Image Schemas in Second Language Learning and Instruction: A Case for a Multidisciplinary Approach

Lukeš D. Image Schemas in Second Language Learning and Instruction: A Case for a Multidisciplinary Approach. In: Image Schemas and Linguistic Relativity, Language Culture and Mind. Portsmouth; 2004.
This paper is based on the premise that there is no meaningful conflict between the theory of image schemas and a view of linguistic universalism based on relativity. While it is not doubted that image schemas derive from “universal aspects of how the human body interacts with its environment”, there is no evidence whatsoever that the particular schemas themselves have any claim to universality. In fact, as evidence from second language acquisition suggests, quite the opposite is true. Not only do these schemas have to be acquired, often at great expense of time and effort on the part of the learner, they have to be acquired in a specific way. What this way is and what evidence language classrooms can offer to practicing linguists is the subject of the present undertaking.
The paper makes two essential claims: 1) Image schemas cannot be divorced from their origin in rich imagery (to use Lakoff 1987’s distinction). In fact, they can only be profitably treated in situ, i.e. as parts of larger cognitive models. To extricate them from these models will confine research on image schemas to the realm of speculation. 2) The ‘schematicity’ of image schemas is a matter of degree. This degree of schematicity will vary between languages and, what is more, may be different between speakers of them same language. This difference is more likely to be ‘idiolectal’ rather than ‘dialectal’ in nature. I.e. two speakers of the same dialect may operate with different image schemas and this may often not impede communication. Subsequently, a search for the exact and ‘correct’ idealized depiction of a given image schema can be detrimental to the understanding of what is in equal measure a process and a product.
None of these claims are new or controversial in themselves but they bear reiterating because it is easy to lose sight of them and elevate image schemas to the level of independent real linguistic entities where they do not belong. What this paper claims is new, is looking for evidence in second language learning as opposed to second language acquisition. Over the past 50 years, this field has witnessed a number of spectacular failures of expectation when it came to rely unduly on results of research in linguistics or psychology. For instance, in the case of generative linguistics, humans proved to be as impervious to acquiring language through access to deep structures as computers continue to do. Image schemas pose a similar dilemma. Exposing learners to image schemas as a way of facilitating acquisition bears only limited success. In fact, it seems that learners need to acquire a bank of rich images along with the concept of a schema to be able to apply it successfully. This paper will present examples from the acquisition of the Czech preposition na and the Albanian verb duhet as evidence of its claims. Supporting examples will be offered from the field of cross-cultural communication and contrastive analyses at all the traditional levels of linguistic description.