Comparing of two corpora from a linguistic journal and a transcribed interview - Essay Example

1.0 Introduction This study compares two small corpora, each comprising 300 words, taken from a linguistic journal and a transcribed interview. The texts are therefore diverse in terms of content, style, register and their proposed audience, and are compared and contrasted in terms of the linguistic properties including the number and length of sentences, the number and type of dependent clauses, and the use of contractions. The focus on sentence constructions begins by ascertaining the number of sentences within each corpus. The purpose of this is that the number of sentences will provide insight as to the length of the sentences, which is measured in terms of the number of words within a punctuated sentence. The higher the number of sentences found within a 300 word corpus, then the shorter the sentences would have to be. The fewer the number of sentences therefore, would realize longer, compound and complex sentences and thus would be expected to render more dependent clauses. Dependent clauses, as being naturally longer and more complex (by supplying extra information) than independent clauses, will affect both the length of sentences and the word count of sentences in each text. Hence, the second property to be measured in this study is the number and type of dependent clauses used within each text. A dependent clause does not convey a complete idea and therefore is unable to stand by itself; it relies on an independent clause for meaning. Tallerman (2005) refers to dependent clauses as subordinate clauses and the independent clause as the matrix clause. She also tells us that subordinate clauses are often embedded within another clause and are “syntactically dependent on the matrix clause” (p.77). A dependent clause is introduced with a subordinator (Oshima & Hogue, 1999) which can take the function of an adverb, adjective or noun. These ‘complementizers’ (Tallerman, 2005) determine the type of dependent clause. If introduced with an adverbial subordinator, such as because, so, if, when and while, the dependent clause is an adverbial clause; modifies the verb in the independent clause (Oshima & Hogue, 1999); for example, 1 I workout regularly at the gym so that I can be fit and healthy. The dependent adverbial clause ‘so that I can be fit and healthy’ modifies the verb ‘workout’ in the independent clause, explaining the reason why I workout. Adverbial clauses may also provide reference to place, time, frequency, distance, reason, result, contrast, concession (results that are not expected), purpose and manner (Oshima & Hogue, 1999). (Can I put this sentence right after “why I workout” or put it like this?I am not sure where to put it.) You can leave it here where it is as it explains what adverbial clauses can do in a dependent preposition. I wouldn’t move it but I would add ‘also’ as above. A dependent clause introduced with a relative adverb (when or where), or a relative pronoun (which, who, that or whose) is called a relative clause and modifies a noun or a pronoun (Oshima & Hogue, 1999); for example, (2) Many aerobic classes, which are scheduled in the evenings, are too crowded. The dependent relative clause ‘which are scheduled in the evenings’ modifies the noun phrase ‘many aerobic classes’. Relative clauses are dependent and add extra information; if introduced by an adverb they refer to either time or place (Oshima & Hogue, 1999) as in the above example. ( same here, I don’t know where to put it) This is fine here because it summarizes relative clauses and relates back to the example (please note the added ‘as in the above example) A dependent clause introduced with if, whether, that or a wh-question word, is called a noun phrase and works as a noun (Oshima & Hogue, 1999); for example, (3) My girlfriend cannot decide whether to join the gym with me. The dependent noun clause ‘whether to join the gym with me’ is used as the object of the independent clause ‘my girlfriend cannot decide’. It is important to note that within most complex sentences there is usually only one independent clause but there could be “infinite subordinate” (Tallerman, 2005, p.79) or dependent clauses. The study will consider the number of dependent clauses in each text and deal with the type of each dependent clause, whether each is a relative, noun or adverbial clause. The third and last property to be measured within this study is contractions; more specifically the number of contractions used within each corpus. A contraction occurs when two word are shortened by omitting letters as in ‘I will’ – ‘I’ll’, ‘do not’ – ‘don’t’, ‘they are’ – ‘they’re’, and ‘can not’ – ‘can’t’. They are not usually considered appropriate in formal writing but are more associated with the spoken word and informal writing. The most common contractions include the word ‘not’ and some form of the verb ‘to be’, such as wouldn’t (would not) or isn’t (is not), or mixtures of auxiliary verbs and pronoun, such as it’s (it is), I’d (I would/should) and necessitate an apostrophe to replace the missing letter/s. There are some exceptions such as ‘won’t’ (will not) and the reason for this is because of historical use, which in old English was ‘wynnot’ then ‘wonnot’ and now ‘won’t’ (Wise Geek, 2010). In summary, the sentence construction of two texts – an extract from a linguistic journal and an extract from a transcribed interview – are examined in relation to five variables. The first variable for measurement is the number of sentences found within each text, determined by punctuation. The second is the average length of each sentence, determined by the number of words. The third variable for examination is the number of dependent clauses provided within each 300 word corpus; Then, the fourth variable will be measured is the types of each dependent clause appear in each text (relative, noun or adverbial clause). Finally, the fifth variable to be measured is the number of contractions, shortened combinations of words, found within each corpus. 2.0 Methodology 2.1 Data The first corpus was extracted from a linguistic article entitled ‘The acoustic characteristics of /hVd/vowels in the speech of some Australian teenagers’, written by Felicity Cox and published in the Australian Journal of Linguistics (2006). The article pertained to a study that examined the frequency and length of stressed monophthongs and diphthongs within the context of hVd, provided by 120 (60 female and 60 male) Sydney teenagers (Cox, 2006). The extract was taken from page 148 in relation to vowels. This text is aimed at an audience of academics, students or those interested in linguistics. The second corpus is an extract taken from a transcribed 1991 interview between Oprah Winfrey and a reporter of the Academy of Achievement. The audience for this text is therefore far wider than the first and does not necessitate any prior knowledge or expertise in any area. 2.2 Analysis and coding of linguistic variables The total number of sentences in each text was determined; a sentence is defined with a period (full stop) or question mark at the end. One notable point within the word count pertains to contractions such as ‘won’t’. These occurrences were each counted as one word even though in reality they are two because by their very definition, contractions become a single entity. A similar counting method was applied to hyphenated words, and they were counted as a single word. The word count of 300 was then divided by the number of sentences in each text to ascertain the average sentence length. Each text was then examined for the number of contractions used in each. Finally, the number of dependent clauses was tallied within each text and a list of each type (relative, adverbial or noun clauses) was extracted. 2.3 Research Hypotheses The following hypotheses are proposed in relation to the variables examined within the two corpora taken from a linguistic journal article and a transcribed interview with a famous TV personality. Hypothesis 1. The transcribed interview corpus will provide more sentences than the linguistic journal corpus and thus shorter sentences on average. Hypothesis 2. The transcribed interview corpus will provide less dependent clauses than the linguistic journal corpus. Hypothesis 3. The transcribed interview corpus and the linguistic journal corpus will provide more dependent noun clauses than relative or adverbial clauses. Hypothesis 4. The transcribed interview corpus will provide more contractions than the linguistic journal corpus. Such hypotheses are based on the assumption that the linguisitic corpus will be more formal and adopt a standardized style (including longer sentences and more dependent clauses). By contrast, the interview corpus is expected to feature shorter sentences and fewer dependent clauses due to its genre as a spoken text. 3.0 Results and Discussion The following data shows the counted results from each corpus. With very different results from the written linguistic corpus and the spoken interview corpus, we can analyze the figures in relations to the hypotheses made. Variable Linguistic corpus Interview Corpus Number of sentences 10 29 Average length of sentences 30 words 10.3 words Number of dependent clauses 3 7 Type of dependent clauses 1 noun clause 1 relative clause 1 adverbial clause 4 noun clauses 2 relative clauses 1 adverbial clause Number of contractions 0 14 Table 1- Results indicate that only two of the four hypotheses are fully supported: Hypothesis 1 and 3. A large variance in the number of sentences within each corpus is evident, with the interview extract providing almost 3 times as many as the linguistic extract. Such findings therefore provide a wide disparity between the two corpora in terms of average sentence length. The average length of sentences within the linguistic extract is 30 words, whereas in the interview extract it is only 10.3 words, amounting to almost a third the number found in the linguistic corpus. Many of the sentences within the interview corpus in fact, are simple sentences containing only three or four words, such as “I really do”, “I just couldn’t.” and “He killed eighty people.”. The longest sentence in the interview corpus is 29 words: “I was in the middle of a show with some white supremacists, skinheads, Ku Klux Klan members and in the middle of that show I just had a flash.” There weren’t however, any dependent clauses evident within this longest sentence. The shortest sentence within the linguistic corpus is “Such vowel spaces, with axes F1 and F2, rely on the concept of the vowel target.” and contains 16 words, while the longest is 70 words and contains only one dependent clause – an adverbial clause – wherein the dependent clause is modifying the verb ‘ensure’: The relationships between the first formant (F1) and the auditory quality of height, and the second formant (F2) and the auditory impression of the front, back dimension, (or, more accurately, degree of constriction and point of maximal constriction) ensures that when the first two formants of a set of vowel targets are plotted on axes with appropriate directional and scaling characteristics, the result closely resembles the traditional auditory vowel map Hypothesis 2 predicted that there would be more dependent clauses found in the linguistic corpus because of the expectation of longer sentences. This hypothesis however, is proved incorrect because findings indicate that the interview corpus provides more than double the number of dependent clauses found in the linguistic corpus at a ratio of 7:3. The types of dependent clauses found within the interview corpus supports Hypothesis 3 because it provides 4 noun clauses, 2 relative clauses and 1 adverbial clause; thus illustrating more dependent noun clauses than relative and adverbial clauses. The linguistic corpus however, provides a different result. Rather than having more dependent noun clauses it provides only one of each of the three types, so in fact does not support Hypotheses 3. Hypothesis 4 proposed the speech corpus would provide more contractions; the linguistic corpus in fact does not contain any contractions, while the interview corpus contains fourteen. This is not surprising because spoken text is more spontaneous and apt to contain more contractions in an effort to speed things up and simplify the communication between the speaker and listener. 4.0 Conclusion This study compared two differing corpora, each 300 words in length, for four different linguistic variables. The corpora were extracted from a linguistic journal and a transcribed interview and the linguistic variables were number of sentences, average length of sentences, number and type of dependent clauses and number of contractions. Results supported Hypotheses 1 and 4, wherein more and shorter sentences and more contractions were found in the transcribed interview corpus than in the linguistic journal extract. The findings did not support Hypothesis 2 because even though the linguistic extract provided less and longer sentences it was the interview extract that provided more dependent clauses within the 300 words. Hypothesis 3 was only partly supported because the linguistic journal corpus did not provide evidence of more dependent noun clauses than other types of dependent clauses. Read More