Content Management: Just what has changed?


The incredible expansion of the World Wide Web in the last two decades (1) has been the subject of much comment. It is difficult to obtain accurate, up-to-date statistics on an object of such magnitude that is evolving so quickly, but in July 2008 Google counted a trillion unique URLs and noted that the number of Web pages was growing by several billion pages (2) – a day! Something that is less frequently commented on, however, is the growing complexity of individual websites, whether this complexity is calculated informally as the quantity of information presented on a given site, or more formally as the website’s depth, i.e. the number of interlinked pages, or the minimal number of clicks required to get from the homepage to some embedded page. But as more and more companies increasingly come to rely on their website as the medium of choice for the distribution of information, not only to their external client base, but also to their own internal staff, the growing complexity of those sites creates significant problems of its own. These problems are amplified, moreover, by the fact that fewer and fewer websites are authored by a single individual; rather, their content may be modified and expanded by any number of geographically dispersed company employees.

The job of Web content management systems (WCMS) is to help users come to grips with the growing complexity of today’s websites. Here, we shall focus on just one important aspect of this daunting challenge. In an officially bilingual country like Canada, the websites of all government departments must be published simultaneously in both official languages and, furthermore, must be of equal quality. (3) While this is a legal obligation for the Canadian government, for many private-sector companies a similar obligation obtains for commercial reasons, and not just in Canada. In today’s global economy, many corporations have understood that clients are far more apt to do business with them if they are served in their own language, and hence they too have developed multilingual websites in which the content must be published in several languages in parallel.

Now suppose that you are the manager of such a company’s translation service, whose responsibility it is to ensure that all the ‘foreign language’ content of your company’s sprawling website accurately reflects the evolving content on its English source side, and to do so as promptly as possible. The following are among the many questions you must resolve: (i) How does the translation service know when and where new content has been added to the website, or existing content modified? (ii) Assuming that you are provided with a list of the webpages that have undergone changes, what exactly must be translated on each page, and in what format is it?

Fortunately, there exist WCMS which can help the translation manager answer these critical questions by periodically scanning the entire company website (including its internal and external links), in order to automatically detect all pages that have undergone changes and provide the manager with a detailed report. Indeed, without such a system, it is difficult to imagine how our translation manager could possibly manage.

But what if those same changes and additions are also used in publications other than the company website, e.g. in instruction manuals or product documentation? The translation manager would surely want to know that as well, so that he could avoid translating the same content two or more times. Here, a simple scan of the company website, no matter how thorough, will not suffice. What is required is a full-blown CMS which stores all corporate content in a central repository and which tracks every modification to it and every instance of content reuse. In most such systems, the content is stored a neutral format, regardless of the ‘channel’ in which it is eventually published, which responds to the second of the questions raised above. What is more, the content tends to be stored in smaller chunks, as opposed to complete documents or pages, which allows the system to pinpoint just those passages where changes have been made. Such CMS have only come into use relatively recently in large multinational corporations, and their impact on translation services is not yet fully understood. For example, translators frequently maintain that they cannot translate sentences in isolation, but need to see the larger context in which each sentence occurs. How will they react to the smaller informational chunks of such CMS? And how will such systems interact with commonly used translation support tools like translation memories?

(1) As startling as it may appear, the first publically available Web server dates back to just 1991, and Berners-Lee original project proposal for the “WorldWideWeb” (as he wrote it) was only submitted in 1990.