Second Language Research Methods. Herbert W. Seliger

is meant by a ‘guess’ and even a ‘good guess’ before the phenomenon can be explored in a systematic way. Secondly, it is necessary to control for other possible causes of successful language learning, such as motivation, before drawing any conclusions about the role of guessing.

      If it can be assumed that research involves planned and systematic inquiry, an additional problem confronting the researcher is knowing when the inquiry has reached a conclusion. A conclusion in research is another way of stating that we know something we did not know before we began the investigation. The conclusion may even be that we do not know something we thought was a fact. It may be that what we know is in the form of a theory that still remains to be tested or examined. However, in all of these forms of research, the final objective is to arrive at some form of knowledge that we did not possess before the investigation began. In the next section, we shall examine the kinds of knowledge brought to light in second language research. It is important for the researcher and the consumer of research to distinguish between these different forms of knowledge.

      Finding answers: How do we know something?

      An important dilemma which research must deal with is how we know when we have found the answer to a question. As we shall see, there are different methods which research can use to answer the questions which it poses. How do we know when either we or another researcher have found an answer to a question posed about second language?

      Research findings in second language studies and bilingualism may be categorized according to four types of knowledge which these findings represent. It is often a good idea for a reader of research to ask what kind of knowledge is represented by different claims made in a study. Not all conclusions are necessarily reached on the basis of the same kind of knowledge. Thus, the conclusions of a study might be based partially on belief (Type 1) and partially on a description of a language phenomenon or the results of an experiment (Type 3 or 4).

Four types of knowledge found in second language research

      Type 1: Knowledge as belief

      When we know something on the basis of belief, it may mean we want to believe something to be true but have never submitted it to an empirical test. Many of the conclusions reached through ‘common sense’ may be put into this category.

      Weinreich (1953) presents these ‘scientific’ conclusions reached by researchers on bilingualism:

      Reis (1910) writing about the trilingualism of those who live in Luxembourg stated, ‘The temperament of the Luxembourger is rather phlegmatic … We have none of the German sentimentalism or the French vivacity … Our bilingual electicism prevents us from consolidating our conception of the world and from becoming strong personalities.’

      Gali (no date) suggested that bilingual persons may be morally depraved because they do not receive effective religious instruction in their mother tongue in childhood.