Internationalisation and tagging
At last night’s Web Standards Group meetup Richard Ishida, the Internationalisation Activity Lead for the W3C presented some guidelines and tips for designing international web–pages. Richard’s talk was followed by a thought provoking presentation on “How to make the most out of eGovernment” by José Manuel Alonso, the eGovernment Lead at the W3C.
Richard’s talk triggered a few thoughts inside me regarding i18n and tagging.
According to Information Architecture for the Word Wide Web, “Free tagging, also known as collaborative categorization, mob indexing, and ethno–classification, is a simple yet powerful tool. Users tag objects with one or more keywords. the tags are public and serve as pivots for social navigation”
. Content is tagged collaboratively. Whereas classification is the process of describing things according to common characteristics, my view on ethno–classification is that the descriptions are related to one’s own culture. There isn’t always a link between culture and language, but it’s often present.
In a situation where tags are created automatically (via a service Reuters’ Open Calais perhaps) tag translation can be kept from being a problem. But in situations, like in Flickr, where people from different cultures and speaking different languages use an international site and can apply tags to common content the issues may not be trivial. In these circumstances is it appropriate to translate tags? I asked Richard this after the presentation, his opinion was that tags shouldn’t be translated. Why wouldn’t tags be translated?
Translations of small, descriptive words may have undesirable results. A translated tag may go from being 6 to 20 characters long, or a certain language’s translation won’t be available due to a lack of context around the word. Tags could be language specific, and this may be related to the culture from which the language came from. Tags may be misspelled. Some tags are proper nouns and many are slang. Some of the issues just mentioned are also completely relevant to localising user interfaces.
Could you attempt to work out what language tags are written in and then translate tags to the current user’s preferences or to the site’s default? Or to hide tags written in languages other than the current language preference? The imagined benefits from translating tags are outweighed by details and trivial issues. Flickr seems to have decided not to translate tags (or comments, titles and so on) and that seems to be a good decision.
