Abstract
The article explores the possibility of adopting a form-to-function perspective when annotating learner corpora in order to get deeper insights about systematic features of interlanguage. A split between forms and functions (or categories) is desirable in order to avoid the "comparative fallacy" and because – especially in basic varieties – forms may precede functions (e.g., what resembles to a "noun" might have a different function or a function may show up in unexpected forms). In the computer-aided error analysis tradition, all items produced by learners are traced to a grid of error tags which is based on the categories of the target language. Differently, we believe it is possible to record and make retrievable both words and sequence of characters independently from their functional-grammatical label in the target language. For this purpose at the University of Pavia we adapted a probabilistic POS tagger designed for L1 on L2 data. Despite the criticism that this operation can raise, we found that it is better to work with "virtual categories" rather than with errors. The article outlines the theoretical background of the project and shows some examples in which some potential of SLA-oriented (non error-based) tagging will be possibly made clearer.
Highlights
The article explores the possibility of adopting a form-to-function perspective when annotating learner corpora in order to get deeper insights about systematic features of interlanguage
It is believed that L1 taggers are useless because they are unable to capture the divergent phenomena occurring in learner corpora (LC)
The first one, which is widely accepted in literature and which is adopted in many European projects is that learner data is best viewed in terms of errors
Summary
The topic of this article is the Part-of-Speech (POS) annotation of learner corpora (LC). The research question is whether it is feasible and convenient to instruct an automatic tagger which is capable of recognizing and annotating the grammatical categories in learner data. Misspelled, badly uttered, incomprehensible and not interpretable items are destined to escape the formal requirements of automatic analyzers and of robust parsers. To face this issue, two different solutions are at hand. The POS errortagging procedure is made up of three steps: (a) collecting learners' typical mistakes all together in a list (typical mistakes/errors with respect to homogeneous groups of learners); (b) turning this list into errors related to traditional linguistic categories (such as errors in nouns, adjectives, verbs etc); (c) tagging the items in the list using a markup language (for instance, XML). After a LC has been tagged with error-tags by using, for instance, a markup language, all occurrences are retrievable with software (for instance, with WordSmith Tools or Xaira)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
