【正文】 en study and research in terms of educational theory and practice in ELT settings.2. Three sources of knowledge二級(jí)及以下各級(jí)小標(biāo)題單獨(dú)占一行，左頂格，Times New Roman小四號(hào)，正體。 Research is one of the three major means for human beings to acquire knowledge of the environment including the natural world and our human society. The other two are “experience” and “reasoning”(Cohen and Manion 1)頁(yè)：1夾注的格式要求：所引文獻(xiàn)的作者在正文中不出現(xiàn)的情形：根據(jù)MLA格式，“作者的姓”加上“引文出現(xiàn)的頁(yè)碼”，不加文獻(xiàn)的出版年份。所有夾注必須跟“參考文獻(xiàn)目錄”中的文獻(xiàn)篇目完全對(duì)應(yīng)。（見(jiàn)《MLA科研論文寫(xiě)作規(guī)范》第212213頁(yè)。）. The role of research in the acquisition of human knowledge can hardly be understood fully without being studied in connection to that of experience and reasoning. For the purpose of achieving a better understanding of research, the role of experience and reasoning will be considered before that of research. ExperienceExperience is a kind of development of personal knowledge of the world. It is regarded as an individually accumulated body of knowledge (Cohen and Manion 1). In a problemsolving situation, people tend to resort to personal experience first. However, where solutions to problems clearly lie beyond this body of personal experience, it is often helpless to resort to personal experience. In the case of foreign language learning, the learner’s native language often interferes with or facilitates the learning of the target language. This can be considered as a clear indication of the learner’s reliance on the personal experience in his or her first language. It is arguable that the personal experience is by no means reliable although it is sometimes helpful because it cannot guarantee smooth progress and success in foreign language learning.As for English language teaching, our experience of English examinations can be resorted to when we help our students prepare for the college entrance examination of English. However, it is difficult for us to resort to our previous personal experience when we are facing the problem of how to motivate middle school students in municative language teaching as many of them can hardly see any chance to municate directly with native speakers of English. ReasoningReasoning is the act of forming conclusions, judgements or inferences by thinking in a logical manner. There are two basic types of reasoning: one is inductive reasoning and the other is deductive reasoning.Inductive reasoning begins with observations and evidence of empirical regularities or empirical relationships (Howard 8). This is a mental process from a number of specific cases to a general idea underlying them. When a learner of English es across expressions such as “three books”, “many ships”, “two minutes”, he or she may form a hypothesis that “s” is used to indicate the idea of “two or more”. It is not difficult to see that inductive reasoning in foreign language learning often leads to hypothesis formation.Deductive reasoning begins with basic beliefs, theories, assumptions, propositions, and so on, the validity of which is assumed and untested (Howard 8)頁(yè)：2夾注的格式要求：所引文獻(xiàn)的作者在正文中不出現(xiàn)的情形：根據(jù)MLA格式，“作者的姓”加上“引文出現(xiàn)的頁(yè)碼”，不加文獻(xiàn)的出版年份。所有夾注必須跟“參考文獻(xiàn)目錄”中的文獻(xiàn)篇目完全對(duì)應(yīng)。（見(jiàn)《MLA科研論文寫(xiě)作規(guī)范》第212213頁(yè)。）. This is a mental activity from a general idea to specific cases. In foreign language learning, if we learn a grammatical rule or a wordformation rule first, then we apply it to make a sentence or to coin a new word. For example, according to the English wordformation rule that the prefix “un” and an adjective may bine to form another adjective with negative or opposite force in it: “un” and “happy” go together to form “unhappy” with the meaning of “not happy”.There is an obvious limitation in reasoning as an activity. According to Cohen and Manion, “it [reasoning] was no longer related to observation and experience and became merely a mental exercise” (3).頁(yè)：2夾注的格式要求：所引文獻(xiàn)的作者在正文中出現(xiàn)的情形：根據(jù)MLA格式，只要列出“引文出現(xiàn)的頁(yè)碼”就可，不用加文獻(xiàn)的出版年份，也不用再列上作者的姓。（見(jiàn)《MLA科研論文寫(xiě)作規(guī)范》第212213，64頁(yè)。） That is to say, the credibility of reasoning, whether inductive or deductive, will be questionable once reasoning is not connected to the reality. Now consider the hypothesis that “s” used with a countable noun indicates the idea of “two or more” again. As noted by Quirk and his coauthors, “unlike some languages where plural implies ‘two or more’, English makes the division after ‘more than one’: one half day, one day But: one and a half days, two days, one or two days” (297).頁(yè)：3夾注的格式要求：所引文獻(xiàn)的作者在正文中出現(xiàn)的情形：根據(jù)MLA格式，只要列出“引文出現(xiàn)的頁(yè)碼”就可，不用加文獻(xiàn)的出版年份，也不用再列上作者的姓。（見(jiàn)《MLA科研論文寫(xiě)作規(guī)范》第212213頁(yè)。） Here, it is clear that reasoning itself cannot guarantee its selfcorrection. Similarly, the application of the wordformation rule in the previous paragraph cannot prevent learners from making unacceptable adjectives such as “*unhonest”, “*unactive”. When such errors occur, they are considered as cases of ove