Week 6 - Danny's Article Review

“Lost in Translation”: cross-lingual communication, and virtual academic communities

This paper displays the communication problems in virtual worlds caused by language barriers, provides a general definition of translation, and describes the status quo of Machine Translation (MT) as well as its problems. Some MT problems vital to virtual communities in academia are discussed and the potentiality of MT is also suggested.

Nowadays, many scholars throughout the world work in a collaborative way. They “share similar goals and interests. In pursuit of these goals and interests, they employ common practices, work with the same tools and express themselves in a common language. Through such common activity, they come to hold similar beliefs and value systems.” (Lave and Wagner, 1991) Their communication relies heavily on English which plays a predominant role in this age as Latin did in Middle Ages.

However, the values of academic materials in languages other than English are belittled as they are scarcely translated. While the importance of high-quality translation is indisputable, good translators are scarce commodities.

Moreover, language barriers in virtual world are stumbling blocks to scholars as electronic documents are becoming increasingly popular given the growing importance of Internet technologies in scholarly communication. Not only technology expertise but also translation assistance is needed in academia. The benefits of MT are as following (Theologists, 1998):

1.      Improved quality due to greater terminological and phraseological consistency.

2.      Enhanced productivity in which no re-translation is needed owing to the advantage of software memory.

3.      Portability of transmitting by electronic means.

4.      Improved translation speed compared with human translators.

5.      Cost reduction for large quantities of texts.

6.      Consistency in a particular manner.

7.      Lack of bias with choice of wording by omitting, inserting or subtly changing the meaning of text.

8.      Availability as it meets the growing need of competent translators.

But the problems with MT are also obvious:

1.      No current operational machine translation systems can produce good quality output without either placing restrictions on input texts or involving human assistance before, during or after translation processes.

2.      Present machine translation systems make “simple” grammatical errors that no human translator would make.

3.      Machine translation cannot as yet deal with culturally-based ontological issues (Akahani et al., 2002), although this has been attempted (Hovy, 1998).

4.      Most of the current efforts in machine translation have not progressed much past the stage of providing low-quality translations based on ever-expanding vocabulary databases. “Most translations fell somewhere between impressive and nonsensical; in general they were surprisingly understandable, if odd and stilted” (Budiansky, 1998).

5.      Machine translation works quite well for translation predictable technical texts – texts that never go beyond the expected domain of discourse.

6.      Computers cannot translate like humans because they lack of agency: “the capacity to make real choice by exercising our will, ethical choices for which we are responsible… A computer has no real choice in what it will do next. Its next action is an unavoidable consequence of the machine language it is executing and the value of data presented to it… Without agency, information is meaningless. So a computer that is to handle language like a human must first be given agency.” (Melby, 1995).

It seems unlikely for MT to replace human translators in near future due to its inaccuracy. But MT can release the burden of large quantities of translations which require little in accuracy.

In my opinion, MT is a very huge and complicated interdisciplinary project which requires IT expertise, knowledge of linguistics, as well as techniques of translation and interpreting. Even after certain MT software is developed, great efforts must also be taken to update its database and improve its system. This work can be really trivial and overloading for MT programmers. But the outcome will be well-worth the perplexity. Take dictionaries or encyclopedias as an example: tens of years are spent on compiling these large books; once they are published, a great many scholars, writers, and ordinary people can be benefited; new entries are continually appended decades after their publishing. The same is true for the development of MT. Although the task looks like a mission impossible, it can be built up brick by brick through the efforts of all programmers and scholars. Besides, the update and revision of MT are more accessible nowadays than the age of hard-covers given the convenience of Internet Technology.

References:

Akahani, J., Hiramatsu, K. and Kogure, K. 2002. Approximate Ontology Translation and its Application to Regional Information Services. Unpublished poster delivered at the First International Semantic Web Conference, Sardinia, 9-12th June, 2002. [Online] Available WWW: http://iswc2002.semanticweb.org/posters/akahani_a4.pdf

Budiansky, S. 1998. Lost in Translation. The Atlantic 282(6). [Online] Available WWW: http://www.theatlantic.com/issues/98dec/computer.htm

Hovy. E. 1998. Combining and standardizing large-scale, practical ontologies for machine translation and other uses. In Proceedings of the First International Conference on Language Resources and Evaluation (LREC). Paris: The European Language Resources Association, pp. 535-542.

Lave, J., & Wenger, E. 1991. Situated learning: Legitimate peripheral participation. Cambridge: Cambridge University Press.

Melby, A. 1995. Why Can’t a Computer Translate More Like a Person? [Online] Available WWW: http://www.ttt.org/theory/barker.html

Theologitis, D. 1998. ...and the Profession? (The Impact of New Technology on the Translator). Terminologie & Traduction 98 (1):342-343.