| This thesis is intended to analyze systematically, with the assistance of computer software SNOBOL4 and statistical package SPSS, the distribution and usage of pro-forms in scientific English, as compared with the other five corpora; namely, BROWNA, LOBA, MTE, NEC and MFC. The study reveals a number of features of pro-forms in T4: the highest frequency of pro-forms in T4 is the result of the much use of pro-nominal one and special pro-form thus, which is determined by the request to explain the objects in question precisely; among complex pro-forms, do so is the most widely used one, since it denotes a general type of action instead of the same action; the relatively less use of pro-VP and pro-clause is due to more lost information involved; pro-complements can also often be found in T4, which is governed by the nature of T4. |