AATSEEL 2023

Vinokurova, V., & Novikov, A. (2023, February). Multiliteracies approach to task design using a learner corpus. AATSEEL 2023 (online).

This presentation demonstrated the task-text-task cycle using the concept of “Available Designs” borrowed from the multiliteracies approach (Cope & Kalantzis, 2009). Applying this framework to our context, the cycle of “Available Designs” starts with a teacher creating a task (which can be located in the MACAWS Repository). Next, learners engage in the process of “Designing”: they respond to the task by designing their texts. The “Redesigned,” or the products of learners’ work, can be located in the MACAWS Corpus and further used to create new tasks for learners. In this way, as per the multiliteracies framework, students’ “Redesigned” becomes someone else’s “Available Designs.” Crucially, by creating tasks on the basis of a learner corpus, teachers can quickly find level-appropriate texts for their students and create powerful and confidence-boosting activities, drawing on the experience of other students. 

Teaching Seminar for Teachers of Portuguese as an Additional Language

Sommer-Farias, B., & Centanin-Bertho, M. (2023, February). MACAWS - The creation of assignments with a learner corpus . UFRGS 2023 (online).

This presentation introduced MACAWS to instructors of Portuguese as an additional language from the Federal Univsersity of Rio Grande do Sul in Brazil. After being introduced to the corpus and the concepts of data-driven learning, attendees could navigate the corpus and discuss possible uses of that resources for the courses they teach.

ACTFL 2021

Sommer-Farias, B., Centanin Bertho, M., Vinokurova, V., & Staples, S. (2021, November). How to teach writing with learner corpus data. ACTFL 2021 (online).

Writing in a foreign language is considered one of the most difficult skills to teach. This challenge can be associated with two factors: 1) difficult access to authentic texts that are suitable to topics and proficiency levels of FL learners, and 2) the need for more concise descriptions of the linguistic resources that represent content across genres focused on FL instruction. Texts produced by learners made available through learner corpora are one alternative to access authentic texts on more relatable topics and a more appropriate level of vocabulary and grammar. This presentation will share examples of activities created to teach writing to novice and intermediate students in Portuguese and Russian classes using one learner corpus. The presentation will contextualize corpus-based activities created for the topic "travel and tourism" focusing on motion verbs, accusative and prepositional phrases for Russian, and past perfect and expressions for recommendations for Portuguese.

AILA World Congress of Applied Linguistics 2021

Sommer-Farias, B., & Picoral, A. (2021, August). Teacher perceptions on Open Educational Resources: Interactive DDL & MACAWS. AILA World Congress of Applied Linguistics, Groningen, the Netherlands (online).

CALICO 2021

Picoral, A., Sommer-Farias, B., Novikov, A., & Staples, S. (2021, May). Interactive Data-Driven Learning (iDDL): A report on creating and using interactive corpus-based activities. CALICO, Seattle, WA, United States (online). 

AATSEEL 2021

Novikov, A. (2021, March). Using a learner corpus as a medium between unfocused and focused tasks in task-based language teaching. American Association of Teachers of Slavic and East European Languages (AATSEEL) (online).

AATSEEL 2021

Novikov, A. (2021, March). Syntactic and morphological complexity measures as markers of L2 Russian development. American Association of Teachers of Slavic and East European Languages (AATSEEL) (online).

Hispanic and Lusophone Linguistics Working Group Colloquium

Centanin-Bertho, M. (2021, February). Working with learner corpus-data: An introduction to MACAWS. Hispanic and Lusophone Linguistics Working Group Colloquium, University of Arizona (online).

Talk at Michigan State University

Staples, S. (2021, February). Using corpora for pedagogy and research. Invited talk at Michigan State University (online).

TaLC 2020

Picoral, A. (2020, July). Copula constructions in a multilingual learner corpus as evidence of L3 development. 14th Teaching and Language Corpora (TaLC), Perpignan, France (online).

TaLC 2020

Novikov, A. (2020, July). Lexico-grammatical development of Russian learners. Teaching and Language Corpora (TaLC), Perpignan, France (online).

TaLC 2020

Sommer-Farias, B., Picoral, A., Novikov, A., Staples, S. (2020, July). Teachers’ Perceptions of Interactive DDL Using a Multilingual Learner Corpus. Teaching and Language Corpora (TaLC), Perpignan, France (online).

AATSEEL 2020

Novikov, A. (2020, February). MACAWS Russian: Corpus design and creation of usage-inspired pedagogical materials. American Association of Teachers of Slavic and East European Languages (AATSEEL), San Diego, CA, United States.

This presentation introduced MACAWS (Russian), and how it represents a valuable resource for creating usage-inspired pedagogical materials. The rationale behind using texts from this corpus for creating pedagogical materials is rooted in the principles of ecological validity and situational context that guided the corpus design. More specifically, the texts in the corpus come from student assignments with information about tasks collected from naturally occurring pedagogical settings.

The situational context also proves useful in designing activities based on the premise that situational characteristics of texts are directly linked to the linguistic features (Biber, 1988; Biber et al, 2015). Although the functionality of linguistic features has been thoroughly investigated in research, its pedagogical application is still lacking. Thus, this presentation discussed using learner corpora as a resource for making meaningful connections between grammar and lexis, grammar and culture, and grammar and task. These connections were demonstrated through the presentation of three activities developed with the use of MACAWS.

NWAV 2019

Picoral, A., & Carvalho, A. (2020, October). The acquisition of preposition+article contractions in L3 Portuguese among different L1- speaking learners: A variationist approach. NWAV 2019, Eugene, OR, United States.

This paper sheds light on differential paths of third language (L3) acquisition of Portuguese by Spanish-English speakers whose first language is Spanish (L1 Spanish), English (L1 English), or both in the case of heritage speakers of Spanish (HL). Specifically, we look at the acquisition of a categorical rule in Portuguese, where some prepositions are invariably contracted with the determiner that follows them. Based on a corpus of 841 written assignments by Portuguese L3 learners, we extracted 10,047 tokens in obligatory contraction contexts. We analyzed the impact of linguistic (type of preposition and lexical frequency) and extra linguistic factors (course level, learner’s L1 and L2), with individual as random factor, using Rbrul (Johnson, 2008). Results point to clear tendencies, albeit abundant individual differences. L1 English and HL speakers acquire contractions at higher rate than L1 Spanish speakers, revealing a non-facilitatory role of a cognate L1 in transfer patterns during L3 acquisition. 

EMEP 2019

Sommer-Farias, B. (2019, August). MACAWS e o ensino da escrita: O uso de um corpus de textos de aprendizes de português para a criação de materiais didáticos. VIII Encontro Mundial sobre o Ensino de Português (EMEP), Princeton University, Princeton, NJ, United States. 

This study introduces a  Portuguese learner corpus, MACAWS (Multilingual Academic Corpus of Assignments Writing and Speech), and demonstrates how the platform can be used to create teaching materials to support writing classes. To date, MACAWS contains texts produced by 193 students in five Portuguese courses at the University of Arizona (UA), covering 41 different tasks, totaling 256,430 tokens, in addition to metadata about each student. Since many University of Arizona students are bilingual English-Spanish, and some have Spanish and Portuguese as their heritage language, this platform has the potential to contribute to the development of Portuguese teaching materials in similar contexts. The use of the corpus is exemplified through two form-focused learning activities, which include reviews and travel reports, to demonstrate how to establish links across linguistic resources, textual genres (Boulton, 2009) and levels of proficiency. Portuguese language teaching and the elaboration of pedagogical tasks on such analyzes are based on data-driven learning principles and on the noticing of linguistic patterns (Boulton; Cobb, 2017). The results offer support for the contextualized teaching of linguistic resources, enabling the creation of tasks appropriate to the learners' prior knowledge and the improvement of writing through the induction of noticing linguistic patterns relevant to the genres of discourse.

AZCL 2018

Sommer-Farias, B., Novikov, A., Picoral, A., & Staples, S. (2018, October).  MACAWS: Multilingual Academic Corpus of Assignments Writing and Speaking. Arizona Corpus Linguistics conference (AZCL), Flagstaff, AZ, United States. 

This paper describes the process of building the Multilingual Academic Corpus of Assignments - Writing and Speech (MACAWS) under the mentorship of Dr. Shelley Staples, who has extensive experience with corpus building, such as CROW (Corpus & Repository of Writing) (Kwon, Partridge & Staples, 2018). MACAWS is designed to be an online platform for researchers and instructors to access students’ assignments and pedagogical materials from language programs at the University of Arizona, starting with Portuguese and Russian. Few corpora are built directly from assignments in language classes, especially in Less Commonly Taught Languages (LCTLs) (Spina, 2017), limiting the ways in which instructors and researchers can use corpora in their classes. In addition to a corpus, MACAWS will also include a repository of pedagogical artifacts (e.g., syllabi, lesson plans) associated with the assignments in the corpus. Hence, this project offers new support to SLA and language teaching research for LCTLs through learner corpora, allowing teachers and researchers to investigate a large amount (256,430 tokens for Portuguese writing; 16,496 tokens for Russian writing) of representative learner data (Granger, 2002). The presentation will discuss: 1) the process of collecting assignments and processing files to build the corpus; 2) involvement of instructors at the initial stage of assignment submission and their participation in Fall 2018 workshops on how to use the corpus and corpus-based materials for language teaching. A preliminary set of pedagogical materials for both languages will be discussed.

AACL 2018

Picoral,  A. (2018,  September). Compilation and automated annotation of a Portuguese learner corpus: Challenges and lessons learned. American Association of Corpus Linguistics, Atlanta, GA, United States.