CHAPTER 4 RESULTS AND DISCUSSION
4.1 EXTRACTED COMMON WORDS
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Chapter 4 Results and Discussion
The previous chapter has reported the research design for compiling a Business Reports Corpus (BRC), extracting common words, and identifying formulaic language composed with extracted words. This procedure was designed for answering the research questions proposed in Section 2.5:
Research Question (1)
To what extent do common words occur in company annual reports?
Research Question (2)
How formulaic language composed with identical common words vary across genres?
Research Question (2) has been refined to explore formulaic language from two different perspectives:
Research Question (2-1)
Do the lengths of identified formulaic language vary across genres?
Research Question (2-2)
Does composition of the identified formulaic language allow syntagmatic variation or paradigmatic variation?
This chapter demonstrates results of research in an attempt to answer the above research questions.
4.1 Extracted Common words
This section gives results to answer Research Question (1) that asks for how common words are distributed in company annual reports. Figure 4.1 below displays
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
results of extracted common words. There are 733 words27 in Business Reports Corpus (BRC) that have unusual frequency when compared to their equivalents in Brown Corpus, match words in the Senior High English Wordlist for Reference [高中英文參考詞彙表]
(Jeng, et al. [鄭恆雄等] 2002), and does not involve technical words or compounds in the accounting field. Words in this list of key-keywords are displayed in descending order of word coverage among texts; the first nine words from annual to significant have the widest coverage among all 105 texts.
27 Complete lists of extracted words are attached as Appendix 4.1.1 and 4.1.2. These two lists may be
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Figure 4.1.1 Results of extracted common words assorted based on text coverage
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Table 4.1.1 Distribution of extracted key-keywords in BRC
Key-keywords in BRC (105 texts) No. of words Percentage (%)
105 texts (100% of 105 texts) 9 1.2
85-104 texts (81-99% of 105 texts) 50 6.8
65-84 texts (61-80% of 105 texts) 27 3.7
44-64 texts (41-60% of 105 texts) 44 6.0
23-43 texts (21-40% of 105 texts) 89 12.1
1-22 texts (1-20% of 105 texts) 514 70.1
Total No. of key-keywords in Business Reports Corpus 733 100.0
Table 4.1.1 presents how the 733 common words are distributed in Business Reports Corpus (BRC). Distribution of occurrence of the 733 words in BRC can be observed in the first and second column of this table, in which the 733 words in BRC are arranged based on their occurrence with different proportion in the total 105 texts. For instance, there are 9 of the extracted common words (1.2% of 733 words) occurring in all the 105 texts.
At this point, the above research results respond to Research Question 1, which enquires about how common words are distributed in company annual reports. The extracted 733 words, which comply with the operational definition of common words of
‧
About 70% of the 733 words (514 words) has occurrence in less than 22 texts of the 105-text collection and only 1.2% of the 733 words (9 words) has the widest coverage across the collection. Since BRC is compiled with annual reports from 35 different industries evenly (Section 3.1.3 & 3.1.5), those figures representing distribution of the 733 words indicate the importance of the extracted common words regarding their degree of practicality in the business genre across various industries. Table 4.1.1 guides ESP practitioners looking for useful vocabulary for learners with business purpose.
As for further scrutinizing the overall practicality of SHEWR [高中英文參考詞 彙表], which is the present generally accepted common-word list in Taiwan, Table 4.1.2 is to be consulted.
Table 4.1.2 Distribution of extracted key-keywords in SHEWR [高中英文參考詞彙表]
Key-keywords in BRC (105 texts) No. of
words
Percentage (%)
Total No. of key-keywords in BRC 733 11.3
SHEWR words that neither bear outstandingness nor occur in BRC
5,747 88.7
Total of SHEWR words 6,480 100.0
Table 4.1.2 tabulates proportion of extracted common words as well as SHEWR [高中英文參考詞彙表] words that neither bear significance in BRC nor have occurrence in BRC. The very bottom of second column of this table presents the total number of words in SHEWR (6,480); the difference between 6,480 and 733 (5,747) is the amount of
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
SHEWR words that neither bear outstandingness nor occur in BRC. As can be seen, the 733 words bear a proportion of 11.3% in SHEWR, indicating that there are about
one-tenth of SHEWR words consistently occur in Business Reports Corpus. Our attention will be focused on the figure 11.3% to discuss practicality of SHEWR. The figure 11.3% cannot be used to claim high practicality of SHEWR words in business genre if viewed with the conventional computation in which practicality of wordlist is evaluated by comparing raw frequency of occurring words to the whole word collection28.
However, a strict interpretation needs to be placed in the figure 11.3% (733 words) since the concept of unusual frequency was applied in the extraction process.
With the research design of this thesis, the 733 words were identified with the notion of unusual frequency. Hence, the 733 words escape the interference of
high-frequency words (such as grammatical words) by comparing each 105 business texts to the Brown Corpus; these extracted words are words with unusual high frequency in BRC (presenting business English) than in Brown Corpus (presenting general English), which means that readers of these business texts are exposed to the 733 words with higher probability than in common situation. With this interpretation, the figure 11.3%
makes a conservative claim that the 733 words, though bearing low proportion in the complete SHEWR [高中英文參考詞彙表], compose a vocabulary bank compulsory for English learners who attempt to advance to English for Business Purpose (EBP) on the basis of English for General Purpose (EGP). The figure of 11.3% (733 words) should not be used to conclude a low practicality of SHEWR words in business context; on the
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
contrary, this figure provides strong assistance for ESP practitioners because it indicates that ESP courses could be arranged with a manageable amount of vocabulary.
Moreover, the 733 words are highly valuable for Taiwan English learners for ESP
courses due to the fact that these words are assumed to be acquired in the period of senior high school, which in a way bridges the gap between EGP and ESP.
The above interpretation about the figure 11.3% (733 words of SHEWR) limits itself in making reference if we are to theorize the link between English for General Purpose (EGP) and English for Specific Purposes (ESP) because of the wordlists and text employed. Wordlists applied in this study include SHEWR [高中英文參考詞彙表] and Chinese-English Translation of Important Accounting Terms [重要會計用語中英對照], both of which are compiled with Taiwan perspectives; texts collected as Business Reports Corpus (BRC) are limited in Chinese corporations involved in issuing securities and these texts cannot be regarded as the only representative of business genre. Hence the above interpretation on 11.3% and inference about link between EGP and ESP holds validity only in the pedagogical context of Taiwan, especially under discussion on the gap between English teaching in the senior high level and College English instruction for learners of the commerce school.
As for identifying formulaic language composed with common words, only the four adjectives annual, applicable, financial, and significant will be discussed due to constraints of space.