Is it reasonable to use language codes in an interface?

Question

In an interface that supports multiple languages, is it reasonable to show them by their language code, or should the full names be used?

eg:

Select Language:
En ❘ De ❘ Es ❘ Fr ❘ Pt ❘ Ru ❘ Zh

Or is it better to use:

Select Language:
English ❘ German | Spanish | French | Portuguese | Russian | Chinese

Personally I don't know what these language codes mean off hand, however - native speakers who are use the internet may be used to recognizing their own language (its often assumed that USD means United States Dollars for example).

---

This is technical documentation, however this may not be important.

Answer 1

It's a common pattern to use both, with following points in mind:

Design for the lowest common denominator

Assume that in the mix of language use and comprehension, many users will get confused with just the two letter codes

Avoid making the user think unnecessarily

In that case, that would mean including as much information as is reasonable. It doesn't take that much extra to use both the language name and the code.

The language selection should be in the language of the user

so 'Español - es', not 'Spanish - es'

The two-letter codes are a standard and should be lower-case

es, not Es or ES

So, ideally, your list should look something like this (formatting is a question of style):

English (en) | Deutsch (de) | Français (fr) | etc.

Answer 2

If you need to abbreviate, they're the abbreviation to use.

If you've a technical audience, of a sort where knowledge of the ISO codes would be expected for their field, they're even more likely.

Any other abbreviations should definitely be avoided. I've come across plain confusing abbreviations the authors deemed "more obvious" before and it's a massive pain.

In all though, it's better to go with full names, but not in another language as you have in your second example. Would you have known to pick Béarla, английский, انگریزی, or 英語 to select English if that was put in front of you? Many English speakers wouldn't. And why should they?

Much better to go with:

English | Deutsch | Español | Français | Português | Ру́сский | 汉语

And of course, never use flags (I know you didn't threaten to, but people often do).

Answer 3

If user needs to input with keyboard, I think that the two letter combination will be efficient but there are also three letters as option. Also check ISO 639-1 here

Using and searching for standards are important.

Answer 4

Is it reasonable to use language codes? Yes. Probably. They're fairly recognizable. Is it better to use the full names...? "Better" is contextual.

It's good of you to mentioned that the context is technical documentation, and since that's the case I would posit that, while the full names may be prettier and slightly more informative, they are not necessarily "better".

Language codes will be recognizable as such in even moderately technical material, a users' manual for instance, and the actual meaning of the codes should be recognizable to the speakers of that language. And if you have any fear that they will be received with any uncertainty you may include the full names (in the native language--not using the English equivalent as in the example from the original question) as a smaller sub-heading or on hover, or some such device, depending on the capabilities of your medium.

Another advantage is that language codes will scan faster than the full names. They can be used almost as an icon would, allowing near-instant recognition of a basic significance, and don't need to convey much meaning of their own.

They're easier to keep out of the way, as they are so concise, which is a plus as once one has found the language he is looking for he doesn't need them anymore.

Finally, in some contexts, it is possible that they are less ambiguous than full names--there are 3 Norwegians as I understand--as they are a standard and thus designed to exact.

The only caveat that I might include is that depending on the perceived literacy of the audience you may need to consider languages which to not use a Romanesque alphabet, but again, you're asking about technical documentation, so likely the user has been well exposed to our characters.