on L2 oral fluency (Segalowitz & Freed, 2004; Shea & Leonard, 2019). Notwithstanding the interrelation of these three categories, utterance fluency was considered for the purposes of the present study, since it permits an objective measurement of L2 oral fluency as a dimension of oral proficiency (Peltonen, 2018; Segalowitz, 2010; Tavakoli & Uchihara, 2020). In this regard, L2 oral fluency is assumed to be the ease, flow, and continuity of speech, measured by acoustic and temporal disturbances as dysfluency indicators (McDonough & Sato, 2019; Tavakoli & Uchihara, 2020). Research investigating L2 utterance fluency concurs that fluent oral performance is described by the least amount of silent and filled pauses, reformulations, hesitations, false starts, and repetitions (Derwing, 2017; Felker et al., 2019; Nergis, 2021; Wood, 2016).

Skehan (2003) categorized L2 utterance fluency measures into three subdimensions: 1) speed fluency (speech rate, articulation rate, and mean length of run), 2) breakdown fluency (silent and filled pauses), and 3) repair fluency (reformulations, repetitions, and false starts). These measures have been largely used to evaluate L2 utterance fluency and have been commonly operationalized into ratios (speech rate, articulation rate, mean length of run, and number of breakdown and repair dysfluencies to time speaking), and/or counts (number of breakdown and repair dysfluencies in speech) (Shea & Leonard, 2019). In the specific case of breakdown dysfluencies (pauses), besides frequency measures, location and duration can also be considered (Huensch & Tracy-Ventura, 2017).

In addition, establishing the length of silent pauses, which includes the threshold and the cap periods, is currently subject to a great deal of discussion in L2 utterance fluency literature (Shea & Leonard, 2019). Some researchers (e.g., Huensch & Tracy-Ventura, 2017; Merlo & Mansur, 2004; Peltonen, 2017) have used a wide variety of thresholds and caps ranging from 0, 1 to 3 seconds, or longer. Other researchers (e.g., Préfontaine & Kormos, 2016; Velásquez-Hoyos, 2021) have not even rated silent pauses quantitatively but perceptibly with qualitative measures of temporary interruption to the stream of speech and stretches of language with a fairly even tempo. In line with Skehan’s (2003) breakdown fluency (pauses) subdimension, the present study assumed that pauses that are dysfluent are the ones which disrupt the normal flow of the speech (silent pauses longer than 4 seconds), and not the natural ones (filled pauses such as mmm, uh, etc., and silent pauses shorter than 5 seconds) that are produced as a cause of breathing or separating and organizing the chunks of ideas. In addition, it was revealed that longer silent pauses (≥ 1 sec.) are better dysfluency markers of L2 utterance fluency in oral proficiency than shorter silent pauses and filled pauses, respectively (Shea & Leonard, 2019).  

Methodology

The research design of this investigation consisted of a longitudinal, and exploratory case study of Chilean EFL preservice teachers’ oral narratives.

Participants

The 26 EFL preservice teachers were from the University of Concepción, Chile. There were 20 females, five males, and one participant who preferred not to state gender. Their ages ranged from 18 to 24 and each participant accepted to participate voluntarily in the study by signing an informed consent form after having been provided with thorough information about the research. Participants’ native language is Spanish, and they had completed the first semester of their EFL major as a prerequisite to participate in the study, so that they could deal with the linguistic and communicative demands of the oral narrative tasks. Participants’ initial level of EFL oral proficiency was not assessed, which constitutes a limitation of this study. Also, it is important to state that there is no entry-level proficiency requirement to be admitted to the EFL major; however, most preservice teachers hold A1 or A2 levels of oral proficiency at the end of their first semester because the courses are designed to reach those levels, according to the Common European Framework of Reference for Languages. 

Data elicitation techniques and procedure

For the elicitation of oral fictional narratives the method of telling a story about a picture was used (McFarland, 2013), in which participants were asked to generate a story using picture stimuli. The pictures depicted familiar topics to the participants in order to facilitate the use of familiar vocabulary, and consequently, spontaneous oral production. The other method was through oral personal experience narratives carried out by asking the participants to recount a particular personal event (McFarland, 2013; Pavlenko, 2008). Both methods, oral fictional and personal experience narrative tasks, were conceived to be meaningful to the participants in accordance with their ages, in order for them to be motivated by the tasks and communicate their ideas coherently and fluently. The whole process of data elicitation was developed in English, including both the tasks and participants’ responses.

Oral fictional and personal experience narrative tasks were submitted to expert judgement agreement to determine their reliability (Refer to Appendix 1 for a detailed description). As a result, for oral fictional narrative tasks, all tasks turned out to be 100% “highly clear” and 100% “highly coherent” in accordance with all raters. In the case of oral personal experience narrative tasks, Task 1 and Task 2 turned out to be 100% “highly clear” and 83% “highly coherent” while Task 3 was exactly the opposite: 83% “highly coherent” and 100% “highly clear”. Task 4 was considered to have 100% of both, “high clarity”, and “high coherence” (Refer to Appendix 2 and Appendix 3).  

Oral narrative samples were collected in four sessions during four consecutive months, that is, Tasks 1 (fictional and personal narratives) in the first month, Tasks 2 in the second month, Tasks 3 in the third month, and Tasks 4 in the fourth month. In the first week of each month, oral narrative tasks were carried out and then participants’ oral samples were collected through audio recordings. During the second and the third weeks of each month, oral samples were transcribed and analyzed through the rubrics described below in the data analysis techniques section. Finally, in the fourth week of each month, the participants were sent their oral recorded narratives, transcripts, and the analytic rubrics with their respective analyses as feedback. The data collection, through the four oral narrative tasks, constituted the intervention of the study, excluding any separate treatment such as eliciting oral narratives in class or a lab during those four months. Participants continued taking their regular English courses, based on B1 and B2 levels of proficiency according to the Common European Framework of Reference for languages.

Data analysis techniques

Oral narratives samples were recorded, and then transcribed in alignment with the procedure used by Brand and Götz (2011). They included all the utterances that participants produced, including reformulations, repetitions, false starts, and fillers such as, er, well, hmm, um, etc. However, nonverbal behavior such as coughing, and sighing were not included in the transcripts.

The Narrative Assessment Profile (McCabe & Bliss, 2003) was used as the analytic rubric to measure oral discourse coherence (topic maintenance, event sequencing and informativeness). This rubric was adapted from McFarland’s (2013) Narrative Assessment Profile Modified Coding Criteria and the structural elements (orientation, complication, and evaluation) of oral narratives (Labov & Waletzky, 1997). The Narrative Assessment Profile evaluates oral discourse coherence as follows:

Dimensions of the NAP are rated on a three-point scale (0-2), to describe behavior that is “appropriate”, “variable”, or “inappropriate”. A designation of appropriate (given 2 points) signifies that the behavior occurs frequently enough to promote and maintain discourse coherence. Variable behavior (given 1 point) in a given dimension indicates that the level of performance occasionally reduces discourse coherence but that the narrator does demonstrate some strengths in that dimension. A behavior is considered inappropriate (and is given 0 points) when its frequency diminishes or compromises discourse coherence. (McFarland, 2013, p. 55)

It is important to establish that there are two types of informativeness: informativeness: police officer (IPO) and informativeness: chef (ICH)—the correct amount of action, description and evaluation) within the Narrative Assessment Profile Modified Coding Criteria, and its dimensions were rated inversely (0=appropriate, 1=variable and 2=inappropriate) in the present study. All four dimensions along with their coding criteria are fully described in the analytic rubric (Refer to Appendix 4). The ANOVA parametric test was applied to carry out the inferential statistical analysis to compare oral discourse coherence improvement through task performance across the four data collection points. Descriptive statistics was implemented to illustrate the occurrences of oral discourse coherence shortcomings. 

The fluency analytic rubric was elaborated on the bases of Skehan’s (2003) measures of fluency, all of which were considered as dysfluency indicators for the purpose of this study. The rubric measured the following dysfluency indicators: 1) silent pauses, 2) reformulations, 3) repetitions and 4) false starts (Refer to Appendix 5 for a detailed description). The dysfluencies were computed to carry out the statistical analysis. To carry out the inferential statistical analysis for oral fluency, Friedman’s ANOVA nonparametric test was applied to determine if there were statistically significant improvement along the four data collection points.

Results

This section of the article presents the results in accordance with the research questions of the present study.

Is it possible to establish a direct relation between oral narrative tasks and oral discourse coherence and fluency improvement over the course of four months?

The ANOVA parametric test carried out for oral discourse coherence showed that the four months of oral narrative tasks implementation had a statistically significant effect on oral discourse coherence improvement, F(2.104)=8.268,p=0.001. Mauchly’s Test of Sphericity indicated that the assumption of sphericity had been violated, χ¹(5)=15.074, p=0.10, and therefore, a Greenhouse-Geisser correction was used. Table 1 displays the Tests of Within-Subjects Effects where the mean difference is significant at the 0.05 level.

Table 1: Tests of within-subjects effects for oral discourse coherence

Hence, the alternative hypothesis-H1 (there is a statistically significant improvement in oral discourse coherence after four months of oral narrative tasks implementation) is accepted and null hypothesis-Ho (there is no statistically significant improvement in oral discourse coherence after four months of oral narrative tasks implementation) is rejected. The Partial Eta Squared (η²_p=0.249) from Table 1 above means that oral narratives tasks are 25% effective on oral discourse coherence improvement, which represents a large effect size in Social Sciences (Field, 2013). Table 2 below reflects the statistical pairwise comparisons between the first and the rest of the four data collection points of the study for oral discourse coherence.

Table 2: Pairwise comparisons for oral discourse coherence

It is observed in Table 2 that data collection point 3 (p=0.045; m=0.43; SD=0.216) and data collection point 4 (p=0.002; m=0.33; SD=0.194) present a statistically significant improvement in comparison with data collection point 1 (m=0.64; SD=0.375). Thus, a statistically significant effectiveness of oral narrative tasks implementation on oral discourse coherence improvement was evidenced from data collection point 3 on.

The Friedman’s ANOVA nonparametric test carried out for oral fluency showed that there was a statistically significant difference regarding the frequency of dysfluencies along data collection points 1 and 4. Table 3 shows the test statistics of Friedman’s ANOVA where p=0.002<0.05.

Table 3: Test statistics for fluency

Hence, H1 (there is a statistically significant improvement in oral language fluency after four months of oral narrative tasks implementation) is accepted and Ho (there is no statistically significant improvement in oral language fluency after four months of oral narrative tasks implementation) is rejected. Furthermore, Table 4 below reflects the statistical pairwise comparisons between the first and the rest of the four data collection points along the study.

Table 4: Pairwise comparisons for fluency

It is observed in the previous table that data collection point 3 (p=0.018; m=5.04; SD=6.791) and data collection point 4 (p=0.003; m=4.27; SD=5.008) present a statistically significant improvement in comparison with data collection point 1 (m=11.08; SD=9.204). Thus, a statistically significant improvement was evidenced in oral fluency from data collection point 3 on. These findings demonstrated that a direct relation between oral narrative tasks and oral discourse coherence and fluency longitudinal improvement might be established.

How do incoherence and dysfluency indicators affect oral narrative performance in participants’ oral narratives?

In addition, a descriptive analysis of the rating scores was carried out for oral discourse coherence considering the four data collection points; where 0 means appropriate discourse coherence, 1 signifies variable discourse coherence, and 2 implies inappropriate discourse coherence. Table 5 shows the mean, the mode, and the mode frequency per each dimension (TM=Topic Maintenance, ES=Event Sequencing, IPO= Informativeness: Police Officer, and ICH=Informativeness: Chef) of the Narrative Assessment Profile Modified Coding Criteria.

Table 5: Descriptive analysis of oral narrative performance

As evidenced in Table 5, oral discourse coherence was appropriate according to topic maintenance (0) and event sequencing (0) for both types of oral narrative tasks. IPO maintained oral discourse coherence in fictional narrative tasks (0=appropriate), but occasionally reduced oral discourse coherence in personal experience narrative tasks (1=variable). The most frequent rating score for ICH was 1 (variable oral discourse coherence) for both narrative tasks.

ICH was the most negatively affected dimension since participants occasionally omitted either the complication (story conflict) or the evaluation (significance of some events) in their narratives. The indicator complication was more frequently missed in oral personal experience narratives, and the indicator evaluation, in oral fictional narrative tasks. Perhaps participants might have had weak management of the linguistic resources required to tailor such kinds of utterances or might still be developing story schema in the target language. When comparing the behavior of oral discourse coherence between the types of oral narratives, oral personal experience narratives were the most negatively affected. As oral personal experience narratives occasionally lacked the complication, that omission implied the absence of the resolution (result of the happenings) of the narratives often, which was the most negatively affected indicator in IPO. Thus, oral personal experience narratives did not always fit the Narrative Assessment Profile. Figure 1 below constitutes a personal experience narrative transcription sample, in which there was neither complication nor resolution.

Figure 1. Oral personal experience transcription from Participant 1

Regarding the descriptive analysis of oral fluency, there was an emerging dysfluency that did not match any of the predetermined dysfluency indicators. The interesting feature about this dysfluency is that it was sort of an in-between mixture of a reformulation and a false start. For instance, Participant 2 produced an utterance which was automatically aborted and restarted on realizing a potential mistake. Therefore, the participant realized they were making a mistake, then they aborted the mistake in progress and replaced it with the correct form. Therefore, it was not a reformulation because no mistake was really made to be reformulated, but it was not a false start either since the abortion-restart was not related to a change of mind or idea, only to the awareness of the mistake to be made. Table 6 provides a comparison among a reformulation, a false start, and the emerging dysfluency.

Table 6: Comparison of dysfluencies

It is observed that in the reformulation the mistake is fully made and then corrected, and in the false start there is an abortion and restart of a similar idea. However, in the emerging dysfluency there is no mistake but the awareness of it, then there is no correction and there is no restart of an idea either, but there is a replaced utterance of a possible mistake which was recognized and aborted while tailoring the utterance. This kind of dysfluency behaved in three different manners in the present study.

1. When the speakers recognized an error in progress and automatically aborted it and replaced it with the correct form (e.g., from Participant 5: “…we had rea… we had already everything…”; e.g., from Participant 6: “…they sh… they they could go out …”).
2. When the speakers believed they were making a mistake which was not a mistake indeed, but it was automatically aborted and replaced anyway (e.g., from Participant 7: “…he felt rea…very bad…”; from Participant 8 “…there were ma… some marshmallows…”).
3. When the speakers believed they were making a mistake, then attempted to start reformulating it, but automatically aborted the reformulation when recognizing it was not a mistake indeed, so the speaker resumed the expression (e.g., from Participant 9: “…one of my friends hid, sorry, yeah and one of my friends hid in...”).  

To date, the features of this emerging dysfluency have not been recognized in the field of oral fluency. Therefore, such an emerging dysfluency is defined in the present study as a false reformulation, which refers to the speaker’s recognition or false recognition of a mistake while producing an utterance which is automatically aborted and restarted with no change of idea in an attempt of self-correction.

Discussion

These findings of this study suggest a relation between oral narrative tasks and oral discourse coherence and fluency longitudinal improvement. This might be explained by the fact that oral narrative performance has been associated with high academic achievement and literacy skills (Pinto et al., 2016); which is why oral discourse coherence and fluency improved longitudinally. Consistent with the present study, the EFL subjects of the study conducted by Saeedi and Kazerooni (2014) significantly improved their oral fluency, because of the influence of narrative task structure and repetition. In a similar way, Javad Ahmadian and Tavakoli (2011) conducted a study that was largely focused on examining the effects of simultaneous use of careful online oral fictional narrative planning and task performance repetition on fluency in EFL learners’ oral proficiency. The results attained from one-way ANOVAs showed that the engagement in careful simultaneously online oral fictional narrative planning and task performance repetition improves fluency significantly, just as in the current study. Likewise, some researchers (e.g., Foster & Tavakoli, 2009; Skehan & Foster, 1999; Tavakoli, 2009; Tavakoli & Foster, 2011) have found that an oral fictional narrative task structure is correlated to oral fluency.

As for oral discourse coherence, a large number of of studies (e.g., Appose & Karuppali, 2018; Kunnari et al., 2016; Mills et al., 2013; Zanchi et al., 2020) have used oral narratives with the purpose of eliciting oral language in order to transversely analyze developmental stages of narrative discourse patterns, linguistic constraints in children and adults with or without language impairment, and sociocultural identities. Yet, no previous research has been found to use oral narratives as a means of improving oral discourse coherence longitudinally in EFL.

The descriptive analysis of the results for oral discourse coherence indicate that oral personal experience narratives do not always fit the Narrative Assessment Profile. The exclusion of the resolution (results of the happenings) in oral personal experience narratives was partly because of the absence of the complication (story conflict). Regarding this issue, it is worth considering that in a narrative of real events (oral personal experience narrative), sometimes the conflict is not necessarily expressed because the events are narrated just the way they happen, no dilemma must be developed to narrate the characters’ experience. On the contrary, an oral fictional narrative (nonreal/imaginary events) is more likely to be embellished with a conflict for the story to fulfill its purpose. In other words, if there is no conflict, there is no possible fictional narrative to be told, and there is no possible resolution for personal experience narratives. Participants demonstrated appropriate topic maintenance and event sequencing, variable IPO and ICH.

Soto and Hartmann (2006) found that topic maintenance was assessed to be appropriate more than any other dimension in all narratives and all subjects of their study, similarly to the present study’s results. Also, participants presented appropriate event sequencing throughout the shared reading and wordless picture book elicitation tasks. In the conversational narrative and the story stem narrative tasks, event sequencing was assessed to be variable. Such results related to event sequencing differed a bit from the ones of the present study, in which this dimension turned out to be the most appropriate considering its mean score. In the case of IPO and ICH, participants from the study of Soto and Hartmann (2006) fluctuated from inappropriate to appropriate rates. On the one hand, some participants “lacked structure… concrete supporting details… evaluative and elaborative comments in the narratives were limited or absent” (Soto & Hartmann, 2006, p. 468); on the other, some other participants “provided details to their narratives in a variety of forms... supporting details was a clear strength (Soto & Hartmann, 2006, p. 469). Thus, IPO and ICH, referred to as explicitness, resulted as variable discourse coherence, in partial congruency with the present study.

The study of Fiestas and Peña (2004) analyzed the narratives of Spanish-English speakers from Hispanic backgrounds and informed that more than 75% of narratives samples incorporated consequences whereas less than 50% possessed an ending. The latter result means that IPO was the most underachieved dimension, in complete conformity with the results of the current research. As for the former finding, it implies that ICH outperformed the other dimensions, unlike what was found in the present investigation.

Nordberg et al. (2015) encountered that 87% of the subjects of their study demonstrated appropriate topic maintenance and reached appropriate event sequencing scores. As for explicitness (IPO and ICH), 60% were incapable of producing fully explicit and elaborate narratives. Likewise, in McFarland’s (2013) study, participants “were fairly evenly distributed in terms of their performance on topic maintenance and event sequencing with notably fewer demonstrating appropriate levels of informativeness” (p. 138). The few existing studies (McFarland, 2013; Nordberg et al., 2015; Soto & Hartmann, 2006; Tsai & Chang, 2008) that were found on the databases of indexed journals with these measures, are congruent with the current study in the fact that IPO and ICH are the most problematic dimensions relating oral discourse coherence in conformity with the Narrative Assessment Profile. 

Limitations of the study

This study had three major limitations. Firstly, participants’ initial level of oral proficiency was not assessed, which constitutes a missing controlling factor in oral discourse coherence and oral fluency improvement over time. However, the first data collection point of the study somewhat served as such, depicting participants’ initial level of oral proficiency. Secondly, after each data collection of oral narrative tasks participants were provided with feedback; feedback could also influence oral discourse coherence and oral fluency improvement along with other confounding variables which were not controlled. Thirdly, as case study research, there was no control group to ensure that findings are generalizable or directly influenced by the intervention. 

Recommendations for future research

Although oral narratives constituted a valid and reliable instrument to improve overall oral proficiency over time, they were supported by feedback; therefore, it is recommended to implement this instrument (oral narratives) by itself since feedback could also influence oral proficiency improvement. Additionally, it would be interesting to investigate oral narratives within a task-based language teaching approach to improve oral narrative performance.

Also, it would be productive to replicate this study with a greater number of participants and with a control group to obtain findings that are generalizable. Furthermore, it is recommended to adjust the Narrative Assessment Profile Modified Criteria for it to be suitable for both types of oral narratives measurement. Finally, the findings of this case study encourage an avenue for further research within this exploratory scope of oral narratives in EFL.   

Conclusions

After analyzing EFL Chilean preservice teachers’ oral discourse coherence and oral fluency through the implementation of oral narrative tasks over the course of four months, it could be concluded that oral narrative tasks might constitute a valuable instrument for enhancing both constructs over time. A significant impact of oral narrative tasks on oral discourse coherence and fluency improvement was evident since data collection point 3, along with a large effect size on oral discourse coherence.

Oral personal experience narratives did not always fit the Narrative Assessment Profile, especially in the case of ICH dimension, in which the complication (story conflict) indicator was occasionally excluded, thus leading to the exclusion of the resolution (result of the happenings) often. In a narrative of real events (oral personal experience narrative), sometimes the conflict is not necessarily expressed because the events are narrated just the way they happened, no dilemma must be developed to narrate the characters’ experience. However, a fictional narrative (nonreal/imaginary events) is more likely to be embellished with a conflict for the story to fulfill its purpose.

Hence, the Narrative Assessment Profile may be a technical cause for oral personal experience narratives to be rated more negatively than oral fictional narratives. In addition, the evaluation (explicit meaning of some events with respect to others) indicator of ICH dimension was the most frequently omitted in both oral personal experience and fictional narratives. These results led to conclude that participants might still be developing story representation in the target language or did not have enough control of the linguistic resources needed to verbalize such utterances.

Finally, an emerging dysfluency that was sort of an in-between mixture of a reformulation and a false start arose from participants’ oral narrative performance and was defined in the present study as false reformulation. This finding represents an important contribution to the state of the art of L2 utterance fluency; false reformulation constitutes a new measure and dysfluency indicator of L2 oral fluency with implications in L2 teaching and research.

References

Appose, A., & Karuppali, S. (2018). Decoding the macrostructural form of oral narratives in typically developing children between 6 – 11 years of age: Using story grammar analysis. Online Journal of Health and Allied Sciences, 17(65), 1-12. https://www.ojhas.org/issue65/2018-1-12.pdf

Arévalo Balboa, F., & Briesmaster, M. (2018). “Claim – Support – Question” Routine to foster coherence within Interactive real communication among EFL students. Profile: Issues in Teachers´ Professional Development, 20(2), 143-160. http://dx.doi.org/10.15446/profile.v20n2.63554

Bliss, L. S., McCabe, A., & Miranda, A. E. (1998). Narrative Assessment Profile: Discourse analysis for school-age children. Journal of Communication Disorders, 31(4), 347-363. https://doi.org/10.1016/S0021-9924(98)00009-4

Boudreau, D. M., & Chapman, R. S. (2000). The relationship between event representation and linguistic skill in narratives of children and adolescents with Down syndrome. Journal of Speech Language and Hearing Research, 43(5), 1146-1159. https://doi.org/10.1044/jslhr.4305.1146

Brand, C., & Götz, S. (2011). Fluency versus accuracy in advanced spoken learner language: A multi-method approach. International Journal of Corpus Linguistics, 16(2), 255-275. https://doi.org/10.1075/ijcl.16.2.05bra 

Bublitz, W. (1989). Topical coherence in spoken discourse. Studia Anglica Posnaniensia, 22, 31-51. http://hdl.handle.net/10593/17938

Chatterjee, R., & Chakraborty, J. (2019). Analyzing discourse coherence in Bengali elementary choras (children’s nursery rhymes). Rupkatha Journal on Interdisciplinary Studies in Humanities, 11(3). https://dx.doi.org/10.21659/rupkatha.v11n3.06

Derwing, T. M. (2017). L2 fluency development. In S. Loewen & M. Sato (Eds.), The Routledge handbook of instructed second language acquisition (pp. 246-259). Routledge.

Felker, E. R., Klockmann, H., & De Jong, N. H. (2019). How conceptualizing influences fluency in first and second language speech production. Applied Psycholinguistics, 40(1), 111-136. https://doi.org/10.1017/S0142716418000474

Field, A. (2013). Discovering statistics using IBM SPSS statistics. Sage.

Fiestas, C. E., & Peña, E. D. (2004). Narrative discourse in bilingual children: Language and task effects. Language, Speech, and Hearing Services in Schools, 35(2), 155-168. https://doi.org/10.1044/0161-1461(2004/016)

Foster, P., & Tavakoli, P. (2009). Native speakers and task performance: Comparing effects on complexity, fluency, and lexical diversity.Language Learning, 59(4), 866-896. https://doi.org/10.1111/j.1467-9922.2009.00528.x

Freed, J., & Cain, K. (2021). Assessment of inference-making in children using comprehension questions and story retelling: Effect of text modality and a story presentation format. International Journal of Language & Communication Disorders, 56(3), 637-652. https://doi.org/10.1111/1460-6984.12620

Granados Ramos, D. E., Castañeda Landa, L. L., & Mora Hernández, P. (2019). [Análisis de narraciones orales en preescolares con signos neurológicos blandos y dificultades de lenguaje [Analysis of oral narratives in preschoolers with soft neurological signs and language difficulties]. Lenguaje y textos, 50, 87-96. https://doi.org/10.4995/lyt.2019.11154

Heilmann, J., Miller, J. F., & Nockerts, A. (2010a). Sensitivity of narrative organization measures using narrative retells produced by young school-age children. Language Testing, 27(4), 603-626. https://doi.org/10.1177/0265532209355669

Heilmann, J., Nockerts, A., & Miller, J. F. (2010b). Using language sample databases. Language, Speech, and Hearing Services in Schools, 41(1), 84-95. https://doi.org/10.1044/0161-1461(2009/08-0075)

Heilmann, J., Miller, J.F., Nockerts, A., & Dunaway, C. (2010c). Properties of the narrative scoring scheme using narrative retells in young school-age children. American Journal of Speech-Language Pathology 19(2), 154-166. https://doi.org/10.1044/1058-0360(2009/08-0024)

Hessling, A., & Brimo, D.M. (2019). Spoken fictional narrative and literacy skills of children with Down syndrome. Journal of Communication Disorders, 79, 76-89. https://doi.org/10.1016/j.jcomdis.2019.03.005

Huensch, A., & Tracy-Ventura, N. (2017). L2 utterance fluency development before, during, and after residence abroad: A multidimensional investigation. The Modern Language Journal, 101(2), 275-293. https://doi.org/10.1111/modl.12395

Hughes, D. L., McGillivray, L., & Schmidek, M. (1997). Guide to narrative language: Procedures for assessment. Thinking.

Janssen, L., Scheper, A., De Groot, M., Daamen, K., Willemsen, M., Vissers, C., & Verhoeven, L. (2020). Narrative group intervention in DLD: Learning to tell the plot. Child Language Teaching and Therapy, 36(3), 181-193. https://doi.org/10.1177/0265659020950386

Javad Ahmadian, M., & Tavakoli, M. (2011). The effects of simultaneous use of careful online planning and task repetition on accuracy, complexity, and fluency in EFL learners’ oral production. Language Teaching Research, 15(1), 35-59. https://doi.org/10.1177/1362168810383329

Kahveci, G., & Güneyli, A. (2020). The effect of personal narrative intervention on late talkers’ personal narrative generation skills. Cogent Education, 7(1), 1-14. https://doi.org/10.1080/2331186X.2020.1758288

Karlsen, J., Hjetland, H. N., Hagtvet, B. E., Braeken, J., & Melby-Lervåg, M. (2021). The concurrent and longitudinal relationship between narrative skills and other language skills in children. First Language, 41(5), 555-572. https://doi.org/10.1177/0142723721995688

Karmiloff-Smith, A. (1985). Language and cognitive processes from a developmental perspective. Language and Cognitive Processes, 1(1), 61-85. https://doi.org/10.1080/01690968508402071

Karusoo-Musumeci, A., Pearce, W. M., & Donaghy, M. (2021). The effect of workshop training on rater variability in children’s oral narrative assessment. Child Language Teaching and Therapy, 38(1). https://doi.org/10.1177/02656590211023839

Kunnari, S., Välimaa, T., & Laukkanen-Nevala, P. (2016). Macrostructure in the narratives of monolingual Finnish and bilingual Finnish–Swedish children. Applied Psycholinguistics, 37(1), 123-144. https://doi.org/10.1017/S0142716415000442

Labov, W., & Waletzky, J. (1997). Narrative analysis: Oral versions of personal experience Journal of Narrative and Life History, 7(1-4), 3-38. https://doi.org/10.1075/jnlh.7.02nar 

Lennon, P. (1990). Investigating fluency in EFL: A quantitative approach. Language Learning, 40(3), 387-417. https://doi.org/10.1111/j.1467-1770.1990.tb00669.x

Lucero, A. (2016). Oral narrative retelling among emergent bilinguals in a dual language immersion program. International Journal of Bilingual Education and Bilingualism, 21(2), 248-264. https://doi.org/10.1080/13670050.2016.1165181

McCabe, A., & Bliss, L. S. (2003). Patterns of narrative discourse: A multicultural, life span approach. Pearson.

McDonough, K., & Sato, M. (2019). Promoting EFL students’ accuracy and fluency through interactive practice activities. Studies in Second Language Learning and Teaching, 9(2), 379-395. https://doi.org/10.14746/ssllt.2019.9.2.6

McFarland, L. (2013). Tell me a story: Scoring and analysis of the English oral narrative skills of second grade Spanish-speaking English-language learners. [Doctoral dissertation, The University of Texas at Austin]. https://repositories.lib.utexas.edu/handle/2152/21173

Merlo, S., & Mansur, L. L. (2004). Descriptive discourse: Topic familiarity and disfluencies. Journal of Communication Disorders, 37(6), 489-503. https://doi.org/10.1016/j.jcomdis.2004.03.002

Miller, J. F., Andriacchi, K., & Nockerts, A. (2015). Assessing language production using SALT software: A clinician’s guide to language sample analysis. SALT Software.

Mills, M. T., Watkins, R. V., & Washington, J. A. (2013). Structural and dialectal characteristics of the fictional and personal narratives of school‐age African American children. Language Speech and Hearing Services in Schools, 44(2), 211-223.https://doi.org/10.1044/0161-1461(2012/12-0021)

Nergis, A. (2021). Can explicit instruction of formulaic sequences enhance L2 oral fluency? Lingua, 255, 1-20. https://doi.org/10.1016/j.lingua.2021.103072

Norbury, C. F., Gemmell, T., & Paul, R. (2014). Pragmatics abilities in narrative production: A cross-disorder comparison. Journal of Child Language, 41(3), 485-510. https://doi.org/10.1017/S030500091300007X

Nordberg, A., Dahlgren Sandberg, A., & Miniscalco, C. (2015). Story retelling and language ability in school-aged children with cerebral palsy and speech impairment. International Journal of Language & Communication Disorders, 50(6), 801-813. https://doi.org/10.1111/1460-6984.12177

Pavlenko, A. (2008). Narrative analysis. In L. Wei & M. G. Moyer (Eds.), The Blackwell guide to research methods in bilingualism and multilingualism (pp. 311-325). Blackwell.

Peltonen, P. (2017). Temporal fluency and problem-solving in interaction: An exploratory study of fluency resources in L2 dialogue. System, 70, 1-13. https://doi.org/10.1016/j.system.2017.08.009

Peltonen, P. (2018). Exploring connections between first and second language fluency: A mixed methods approach. The Modern Language Journal, 102(4), 676-692. https://doi.org/10.1111/modl.12516

Pinto, G., Tarchi, C., & Bigozzi, L. (2016). Development in narrative competences from oral to written stories in five-to seven-year-old children. Early Childhood Research Quarterly, 36(3^rd Quarter), 1-10. https://doi.org/10.1016/j.ecresq.2015.12.001

Pinto, G., Tarchi, Ch., & Bigozzi, L. (2019). Promoting narrative competence in kindergarten: An intervention study. Early Childhood Research Quarterly, 47(2^nd Quarter), 20-29. https://doi.org/10.1016/j.ecresq.2018.09.003

Préfontaine, Y., & Kormos, J. (2016). A qualitative analysis of perceptions of fluency in second language French. International Review of Applied Linguistics in Language Teaching, 54(2), 151-169. https://doi.org/10.1515/iral-2016-9995

Saeedi, M., & Kazerooni, S. R. (2014). The influence of task repetition and task structure on EFL learners' oral narrative retellings. Innovation in Language Learning and Teaching, 8(2), 116-131. https://doi.org/10.1080/17501229.2013.770860

Salamanca González, F. O. (2015). Personal narratives: A pedagogical proposal to stimulate language students’ writing. HOW, 22(1), 65-79. https://doi.org/10.19183/how.22.1.134

Schaughency, E., Suggate, S., & Reese, E. (2017). Links between early oral narrative and decoding skills and later reading in a New Zealand sample. Australian Journal of Learning Difficulties 22(2), 109-32. https://doi.org/10.1080/19404158.2017.1399914

Segalowitz, N. (2010). Cognitive bases of second language fluency. Routledge.

Segalowitz, N. (2016). Second language fluency and its underlying cognitive and social determinants. International Review of Applied Linguistics in Language Teaching, 54(2), 79-95. https://doi.org/10.1515/iral-2016-9991

Segalowitz, N., & Freed, B. F. (2004). Context, contact, and cognition in oral fluency acquisition: Learning Spanish in at home and study Abroad contexts. Studies in Second Language Acquisition, 26(2), 173-199. https://doi.org/10.1017/S0272263104262027

Shea, C., & Leonard, K. (2019). Evaluating measures of pausing for second language fluency research. The Canadian Modern Language Review, 75(3), 216-235. https://doi.org/10.3138/cmlr.2018-0258

Skehan, P. (2003). Task-based instruction. Language Teaching, 36(1), 1-14. http://dx.doi.org/10.1017/S026144480200188X

Skehan, P., & Foster, P. (1999). The influence of task structure and processing conditions on narrative retellings. Language Learning, 49(1), 93-120. https://doi.org/10.1111/1467-9922.00071

Soto, G., & Hartmann, E. (2006). Analysis of narratives produced by four children who use augmentative and alternative communication. Journal of Communication Disorders, 39(6), 456-480. https://doi.org/10.1016/j.jcomdis.2006.04.005

Spencer, T. D., & Petersen, D. B. (2018). Bridging oral and written language: An oral narrative language intervention study with writing outcomes. Language, Speech, and Hearing Services in Schools, 49(3), 569-581. https://doi.org/10.1044/2018_LSHSS-17-0030

Stein, N., & Glenn, C. (1979). An analysis of story comprehension in elementary school children. In R. O. Freedle (Ed.), New directions in discourse processing (pp. 53-120). Ablex.

Suggate, S., Schaughency, E., McAnally, H., & Reese, E. (2018). From infancy to adolescence: The longitudinal links between vocabulary, early literacy skills, oral narrative, and reading comprehension. Cognitive Development, 47, 82-95. https://doi.org/10.1016/j.cogdev.2018.04.005

Tavakoli, P. (2009). Assessing L2 task performance: Understanding effects of task design. System, 37(3), 482-495. https://doi.org/10.1016/j.system.2009.02.013

Tavakoli, P., & Foster, P. (2011). Task design and second language performance: The effect of narrative type on learner output. Language Learning, 61(s1), 37-72. https://doi.org/10.1111/j.1467-9922.2011.00642.x

Tavakoli, P., & Uchihara, T. (2020). To what extent are multiword sequences associated with oral fluency? Language Learning, 70(2), 506-547. https://doi.org/10.1111/lang.12384

Tsai, W., & Chang, C. (2008). “But I first… and then he kept picking” Narrative skill in Mandarin-speaking children with language impairment. Narrative Inquiry 18(2), 349-377. https://doi.org/10.1075/ni.18.1.09tsa

Velásquez-Hoyos, A. P. (2021). Theme-based teaching to promote oral fluency in a university in Colombia. HOW, 28(2), 97-119.https://doi.org/10.19183/how.28.2.555

Westerveld, M. F., & Roberts, J. M. A. (2017). The oral narrative comprehension and production abilities of verbal preschoolers on the autism spectrum. Language, Speech, and Hearing Services in Schools, 48(4), 260-272. https://doi.org/10.1044/2017_LSHSS-17-0003

Wood, C., Wofford, M. C., & Schatschneider, C. (2018). Relationship between performance on oral narrative retells and vocabulary assessments for Spanish-English speaking children. Communication Disorders Quarterly, 39(3), 402-414. https://doi.org/10.1177/1525740117722507

Wood, D. (2016). Willingness to communicate and second language speech fluency: An idiodynamic investigation. System, 60, 11-28. https://doi.org/10.1016/j.system.2016.05.003

Zanchi, P., Zampini, L., & Fasolo, M. (2020). Oral narrative competence and literacy skills. Early Child Development and Care, 1(14), 2309-2320. https://doi.org/10.1080/03004430.2019.1572132