英语口语中的情态动词Must, Should, May and Might.doc

Silvie RisingerováMasarykova univerzita v BrnFilozofická fakulta Katedra anglistiky a amerikanistiky30.6.2006Magisterská diplomová práce z oboru anglická lingvistikaThe use of pronoun subjects I and you in connection with the modal verbs Must, Should, May and Might in the spoken English produced by men and womenI proclaim that I have written this thesis by myself and that the research presented is my own work.I would like to thank PhDr. Nadezda Kudrnacova, CSc. for her support and professional guidance. Contents:Introduction.5Aim and scope.5Corpus Linguistics.7The British National Corpus (the BNC).9Method.11Analysis and discussion22Conclusion.72Sources.76References.76Introduction:Sociology is becoming more and more important for our multicultural society. We have to learn more about each other and only then can we understand our differences and be more tolerant. This understanding can facilitate communication in my opinion. My research is a sociolinguistic one. I have been interested in gender linguistics. Many books have been written on this subject and most of them state that there are relatively great differences between the language of women and the language of men. I would like to find out if this is true and, if so, what these differences are in the field of modal verbs. When we read the list of the most common words occurring in the corpora, we see that modal verbs and pronouns are at the top of the list. Their frequencies are one of the highest ones in the corpora. If we presume that there is a significant difference between the female and male languages, then it must be visible in the way women and men connect the pronouns I and you in the subject position with the modal verbs. The research is carried out only on the spoken part of the corpus as this is a more natural form of language than the written language and it is connected with the common daily language.Aim and scope:Firstly, as the length of this paper is limited, decision has been taken to study the verbs must, should, may and might in detail after having briefly looked at all the main modal verbs. The first two verbs form a semantically related group and the third and fourth verb are also semantically related ones. The reader will be presented a table of all the most frequent modal verbs and their occurrences with I and you, however the further investigation which is carried out in detail is limited to must, should, may and might. The alternative variants of the modal verbs are left out with the exception of have to as this would lead to an extensive study that cannot be carried out in a Master thesis and thus it has to be dealt with elsewhere.It has to be noted down that the pronoun you is a problematic one. It was not possible to find out whether the pronoun was used for the second person singular or second person plural. This could be probably found out from the context nevertheless many of the examples are ambiguous. I have decided not to differentiate between these two uses. Secondly, the paper does not contain research on the interrogative uses of the verbs must, should, may and might; only declarative sentences have been studied, including the negation. The questions have a different meaning and this has to be dealt with elsewhere. Thirdly, the research is carried out on two age groups. The age range is from 25 to 44 years of age. The reason for this is that this group of people can be considered as the productive age group. These people are most probably the ones who pass the language to their children. Thus there is certain continuity and we can expect the children to acquire the linguistic habits of their parents. Coates (1993:125) describes Engles study that he made in 1980 of the language of parents when they play with their children and points out the differences between mothers language and fathers language. Fathers favour directions such as “Why dont you make a chimney?”, yet mothers tend to consult childs wishes such as in “What else shall we put on the truck?”. There is a difference when fathers talk to a daughter or to a son. Men are inclined to use more directives with sons than with daughters. In general, fathers introduce new ideas while mothers put emphasis on interaction and help their children learn how to choose. The people are divided into two groups, one group ranges from 25 to 34 and the other one from 35 to 44. There was not any special intention in dividing the sample. However, interestingly enough, we can observe differences between these two categories. The aim of this research is to collect material for each verb in all of the needed categories (age, gender) and to investigate the instances where we could see significant differences or similarities, these will be further investigated and theoretical conclusions drawn. “One of the purposes of corpus-based research is the testing of theoretical claims made in literature” (de Haan, Ferdinand. 2002:91). The purpose of this investigation is to study the theoretical background of modal verbs and gender linguistics and to find out if the results of the analysis support the theories or, if they do not, to propose new theories based on the results drawn from the actual spoken data from the BNC. Corpus Linguistics:The corpus linguistics has many supporters but there are also many linguists who are against the use of this method for example Chomsky (Aarts, Bas. 1999:5-13). The problem is the same as between the formalist and the functionalist approach in linguistics. Formalists see the importance in the study of competence/langue and functionalists prefer to study the real language production, that is to say the performance/parole. Corpus linguists study the latter. Chafe 1992 formulates the research process that the corpus linguists should follow. The scientist proceeds from observation of language data through description to explanation. He warns against the danger of using corpora only for collection of data and forgetting to analyse it and to draw conclusions out of it. Nowadays also grammarians use corpora as a basis of modern grammar books. Grammar has become, thanks to it, alive. It is based on language that is used in daily conversation and therefore it is up-to-date. Longman Grammar of Spoken and Written English 1999 (Biber, Johansson, Conrad, Finegan) and A Comprehensive Grammar of the English Language 1985 (Quirk, Greenbaum, Leech Svartvik) are just two examples of this kind of grammar books.One of the supporters of the corpus linguistics is Halliday (Cognitive linguist). He is particularly glad for the creation of spoken language corpora. “Precisely because there are patterns which dont occur in writing, we need a corpus of spoken language to reveal them” (Halliday. M.A.K. 2004:11-38). Halliday sees the importance of these corpora in possibility of studying language in progress. Written language can create systemic patterns by an exceptional entry. This is what Halliday calls the Hamlet factor. However, by studying the spoken language, we can find systemic patterns formed unconsciously by native speakers. The problem is that the spoken language still has a low value in culture but because of its essential nature (its semiogenic or meaning-creating potential as Halliday calls it) it reveals language patterns unconsciously. Halliday says. “I would put a high priority on quantitative research into spoken language, establishing the large-scale frequency patterns that give a language its characteristic profile- its “characterology”, as the Prague linguists used to call it.” (ibid). This approach is also supported by Coates (1993:56) who finds this method helpful in the gender linguistics. “It is only with the advent of quantitative sociolinguistic studies that we have reliable data on gender differences in language”. In earlier times research material could be, even unintentionally, manipulated because the researcher expected certain results before even beginning the analysis. There are different types of corpora described in Meyer 2002 and a linguist must find the most suitable one for his particular linguistic investigation. Before the corpora could be digitalised and read as Internet pages, there had been written corpora. The last major pre-electronic corpus of English was the Survey of English Usage. Now we have access to several electronic corpora. Some of them are accessible without any restrictions on-line nonetheless many are fully accessible only after purchasing them or only from PCs at certain institutions. Meyer differentiates corpora according to size (rare features need to be studied on a large corpus and the frequent ones on a smaller corpus), regional variety, diachronic variety, text types included (spoken/written, news, fiction, legal texts), general reference or specific purpose, degree of annotation/mark-up, static vs. dynamic corpus (corpus in progress or in final version) and corpora can be monolingual or multilingual which is helpful for translation or for improving teaching techniques of non-native speakers of English (Norwegian-English parallel corpus which is being extended to other languages, Learners corpus available in Czech as well). The British National Corpus (The BNC):I have chosen the British National Corpus (The BNC) to search for the material to be studied. This corpus was created by cooperation between descriptive linguists, computational linguists, major British universities, and the publishing industry and it is a static corpus. This is why, as Charles Meyer puts it, “the BNC is the culmination of all the knowledge we have gained since the 1960s about what makes a good corpus” (Meyer 2002:138). The BNC is composed of 100 million words and therefore it was the largest structured corpus ever compiled when it was made available in 1995. The texts are taken mainly from the last two decades of the 20th century. The corpus has two parts. The larger one is the written part, which has 90 million words. This part of the BNC is further divided into 80% informative prose and 20% imaginative prose. The former is subdivided into genres. These are Natural and Pure Science, Applied Science, Social Science, World Affaires, Commerce and Finance, Arts, Leisure, Belief and Thought. The BNC offers a spoken part of 10 million words. 50% of the material is the demographic part, which contains transcriptions of spontaneous natural conversations made by members of the public, and 50% is dedicated to a context-governed part, which contains transcriptions of recordings made at specific types of meeting and event. As the informative Internet pages of the BNC tells us the context-governed part has four main categories of interest, Educational and informative events, that is to say, classroom discussions, lectures, news broadcasting, etc The other three categories are Business, Institutional and Public events (sermons, political speech), and Leisure events, such as sport and free time activities. The spoken part was collected according to the random location sampling procedures. This means that Great Britain was divided into 12 regions. 38 points were selected within these regions based on their ACORN profile. ACORN is the Classification of Regional Neighbourhoods. “This profile provides demographic information about the types of people likely to live in certain regions of the Great Britain” (Meyer 2002:25), writes Charles Meyer in his English Corpus Linguistics. According to this profile, speakers of various social classes could be chosen for recording. The spoken material is well balanced as 50% of the speech material is produced by women, and 50% is produced by men. The material has an approximately equal distribution when it comes to the demographic aspect, age groups, social groups (four social groups), and to the type of spoken language. “Recruits used a personal stereo to record all their conversations unobtrusively over two or three days, and logged details of each conversation in a special notebook. Those who took part in the recordings were asked after the conversation to give permission for their speech to be included in the corpus (http:/www.natcorp.ox.ac.uk/).” This method should minimise what corpus linguists call the Observers paradox says Meyer. This occurs when the scientist changes the behaviour of the object that is observed by observing it. It has been shown that the method by which the BNC was created ensures that the 2000-word speech sample is natural and coherent. That is why this corpus is made as good as possible and the research carried out on it is relatively well balanced. All the recordings transcribed to be used in the BNC are preserved in the National Sound Archives of the British Library. The only problem with the BNC is the search for the negative or other contracted forms. The instructions are not easily found and thus the research is complicated. The BNC is, in my opinion, well suited for different kind of sociolinguistic studies. The amount of material is great but the search page has many possibilities how to limit the material. This means that the research can be more detailed if needed. If the search is carried out on a rare word or phrase, the BNC offers such a great amount of material that this makes it possible to search for this as well.Unfortunately the corpus is not available for free on the Internet. However there is a free demo-version of the BNC found at http:/sara.natcorp.ox.ac.uk/lookup.html. This page offers a million-word version on-line.Method:Modal verbs and the BNC:The analysis is carried out, as mentioned above, on the spoken part of the BNC. Firstly a selection was made choosing either male or female speaker in each age group that means in the group 25-34, and then in the group aged 35-44. The simple search was done for the pronouns I and you in the subject position. The verbs investigated at this point were the basic modals must, should, can, could, may, and might. I have decided to include just real modals. Modal verbs are divided into several groups. There are the central nine modals will, would, can, could, may, might, shall, should and must. Verbs ought to, need to, used to and dare are called marginal modals. They behave in some ways as regular modals and in some situations they behave as main verbs. There is, as well, a group called semi-modals which function like modals nevertheless they are rather called modal expressions. (Had)better to, (have) go