Above: the main square
in Tuscania; the Post
|Discussion Topic 7:
Individual differences in L2 acquisition
Most studies of L2 speech learning have revealed important
differences between individual participants (Ss), especially late
leaners who began learning their L2 in adolescence or adulthood.
No one has as yet managed to provide a convincing account for the
substantial inter-subject variability that is a hallmark of late L2
learners. To my mind, doing so should be considered the Holy Grail
of L2 acquisition research in general, and of L2 speech learning in
|Write if you have something to say on this topic. Please send a carefully edited text and
permission to publish on this site if you want to make your comments public. Thanks. JEF
Most of the research carried out in my lab in Birmingham employed
a randomized block design, comparing groups of Ss who had been
selected according to some variable (e.g., AOA, age of arrival in an
L2 speaking country). We then used traditional ANOVA techniques
to compare groups. For studies in which large numbers of Ss had
been recruited, we used a subgroup matching technique to control
for variables that were confounded with the variable used in
participant selection (see, e.g., Age constraints on second language
A focus on between-group differences does not imply that individual
differences are unimportant. Indeed, I consider the individual
participant as the primary unit of analysis in L2 speech acquisition
research. Our usual practice was to summarize the performance of
individual Ss following analyses of the effect of Gropu. We often did
this by indicating the percentage of Ss in each group who met some
criterion. In the long run, more is needed.
I'll begin this discussion by showing some data presented during a
talk I gave in Poland in 2012. The talk dealt with the role of input in
L2 speech learning. Fig. 1 below shows the results of a 5-year
longitudinal study examining overall degree of foreign accent in
English sentences produced by native speakers of Spanish who
were living in Birmingham. All were late learners who had arrived in
the U.S. after the age of 18 years.
The aim of the study was to examine change in L2 pronunciation over time. As can be
seen here, there was very little such change for any of the native Spanish participants. In
the talk, I attribute this to the fact that the Ss who completed the study had already been
living in the US for over 4 years on average by the time they were recruited for the
longitudinal study. (We also recruited Ss having far less experience in English but, alas,
all of these dropped out of the study before it could be completed.)
The study did yield one useful finding, however. It demonstrated that three of the late learners who
completed the study had a substantially better overall pronunciation of English than did the remaining 12 late
learners. I suspect that the quantity and quality of input may explain, at least in part, the success of the three
Figure 2 shows the average percent English use estimates obtained for the three Ss having the best
pronunciation of English ("best" for short) and the three Ss having the worst pronunciation ("worst"). The
"best" Ss reported using English more often, on average, than did the "worst" Ss. Indeed, the "best" Ss
reported using English with friends far more often than did the "worst" Ss. Although the numbers are very
small - three per subgroup - I doubt that this could have been an accident.
The difference between the "best" and "worst" Ss in self-reported use of English with friends diminished over
time. I speculate that the "best" Ss worked hard to become integrated into the English-speaking community
when they arrived in Birmingham, forming friendships with native English speakers. As they began mastering
English and getting settled into their careers, the "best" Ss may have begun socializing more other with other
native speakers of Spanish while at the same time maintaining their existing relationships with native English
This speculative account may be true, but it is hardly convincing and will need to be verified in a prospective
study. I'll lay out a possible study that I personally would find interesting. I'll call it the "Prospective Study" (or
"PS", for short).
Subject screening will be carried out at "Time 0". The overall degree of FA of Ss who meet the selection
criterion will be rated by a panel of 20 native English-speaking listeners. Ss showing substantial differences
from the overall mean FA will be removed from the study. The remaining Ss will be invited to return for six
additional sessions at 6-month intervals (Time 1 to Time 6). The Ss' language use will be evaluated thoroughly
at each session in order to provide a quantitative estimate of what kind of speech input the Ss typically receive.
The questions to be adddressed will be: How much do the Ss use the L1, and with whom? How much and with
whom do they use the L2? What part of their L2 input comes from native speakers, and what part is foreign
accented? Of the foreign-accented input, how much is provided by fellow native L1 speakers and how much
from other people who speak the L2 with a different kind of accent?
Ss for the PS would be drawn from a larger pool of immigrants who have all arrived in an L2-speaking
environment after the age of 18 years and have resided there for 6 to 12 months. An attempt will be made to
recruit Ss whose living contexts afford differing amounts of access to native-produced English. Among the Ss
recruited will be individuals who interact frequently with a large number of native speakers and others who have
few such interactions. The longitudinal study will last for 3 years.
At the conclusion of the study, L2 pronunciation will be evauated in two ways by native English-speaking
listeners. First, FA will be scaled for all of the speech samples gathered at both Time 1 (start of the study) and
Time 2 (end of the study) by a new group of 20 native English-speaking listeners. Second, the Time 1 and Time
2 samples of all individual Ss will be presented in pairs to native English-speaking listeners, whose task will be
to decide which member was "pronounced more accurately". Flege et al.(1995) showed that listeners differ in
sensitivity to the presence of FA. The dependent variable of the paired-comparison test will be how many of 20
listeners successfully chose the Time 2 sample as "pronounced more accurately". Result of the paired
comparison test will indicate which Ss showed a significant improvement in pronunciation from Time 1 to Time 2.
The FA ratings will be examined in an attempt to understand to what extent and why
The average FA ratings obtained at Time 1 and Time 2 will be examined to provide a better understand of
how much improvement occurred over 3 years, and why. Here the goal is to account for as much variance as
possible in the Time 2 FA ratings, as well as the Time 1-Time 2 difference scores. I suspect that variations in
the kind and amount of L2 input will account for substantially more variance in the outcome measures than
will age of L2 learning (AOA) because all of the Ss will have begun learning the L2 after the age of 18.
Outline of a prospective study (PS) examining the role of L2 input
The question addressed in analyzing the paired comparison data is whether the Time 2 samples for each
individual participant has been chosen at a rate that significantly exceeds a chance level of response (50%). I
think that virtually all Ss will demonstrate a perceptually relevant improvement in their pronunciation over the 3
years of the study. Ss who fail to do so will likely be those who are simply not using the L2 to communicate on a
regular basis, or else have received most of their L2 input from fellow L1 speakers.
Mean foreign accent ratings obtained for male and female native speakers of Italian differing according to
age of arrival in Canada. Flege et al. (1995), Journal of the Acoustical Society of America, 97, p. 3129
The role of foreign-accented L2 input was raised by Flege et al.(1995). The existing literature suggested that
female L2 learners will have a better pronunciation than will male L2 learners. As shown above, this is indeed
what was found for the native Italian Ss who arrived in Canada up to about the age of 12 years. For those
who arrived after about the age of 17 years, on the other hand, the reverse held true. An important gender
difference existed among the native Italian subjects who arrived in Canada as young adults, At the beginning
of their new life in Canada, the women tended to work in the home whereas nearly all of the men began
working outside the home almost immediately. I speculate that most of the English input the women received
during their fist few years in Canada was the foreign-accented pronunciation of English provided by
husbands and other male relatives.
It is unlikely that input will account for all of the variance. I recommend that the PSTM (phonological
short-term memory) be evaluated at Time 1 using L1 speech materials.The results of MacKay et al. (1991)
suggest that PSTM may help account for inter-subject variability in L2 learning. It would also be valuable to
determine if the FA ratings obtained from one group of listeners at Time 1 will account for variance in the FA
ratings obtained from another group of listeners at Time 2. If such a finding is obtained, it might point to
differences in some form of individual aptitude that is useful for L2 learning.