Psycholinguistics and electronic publishing : relevant experimental data from reading, writing and translating technical texts
JM Jan Ulijn · TU/e Research Portal · 1992
Although electronic publishing (EP) does not manifest a more positive impact on human communication (Bradley, 1982, Kleimann & Humphrey, 1982 and Haas & Hayes, 1985 a and b), it has at least one very attractive aspect: the constant option of interaction and revision makes the writing process most flexible, natural and in line with current psycholinguistic research (Flower & Hayes, 1981).Interactive manmachine communication.however, requires more than the traditional ways of efficient communication between human beings.The present information technology (IT) allows for a direct access of reader and writer to each other and an increasing rate of interaction.This puts high constraints on quality of writing and publishing and, hence, on the language, carrier of most of this information.Size and shape of information are no longer fixed for a long time.The end users have such a lljl'ge variety of choice, that they will consult/purchase only the best reapable documents.TherFfore, the task of linguistic research for EP cannot just be to replace the human functions, such as reading, writing, editing, translating and publishing by IT systems.On the other hand Artificial Intelligence (AI) and Machine Translation (MT) projects (e.g. the Eindhoven Rosetta project for English and Spanish, Sanders, 1988) can only succeed if we know how people read, write, edit, translate or publish.This research, of course, can be supported by IT, as has been proven by recent reading research (Thibadeau, Just & Carpenter, 1982) I and the Dutch Language Technology ESPRIT project by Kempen. 1986 i and 1988), based upon careful psycholinguistic research.EP in today's international technical & business communication seems to benefit from SGML (Standard Generalised Mark-up 2 JanM.Ulljn Language) which could include a standardised, limited, simplified natural language system (cfr.the initiative taken by 12 multinational business concerns, Cingras, 1987), but psycholinguistically validated in order to make interaction more efficient and the omnipresent dataoverload truly convertible into useful information not only through the shortest, but also through the most comprehensible representation.Moreover, the international scope of EP demands a multilingual approach.We will present some results of an ongoing research program at the Eindhoven University of Technology about the psycholinguistic verification of hypotheses on cross-linguistic & cross-cultural communication through texts in technology and business, which could provide EP with elements for the above language standard, since it is completely directed towards the end user: novice or specialist.It focuses on general user friendliness, reading, writing and translation and includes lexical (terminological), syntactic and textual aspects for English, French, Dutch and Polish (Ulijn & Kempen, 1976, Strother & Ulijn, 1988, Lankamp, 1988 and Wojnicki & Ulijn, 1987).These findings and the ones for German, Spanish, Russian, Arabic, Japanese and Chinese (Kaplan, 1966, Clyne, 1981, Hinds, 1983and Tsao, 1983) suggest to American, European and Asian manufacturers to use an EP system which translates their product documentation both linguistically and culturally for their respective foreign end users, because of a cross-cultural difference in learning and reading styles (ascertained for English and Japanese).What is the position of Spanish culture to this respect?Finally, we present briefly the Eindhoven Psycholinguistic Communication Program for European Engineers (general MS degree and post-graduate), including research for EP and offer some research suggestions for future EP.