I do not think the language would be much of a problem. I'm not sure whether it was a later invention, but I'd think trumpets or other instruments would be able to give signals that every man would be able to memorize the meaning of. Even with only voice, at least the basic commands would be drilled into the soldiers so they'd know when to charge, retreat, shoot, flank or halt even if there's a foreigner shouting it and waving his sword at the enemy. The language barrier is a lot worse today, using telecommunication with no common visual reference or body language involved (although most people are certainly more acquianted with foreign languages these days).

But a heterogenic army (non-faction troops) might have morale problems, for instance as they are fighting far from their home for a foreign cause they care nothing about or even dislike, or mercenaries who may not be keen to fight on when the one who pays them kicks the bucket.