A second point that has been emphasized by advocates of a dynamic systems
or complexity approach to L2 development is that most studies making
strong claims about fixed orders of acquisition have only looked at the overall
developmental trend for groups of learners that may not be representative of
any of the individual learners in the group. This problem is clearly illustrated
by the results of a longitudinal study on L2 accent (Derwing & Munro, 2013;
Derwing, Thomson, & Munro, 2006). These authors investigated the development
of foreign accentedness and comprehensibility of groups of English
L2 learners with a Slavic or Mandarin L1 background in a naturalistic environment.
A comparison of the data at the onset of the study to the data after
2 years and 7 years of exposure showed that there was a significant improvement
in accent for the Slavic participants, but not for the Mandarin participants
(see Figure 1). However, as Derwing and Munro (2013) showed in a closer
examination of individual development, some learners became more comprehensible
while others had a stronger foreign accent after 12 years and became
less comprehensible. None of the individual participants showed the mean pattern
of the group (see Figure 2). In cases like this, we could and should doubt
the relevance and meaningfulness of the mean trend, in spite of the fact that
this pattern was found to be significant.
The problem for the generalization of longitudinal group data is explained
by van Geert (2014). He argues that different dimensions can be distinguished
in studying development of a L2. The first two dimensions comprise
the interaction of several variables (dimension 1) as they shape the developmental
trajectory of an individual learner (dimension 2). Taken together,
these dimensions can be represented as a dynamic causal network of positive
and negative couplings of the variables as they change over time.
A second point that has been emphasized by advocates of a dynamic systemsor complexity approach to L2 development is that most studies makingstrong claims about fixed orders of acquisition have only looked at the overalldevelopmental trend for groups of learners that may not be representative ofany of the individual learners in the group. This problem is clearly illustratedby the results of a longitudinal study on L2 accent (Derwing & Munro, 2013;Derwing, Thomson, & Munro, 2006). These authors investigated the developmentof foreign accentedness and comprehensibility of groups of EnglishL2 learners with a Slavic or Mandarin L1 background in a naturalistic environment.A comparison of the data at the onset of the study to the data after2 years and 7 years of exposure showed that there was a significant improvementin accent for the Slavic participants, but not for the Mandarin participants(see Figure 1). However, as Derwing and Munro (2013) showed in a closerexamination of individual development, some learners became more comprehensiblewhile others had a stronger foreign accent after 12 years and becameless comprehensible. None of the individual participants showed the mean patternof the group (see Figure 2). In cases like this, we could and should doubtthe relevance and meaningfulness of the mean trend, in spite of the fact thatthis pattern was found to be significant.The problem for the generalization of longitudinal group data is explainedby van Geert (2014). He argues that different dimensions can be distinguishedin studying development of a L2. The first two dimensions comprisethe interaction of several variables (dimension 1) as they shape the developmentaltrajectory of an individual learner (dimension 2). Taken together,these dimensions can be represented as a dynamic causal network of positiveand negative couplings of the variables as they change over time.
การแปล กรุณารอสักครู่..
