Considering Avoidance In Sla

This paper is reprinted from Student Occasional Papers in Hard-Science / Human Linguistics, posted on the Hard-Science Linguistics Web Page, http://humanlinguistics.utoledo.edu, September 2004.

CONSIDERING AVOIDANCE IN L2 ACQUISITION1 AbdulRahman Congreve Abstract In this paper, framed in human linguistics, the writer argues that the concept avoidance in the field of second language acquisition (SLA) can be moved from the logical domain of grammar into the physical domain, accessible to science. After presenting his main arguments, he briefly lays out a plan for testing avoidance as a real-world phenomena involving people communicating. Although he refers to avoidance as a "deliberate" strategy, we could see the relevant properties of the person in terms of real-world variables affecting conditional properties in the plex of a communicating individual.

O'Grady, et al. (1989) discuss avoidance as a subset of an second-language (L2) learner's communication strategies. They also state that a communication strategy is a conscious plan (313). Gass and Selinker (2001) define avoidance as a type of interlanguage (IL) communication strategy (241) where the NNS chooses to use or not to use a particular structure (119). They also state that the difference between first-language (L1) and L2 production is the only variable that consistently predicts avoidance (120). Our task, then, is to examine avoidance in light of a difference between the L1 and L2 production. To examine avoidance in the physical domain we must observe people in the real world (Yngve, 1996:80). Avoidance is seen as a deliberate strategy — different from simple ignorance. The non-native speaker (NNS) deliberately chooses to perform one communicative task rather than another. In order to investigate avoidance empirically, we need to show that the NNS performs a particular behavior in L2 because s/he is deliberately substituting an easier or more familiar behavior, intentionally avoiding the one s/he finds difficult. As a NNS of Arabic, I use avoidance when speaking because my articulations, associated with certain task properties, are different from the articulations associated with similar task properties of native speakers. So I avoid certain articulations in favor of simpler ones so that the articulations I do make will be more similar to those of a NS — and I won't sound so foreign. Because this is one of my avoidance strategies, I know that there is a task property I am lacking, and [intuitively] "I know that I know". In performing empirical linguistic research however, I have to design an experiment demonstrating deliberate avoidance of a certain structure and substitution of an alternate. So as to not skew the results of the experiment, we need to observe the avoidance in as natural a setting as possible. Take, for example the study of avoidance in an Arabic speaker learning English as L2. Many Arabic speakers are unable to duplicate an articulation behavior that English speakers perform, because they possess the relevant task property. "In Arabic there is no contrast between [p] and [b] in word initial position" (Gass and Selinker, 2001:162). In the following hypothetical assemblage we have a NS, a NNS, and some physical props in a setting. A NS police officer, a NNS Arab male on a street in a small midwestern town in the United States. The Arab man has just parked 1

This paper was originally written as a "position paper" as part of the course requirements for Applied Linguistics II, May 2002. Applied Linguistic II is part of the core for the Master of Arts in English — Concentration in ESL, at the University of Toledo, Toledo, Ohio, USA.

his car on the side of the street. The NS and NNS share certain task properties that describe a particular act in a particular setting (parking the car). However, with the NNS, the articulations made and the subsequent sound waves produced are not appropriately processed by the NS. The Arab man has just parked his car in a spot that he thinks may be prohibited. He spies the policeman and asks, "Can I bark here?" whereupon the policeman answers, "You can bark anywhere you want but you can't park here." To prevent this the Arab may use avoidance. He knows he lacks an ability (a certain task property) that English speakers possess. He therefore focuses conscious awareness on his production via the Monitor. Some of his higher task properties are looking ahead to monitor feedback from his production. They detect that the task properties for a [p] articulation are about to be activated. A dampening activation prevents this and forces an alternative set of task properties that can be activated with almost the same input activations. For example, he may avoid saying "parking lot" and say "car lot" instead. Because evidence shows that Arab ESL speakers are aware of this problem, and because of the differences between English and Arabic, this case of avoidance could be examined in light of the L1 — L2 difference. One hypothesis is that there is a correlation between avoidance strategy and L2 proficiency. At low levels of English proficiency, ignorance dominates — there is low avoidance and no relevant task properties (more cases of [b] where the NS would articulate [p] and no substitutions of other behaviors). At intermediate English proficiency, ignorance is lower and avoidance increases — there are relevant task properties for decoding only. Finally, at high English proficiency levels, there is a decrease in the avoidance strategy as the Arabic NS becomes more proficient in English — there are relevant task properties for decoding and encoding (fewer cases of the [b] articulation where the NS produces [p] and fewer substitutions of other behaviors). To test this hypothesis, the most valid method would be to observe the occurrence of the [p] avoidance in a natural setting, say some public place frequented by Arabic ESL speakers — in a fast food restaurant, or parking garage — someplace where avoidance (with or without substitution), and L2 proficiency can be observed. However, the reliability may be affected by difficulty in hearing or recording the speech sounds. A possibly less valid but more reliable way to conduct the research would be to set up a computer mediated activity where the NNS would have to complete different reading and speaking tasks, the outcome of which would depend, in part, on the computer's recognition of the [p] and in part upon the NNS L2 proficiency. This could be done with a game where the NNS has to complete L2 tasks and articulations to travel along a number of alternate paths. Successful completion of a certain percentage of L2 tasks and the computer's recognition of a certain number of [p] articulations would lead the NNS down one of several paths toward one of several goals representing a measure of L2 proficiency and a quantity of [p] articulations. In an assemblage with a NS, a NNS, and props in a particular setting, consider a task property shared by the NS and NNS. Avoidance occurs when the NNS deliberately uses a simpler articulation than the NS to represent the same task property. To investigate this phenomenon in a scientific way, we need evidence of deliberation by the NNS, and we have to observe communicating individuals (NS and NNS) in a natural setting where the appropriate task properties will activate, or not activate certain articulations and consequent sound patterns. In this way we can objectively evaluate our hypothesis: whether avoidance is a function of L2 speech proficiency.

References Gass, Susan M. & Selinker, Larry. (2001). Second language acquisition. Mahwah, New Jersey: Lawrence Erlbaum Associates. O'Grady, W., M. Dobrovolsky, & M. Aronoff. (1989). Contemporary linguistics: An introduction. New York: St. Martin's Press. Yngve, Victor H. (1996). From grammar to science: New foundations for general linguistics. Philadelphia: John Benjamins.