Replicating Dulay and Burt's study on second language syntax acquisition errors

¶ … Syntax Analysis

There is conflict whether syntax is innate or whether it is learned. Based on a 1973 study that investigated whether L2 (second language) errors that children make are created by 'creative construction" or "habit formation," this study plans to conduct a replication of that study with the difference that the phenomenon of "unique errors" (explained later) will be the syntactic element that will be investigated.

At one time, behaviorism held the center place positing that syntax was an element that was picked up from one's environment. In the 1960s, however, Chomsky showed that rather than picked up from environmental factors, syntax / grammar was innate in a person. Dulay, & Burt (1973) conducted research investigating whether advanced syntax of L2 was also innate or whether it was something that was arduously acquired by the child by particular taught linguistic rules.

According to the habit formation theory, certain errors are predicted in the syntax of the child learning a new language and these errors will appear whenever the second language and the child's natural language diverge. The errors are called 'interference' errors since they persist with the child's (or individual's) attempt to learn the new language.

The authors, thereupon, conducted several studies in which they tested their question on 5- to 8-year-olds.

The "creative construction theory" however which is a successor of Chomsky's innate hypothesis says that different errors should appear in learning of advanced syntax of L2 since the first language is innate and the child is learning a new one which is not impacted by his straining to remember the syntax rules of the first. In other words, since L1 is innate, it exists independent of L2 and all the child strains to internalize is the syntactic rules of L2. The errors would be 'developmental" errors since their pattern would be the same of children in the early stages of acquiring their first language.

The differences in semantic, syntactic and phonological errors between the two are, according to the authors, evident.

Method

The authors tested outcome on a group of 145 Spanish-speaking 8-year-olds and 388 unambiguous errors were tabulated. The children were gathered from two different school areas -- two form California and one in NY. The Bilingual Syntax measure (BSM) was used as tool by their teachers. The BSM measures proficiency of L2 in children in the context of eliciting natural speech form the child.

Errors were then classified into three types -- (1) developmental (the errors that were similar to L1 acquisition errors); (2) interference (the errors that reflected child arduously merging one language with another); (3) unique -- those that appeared in neither of the two categories.

Results showed that tabulation of errors into each category revealed that 3% of the 388 error types were interference whilst 85% were developmental and 12% were unique.

Whilst the authors came to conclusion about the derivation and source of errors in advanced syntax, they still could not explain the 12% errors that were unique. They left that, as they said, to 'future research', but as yet no future research seemingly has been done on that subject.

This is the contribution of this study.

Research Objective

The first step is to replicate Dulay & Burt's (1973) previous study. Times have changed. The educational system has changed. Our aim will be to test whether the 12% of syntactic errors are indeed unique, or whether their appearance is due to educational standards and particular curriculum.

Contributions of research

Implication of research will be the following: If the amount of unique errors found as outcome of study approximate 12% which is the number of Dulay & Burt's (1973) study, then Dulay, & Burt are correct in assuming that reason for this is a conundrum and further research is warranted in creating a theory what would explain this phenomenon.

However, if we discover, as presumed that the discovered outcome of unique errors is significantly more or significantly less than those in Dulay & Burt's (1973) case, we may plausibly conclude (although other reasons may as satisfactorily explain the situation) that it is differences in teaching that introduce the quantity and type of errors.

The study will follow that of Dulay & Burt (1973) in all details. Schools will be solicited for permission, and the data will be evaluated by SPSS. Parents and teachers will also be told about experiment and parents too will be solicited permission. The tabulation of errors and summarization will be done by 2 objective associates. Teachers will review results.

Source

Dulay, HC & Burt MK ( 1973) Should we teach children syntax? Language learning, 23, 245-261