Second Language Research Methods. Herbert W. Seliger

again. They were saying eze kef, eze kef. At first, she thought that they were saying, ‘Oh what rocks, oh what rocks!’ But then, she reported, she reasoned to herself, ‘Why would they be laughing and talking so much about rocks?’ She theorized that they must be saying something that expressed their excitement about the climbing. She concluded that they must be saying something like the English, ‘Oh, what fun!’ Subsequently, she approached a friend who knew both English and Hebrew and confirmed her interpretation.

      These examples demonstrate that research in the broader sense of observing, and forming and testing hypotheses is a process that can take place unconsciously as well as consciously, externally as well as internally and is a basic learning mechanism from infancy through adulthood. As will be seen in this chapter and the following chapters, these same processes are also part of formal or scientific research. Scientific research is what is usually associated with the word ‘research’ and while we do not usually think of the child acquiring a first or second language as carrying out research, the same basic components are present in scientific research as well as in child language acquisition and in other everyday activities.

      However, even though the processes identified with research are also part of normal learning activity, it does not mean that research is simply a matter of doing what comes naturally or acting on the basis of intuition. In order to conduct formal research, it is important to be aware of some important differences between conclusions reached on the basis of natural intuitive learning processes and those arrived at through systematic scientific investigation. It is also important to recognize when we really have reached a reliable conclusion about something and when we have not, that is, when we really know something and when that conclusion may be false. The following two sections will deal with these two important issues.

      Scientific research and common sense

      While we have said that the basic components of research are found in everyday experience, there are some important differences that should be kept in mind between research and everyday activities. If research is so simple and so natural, why read a book on research methods and design? Why not simply rely on our intuitions and common sense to reach conclusions about the language phenomena that we observe?

      People who do not do research but have intuitions about language learning and teaching typically react to second language research by regarding it as a confirmation of their common sense. (‘I knew that.’ ‘That doesn’t make sense.’ ‘According to my experience …’)

      While common sense, intuition, and introspection about experience are useful, they are of limited value unless used appropriately. It has been cynically suggested that a guiding rule for common sense is that new ideas should look like old ones. That is, new ideas should confirm what we already know or believe. Some feel that the purpose of science and scientific research is to confirm the beliefs of common sense. If this were the case, science would be involved in supporting superstition and prejudice as well as ‘good’ common sense. For many years, it was believed that children learn a first language by simply imitating their parents. We now know, after many years of scientific research, that what appears to be a common sense conclusion about language acquisition is not true and that language acquisition is much more complicated.

      The conclusions of common sense might become the starting point for scientific research but should not become the end point. Science might begin with the question: Do children really learn a language by imitation, as seems to be suggested by common sense? In this section, we will explore some important differences between a scientific approach to such a question and one based on common sense; One of the functions of scientific research can be to provide empirical or factual support for common sense or to disprove what has become accepted as common sense. Of course, methods and knowledge that masquerade as science have been used to support prejudices and superstitition – as we have witnessed in our own century – and the scientist must be alert to the misuse of science.

      The differences between knowledge arrived at through common sense and intuition on the one hand, and scientific research on the other, can be expressed by concepts such as ‘organized’, ‘structured’, ‘methodical’, ‘systematic’, ‘testable’, and especially by the notion of disciplined inquiry.

Scientific research is conducted systematically

      Common sense conclusions are reached on the basis of superstition, superficial responses to problems, and unexamined beliefs. Recently, one of the authors asked a foreign language teacher why he had the children in a fourth grade class memorize all of the intricate rules of grammar and spent little or no time on having them use the language for communication. His reply was that he could not see how they could be expected to use the language before they had committed to memory what he considered to be the ‘grammar rules’ of the language. When he was told that native speakers of the language no longer observed some of these rules, he replied, ‘That’s even more of a reason to learn the rules. Native speakers shouldn’t be our standard.’

      A scientist might begin with the question: Can children utilize explicitly taught formal grammar in the acquisition of a second language? This same scientist would then systematically observe children learning second languages in natural contexts; try to discover the role of grammar learning for children in a school context; and then plan a program of research to see how children fared under foreign language teaching methodologies employing grammar teaching and under those which did not.

Science builds theories by systematically testing an interrelated body of hypotheses

      Common sense conclusions are often based on superficial evidence which supports something which someone wants to believe. Laymen test the theories held about the world in a selective and subjective way. While it may be said that scientific research is also selective – all possible hypotheses cannot be tested at once – a scientist selects hypotheses for research in a systematic way. The layman often ignores or explains away any disconfirming evidence. In the example above, the foreign language teacher believed that languages are learned by first committing prescriptive grammar to memory. The supporting evidence for this teacher’s belief may have been his own limited experience as an adult. It might also be based on a belief that the body of prescriptive grammar rules is the language itself. That might have been the way he was taught, which led him to believe that it must be a good method. He may not have taken the time to consider all the failures that have resulted from this particular approach to language teaching.

      Theories and hypotheses which are used in research are formulated so that they can be tested by the researcher and others who wish to replicate the research. We say that hypotheses must be falsifiable. They must be amenable to some kind of test which might ultimately disprove them.

      In our example, it would not be enough to say that children learn a language better without grammatical explanation because we believe it to be so. If we say that children learn foreign languages more effectively without an emphasis on formal grammar, we would have to formulate this idea in a manner which would allow us to gather evidence confirming or rejecting it as a hypothesis. The research on this question would have to be carried out so that others who might wish to test our hypothesis under slightly different conditions (with, say, older or younger subjects) would be able to follow our procedures.

Scientific research seeks to describe, identify, and control relationships among phenomena in order to study them

      In the previous section, we noted that it is a very natural response to look for and notice relationships among the phenomena we see around us. Parents do this when they try to interpret what their children are saying. Unfortunately, we sometimes arrive at conclusions which appeal to our common sense but which may be incorrect. We observe that children sometimes imitate words and phrases which they hear from other speakers before they understand what these words mean. From there, it is only a short step to pointing to imitation as the primary manner in which children acquire the whole language system. And since common sense tells us that this is how first languages are learned, it is only another short step to concluding that second languages may also be learned by imitation and repetition of sentences in the second language even though the learner may not fully understand them.

      Research, on the other hand, seeks to identify relationships and reach conclusions about them after ruling out alternative explanations. It might be concluded on the basis of observation