Proposed general model for Serbo-Croatian continuum

[Phillips, Addison]

> "Serbian" without a script subtag shouldn't necessarily cause suspicion.
> Sometimes things are obvious:
> 
> <span lang="sr">Cвa људскa бићa рaђajу сe слoбoднa и jeднaкa у
> дoстojaнству и прaвимa.</span>
> 

Obvious to whom? Yes, you can analyze the text to find out what script it (probably) is in--but that's expensive to do in software. Obviously, humans can do this analysis instantly/visually, but that's not the point of metadata.