Apr 14, 2011

Coming to terms with Kilgray

The pre-conference day for memoQfest 2011 offered dual tracks in the morning session for the TM Repository and qTerm, the advanced, server-based terminology module for memoQ. Ever since attending a webinar when qTerm was released last October, I've been intending to blog on it, and since I already had an overview of what I needed to know about the Repository from last year, I chose the qTerm track. However the Twitter feed from Polish translator @wasaty made it clear that I was missing a lot of interesting news about Kilgray's TM management technologies.

István Lengyel and Gergely Vandor of Kilgray served a lot of meaty technical details on qTerm, many of which could be the subject of an entire blog post. It's a product with great potential I think, and I expect it will evolve considerably in the course of the next year. Gergely's technical insights on problems often encountered in data migration and how non-standard the TBX "standard" truly is were particularly interesting to me personally, revealing some useful and interesting information about data exports in that format from SDL's MultiTerm. And I thought the world of TMX was a big mess....

István's telling of the history of term management at Kilgray offered me a look at the very different world of translators of technical information working with small distribution languages. In some respects, that is a very different world from mine, and I very much appreciate how being shaken out of my comfortable German/English language pair perspective can sometimes help to flush my hardened arteries and get a little more oxygen to my brain. And the repeated reference to the role of terminology in branding by all three speakers in that morning session gave me some new ideas for how to help my direct clients and agencies understand even more clearly the importance of getting terminology right.

But like yesterday's train-the-trainer session, the real highlight of the day for me was not a technical presentation with specific details of a product I find interesting. It was a general discussion of purpose and philosophy in terminology management, by expert terminologist and consultant Barbara Karsch, who was deeply involved with terminology at JD Edwards and Microsoft before becoming an independent service provider to LSPs and corporate translation consumers. Her web site offers a lot of interesting information and definitions that are well worth reading. Her methodical presentation of the real costs involved in terminology mis- or non-management left little room for excuses and made a strong, objective case that any sober business person can appreciate. What I learned from her will help me make a better case to clients I value so they can help themselves. I very much look forward to getting a copy of the slides from that presentation.

Among all her valuable advice, however, one particular point stands out for me, an obvious one that I know well from my own experience. It applies to both terminologies and the collections of data often used inefficiently for terminology: translation memories. Without maintenance and updating, terminologies (and TMs) eventually become worthless. There is a definite life cycle which applies to a lot of this data, and all of the babbling about matches, fuzzy matches, etc. and the inertial complacency of many of us and our clients with regard to existing collections of data can too easily cause us to lose perspective and sacrifice future quality and reputation while indulging in the illusion of cost savings.

Those who make an effort to think clearly about the real costs of processes and decisions very often discover truths at odds with lazy common wisdom and benefit accordingly. When we get beyond the fear, uncertainty and doubt invoked by dishonest companies and "experts" with an agenda at odds with the interests of freelancers, public bodies and LSPs of acceptable size, we are very likely to arrive at decisions that seem risky to those blinded by propaganda.

Viewed objectively, the case is very clear for efficient, modern management of terminology with technologies such as those offered by Kilgray. And considering the larger technological context of the integrated environment in which tools such as qTerm work, the case for memoQ as a mother lode of value for translation management is just as clear. That is perhaps why I will be able to greet esteemed colleagues from SDL and other important contributors to the translation tools industry who will be attending memoQfest this year to divine the future directions of translation technology :-)


  1. Good point about resource life cycle. Apart from "set in stone" TMs, basically legislative or equivalent texts, e.g. standards, that sort of thing (until they get amended, of course), we reckon that TMs have a maximum shelf life of around 18 months, and often far less. Their value decreases quite sharply over time to the point where they contain so much noise it's not worth keeping them any longer. But no doubt different rules apply in other subject areas.

  2. Thank you for pointing out the greater longevity of those legislative texts, Robin - that will be a nice example for contrast in discussions.

    I'll have to take your word for the shelf life of text and expressions in your financial world; most of the technical disciplines I translate will have a longer, but still limited, usefulness I think. Management texts on the other hand are probably not far from what you cite. When I think of the popular buzzwords of the 1980s and how those same ideas are often expressed today it's disconcerting.


Notice to spammers: your locations are being traced and fed to the recreational target list for my new line of chemical weapon drones :-)