Dr Anne O'Keeffe: Valorising learner competence: how can we see beyond errors in learner language and value growth in competence?

 Dr Anne O'Keeffe

TESOL Spain online 2021 plenary

Valorising learner competence: how can we see beyond errors in learner language and value growth in competence?

To introduce the topic, Anne asked us to consider the following image. What do you see first – the duck or the rabbit? These two images co-exist, but it’s very difficult to see both at the same time: you need to defocus one in order to concentrate on the other.

In this analogy the duck = errors and the rabbit = competence. Typically, as teachers, we scan a learner’s text, like this one at B1 below, looking for errors. In this case, you might pick up on lexical errors, e.g. came v went, or syntactic errors e.g ‘had not have’. Teachers are trained to analyse learner output via this ‘deficit’ model.

Only recently has research started to focus on the positive areas of learner production: evidence of competence. For example, taking the same text, but analyzing it through a different lens, we can highlight what the learner has done well, e.g. examples of clausal complexity, fronting for emphasis or summary as a rhetorical device.

In order to place focus on errors within a historical context, in the next section of the presentation Anne took us through an overview of the key areas of research in this field.

Contrastive Analysis.

In the 1930s and 40s, structuralists such as Robert Lado studied how languages differed. The Contrastive Analysis hypothesis proposed that similarity between two languages would result in positive transfer and facilitate the learning process. Conversely, if an aspect of the languages was very different, the negative transfer would make this more difficult to learn. Within the context of the prevailing audiolingual approach, based on the belief in learning as habit formation, the aim was to predict errors in order to eliminate them.

Error Analysis

Continuing into the 50s and 60s, the work of Jean Piaget in cognitive development led the field in the direction of analysing and categorizing errors in order to understand them as an inevitable part of learning. Prominent figures such as Pit Corder were examining errors, not as a pre-emptive measure, but rather as a window into the learning process. Around this time the term ‘Interlanguage’ was coined (Selinker, 1972), which afforded the learner’s language an independent status along a continuum of progression, with recognizable stages of development.

Corpus Linguistics

By the 1990s, developments in computer science enabled work by pioneers such as Sylviane Granger at the University of Louvain, Belgium. The analysis of large samples of learner English (corpora) allowed the examination of errors from both a Contrastive and Error Analysis perspective. Anne highlighted the following contributions to the field:

Relatively recently, in line with the development of the CEFR, Corpus Linguistics has turned its focus towards the analysis of competency: what learners can do, rather than what they can’t. While CEFR statements are deliberately vague, two projects: The English Vocabulary Profile and the English Grammar Profile, have aimed to better define what it means to be a learner at a particular level of the CEFR. Using access to the Cambridge Learner Corpus (data from Cambridge exam scripts over 19 years), they describe what vocabulary or structures a learner would typically know at a given level. If you enter the word LIVE into the EVP, it will give you information on which uses of this word are typical of learners at different CEFR levels.

Anne returned to the example of the German B1 level learner from the beginning of the talk. Items such as lean or row are indicative of a B2, rather than a B1 level. Texts like this can be analysed using tools such as Text Inspector (NB not all functions are free), to see a breakdown of the lexis used according to its representative CEFR level.
Regarding grammar competence, the English Grammar Profile describes what learners can typically do at each CEFR level. This works in combination with the Vocabulary Profile to give an accurate picture. An example structure such as adverb + adjective will be within a learner’s competence from A1 level (e.g. very good). As they progress through the levels, the structure remains the same, but the lexis that is slotted in becomes increasingly sophisticated.

Anne stressed that these tools are descriptive rather than prescriptive. The aim is not to suggest what learners should know at this level, but to provide a picture of what they probably know. In this example below, we can see the typical development route with the word MUST, which also shows that there is a substantial leap in competence around B1 level. As learners are coming to terms with so many different uses of MUST at B1 level, the duck and the rabbit are still co-exisiting.

We know that learners are still making errors with this structure even at C2 level, so there is no ‘end point’ – you don’t simply come to the end of the continuum. The accuracy-complexity trade off (Thewissen, 2013) means that learners will take more risks as they progress, and therefore produce more errors. Outwardly, this may give the appearance of stabilization, but is not actually a sign that progress is not happening, or as Diane Larsen-Freeman says (2006) this stabilization is not ‘linguistic rigor mortis’.
In conclusion, competence and errors co-exist, but it’s difficult to see the positives or progress when we focus exclusively on the errors (the rabbit and the duck again). Anne emphasizes the need to give feedback also on the positive aspects of your learners’ production, e.g. “You’ve just used a word students don’t normally use until C1 level!”. In other words, make equal use of the green pen as well as the red pen.

*Many thanks to Anne for sharing her notes with me and allowing me to publish this summary

References

Corder, S. P. (1967) The Significance of Learner’s Errors. International Review of Applied Linguistics, 5 161-169.
Larsen-Freeman, D. (2006). The emergence of complexity, fluency and accuracy in the oral and written production of five Chinese learners of English. Applied Linguistics, 27 590-619
Thewissen, J. (2013) Capturing L2 accuracy developmental patterns: Insights from an error-tagged learner corpus. The Modern Language Jornal, 97 77-101