Abstract
Application of computational linguistics, i.e., language translation by computer, has been proposed as a means of producing readable translations of technical English-to-Vietnamese.
This report is about an experimental study of the readability of translations that could be used for training or equipment maintenance.
The experiments involved assessing the readability of Vietnamese that had been translated from English by three methods: (1) expert human translators, (2) un-edited translation by computer, and (3) edited computer translation. English was a control condition. Readers included two groups of student pilots : 168 in the Vietnamese Air Force (VNAF) and 88 in the USAF. Material that was translated consisted of three 500-word passages sampled from a standard Air Force text, Instrument Flying. Readability was measured by : (1) reading comprehension tests, (2) cloze procedure, and (3) clarity ratings. Time to complete each of these tasks was also measured.
Major conclusions of the study are : (1) expert human translators produce more readable translations of technical English-to-Vietnamese than is done by computer; (2) Vietnamese readers, trained in English, show the highest comprehension when dealing with that language; (3) comprehension loss becomes relatively greater, as more and more difficult material is read, for computer-based translations than for human translations; (4) method of translation does not affect reading speed.
Scrambled Cloze Procedure: Does It Influence EFL Learners’ Reading Comprehension and Writing Performance?
