What’s aptitude got to do with it?

Ljiljana’s and Aleksandra’s recent posts here were responses to my invitation to NNS teachers to tell us how they learnt English. My original question was to do with the relative importance thay gave to implicit and explicit learning, but behind this lies the more fundamental question of why Ljiljana and Aleksandra (like Hana Tichá, Svetlana Kandybovich, Rose Bard, Thom (who often comments here), Marek Kiczkowiak and other bloggers) succeeded in an endeavour where the majority fail.

One of the key phenomena in SLA research is “incompleteness”, which refers to the fact that most learners fail to reach a high level of ultimate attainment. As Sawyer and Ranta said:

the clearest fact about SLA that we currently have is that L2 learners differ dramatically in their rates of acquisition and in their ultimate attainment (Sawyer and Ranta, 2001: 319).

While it might be “the clearest fact” we have, it is not, alas, the area of SLA research that’s made the most progress; rather the opposite is the case. Explaining the role that individual differences make to SLA has proved very difficult, partly because the theoretical constructs involved are so slippery.

Aptitude would seem to be the individual difference most likely to explain the huge differences in results of foreign language learning, wouldn’t it? “I’m no good at languages” we say, by which we mean that we have no aptitude for it, just as we might say we’re no good at maths, or singing, or cooking. But when we try to define aptitude for language learning so as to get a clear idea of what we’re talking about, it turns out to be notoriously hard to pin down.  Until quite recently, the best attempt to define language aptitude was made by Carroll and Sapon (1959), who produced the Modern Languages Aptitude Test (MLAT). This divided language aptitude into four components:

1. Phonemic Coding Ability
2. Grammatical Sensitivity
3. Inductive Language Learning Ability
4. Rote learning activity for foreign language materials.

Scores on the MLAT subsequently yielded multiple correlations of between 0.40  and 0.60, which is quite impressive. As Sawyer and Ranta commented:

These are considered moderate to strong correlations, and although they imply that considerable learner variation remains to be explained by additional factors, they also demonstrate that language aptitude has consistently been the single best predictor of subsequent language learning achievement (Sawyer and Ranta, 2001: 322).

But there are a number of problems with this view.

1. The biggest problem is that defining aptitude in terms of a bank of tests falls into the trap of being circular; we say that those who have an aptitude for language learning are those who do well at language aptitude tests. But as with IQ tests, we can turn this round and say that a high score simply shows that you are good at the tests.
2. Related to this, Oller (1983) argued that the MLAT treats language attitude as simply general intelligence applied to the task of foreign language learning, and there is therefore little point in studying it as a special learner trait.
3. Then there are unanswered questions. Is aptitude innate, declining with age (linked to the “critical period” argument), or is it a matter of skill development?
4. Last but not least, the aptitude being defined here seems to relate only to formal instruction, and, what’s more, to a particular type of formal instruction. Cook (1996) argues that the aptitude tests are not relevant to current L2 teaching methodology:

•  Such tests are not neutral about what happens in the classroom nor about the goals of language teaching.  They assume that learning words by heart is an important part of L2 learning ability, that the spoken language is crucial, and that grammar consists of structural patterns. In short, MLAT mostly predicts how well a student will do in a course that is predominantly audiolingual in methodology rather than in a course taught by other methods.  (Cook, 1996: 101)

Long (2017) seems to agree with Cook when he notes that the MLAT is heavily weighted towards aptitude for analytic, explicit language learning.  He says that it’s not surprising to find moderate to high correlations between scores on such measures and scores on tests of foreign language abilities learned and measured in the same way. Furthermore, although the tests are still good at predicting rate of progress in the early stages of foreign languages taught explicitly, they don’t address questions about success at later stages, or very advanced proficiency, or learners with higher aptitude for implicit language learning, or classroom learners taught by teachers using a communicative methodology.

Hi-LAB 

The Hi-LAB (High-Level Language Aptitude Battery) tests address some of these questions.  Linck et al. (2013) explain:

Hi-LAB was designed to predict the attainment of high-level proficiency—rather than initial rate of learning—with the expectation that high-level acquisition requires going beyond the classroom setting, for instance by participating in an immersion experience. Most of the constructs in Hi-LAB were designed to capture potential for language learning processes that operate in such non-instructional settings, where superior cognitive and perceptual abilities of the learner may enhance the processing of language input and facilitate the mapping in memory of apperceived forms, meaning and function. Hi-LAB does not measure phonetic coding ability (important in learning to read in a foreign language) or grammatical sensitivity (important in explicit language instruction), since those measures already exist and are hypothesized to be more relevant at initial stages. Instead, the focus was on measuring potential for dealing with the remaining language learning problems, such as mastering complex linguistic systems and perceiving non-salient language features. Therefore, we have focused on cognitive and perceptual abilities that are hypothesized to support more advanced aspects of L2 learning that are required to attain high-level proficiency.

The results of the Linck et al. (2013) study are tentative and rather general, but, nevertheless, it’s a promising development. I would love to see how the NNS teachers and bloggers I mentioned at the start of this post would do in Hi-LAB. Both Ljiljana and Aleksandra emphasised the importance of explicit learning, and Aleksandra in particular stressed how much hard work she’d put in to reaching such a high level. Still, I can’t help thinking that there’s more to it than working hard on explicit learning. On top of all their hard work, do they have more than just the ability to get high scores on tests? Do they have exceptional cognitive abilities related to attention and memory that make both explicit and implicit learning easier and faster for them than for most people?

Another thing, of course, is motivation, but we’ll have to leave that for another day.

 

References 

Carroll J. and Sapon S. (1959). The Modern Languages Aptitude Test. San Antonio, Tx.: The Psychological Corporation.

Linck, J. A., Hughes, M. M., Campbell, S. G., Silbert, N. H., Tare, M., Jackson, S. R., Smith, B. K., Bunting, M. F. and Doughty, C. J. (2013) Hi-LAB: a new measure of aptitude for high-level language proficiency. Language Learning 63(3): 530–66.

Long, M.H. (201) Instructed second language acquisition (ISLA): geopolitics, methodological issues, and some major research questions. Instructed Second Language Acquisition, 1.1, pp.7 to 44.