Software localization content is unique, in the sense that it often involves two very different types of syntax – human language, and computer language. Programming languages are based on breaking down tasks into a series of logical commands, that the computer can then run and implement. Human languages, on the other hand, are far more complex and are based on what are often random connections, emotions and a whole range of other psychological and sociological factors.
The interface between those two types of communication often leads to strings that look more like lines of code than an actual sentence. Not to mention that in many cases, content is written by non-native English speakers, resulting in text that’s somewhat hard to understand. Computer-oriented and full of placeholders, it’s then sent over for translation, where linguists must battle all sorts of language specific issues to create a final product that actually looks, and reads, perfect.
Not as easy as it looks, ha? And I haven’t yet mentioned how each software products often contain thousands of strings, translated by many linguists over a very, very long period of time. These challenges make software localization a very delicate and complicated process. But fear not – I’m happy to share with you three of our best tips, so you can improve your localization efforts and go have some ice tea on the porch.
If you speak a gendered language, I’m sure you know how hard it is to maintain gender neutrality when translating content from English. In many languages, it’s customary to use male by default – a practice that might have worked well in the past, but doesn’t fit well with modern businesses that take pride in maintaining equality in all aspects. A good option is to try and twist the original sentence to avoid using any gendered words. It’s cost effective, but might sometime leave you with some oddly-phrased sentences, and isn’t always possible. Using both genders in a single sentence is another way to go, albeit a sometimes-clumsy result. Our best suggestion is showing the user customized text based on their gender. This will give you a very natural-feeling translation, while still welcoming all users. While it does often increase the amount of strings to translate, you will be rewarded, as users will appreciate the extra effort and feel much more comfortable using your product.
Why put so much effort into localizing your product just to end up with “thank you, 3 file are saved”? The plural-singular dilemma is real. As in the issue above, the best and easiest solution is to translate two instances of each string – one for plural, and one for singular. But if that’s not possible, there are several other options you can try. Using a slanted line or brackets (“thank you, 3 file(s) are saved”) will keep you covered in the grammar department, even if it’s not the most stylish approach. Rephrasing the strings could also be a good way to go, leaving you with something like “thank you. Number of files saved: 3”. Whichever way you choose to go, don’t forget to check strings with numeral placeholders after translation and make sure they work.
They might look straightforward enough in English, but placeholders can be very confusing when included in translated strings. “In” could be followed by both months and locations in source, but each instance might require a completely different translation. Our suggestion is to try and make sure your linguists receive as much information as you can. Give your placeholders informative names, try to use unique syntax when possible, or add comments. Besides, it never hurts to try and learn a bit about the languages you work with (or consult with your friendly neighborhood language services provider!)
