Well, one limitation is that a language (formally described) is a countable set of strings. This means that there is some way to write down all the sentences in the language, one by one, so that every sentence in the language will appear at some definite position in your list. Your list will have a beginning, but not an end; you will keep writing it forever, much like the digits of pi.
Since language is a countable set of strings, it can at most name a countable set of phenomena.
But some phenomena are not countable. The set of real numbers between 0 and 1, which we can name by the interval (0, 1), is not a countable set; there are "too many" to write them in a list like that. See Cantor's diagonal argument.
As a consequence, there are real numbers between 0 and 1, that have no sentence or formula naming that specific real number. In fact, another theorem says the set of specific real numbers that we can name has measure 0. This means that greater than 99.999999999% (as many 9s as you like) of the interval (0, 1) is not specifically nameable by any language. We can talk about sets of numbers that happen to include these un-nameable numbers, but we can't name each un-nameable number individually.
Many natural phenomena may be continuous. This means that language can never precisely say where an electron is, or even what the electron's probability distribution of locations is; language can be as precise as desired, but never perfectly precise.