Apparently there is an entire blog dedicated to answering their contradictions, and its take on the above is this:
In chapter 10 of Genesis, genealogies are given of Noah’s sons and their generations. Each lineage developed their own language, as would be expected when they divided and diversified in land settlement. However, at some point in the last generations of Chapter 10, and at the beginning of Chapter 11, the tribes of the world decided to join back together under one head, or government. It was at that time that they agreed upon one language, just before the tower was built.So there was a single language which naturally split into many, was purposely compressed back to one, then were split into many again? I find that a little unlikely for at least the reason that if mankind could agree on a single language before the tower, what's to keep them from simply doing it again? It also undermines the nature of the judgment at Babel, which presumes that a plethora of languages is an insurmountable obstacle that will force separation of peoples (as it has proven to be in history). To propose that everyone in the world decided to speak a single language is not only unsupported by the text or history, but by what we know of people - they simply don't do that sort of thing.
No, there's a simpler answer: Gen 10 and 11 form a single narrative separate from the rest of Genesis, and the genealogy is its main purpose and the narration is simply explanatory material. To see it laid out in outline form without those distracting verse divisions, click here.
I mentioned before the "toledoth" (for a thorough explanation, try here), a separating comment that divides Genesis into books, probably with different original authors, which books were combined by Moses at a later date. There is a toledoth in 10:1 ("Now these are the generations of the sons of Noah...") which rather than belonging to the material following, belongs to the material preceeding. The story (generations) of the sons of Noah is the flood narrative which begins at the toledoth in 6:9 and ends in 10:1. 10:2 thru 11:10 are one account ending with the toledoth "These are the generations of Shem."
Now, if we take 10:2-11:10 as a single "story" we find that it has 2 natural parts, genealogy (10:1b-32) and narration (11:1-9), and the author has chosen to arrange his book topically with the narration explaining the genealogy. He lays out the families of mankind as they existed in his day, then he narrates the events that caused them to be as they are.
Catherine Murphy of Santa Clara University explains how the ancients did it this way:
The historian usually began by writing out a sketch of the events, then inserted secondary material (speeches, dramatic episodes, digressions) and finally arranged the collected material in a systematic way (chronologically, topically)...What we have in Gen 10-11 is a case where our historiographer arranged his material topically rather than chronologically and even followed the pattern above - except that he mixes travel descriptions in the genealogy - and then relied on his audience to understand the cause/effect relationship. It's only because we insist on treating the whole passage as a chronology that we have troubles, because then we have to propose unworkable solutions to get around our own bad assumption.
Ancient historiographers describe five types of historiography:
1. genealogy or mythography
2. travel descriptions (geography and ethnography)
3. local history (horography, annalistic writing)
4. chronography (chronicles)