Reinhardt (2010): Summary
This is a paper that reviews the theoretical debate behind the pedagogical applications of concordance such as the use of online corpora in L2 teaching with a mind to promote a wider spread use of such corpora. The first issue the paper tackles is some of the criticism directed at corpus use, especially in regard to the theoretical underpinnings of grammar learning or acquisition. At the core of the debate was weather or not statistical data such as what concordance typically uncovers is truly applicable to actual language learning. Some famous linguists such as Chomsky (1957) who where challenging the behaviorist model of grammar learning were promoting ideas of universal grammar or grammar construction theory which were processes that depended less on brute association and more on creation of subconscious rules in the learners mind. As a result the study of corpora was somewhat disdained for a variety of reasons of varying strengths. For example, one criticism was that the scope of a corpus is biased (weak). Another example is that native speaker language use is a unrealistic goal generally speaking, hence the L1 corpus is redundant or limited to students of the highest levels. (medium) Or an argument is that such conscious focused study is unnecessary if grammar is more effectively acquired rather than learned. (strong) Yet despite these arguments Reinhardt points out the value of such information, such as the notion that language is actually not as strictly rules based as it might appear. For example, we can see this with lexical chunks or phraseology. We can see the degree to which words are attracted to one another: a strong attraction is idiomatic while a weak attraction is open. The lines between grammar and vocabulary start to blur to the point where terms like lexicogrammar start making sense. It is here in this space that a corpus starts to show promise as a tool for learning. The second issue is to what extent can such data be leveraged in an L2 classroom. The idea that students can become autonomous researchers in charge of their own learning is one way this tool could have a powerful effect. Or even the idea of creating a corpus of learner data could bring about a much higher awareness of how each student uses language. Of course the tools seem to have a steep learning curve and a certain messiness that makes these tools hard to implement. The third issue is how to bypass these weakness and it is suggested that such tools be a bigger part of teacher training or SLA or simply be more widely used in teaching material or curricula. In fact Conrad (2000) thinks all these methods should be integrated as they are still very compatible with communicative and constructionist approaches to pedagogy.
Perez-Paredes et al (2011): Summary
This paper seeks to quantitatively investigate the claims that other researchers have made about the use of corpora. In particular, they want to showcase the new technology that has made their methodology more empirical than the indirect methods used previously. The research question they sought to answer is to do with whether or not instruction on the use of corpora would make a big difference in completing tasks that would benefit from the use of such a corpora. And how such instruction would affect students online behavior.
They set up 2 groups of university EFL learners: a control group and an experimental group. Each group were given identical tasks but the experimental groups received instructions on how to use a corpus. Both groups were allowed to use the internet as a resource to help complete the tasks and in both cases a plug-in called “fiddler” was used to track students activities including what sites where visited and input typed.
The results didn't show any particular benefit for using using corpus to complete the tasks as both groups had a similar completion rate. The most positive way they could spin this was that using he corpus didn't negatively affect output so it wasn't a complete waste of time. Along these lines it seems likely that Parez-Paredes et al decided to present their information in terms of the efficiency of their instructions in terms of how well students were able to use the corpus. (Not exactly a riveting or revealing topic.) They came to the conclusion that their instructions did indeed positively affect students use of the corpus, but it seems to me that this is simply an exploratory research for the purpose of showcasing the technical recording tools that are now available to researchers.
Compare and Contrast
I found Reinhardt (2010) to be a fascinating treatment of the theory, while Perez-Paredes et al (2011) presented research information that though in some ways useful was also rather disappointing in that they were not successful in showing corpora could have a measurable positive impact on learning outcomes. This seems to reinforce some of the criticism noted in Reinhardt. So even though both papers take a positive attitude towards this technology in neither case is it really clear that such positivity is justified.
In my opinion, in having some acquaintance in using corpora, I would say that the sites I've seen are very clunky and unintuitive. Most student would have a difficult time using such resource as is is both tedious and complex, requiring a lot of instruction and a task that couldn't be done more efficiently some other way. Two of students in Perez-Parendes et al's study (2011) were able to complete the tasks without any use of the corpus at all. Even the students who did use the corpus used it much less often than other websites. Google comes to mind as in many ways this search engine employs similar statistical matching with an considerably more convenient access to relevant information. (At least for the information students where looking for.)
I would say that this technology is probably better used by teachers and material designers at this point though with a more refined interface it could become a tool as useful as a dictionary for students. Or perhaps the functions of such a tool could be more completely integrated into an online dictionary. In some cases the functions already overlap with lists of common idioms under entries. Therefore, it wouldn't be unreasonable for a dictionary to include a list of strong collations, or even the opposite, words that never collate perhaps as way to help learners avoid common mistakes.
Clarification Question
I would like to know what exactly is the difference between concordance and corpora. I couldn't figure it out contextually.
Application Question
The research presented didn't give a detailed account of how the corpus was used to resolve the learning tasks at hand. So I want to know exactly what information the students are able to get from the corpora and how that information applied directly to resolving a task.
References
Pascual Perez-Paredes, Maria Sanchez-Tornel, Jose Maria Alacaraz & Pilar Aguado Jimenez (2001):
Tracking learners’ actual uses of corpora: guided vs non guided corpus consultation, Computer Assisted Language Learning, 24:3, 233-253
Reinhardt Jonathon (2010): The Potential of Corpus-Informed L2 Pedagogy, Studies in Hispanic and Lusophone Linguistics, Volume 3, Issue 1, Spring 2010
Choamsky, Noam. 1957. Syntactic structures. The Hague: Mouton
Conrad, Susan. 2000. Will corpus revolutionize grammar teaching in the 21st century? TESOL Quarterly 34,548-560
