> Over a couple of generations, the pidgin evolves into a full language. Yes, and full creole languages, ones with relatively little of the grammar of their pa

> Over a couple of generations, the pidgin evolves into a full language. Yes, and full creole languages, ones with relatively little of the grammar of their parent languages, have a number of similarities in their grammar. These similarities persist whatever the parent languages. Different continents, different parent languages, grammatical similarities. For example, plurals are often formed by reduplication (dog -> dog-dog = dogs), word order is subject-verb-object or subject-object-verb, suffixes follow the morpheme they're attached to. The grammars are extremely regular, almost no exceptions, no complex sound-fusion rules. It's generally thought that creole languages, immediately after formation, are actually simpler than most other languages grammatically. Which may have been like creoles at one point but then picked up a bunch of useless nonsense you just have to memorize. (This view is sometimes challenged, and it must be admitted, it is hard to define and measure complexity in language.) That is one of those small hints that language may be physiologically innate, or an emergent phenomenon that recreates itself in similar ways each time it is invented by people who need to communicate -- i.e., universal grammar, that ever-so-tempting yet ever-so-unprovable idea that got everyone tangled up last century.