Biological evolution is driven by selection pressures of the environment on random non-directed changes in the variation present in genes in a species.
I'm sure that some objection could be raised in a technical manner. But as a single sentence it sums it up well enough.
Language evolution happens through the gradual accumulation of random non-directed changes in the variation present in the vocabulary, syntax, and grammar of a language. I wanted to say "selection pressures on languages" but I don't know if that's true.
There are some selection pressures. If a new word tried to gain fixation it would have to appeal to the intended speakers as a good word. I might introduce the word 'Shoeburyness' to the vocabulary and get people to use it to refer to that uncomfortable feeling of sitting in a seat that has been warmed by someone else's bottom. If it doesn't appeal to other human English-speaking minds, it will be selected against. Some minds might find it an appealing word and adopt it, so that it becomes part of some local dialect but gains no more fixation.
Likewise, extremely long words are selected against in most populations of minds. Other populations might positively select for them (for instance in technical contexts). Most people prefer their words to keep a low syllable count so there is a selection pressure against words becoming very long. This isn't so much the case in the German population, who don't mind creating comparitively long words. Still, there are constraints. It's unlikely that a 400 letter long word will get a lot of use.
Disclaimer: I am not a linguist. I am an evolutionist though, I guess.
™ Version 4.2