Page 11 - elgazette/ELG1812

NEWS
Two cheers for
DAVEYNIN, FLICKR CREATIVE COMMONS TBLT...

By Gillian Ragsdale
Task-based language teaching (TBLT) is an
effective method for language development,
the first ever meta-analysis suggests.
However, the results from the Georgetown
University study raise questions, too.
Lara Bryfonski and Todd McKay used
data from 52 studies from around the world,
published from 1998 to 2016. The Middle
East, however, was over-represented –
contributing 17 of the studies. The L2 was
English in 40 of the 47 studies.
Overall results supported TBLT with a
strong effect size of d = 0.93 (0.4=small,
0.7=moderate). Nearly 80 per cent of
stakeholders were positive about the method.
The strongest effect was seen in the
Middle East and effects in East Asia were
Code-switching disentangled moderate. However Europe, North America
and SE Asia showed lack-lustre effect sizes
under 0.5.
By Gillian Ragsdale Until now, no-one knew which part of TBLT also appeared to be more effective
Suppressing L1 takes more effort than using the process of language switching was more in schools and universities than in the two
two languages at the same time, ground- demanding: switching the new language on private language institutes studied – though
breaking research has shown or switching the old language off. And how a sample of two is too small to be confident
When bilinguals switch from one could you even study this when both things about the difference.
language to another, the burden of effort is have to happen at the same time? For now, teachers would do best to assess
on switching off the language in use, while The team at New York solved this dilemma the needs and context of their own students
switching on the new language is relatively by studying bilinguals whose languages can when deciding whether to use TBLT.
effortless, researchers at New York be used simultaneously: spoken English
University have found. This is the result and American Sign Language (ASL). The n Bryfonski, L. and McKay, T. H. (2017)
of an ingenious study that has managed to participants were shown pictures in rapid ‘TBLT implementation and evaluation: A meta-
unravel this process for the first time, as succession and instructed to name the analysis’. Language Teaching Research, DOI:
reported in PNAS. pictures in one or both languages while their 10.1177/1362168817744389
I am surely not alone in feeling that my brain activity was recorded.
brain is working extra hard when switching It seems intuitive that using two languages …but mind your
between two languages and this kind of at once must surely be more work than using
language switching is generally associated one – but that is not the case. In fact, the meta-analyses
with increased activity in the prefrontal effort of using ASL is actually decreased
cortex of the brain. when English is used at the same time. Linguistic researchers (and teachers) should
However, when switching from using both tread cautiously when using meta-analyses,
languages to only one, activity in key parts writes Rod Ellis, Research Professor of
of the prefrontal cortex increased, reflecting Education at Curtin University. With the
the cognitive effort required. No such exponential growth in published research,
activity was recorded when going from one meta-analyses are increasingly popular.
language to two. Results, however, are often underwhelming
This explains the common phenomenon and may disappoint if the individual studies
of code-blending, where spoken and signed lack power and compatible methodologies.
language are often used simultaneously. The problems start with collecting the
However, one tends to dominate, especially studies: those with positive findings are
in determining syntax. generally over-represented while those with
For language learners the implication negative results often remain unpublished.
is that the burden of effort is in switching Fortunately, Bryfonski and Todd (see above)
off the dominant language in order to are meticulous in outlining this publication
produce only the second language. Once the bias in their own study.
dominant language is switched off, switching The increasing diversity of research
the second language on should be relatively methods makes pooling data difficult. Studies
effortless – which perhaps adds yet another that seem to ask the same questions turn out
strand to the debate on using L1 in the L2 to be either measuring different constructs or
classroom. using different measures. So while the overall
results of meta-analyses can be eye-catching
n Blanco-Elorrieta, E., Emmorey, K., & – practitioners should interpret with caution.
Pylkkänen, L. (2018). Language switching
decomposed through MEG and evidence from n Ellis, R. (2018) ‘Meta-analysis in second
bimodal bilinguals. Proceedings of the National language acquisition research.’ Second Language
Academy of Sciences, 115(39), 9708-9713. Research, 1(2): 231-253.

editorial@elgazette.com 11

ELG1812_p10-11.indd 3 05/12/2018 22:13:27