Journal of Linguistics and Language Teaching (JLLT)

Journal of Linguistics and Language Teaching
Volume 10 (2019) Issue 2, pp. 157-181

The Language-Learning Potential of Writing for Learners of

Spanish: Languaging and Task Complexity 

in Foreign Language Writing

Marcela Ruiz-Funes (Statesboro (GA), USA)

Abstract (English)

This investigation explored the processes involved in languaging during composing and its relationship to task complexity in foreign language writing. It focused on the language-learning potential of writing for learners of Spanish as a foreign language (FL) at a southeast university in the United States. A multiple case study design was adopted. The students worked individually on two writing tasks of different levels of cognitive complexity and were asked to think-aloud while composing. The languaging processing was studied through the analysis of language-related episodes (LREs) based on the think-aloud protocols. The keystroke-logging software Inputlog was also used to gather process data about the writers' behavior to solve the LREs. In addition, students were asked to complete a questionnaire to gauge their perception of task complexity. The data analysis employed a mixed-methods approach. The LREs, records from Inputlog, and the data from the task perception questionnaire were analyzed quantitatively and qualitatively. Patterns regarding the interplay between task complexity and languaging started to emerge. These indicated that the learners were involved in languaging processes, mainly with a focus on lexis followed by form, yet with some variation depending on task demands and level of proficiency of the learner. The paper ends with a discussion of the benefits and challenges of the methodological approach employed.

Keywords: Foreign language writin, languaging, language-related episodes, task complexity, think-aloud protocol, Inputlog

Abstract (Español)

Esta investigación exploró los procesos involucrados en "languaging" (el modo de emplear el lenguaje) durante la composición y su relación con la complejidad de la tarea en la escritura en una segunda lengua. Se enfoca en el potencial de aprendizaje de idiomas de la escritura para estudiantes de español como lengua extranjera (LE) en una universidad del sureste de los Estados Unidos. Se adoptó un diseño de estudio de casos múltiples. Los estudiantes trabajaron individualmente en dos tareas de escritura de diferentes niveles de complejidad cognitiva y se les pidió que pensaran en voz alta mientras componían. El procesamiento de languaging se estudió mediante el análisis de episodios relacionados con el lenguaje (LRE) basados​en los protocolos de pensamiento en voz alta. El software de registro de pulsaciones de teclas Inputlog se utilizó para recopilar datos de proceso sobre el comportamiento de los escritores para resolver problemas relacionados con los LREs. Además, se pidió a los estudiantes que completaran un cuestionario para evaluar su percepción de la complejidad de la tarea. El análisis de datos empleó un enfoque de métodos mixtos. Los episodios relacionados con el lenguaje, los registros de Inputlog y los datos del cuestionario de percepción de tareas se analizaron cuantitativamente y cualitativamente. Comenzaron a surgir patrones con respecto a la interacción entre la complejidad de la tarea y el lenguaje. Estos indicaron que los aprendices estaban involucrados en procesos de lenguaje, principalmente con un enfoque en el aprendizaje seguido por la forma, pero con algunas variaciones dependiendo de las demandas de la tarea y el nivel de competencia del aprendiz. El documento termina con una discusión de los beneficios y desafíos del enfoque metodológico empleado.

Palabras claves: Escritura en un idioma extranjero, modo de emplear el lenguaje, episodios relacionados con el lenguaje, complejidad de la tarea, protocolo de pensar en voz alta, Inputlog

1  Introduction

The potential of writing for second / foreign language (L2/FL) learning has gained renewed attention in recent years (Byrnes & Manchón 2014, Manchón 2011, Manchón & Roca de Larios 2011, Ortega 2012, Williams 2012). It is claimed that writing fosters language development because of the availability of time it offers and the attention to language use and processing paid by learners in meaning-making tasks (Byrnes & Manchón 2014). Primarily, it has been proposed that writing helps in L2 / FL learning and development due to its cyclical nature (Kormos 2014). This cyclical process not only leads to closer attention to linguistic elements (Manchón & Roca de Larios 2007a), but also engages learners in languaging, i.e. the "dynamic, never-ending process to make meaning" (Swain 2006: 96) that leads to conscious reflection on language use and noticing of gaps in linguistic knowledge.

The notion of languaging, as coined by Swain (2006) and used in this study, emphasizes the role of language in mediating cognition, as an agent in the making of meaning, and a way to push L2 / FL learning to an advanced level. In languaging about language, be it in the form of dialogue as in collaborative writing or private speech as in individual composing, "we can observe learners operating on linguistic data and coming to an understanding of previously less well-understood material" (Swain 2006: 98). Some examples of languaging include using language to find which word would be best to use or how to structure a sentence to mean what the writer wants to express (Swain 2006).

In addition, in writing as a meaning-making, cognitive phenomenon (Byrnes & Manchón 2014), a factor that holds a prominent role in language pedagogy is task complexity, i.e.  “the result of attentional, memory, and other information processing demands imposed by the structure of the task on the language learner” (Robinson 2001: 29). Research has shown that task complexity affects how students balance their attention to cognitive processes and linguistic demands (Byrnes & Manchón 2014, Kormos 2011, Kuiken & Vedder 2007, 2008, 2011, 2012, Révész et al. 2017, Ruiz-Funes 2014a, 2014b, Zalbidea 2017.

A connection between languaging and task complexity in L2 / FL writing is expected; however, no empirical studies have been conducted to investigate this link. To help close this gap and to bring further understanding on the notion of writing as a possible context for L2 / FL learning, the present study focuses on the processes of languaging and its relationship to task complexity in FL writing based on the data from multiple case studies. Six learners of Spanish at the university level in the United States were the participants. They individually completed two writing tasks of different levels of cognitive complexity and were asked to think aloud while composing.  The languaging processing was explored through the analysis of language-related episodes (LREs) based on learners’ think-aloud protocols. Following earlier work (Kowal & Swain 1994, Swain & Lapkin 1995, Storch 1998), an LRE was defined as:

... segments in the transcript where students discussed, offered repairs or sought confirmation on an item of language, be it vocabulary, grammar... Each LRE dealt with one linguistic item, although it was possible for one episode to be embedded in another or for an episode to deal with two items simultaneously. An LRE generally began with some form of identification of a problematic item either by way of a question, confirmation request or repair and ended when the issue was resolved either successfully or unsuccessfully. (Storch 1998: 179)

The keystroke-logging software Inputlog (Leijten & Van Waes 2013) was used to complement the data from the think-aloud protocols on the strategies the participants used to resolve their LREs. In addition, participants were asked to fill out a questionnaire to gauge their perception of task complexity, namely which task was more cognitively demanding and required greater mental effort (Norris & Ortega 2003, Révész 2014). The data analysis employed a mixed-methods approach. Data from the LREs, the Inputlog source analysis, and the task perception questionnaire were subjected to both quantitative and qualitative analysis.

2  Literature Review

2.1 Languaging in Second / Foreign Language (L2 / FL) Writing

Most of the work on languaging in L2/FL writing was conducted using collaborative writing in which students worked in pairs on tasks such as text reconstruction, short compositions (Garcia Mayo 2002a, 2002b, Storch 1998, 2001, 2008, Wigglesworth & Storch 2009), and translations (Källkvist 2013). This line of research explored LREs with a focus on lexis, form, and mechanics as well as identified strategies used by L2 writers in solving LREs (Storch, 1998).

In Storch's studies (1998, 2001, 2008), ESL learners of different Asian L1 backgrounds at an Australian university performed several tasks, including multiple choice, text editing, cloze test, text reconstruction, and a short composition. Findings indicated that the multiple-choice task elicited the highest proportion of LREs, followed by the text reconstruction and the editing tasks. The lowest proportion for grammar LREs was found in cloze tests and short compositions. Results also showed that as the percentages of grammar-related LREs decreased, vocabulary LREs increased. Regarding the strategies used to resolve LREs, some of the more frequent ones included resorting to knowledge sources such as considering meaning, rehearsing grammatical rules, searching the text or the learner's memory or intuition. Participants drew on a combination of these knowledge sources on multiple choice, text reconstruction, and short composition. Storch’s findings aligned with those in Garcia Mayo (2002a, 2002b), who found that text editing, text-reconstruction, and multiple choice tasks generated more grammar LREs than cloze and dictogloss tasks. Källkvist (2013) explored languaging in translation tasks with ESL students at a Swedish university. Results showed high levels of student-initiated interactions with a focus on vocabulary LREs rather than on grammar.

Few studies investigated LREs in individual writing. Comparing LREs in collaborative versus individual writing tasks, Wigglesworth and Storch (2009) identified a positive impact of the collaboration for accuracy with no effect on complexity or fluency. Murphy and Roca de Larios (2010) explored how advanced EFL learners of Spanish background used their mother tongue to solve lexical problems in two individual tasks - a narration and an argumentation task. Based on Grabe's (2001) proposed hierarchy of difficulty of writing purpose, the narrative task was considered as ‘simple’, and the argumentative task, as ‘complex’. Findings showed that the argumentative task resulted in more lexical searches and the use of the participants' L1, which, in turn, triggered increased linguistic complexity. The authors explained these results, based on the tenets of Robinson's (2001) cognition hypothesis. Their study highlights how task features may influence the incidence and nature of language elements during individual writing which is discussed in more detail in the next section.

2.2 Task Complexity and Linguistic Complexity

In the last few years, research on task complexity in L2 / FL writing has gained impetus. Some studies explored the relationship between task complexity and written output (Kormos 2011, Kuiken & Vedder 2007, 2008, 2011, 2012, Révész et al. 2017, Ruiz-Funes 2015, Zalbidea 2017). Findings about the relationship between task complexity and measures of syntactic complexity, accuracy, lexical variation, and fluency varied considerably. One of the earliest studies were conducted by Kuiken & Vedder (2007, 2008, 2011, 2012), who showed a decrease in the number of errors and an increase in lexical variation on more complex tasks, operationalized as more or fewer requirements and types of decision to make in letter writing. They did not find a direct effect between task complexity and syntactic complexity or students’ proficiency levels. Kormos' (2011) study on task complexity in narration showed no significant effect between task complexity and linguistic performance except on the elicitation of more abstract words in the more complex task. Ruiz-Funes (2015) investigated task complexity in essay writing. Findings indicated a positive effect of task complexity on syntactic complexity, accuracy, and fluency for students with higher levels of language proficiency and performance on tasks.

One of the most current and comprehensive studies on task complexity in L2 writing is by Révész et al. (2017) who explored the effects of task complexity on fluency, pausing, and revision behaviors of L2 writers, the cognitive processes underlying such behaviors, and linguistic complexity of the resulting output. Participants worked on argumentative essay tasks, and task complexity was operationalized as absence (‘complex’) and presence (‘simple’) of content support. The keystroke-logging software, Inputlog (Leijten & Van Waes 2013), was used to record and analyze the writing behaviors of the participants. Findings indicated that content support led to significantly less pausing as well as increased revision and linguistic complexity, which "suggest that content support likely reduced processing burden on planning processes, facilitating attention to linguistic encoding" (Leijten & Van Waes 2013: 208).

As was indicated before (Ruiz-Funes 2015), the variability of findings from these studies makes it challenging to establish a coherent body of knowledge about the presumed link between task complexity and attention to language-related elements. This may  in large part be due to the emerging nature of this type of research within the domain of L2 / FL writing. The present study intends to contribute further empirical data to help explore what and to what extent task characteristics in writing account for variation in attention to linguistic factors in the form of languaging. As in the author's previous work (Ruiz-Funes 2014a, 2014b, 2015), in this study, task is defined as a meaning-making, genre-based activity, and cognitive complexity is varied in terms of familiarity of topic, genre, and reasoning demands. Specific features about the tasks used are presented in the next section below.

3  The Study

3.1 Research Questions

The present study explored the processes involved in languaging operationalized by LREs during individual writing and its relationship with task complexity. The following research questions guided the investigation:

Research Question 1

In what types of LREs do FL learners engage as they perform writing tasks of different levels of cognitive complexity? What strategies do they use to resolve the LREs?

Research question 2:

What are the perceptions of FL learners about task complexity and its effect on their attention to linguistic elements while composing?

3.2 Participants

Eight students participated in this study, but only six completed both tasks. As such, findings from the six case studies are reported and allowed for the predominantly qualitative nature of the study. The learners were majors or minors of Spanish at a large southeastern university in the United States enrolled in the same upper-division Spanish course where data collection took place. Table 1 shows the basic demographics of the participants with their assigned pseudonyms used to protect their identity. Their experience studying Spanish (range = 1.5 to 9 years) and their proficiency levels in Spanish (intermediate to advanced) were the two factors that showed the most variation within the group. The proficiency level in writing in Spanish was assessed according to the ACTFL Proficiency Guidelines (ACTFL 2012) based on a writing sample administered before the data collection process started:

Case Study	Age	Gender	First language	Length of time studying Spanish	Proficiency level in Spanish	Spanish Major or Minor
Case 1: Emma	22	Female	English	9 years	advanced low	Minor
Case 2: Meagan	23	Female	English	3 years	intermediate low / mid	Major
Case 3: Kate	21	Female	Spanish Heritage Learner	2 years	advanced low	Major
Case 4: Paul	21	Male	Spanish Heritage Learner	1 1/2 years	intermediate mid	Minor
Case 5: John	21	Male	English	3  1/2 years	intermediate mid	Major
Case 6: Robert	22	Male	English	2 years	intermediate low / mid	Major

Table 1: Participants' Demographics

3.3 Procedure

The students worked individually on each task and were asked to think-aloud while composing. Before the data collection process started, they participated in a training session to get introduced to the think-aloud technique. The training consisted of an explanation of the procedures for thinking aloud while composing, and a short practice session on a mock text about their families. No modeling for the think-aloud method was provided to avoid possible influence on the participants (Manchón et al. 2000). The participants completed the tasks in two class sessions (one on each day in two separate weeks) in a language lab. A graduate assistant and the researcher were present to explain the tasks and the think-aloud procedure. They were allotted 40 minutes to complete each task.

The think-aloud sessions were recorded using the Audacity software, saved as MP3 files and later transcribed in MS Word for analysis. In addition to the think-aloud protocols, the keystroke-logging software Inputlog (Leijten & Van Waes 2013), was used to record the composing process of the writers so as to gather process data on strategy use apt to resolve LREs. Upon completion of the second task, the participants filled out a questionnaire used to gauge their perception of task complexity and its effect on their attention to language-related elements.

3.4 Tasks and Task Complexity

The tasks in this investigation consisted of individual essay writing commonly used in FL programs at the university level in the United States. They were genre-based and emphasized context, audience, purpose, and culturally and socially determined rhetorical and discourse patterns. To vary task complexity in writing, tenets in L2 writing theory and research, namely the role of long-term and working-memory capacity in writing processes, were taken into account (Kormos 2012). Based on the claims posed by writing models (Kellogg 1996), particularly in regards to the role of long-term and working-memory capacity in composing processes, writers’ familiarity with the subject matter, genre, and task type were considered to have a prominent function in how they allocate their attentional resources while composing. As such, task complexity in this study was controlled, taking these features into account, as described below.

Both tasks were framed and contextualized around the theme of study abroad. Task 1 (‘simple’) required the participants to write a personal essay about themselves, their interests, and expected goals for studying abroad using both description and narration. In Task 2 (‘complex’) learners were asked to write an expository essay, formally stating college students’ benefits and challenges of studying abroad. Eight students participated in this study, but only six completed both tasks. As such, findings from the six case studies are reported and allowed for the predominantly qualitative nature of the study.

3.5 Inputlog

From the analysis module which the keystroke-logging software Inputlog (Leijten & Van Waes 2013) offers, the source-analysis report was used to track how the participants accessed digital sources to solve LREs. The report also provided the amount of time in seconds (total and relative) which the writers spent on each source search. An excerpt from one of the participants is provided below that includes two examples of words searched (disappointed  and  tener): 

Window Title	Total Time (s)	Total Time (relative)
google translate - Google Search - Google Chrome	2.730	0.001
disappointed - English-Spanish Dictionary - WordReference.com - Google Chrome	20.796	0.009
conjugation of tener - Diccionario Inglés-Español WordReference.com - Google Chrome	2.808	0.001

Table 2: Excerpt from Inputlog Source Analysis Output 

3.6 Task Perception Questionnaire

The questionnaire consisted of four open-ended questions in which participants were requested to identify and explain which task was more / less difficult and required more / less mental effort: (1) to write overall, (2) to plan and generate ideas, or (3) to revise and edit. The fourth question related to linguistic aspects. Learners were asked to identify and explain which task helped them pay more / less attention to: (a) grammar, (b) sentence structure, (c) word order, (d) vocabulary, (e) cohesive devices, (f) coherence, and (g) other.   

4  Data Analysis

The data analysis employed a mixed-methods approach. Data from the LREs, the Inputlog source analysis, and the task perception questionnaire were subjected to both quantitative and qualitative analysis.

4.1 LREs

The processes of languaging were operationalized with the identification of LREs which were captured through the self-speech that resulted from the think-aloud sessions. Following Wigglesworth & Storch’s (2009) data analysis procedure, the transcripts of the think-aloud protocols were analyzed at two levels:

● identification of episodes in transcripts, and

● analysis of LREs.

For the first level, the think-aloud transcripts were segmented into episodes, and the resulting events were assigned into one of the following categories:  

● Task-related: Task clarification and management (for example, students read / discusses instructions on tasks, clarifie tasks requirements, etc.)

● Content: Idea generation and discussion of content (for example, student generate and reformulate ideas / information to use)

● Structure: Organization and ordering of ideas in the text, in a paragraph, sentence, etc.

● Revision: Revision activities (for example, students read or re-read the text they had composed and make comments)

● LREs: Students focuse on some aspect of the language: form, lexis, or mechanics

● Other: Off-task activities

The number of episodes per category and corresponding percentages were calculated per task for each participant separately.

For the second level, a closer analysis of the LREs was conducted. LREs provide insights into learners’ understanding of language as they talk about the language that they produce and self-correct (Wigglesworth & Storch 2009). LREs may capture, for example, how students deliberate over a word choice, the correct conjugation of a verb form, or the correct verb tense to use. LREs are categorized by focus: lexis-focus (L-LRE), form-focus (F-LRE) and mechanics-focus (M-LRE). L-LREs were defined as episodes in which students searched for words or considered alternative expressions, for example. F-LREs involved incidents in which students deliberated over grammar-related issues such as word forms, tense or aspect, or sentence structure. M-LREs focused on the spelling of words or punctuation (Wigglesworth & Storch 2009). The LREs were coded based on their focus as lexis-focus (L-LRE), form-focus (F-LRE) and mechanics-focus (M-LRE) by the researcher and a trained graduate research assistant. Inter-coder agreement for all episodes in the case studies was 92%. Those  episodes that the two coders disagreed upon, were discussed, and a consensus was reached. Due to the small number of participants, only descriptive statistics including count, percentage, mean, standard deviation, and the 95% confidence intervals (CIs) above and below the means were calculated.   

The final step in the LRE analysis was to consider how the participants resolved the LREs. This analysis was guided by Storch's (1998) taxonomy of knowledge source. The taxonomy includes the following strategies:

● application of a grammatical rule: students cite or refer to a grammatical rule;

●meaning of words or phrases or knowledge of the topic: students resort to their understanding of lexis or the topic area and call upon that information to assist in their grammatical choice;

● analogy or memory: students make reference to previously learned or memorized chunks or phrases;

● intuition. students resort to what sounds or seems correct or appropriate (this often presented itself as students stating that something makes / does not make sense or students repeating phrases a few times to note how they sounded or whether they 'sounded right');

● use of contextual clues: students use contextual clues, such as preceding or following words, phrases, sentences or the entire text as the basis for decision;

● no reason given: students do not provide the reason for their choices;

● other: other solving strategy not mentioned above.

In addition, source searches related to solving LREs were gathered from the reports generated by the Inpulog source analysis (Leijten & Van Waes 2013) for each case separately. From these reports, the following features were calculated manually: type of sources accessed (e.g. Word Reference), count (total times sources were accessed), total time spent on searches in minutes, purpose (e.g. word meaning, conjugation of verbs, etc.), and an example to illustrate the search. To prepare the raw data for analysis, any entry unrelated to the LREs was removed from the data files (Leijten et al. 2014). For example, an entry for access to the Audacity software did not relate to LREs. As such, the Audacity entry and others similar to it were removed from the final dataset.

4.2 Task perception questionnaire

Participants' responses regarding task complexity for each of the questions were recorded and tallied. Also, patterns were identified about the reasons they posed to justify their selection for the more or less difficult task to write overall, to plan and generate ideas, and for the task that helped them focus more or less on linguistic aspects.

5  Results

5.1 LREs and Strategy Use

This section first presents an overview of the findings about LREs and strategy use for all cases combined. Then patterns for each individual case study and relevant comparisons among the cases are provided.

Table 3 summarizes the counts and percentages for the different types of episode per task per case. The highest proportion of types of episode was in LREs for most cases on both tasks followed by Content and Revision. The exception was Meagan who had the largest percentage in Content followed by LREs and Revision under Task 2 ('complex'). 

Type of Episode	Emma		Meagan		Kate		Paul		John		Robert
	T1	T2	T1	T2	T1	T2	T1	T2	T1	T2	T1	T2
Task	5	8	1	4	12	3	4	1	4	1	5	3
	6%	8%	2%	5%	15%	4%	6%	2%	12%	2%	10%	12%
Content	21	23	16	28	19	16	27	17	9	12	16	14
	26%	24%	31%	38%	23%	24%	38%	33%	27%	23%	33%	56%
Structure	7	4	0	0	0	0	0	0	0	0	0	0
	9%	4%	0%	0%	0%	0%	0%	0%	0%	0%	0%	0%
Revision	13	34	14	19	18	19	11	9	9	13	9	4
	16%	36%	27%	26%	22%	28%	15%	17%	27%	25%	19%	16%
LREs	30	24	20	22	32	29	28	25	11	27	16	4
	38%	25%	39%	30%	39%	43%	39%	48%	33%	51%	33%	16%
Other	4	2	0	0	0	0	2	0	0	0	2	0
	5%	2%	0%	0%	0%	0%	3%	0%	0%	0%	4%	0%
Total	80	95	51	73	81	67	72	52	33	53	48	25

Table 3: Count and Percentages of Type of Episodes per Task per Case Study

(Note. T1 = Task1; T2 = Task 2)

Table 4 shows the summary for the different types of LRE per task per case. Higher values of L-LREs resulted in both tasks for all cases except Meagan with the larger proportion in F-LREs. Table 5 includes the descriptive statistics for LREs per task for all cases combined. On both tasks, the L-LREs had higher mean values with a considerable margin of difference compared to the F-LREs and M-LREs. Task 1 (‘simple’) resulted in slightly higher values for total LREs, F-LREs, and M-LREs;  under Task 2 (‘complex’), the mean value for L-LREs was slightly greater. Analysis of variance to compare group mean values for these measures calculated using the 95% confidence intervals (CIs p< .05) indicated no statistically significant difference which may be due to the small sample size.

Type of LREs	Emma		Meagan		Kate		Paul		John		Robert
	T1	T2	T1	T2	T1	T2	T1	T2	T1	T2	T1	T2
L-LREs
	13	12	7	9	25	21	14	15	7	22	13	3
	43%	50%	35%	41%	78%	72%	50%	60%	64%	81%	81%	75%
F-LREs
	10	6	12	12	4	5	4	4	4	5	2	1
	33%	25%	60%	55%	13%	17%	14%	16%	36%	19%	13%	25%
M-LREs
	8	6	1	1	3	3	10	6	0	0	1	0
	27%	25%	0.5%	0.5%	0.9%	0.1%	36%	24%	0%	0%	0.6%	0%
Total	30	24	20	22	32	29	28	25	11	27	16	4

Table 4: Count and Percentages of Type of LREs per Task per Case

Note. T1 = Task1; T2 = Task 2

Types of LREs	Task 1 (n=6)			Task 2 (n=6)
	M	SD	CI (95%)	M	SD	CI (95%)
L-LREs	13.17	6.59	+/− 5.27	13.67	7.26	+/− 5.80
F-LREs	6	4	+/− 3.2	5.5	3.62	+/− 2.89
M-LREs	3.83	4.17	+/− 3.33	2.67	2.8	+/− 2.24
Total LREs	22.83	8.44	+/− 6.75	21.83	9.06	+/− 7.24
Total LREs (%)	38%			36%

Table 5: Means, Standard Deviations, and Percentages of LREs per Task all Cases Combined

Note. M = Mean; SD = Standard Deviation; CI = Confidence Interval

Table 6 provides a breakdown from the Inputlog source analysis per case per task. The following patterns were identified: (a) not every participant accessed digital sources to solve LREs; (b) a higher count of sources accessed was recorded under Task 2 (complex) for most cases; (c) Word Reference was used more frequently over other search sources; (d) the purposes for accessing sources included primarily word translation from English to Spanish, followed by verb conjugation and translation of a few phrases also from L1 to Spanish; and (e) more time was spent on source searches under Task 2.   

Case	Source searched	Count	Time (min.)	Count and Purpose	Example
Emma
Task 1	Google Translate	2	2.56	7-Word 5-Verb Conj. 3- Not Provided	accustomed conj. of tener n/a
	Word Reference	13
Task 2	Word Reference	20	3.23	16-Word 4-Verb Conj.	autumn conj. of realizar
Meagan
Task 1	None	0	0	n/a	n/a
Task 2	Google Search	11	7.56	14-Word 6-Phrase 3-Verb conj. 1-Not provided	lifestyle the way of eating would have n/a
	Word Reference	13
Kate
Task 1	None	0	0	n/a	n/a
Task 2	None	0	0	n/a	n/a
Paul
Task 1	Word Reference	18	3.14	18-Word	abroad
Task 2	Word Reference	27	3.64	23-Word 1-Verb conj. 3-Not provided	challenge prepares n/a
John
Task 1	Word Reference	12	1.74	9-Word 2-Phrase 1-Not provided	gain college major n/a
Task 2	Google Search	8	1.96	6-Word 1-Phrase 1-Nor provided	welcome study abroad n/a
	Word Reference
Robert
Task 1	None	0	0	n/a	n/a
Task 2	None	0	0	n/a	n/a

Table 6: Inputlog Source Analysis Report by Case per Task

A separate analysis of each case revealed the following patterns about types of LRES and strategies used to solve them:   

Emma

Emma is a skillful writer with an advanced proficiency level in Spanish. As shown in Table 4 above, in both of her tasks, the L-LREs made up the largest proportion (43.3% in Task 1 and 50% in Task 2). Under Task 2, the difference between L-LREs and the other types was more prominent. These findings reveal a stronger focus on lexis than on grammar or mechanics on both tasks.

The following example provides insight into the kind of issues that were of concern to Emma and how she resolved them:  

(1)   ...how do you say to experience? Y hacer parte…I need to be a part of… y no y ser...How do you say that? Yo siempre querido estudiar una otra language and to experience y experi- no y tener la experiencia de una cultura diferente de la mía. That’s it, una otra lengua y tener la experiencia, una otra lengua y tener la experiencia de una cultura diferente de la mía.  Estudiar otra lengua y tener la experiencia de una cultura diferente de la mía. (L-LRE Task1)

In this episode, Emma is deliberating over how best to express an idea that she is formulating in English and stumbles over how to say 'experience' in Spanish. She rereads her previous sentence and tries to continue with the idea inserting the word in question in English, she offers alternatives and finally comes up with the correct word. Emma is satisfied and writes it down to complete her thought in Spanish, rereading the lines she had produced for confirmation. In these examples, she solves the issue resorting to her previous knowledge.

The primary focus on the L-LREs was word choice particularly of less commonly used nouns, verbs, and adjectives. The episodes were of several lines in length, which demonstrates that she spent a considerable amount of time to solve her issues and did not move forward until she found a satisfactory answer. In most cases, Emma thought of the word in English and explored options to find the solution. She either looked up the words in the dictionary or resorted to her previous knowledge. She took the time to consider the multiple meanings and the parts of speech offered in the dictionary entry (nouns versus verb as in manivela or manejar for ‘handle’). Some examples also revealed a focus on word choice of more frequently used words such as preparatoria or otoño which she handled in the same way as described above. The same patterns were seen on both tasks. 

  

Most F-LREs (examples (2) and (3) below) show that Emma deliberated primarily over verb tense, verb conjugation, and gender or number agreement. In most instances, she sought confirmation for choices resorting to her prior knowledge or what sounded right to her; at other times, she used the dictionary to solve the issue. Also, there were fewer and shorter (one or two lines) F-LREs in Task 2 than in Task 1:

(2)   um ok no tener…tuve, tuviste, tuvo, tuvimos, tuve? Is that ok I’m gonna look it up because I feel like I’m going crazy because I don’t understand (pause) or cuz I cannot remember the past tense, preterito, tuvieron, oh ok (F-LRE Task1) 

(3)  como un pol- I feel like I am using the wrong tenses in here but I just want to type um que no sabía lo que quería, no sabía lo que, lo que estaba haciendo... (F-LRE Task 2)

Report from the Inputlog source analysis (Table 6) shows that Emma used Word Reference most often. She did 15 searches under Task 1 and 20 in Task 2. Her searches were for word translation (count = 7 in Task 1; count = 16 in Task 2) and verb conjugation (count = 5 in Task 1; count = 4 in Task 2). These results confirm her focus on lexis in Task 2.

Kate

Kate is a heritage learner of Spanish with strong language and writing skills in Spanish also at the advanced level. The L-LREs on both of her tasks accounted for over 70% of all LREs (Table 4). As in Emma's case, this indicates a greater focus on lexis than on grammar or mechanics on both tasks. Below are examples of L-LREs that illustrate the patterns she used on both tasks:    

(4)  sería bueno, how can I say that? sería bueno, también sería bueno, say exposed in Spanish? exponernos, no, you don't say exponernos, exposed, exposed. También sería útil um (pause) uh sería útil, to be exposed to other cultures. También sería bueno conocer otras culturas latinas o hispanas que son di-ferentes a la mía ... (L-LRE Task 1) 

(5) they can practice, how do you say speaking skills? sus habilidades de, no, de comunicación, no, al estudiar en el extranjero, los estudiantes también podrán practicar (pause) to español, con (pause) Mmm practicar la lengua que estén aprendiendo con hablantes nativos. (inhale) (L-LRE Task 2)

Kate found some problems expressing more elaborate ideas such as 'to be exposed to other cultures' in the first example above. As Emma, Kate formulated an idea in English, reread what she wrote before, tried alternative ways to express her thought, and came up with an option in Spanish that seemed satisfactory to her. Kate went back and reread line by line what she had written previously to make sure it conveyed her intended meaning. She solved these issues resorting to her previous knowledge. There is no indication from the think-aloud protocol or Inputlog source analysis report (Table 6) of the use of any digital resource such as an online dictionary. Kate's episodes were long which shows her effort in dealing with the language issues she encountered.

  

Very few F-LREs were identified on both tasks; they were brief of only one or two lines long as in the example below. Kate resorted to her previous knowledge to solve them:

(6) ...es posible que los estudiantes aprendan cosas valiosas en el país extranjero que no podrá, que no hubiera...aprendido en una aula. (F-LRE Task 2)

Paul

Paul is another heritage language speaker of Spanish but with an intermediate-mid level in the language. The L-LREs on both of his tasks had the highest proportions (50% in Task 1 and 60% in Task 2) followed by M-LREs (28% in Task 1 and 25% in Task 2) and F-LREs (14% in Task 1 and 16%  in task 2) (Table 4). As in Emma's and Kate's profiles described above, this indicates a marked focus on lexis. However, in Paul's protocol, an important consideration for mechanics is also noted. This pattern manifested on both tasks.

Below are examples of L-LRE and M-LRE that illustrate the issues that were of concern to Paul and how he resolved them:

(7)  ... quiero saber que quiere decir, que significa la palaba committee en español porque no sé y quiero tener el conocimiento. Entonces, lo estoy buscando en línea...um comisión es la palabra que se usa, entonces, tengo un conocimiento más. (L-LRE Task 1) 

(8)  Mientras espilibro espelea la palabra extranjero, tengo problemas y tengo que refirirme a las oraciones anteriores para poder escribirlas correctamente. (M-LRE Task 2)

As seen in (7) and (8), Paul had problems with specific words like 'committee' in Spanish. He thought of the word in English, and as soon as he realized he did not know its meaning in Spanish, he looked it up in a dictionary without hesitation or delay. In the (8), Paul’s difficulty with spelling words in Spanish is documented. He acted quickly and referred back to how he had spelled it before in a previous section of his essay. His entries on both tasks were brief (two or three lines), which indicates that he had a fast approach to dealing with the issues he encountered. He tended to access resources to solve the LREs instead of relying on his previous knowledge. Similar to Emma's case, Paul accessed Word Reference quite extensively (count = 18 in Task 1; count = 27 Task 2) (Table 6) with a primary focus on word searches.

John and Robert

Findings for John and Robert present some commonalities and some contrasts. Both participants are English speakers with an intermediate level of proficiency in Spanish. As Table 4 shows, the L-LREs had the highest percentages on both of their tasks (John: 65% in Task 1 and 81% in Task 2; for Robert: 81% in Task 1 and 75% in Task 2) followed by F-LREs (John: 36% in Task 1 and 19% in Task 2; Robert: 13% in Task 1 and 25% in Task 2). However, some differences are also noted. For John, Task 2 resulted in a higher total count of LREs, a higher value of L-LREs with a lower percentage of F-LREs. Conversely, for Robert, this pattern in values was seen in Task 1. The following examples illustrate the LREs and strategies they used:

John:   

(9)   beneficio beneficiosa. No sé beneficial (looks up) beneficioso beneficiosa, beneficiosa... (L-LRE Task1) 

(10)  que no es lista que no está lista que no está lista... (F-LRE Task 2) 

Robert: 

(11) puedo mis sonidos o orejas de no sé la palabra para ears ears pero... (L-LRE Task1) 

(12) los pegan y los ellos me pegan pagan no sé pegan dólares (F-LRE Task 1) 

(13) ... una que había I would change I would change que una cosa que cambiar (F-LRE Task 2)

 John and Robert had problems with individual words (L-LRE) and verb conjugation (F-LRE) on both tasks. The difference between the two participants is the approach they used to resolve the issues. John used both the dictionary (9) and resorted to his recollection repeating the phrases and settling for what sounded correct (10). Robert, however, did not seem to worry much about finding a solution to the issues he stumbled upon as he either kept the word in question in English (11) or resorted to his recollection of previous knowledge settling for a quick fix (12) and (13). 

Report from the Inputlog source analysis (Table 6) corroborates these findings. John used Word Reference 12 times for Task 1 and a combination of Google Search and Word Reference (8 times) in Task 2. His searches mainly focused on words (count = 9 in Task1; count = 6 in Task 2) as well as on a few phrases. Robert's Inputlog source analysis, however, indicates no access to any digital resource while completing either task.

Meagan

Meagan had a weaker command of the language and struggled with writing in Spanish. Her level of proficiency was within the intermediate range. Unlike the cases presented above, on both of her tasks, the F-LREs made up the largest proportion (60% in Task 1 and 54.5% in Task 2) followed by L-LREs (Table 4). These findings indicate a greater focus on form than on lexis on both tasks. The following examples illustrate the issues that were of concern to Meagan.  

(14)  ... estoy, no, no, no, um, umm será? yo será? (pause) seré?...yo sere?...Mmm no sé (F-LRE Task 19  

(15) ...algunas de las tiendas habría cerrada? Mmmm no sé...uh de (España) no no, cerrade, cerradia, cerrarr-, ce-rrar-dia... (F-LRE Task 2)

Meagan hesitated over a grammatical issue, paused, offered possible alternatives, settled for one, and continued with her writing. The types of grammar problems she questioned included verb conjugation and verb tense on both tasks.

Meagan followed the same pattern for L-LREs (c). The types of words in question included low-frequency used words (cheerleading, challenging, grits) and some of more high frequency (holiday, start).

no sé...challenging (humming a tune) ah!...challenging...dis-, uh desa-, de-sa-fi-an-te?...uh no, no, no, no, no, no, no...es desfi-, desfi-ante (pause) (L-LRE Task 2)

Her episodes were brief and seemed to indicate that she tried to find a quick solution and move on with her writing. They also showed her frustration and impatience when she encountered these language dilemmas. Based on the Inputlog Source Analysis (Table 6), Meagan did not access any digital resource for Task 1, but she did for Task 2 with 24 searches that included words (count = 14), verb conjugations (count = 3), and some phrases (count = 6).  

5.2 Participants' Perception of Task Complexity

The second goal of this investigation was to explore the perceptions of FL learners on task complexity and its effect on their attention to linguistic elements. As seen in Table 7 below, students' perception was divided. On the one hand, for Emma, Meagan, and Paul, Task 1 was more demanding to write, plan, and generate ideas. Some of the reasons were that they did not know how to approach the topic (Emma) or felt it was challenging to come up with personal plans to build their rationale to study abroad (Meagan and Paul). On the other hand, Kate, John, and Robert viewed Task 2 as more complex. Some of the reasons were lack of personal experience related to the subject (Kate), the topic was too specific (John), or it was too broad (Robert). Participants also differed in which task was more difficult to revise and edit. Emma and John stated that Task 1 was more challenging to revise for precisely the opposite reasons: limited information (Emma) versus more material that could lead to more errors (John). For Kate, Paul, and Robert Task 2 was more demanding to revise and edit because they wrote more and with more specific details.  

The participants' responses about which task helped them focus more on specific linguistic elements were also divided. An interesting observation is that in some cases, the task they viewed as more challenging led them to pay more attention to linguistic features; for others, it was the opposite. For example, for Emma, Task 1 was more complex and the one that helped her focus more on language elements. Kate and Robert found Task 2 as more demanding while Task 1 led to more attention to language features: 

Question	Emma	Meagan	Kate	Paul	John	Robert
Task more difficult:
to write	T1	T1	T2	T1	T2	T2
to plan/generate ideas	T1	T1	T2	T1	T2	T2
to revise and edit	T1	NS	T2	T2	T1	T2
Task that helped focus more on
grammar	T1	NS	T1	T2	T2	T2
sentence structure	NS	NS	T1	T2	T1	T2
word order	NS	NS	T1	T2	T2	T1
vocabulary	T1	NS	T2	T2	T2	T1
cohesive devices	NS	NS	T1	NS	T1	NS
coherence	T1	NS	T2	NS	T1	T1

Table 7: Perception on Task Complexity and Attention to Linguistic Elements

Note. T1 = Task 1; T2 = Task 2; NS = No selection made

6  Discussion

The present study explored the processes involved in languaging during composing and its relationship to task complexity to shed further light on the language-learning potential of L2/ FL writing. Its goal was twofold: first, to identify the LREs and the strategies used by FL learners to resolve them as affected by task complexity, and second, to gain insight into the perceptions of FL learners on task complexity in writing and its role in attention to linguistic elements.

Findings provided evidence of students' attention to linguistic elements in writing which support the notion of writing as a possible context for L2 / FL learning (Byrnes & Manchón 2014, Manchón 2011, Manchón & Roca de Larios 2011, Ortega 2012, and Williams 2012). Although not to the same extent, participants deliberated over language-specific issues, considered possible alternatives, sought solutions, and settled for what they thought was the correct option

Patterns of languaging affected by task complexity were identified. Slightly higher total values for LREs were found under the ‘simple’ condition (Task 1) for all cases combined. Task 1 may have allowed writers to pay attention to processes of formulation of content and revision while, at the same time, focussing on vocabulary searches and morphosyntactic considerations to refine the expression of their ideas (Kellogg 1996, 2001, McCutchen 1996). In Task 2, the ‘complex’ task, however, the lower value for LREs might be an indication that the demands of the task led writers to place more attention to the generation of content and revision processes to prioritize formulation and monitoring over attention to linguistic issues. This effect was more pronounced in Meagan who had a lower level of proficiency compared to the other cases. We may explain this trade-off effect based on Skehan' s Limited Capacity Hypothesis that claims that tasks that are cognitively more demanding "consume more attentional resources … with the result that less attention is available to focus on form" (Skehan (1998a: 97).

Regarding the types of LREs across cases, findings showed that the largest proportion was in L-LREs on both tasks for most participants. This focus on lexis goes in hand with results in Källkvist (2013) and Storch (2001, 2005, 2011) who found that meaning-making tasks generated more attention to lexis than to grammar. As Källkvist (2013) claimed

... whenever students are expected either to generate vocabulary and expressions larger than one lexical item (as in translation and composition writing) [...] a great deal of attention is directed toward vocabulary and expressions (Källkvist 2013: 97).

Only Meagan had a higher proportion of F-LREs on both tasks. This difference in focus may be related to the proficiency level of the student. According to Wigglesworth & Storch (2009), lower proficiency learners tend to concentrate more on grammatical issues than more advanced students. However, this was not the case for the other participants with lower levels of proficiency in the present study.

For types of LREs between tasks, Task 1 (‘simple’) resulted in slightly more attention to grammar and mechanics than Task 2 (‘complex’). In turn, Task 2, led to a greater focus on lexis as compared to Task 1. We can see these results in light of Storch's (1998, 2001, 2008) findings that as F-LREs declined, L-LREs increased, and vice versa. Moreover, the expository essay (Task 2), which is more formal than the personal essay (Task 1), may have pushed writers to think of ideas that were more abstract. This could have required the use of vocabulary that was beyond their frequent use and control and led to a higher number of lexical issues.

The strategies the participants used to solve LREs were affected by learners’ proficiency level in Spanish rather than task complexity. The more advanced students took more time to address the language issues they encountered and resorted to knowledge sources that included considering the meaning of words and phrases, relying on their memory or intuition, rehearsing grammatical forms, and (some of them) using digital sources such as an online dictionary. Students within the intermediate level tended to seek for a quick solution and either accessed a digital source to solve the language issue or left it unattended. The use of a combination of sources identified for the advanced learners aligns with the respective finding in Storch (1998).  

Results also indicated that those participants who accessed an online dictionary searched primarily for translations of words and phrases from English to Spanish and, to a lesser degree, verb conjugations. In addition, findings showed that under Task 2 (‘complex’), participants spent more time accessing sources and engaged in more lexical searches. This supports Murphy & Roca de Larios (2010), who found that the more complex task led to more use of the participants' L1 and a higher number of word searches.

Concerning the participants' perception of task complexity, the findings from the questionnaire were divided and provided insight into the variability and subjectivity of opinions among students about which task required more or less mental effort. Some viewed Task 1 (‘simple’, i.e. personal essay) as more difficult to address because they did not know how to approach the topic from a personal perspective. Conversely, others considered Task 2 (‘complex’, i.e. expository essay) more demanding because they lacked personal experience related to the subject. In addition, the more complex task incited some students to focus more on linguistic elements; for others it was the reverse. In spite of the variation in the outcome about student perceptions of task complexity,  some factors apparently weighted more in their opinions about which task was more or less complex. One such factor was familiarity with or personal experience related to the topic. The other seemed to relate to structure or content support that led participants to regard a task as complex because it was too general or too specific.

7  Conclusion

In the present study, a mixed-method approach was employed which allowed for a comprehensive exploration of languaging processes and task complexity in FL writing. Data was gathered from multiple sources including think-aloud protocols, keystroke logging, and a perception questionnaire. Patterns regarding the interplay between task complexity and languaging started to emerge. Learners were involved in languaging processes, mainly with a focus on lexis (L-LREs), followed by form (F-LREs) with some variation depending on task demands and learners’ respective levels of proficiency.

The following considerations may provide further direction to guide what next steps may be taken to gain a better grasp on languaging and task complexity in L2 / FL writing:

● The level of learner proficiency should become a crucial factor in the planning and design of future studies on task complexity and languaging processes in L2 / FL writing (Ortega 2010). In this study, differences in strategy use to solve LREs seemed to be related to learner proficiency level rather than task complexity.

● Analysis of the content and revision episodes may shed light on the connection between LREs and other process behaviors in L2 / FL writing as affected by task complexity. An exploration of how attention to linguistic processing as evidenced by a focus on LREs relates to other conceptual processes in writing such as content generation and revision may prove revealing. Future research may consider this matter based on the claims by Schoonen et al.:

...in a FL, writers can become so absorbed in linguistic processing, that is, in searching for the right words, the right sentence structures and the right spelling, that they have little eye for conceptual processing, that is for the global content and structure of the text (...). (Schoonen et al. (2009: 86)

● An innovative aspect in this study was the use of the keystroke-logging software Inputlog (Leijten & Van Waes 2013). Particularly, the source analysis report complemented the think-aloud protocols and helped track how participants interacted with digital sources to solve LREs. As Leijten et al. stated, this tool is very informative for researchers as

"[w]ith Inputlog’s 'source analysis', all of the windows that the user opens are identified, and it logs them accordingly (e.g., as different Word documents, webpage URLs, or graphical applications)" (Leijten et al. 2014: 294).

The potential of this type of software for collecting quantitative data about composing processes in which writers engage is promising and can undoubtedly help advance the inquiry on the cognition of L2 / FL writing.

Although the findings are revealing, some limitations to this study are noted. First, the use of the think-aloud method has its shortcomings, including a possible reactivity effect and the challenges it poses to writers to think aloud while composing (Murphy & Roca de Larios 2010). Concurrent protocols could be complemented with stimulated-recall interviews to capture a more thorough account of the languaging processes as learners complete writing tasks of different levels of cognitive demands. Second, a multiple case study as a research design provides the opportunity for an in-depth exploration of a case in question; yet, it has limitations concerning generalizability (Hamel, 1993). And third, the time elapsed between the two tasks might have influenced the results of the perception questionnaire as learners could have remembered more the features of one task over the other.   

References

ACTFL (2012). ACTFL Proficiency Guidelines Writing. American Council on the Teaching of Foreign Languages. (http://www.actfl.org/publications/guidelines-and-manuals/actfl-proficiency-guidelines-2012/english/writing; PDF).

Byrnes, H. & R. M. Manchón (Eds.) (2014). Task-based language learning: Insights from and for L2 writing. Philadelphia & Amsterdam: John Benjamins.

Chenoweth, N. A. & J. R. Hayes (2001). Fluency in writing: Generating text in L1 and L2. In: Written Communication 18, 80-98.

Ellis, R. & F. Yuan (2004). The effects of planning on fluency, complexity, and accuracy in second language narrative writing. In: Studies in Second Language Acquisition 26, 59-84.

García Mayo, M. P. (2002a). Interaction in advanced EFL pedagogy: A comparison of form‐focused activities. In: International Journal of Educational Research 37, 323-341.

García Mayo, M. P. (2002b). The effectiveness of two form‐focused tasks in advanced EFL pedagogy. In: International Journal of Applied Linguistics 12, 156-175.

Grabe, W. (2001). Notes toward a theory of second language writing. In: Silva, T. & P. K. Matsuda (Eds.). On second language writing. Mahwah, NJ: Erlbaum, 39-57.

Hamel, J. (1993). Case Study methods. Qualitative Research Methods. Vol. 32. Thousand Oaks, CA: Sage.

Leijten, M. & L. Van Waes (2013). Keystroke Logging in Writing Research: Using Inputlog to Analyze and Visualize Writing Processes. In: Written Communication 30 (3), 358-392. (DOI: 10.1177/0741088313491692).

Ishikawa, T. (2006). The effect of manipulating task complexity along the (± Here-and-Now) dimension on L2 written narrative discourse. In: García Mayo, C. M. (Ed.). Investigating tasks in formal language learning. Clevedon, UK: Multilingual Matters, 136-156.      

Källkvist, M. (2013). Languaging in Translation Tasks Used in a University Setting: Particular Potential for Student Agency? In: Modern Language Journal 97, 217-238.

Kellogg, R. T. (1996). A model of working memory in writing. In: Levy, C. M. & S. Ransdell (Eds.). The science of writing: Theories, methods, individual differences and applications. Mahwah, NJ: Lawrence Erlbaum, 57-71.                                         

              

Kellogg, R. T. (2001). Competition for working memory among writing processes. In: American Journal of Psychology 114, 175-191.

Kormos, J. (2011). Task complexity and linguistic and discourse features of narrative writing performance. In: Journal of Second Language Writing 20, 148-161.

Kormos, J. (2012). The role of individual differences in L2 writing. In: Journal of Second Language Writing 21(4), 390-403.

Kormos, J. (2014). Differences across modalities of performance: An investigation of linguistic and discourse complexity in narrative tasks. In: Byrnes, H. & R. M. Manchón (Eds.). Task-based language learning: Insights from and for L2 writing. Philadelphia & Amsterdam: John Benjamins, 193-216.

Kowal, M. & M. Swain (1994). Using collaborative language production tasks to promote students’ language awareness. In: Language Awareness 3(2), 73-93.

Kuiken, F. & I. Vedder (2007). Task complexity and measures of linguistic performance in L2 writing. In: International Review of Applied Linguistics in Language Teaching 45, 261-284.

Kuiken, F. & I. Vedder (2008). Cognitive task complexity and written output in Italian and French as a foreign language. In: Journal of Second Language Writing 17, 48-60.

Kuiken, F. & I. Vedder (2011). Task complexity and linguistic performance in L2 writing and speaking: The effect of mode. In: Robinson, P. (Ed.). Second language task complexity: Researching the cognition hypothesis of language learning and performance. Philadelphia & Amsterdam: John Benjamins, 91-104.

Kuiken, F. & I. Vedder (2012). Syntactic complexity, lexical variation and accuracy as a function of task complexity and proficiency level in L2 writing and speaking. In: Housen, A., F. Kuiken & I. Vedder (Eds.). Dimensions of L2 performance and proficiency: Complexity, accuracy and fluency in SLA. Philadelphia & Amsterdam: John Benjamins, 143-170.

Manchón, R. M. (2011). Writing to learn the language: Issues in theory and research. In: Manchón, R. M. (Ed.). Learning-to-write and writing-to-learn in an additional language. Philadelphia & Amsterdam: John Benjamins, 61-82.

Manchón, R. M. (2014). The internal dimension of tasks: The interaction between task factors and learner factors in bringing about learning through writing. In: Byrnes, H. & R. M. Manchón (Eds.). Task-based language learning: Insights from and for L2 writing. Philadelphia & Amsterdam: John Benjamins, 27-52.

Manchón, R. M. & J. Roca de Larios (2007a). Writing-to-learn in instructed language contexts. In: Alcón, E. & P. Safont (Eds.). The intercultural speaker: Using and acquiring English in instructed language contexts. Dordrecht: Springer, 101-121.

Manchón, R. M. & J. Roca de Larios (2007b). On the temporal nature of planning in L1 and L2 composing. In: Language Learning 57, 549-593.

Manchón, R. M. & J. Roca de Larios (2011). Writing to learn in FL contexts: Exploring learners’ perceptions of the language learning potential of L2 writing. In: Manchón, R. M. (Ed.). Learning-to-write and writing-to-learn in an additional language. Philadelphia & Amsterdam: John Benjamins, 181-207.

McCutchen, D. (1996). A capacity theory of writing: Working memory in composition. In: Educational Psychology Review 8, 299-324.

Murphy, L. & J. Roca de Larios (2010). Searching for words: One strategic use of the mother tongue by advanced Spanish EFL writers. In: Journal of Second Language Writing 19, 61-81.

Norris, J. & L. Ortega (2003). Defining and measuring L2 acquisition. In: Doughty, C. J. & M. H. Long (Eds.). The handbook of second language acquisition. Malden, MA: Blackwell, 717-761.

Ong, J. (2014). How do planning time and task conditions affect metacognitive processes of L2 writers? In: Journal of Second Language Writing 23, 17-30.

Ong, J. & L. J. Zhang (2010). Effects of task complexity on the fluency and lexical complexity in EFL students’ argumentative writing. In: Journal of Second Language Writing 19, 218-233.                                                               

                                            

Ortega, L. (2003). Syntactic complexity measures and their relationship to L2 proficiency: A research synthesis of college-level L2 writing. In: Applied Linguistics 24, 492-518.

Ortega, L. (2010). Exploring Interfaces between Second Language Writing and Second Language Acquisition. Plenary address delivered at the 9th Symposium on Second Language Writing. University of Murcia, Spain, May 20-22.

Ortega, L. (2012). Epilogue: Exploring L2 writing–SLA interfaces. In: Journal of Second Language Writing 21, 404-415.

Polio, C. & H. Yoon (2016). Task and genre differences in L2 writing research. Paper presented at the Invited Colloquium, Zhang, L. (Organizer). Researching written task complexity in diverse contexts. AAAL, Orlando, FL, April 9-12.

Ransdell, S. E. & C. M. Levy (1999). Writing, reading, and speaking memory spans and the importance of resource flexibility. In: Torrance, M. & G. Jeffrey (Eds.). The cognitive demands of writing: Processing capacity and working memory in text production. Amsterdam: Amsterdam University Press, 99-113.

Ransdell, S. E. & C. M. Levy (1996). Working memory constraints on writing performance. In: Levy, C. M. & S. E. Ransdell (Eds.). The science of writing. Mahwah, New Jersey: Lawrence Erlbaum Associates, 93-101.

Révész, A. (2014). Towards a fuller assessment of cognitive models of task-based learning: Investigating task-generated cognitive demands and processes. In: Applied Linguistics 35, 87-92.

Révész, A., N.-E. Kourtali & D. Mazgutova (2017). Effects of task complexity on L2 writing behaviors and linguistic complexity. In: Language Learning 67, 208-241.

Robinson, P. (2001a). Task complexity, cognitive resources, and syllabus design: A triadic framework for examining task influences on SLA. In: Robinson, P. (Ed.). Cognition and second language instruction. Cambridge: Cambridge University Press, 287-318.                                    

Robinson, P. (2001b). Task complexity, task difficulty, and task production: Exploring interactions in a componential framework. In: Applied Linguistics 22, 27-57.                            

Robinson, P. (2003). The cognition hypothesis, task design, and adult task-based language learning. In: Second Language Studies 21, 45-105.  

  

Robinson, P. (2005). Cognitive complexity and task sequencing: Studies in a componential framework for second language task design. In: International Review of Applied Linguistics in Language Teaching 43, 1-32.

                                                                                                                

Robinson, P. (2007, September). Re-thinking-for-speaking and L2 task demands: The cognition hypothesis, task classification, and sequencing. Plenary address at the Second International Conference on Task-Based Language Teaching. University of Hawai’i. 

                                    

Robinson, P. (2011). Second language task complexity, the cognition hypothesis, language learning, and performance. In: Robinson, P. (Ed.). Second language task complexity: Researching the cognition hypothesis of language learning and performance. Philadelphia & Amsterdam: John Benjamins, 3-38.

Robinson, P. & R. Gilabert (2013). Task-based learning: Cognitive underpinnings. In: Chapelle, C. (Ed.). Encyclopedia of applied linguistics. Malden, MA: Wiley-Blackwell.

Ruiz-Funes, M. (2015). Exploring the potential of second/foreign language writing for language learning: The effects of task factors and learner variables. In: Journal of Second Language Writing, 28, 1-19.

Ruiz-Funes, M. (2014b). An exploration of task complexity in foreign language writing at the intermediate level and measures of linguistic production. In: Journal of Linguistics and Language Teaching, 5.2, 181-206. 

Ruiz-Funes, M. (2014a). Task complexity and linguistic performance in advanced college-level foreign language writing. In H. Byrnes and R. M. Manchón (Eds.), Task-based language learning and teaching: Insights from and for L2 writing (pp. 163-191). John Benjamins Publishing Company, Philadelphia, PA.

Schoonen, R., van Gelderen, A., De Glopper, K., Hulstijn, J., Simis, A., Snellings, P., & Stevenson, M. (2003). First language and second language writing: The role of linguistic fluency, linguistic knowledge, and metacognitive knowledge. Language Learning, 53, 165-202.

Schoonen, R., Snellings, P., Stevenson, M., & van Gelderen, A. (2009). Towards a blueprint of the foreign language writer: The linguistic and cognitive demands of foreign language writing. In R. M. Manchón (Ed.), Writing in foreign language contexts: Learning, teaching, and research (pp. 77-101). Clevedon, UK: Multilingual Matters.

Skehan, P. (1998a). A cognitive approach to language learning. Oxford: Oxford University Press.

Skehan, P. (1998b). Task-based instruction. Annual Review of Applied Linguistics, 18, 268-286.                                                                      

                                                        

Skehan, P. (2001). Tasks and language performance assessment. In M. Bygate, P. Skehan, & M. Swain (Eds.), Researching pedagogic tasks: Second language learning, teaching, and testing (pp. 167–185). London: Longman.                           

Skehan, P. (2003). Task-based instruction. Language Teaching, 36, 1–14.   

          

Skehan, P. (2014). The context for researching a processing perspective on task performance.  In P. Skehan (Ed.), Processing perspective on task performance (pp. 1-26).  Philadelphia/Amsterdam: John Benjamins.        

    

Skehan, P., & Foster, P. (1999). The influence of task structure and processing conditions on narrative retellings. Language Learning, 49, 93–120. 

                                          

Skehan, P., & Foster, P. (2001). Cognition and tasks. In P. Robinson (Ed.), Cognition and second language instruction (pp. 183–205). Cambridge: Cambridge University Press.

Storch, N. (1998). Comparing second language learners’ attention to form across tasks. Language Awareness, 7 (4), 176–191.

Storch, N. (2001). Comparing ESL learners’ attention to grammar on three different classroom tasks. RELC Journal, 32, 104–124.

Storch, N. (2005). Collaborative writing: Product, process, and students' reflections. Journal of Second Language Writing,14(3), 153-173.

Storch, N. (2011). Collaborative Writing in L2 Contexts: Processes, Outcomes, and Future Directions. Annual Review of Applied Linguistics, 31(1), 275-288.

Swain, M. (1985). Communicative competence: Some roles of comprehensible input and comprehensible output in its development. In S. M. Gass & C. G. Madden (Eds.), Input in second language acquisition (pp. 235-253). Rowley, MA: Newbury House.

Swain, M. (1995). Three functions of output in second language learning. In G. Cook & B. Seidlhofer (Eds.), Principles and practice in applied linguistics (pp. 125–144). Oxford: Oxford University Press.

Swain, M. (2005). The output hypothesis: Theory and research. In E. Hinkel (Ed.). Handbook of research in second language teaching and learning (pp. 471–483). Mahwah, NJ: Lawrence Erlbaum.

Swain, M. (2006). Languaging, agency, and collaboration in advanced second language proficiency. In H. Byrnes (Ed.), Advanced language learning: The contribution of Halliday and Vygotsky (pp. 95–108). London: Continuum.

Swain, M., & Lapkin, S. (1995). Problems in output and the cognitive processes they generate: A step toward second language learning. Applied Linguistics, 16, 371–391.

Williams, J. (2012). The potential role(s) of writing in second language development. Journal of Second Language Writing, 21, 321–331.

Wigglesworth, G. and Storch, N. (2009). Pair versus individual writing: Effects on fluency, complexity and accuracy. Language Testing, 26, 445–466.

Zalbidea, J. (2017), ‘One Task Fits All’? The roles of task complexity, modality, and working memory capacity in l2 performance. The Modern Language Journal, 101, 335–352.

Author:
Marcela Ruiz-Funes, PhD
Assistant Professor
Department of Foreign Languages
Georgia Southern University
Statesboro
Ga 30458
E-mail: mruizfunes@georgiasouthern.edu