There are a number of positive things that can be said about Wolfe's book. First of all, I think Wolfe should be commended for writing about a seemingly arcane topic like the origin of human language. It is unlikely that an author less famous (or controversial) than Wolfe would have been as successful in capturing the attention of the general reading public, so it could be argued that Wolfe has done the field of linguistics a service. Second, I think Wolfe does a good job of demonstrating that Darwin's theory of evolution is, in some ways, less a scientific theory than a cosmogony, a sort of atheistic "creation story." That he should make such a point is particularly interesting as it is my understanding that Wolfe himself is an atheist (who presumably believes in evolution of some sort). Third, Wolfe also does a fairly good job of puncturing Noam Chomsky's overinflated reputation. When I was in graduate school pursuing a master's degree in linguistics, all of my professors were ardent Chomskyans (I suspect this was not only because of his linguistic theories but also because of his radical politics). Consequently, it was rarely (if ever) acknowledged that there might be a different approach to linguistics from Chomsky's--it was always assumed that Chomsky must be right. It was refreshing to see a less reverent attitude toward Chomsky displayed in Wolfe's book, especially given how fallible Chomsky has proven at times. Fourth, Wolfe isn't afraid to point out Chomsky's tendency to deal with criticism by engaging in in hominem attacks,, a less-than-admirable trait.
On the other hand, for me, there are some weaknesses in Wolfe's book. For one thing, at some points in the book Wolfe indulges in some rather extensive digressions that seem to veer far from his main point (his long recounting of Everett's adventures in the Amazon come to mind). Still, this is mainly just a matter of style. What I found more problematic were certain matters of substance.
First, Wolfe appears to believe that the evidence for a lack of recursion (embedding) in the Piraha language (contrary to Chomsky's claim that all languages exhibit recursion) is a "slam dunk," but I am not convinced that this is necessarily the case. I certainly don't claim any expertise with the language, but a little online research suggests to me that the question of whether Piraha has recursion has not been totally resolved. According to the Wikipedia article on the language (see here), Piraha has a sentence-final suffix (-sai) that Everett originally claimed was needed in order to embed a clause within a sentence. Later, Everett claimed he had been mistaken in his original analysis--that what appear to be embedded clauses in Piraha are actually separate sentences. However, according to Wikipedia, the German linguist Uli Sauerland reanalyzed some of Everett's data and concluded that Piraha does in fact show evidence of recursion. In short, the debate over recursion in Piraha hardly seems settled.
Second, Wolfe apparently believes that Piraha is an extremely simple language, reflecting the extremely simple culture of its speakers. For example, he writes that "Piraha [is] a language with only three vowels...and eight consonants" (p. 115). He tells us that Piraha lacks numbers ("not even 1 and 2"), and that Piraha speakers "[have] no words for colors" (p. 116) and "[speak] only in the present tense" (p. 114). Indeed, he calls Piraha "the simplest language on earth" (p. 116). The problem is that Piraha is not quite as "simple" as Wolfe supposes. In particular, despite a lack of tenses, its verb system appears rather complex. According to Wikipedia, verbs in Piraha must include suffixes that indicate the speaker's source of information--whether from personal observation, circumstantial evidence, or hearsay (incidentally, Piraha shares this feature with Kwakiutl, a Native American language spoken in British Columbia--see Geoffrey Sampson, Schools of Linguistics, p. 61). Moreover, in Piraha there are "suffixes that indicate a desire to perform an action, frustration in completing an action, and frustration in even starting an action" (Wikipedia). Verbs in Piraha also feature contrasts between completed and uncompleted actions (as is the case in Mandarin Chinese) and between reaching a goal and repeating, continuing, or starting an action. In other words, while Piraha may appear "simple" in some aspects to a nonnative speaker, in other aspects it is rather complex. This clearly illustrates the truism among linguists that there are really no "primitive languages"--to the extent that by "primitive" we mean "simple." In fact, simplicity or complexity in language is relative. For example, Piraha appears to have only one term for human relationships, similar to the English word parent. This makes Piraha seem very simple compared to English. On the other hand, in Mandarin Chinese there are five words for "uncle"--one for "the older brother of one's father," one for "the younger brother of one's father," one for "the brother of one's mother," one for "the husband of the sister of one's father," and one for "the husband of the sister of one's mother." Thus, in comparison with Chinese, English seems "simple" in terms of its kinship vocabulary.
Third, remember that Wolfe claims that language is an artifact--"the first instance in which a creature, man, had removed elements from nature...in this case, sounds...and turned them into something entirely new and man-made" (p. 163)--a view he shares with Everett. However, there is a logical problem with the idea of language as an artifact. At the most basic level, language consists of words, and words are (to a large extent) used for communication--that is, in effect, to transfer an idea from the mind of a speaker to the mind of a listener. For this to work, the speaker and listener must have a common understanding of what a word means (refers to). However, if language were a human invention, the only way speakers and listeners could have a common understanding of a word would be through agreement, which could only be achieved through the use of other words, whose meanings would also, in turn, have to be determined by agreement. In short, the theory that language is a human invention is logically implausible.
Fourth, Wolfe argues that words originated as mnemonics--devices used to jog the memory. There are a number of problems with this idea. For one thing, it's not clear to me what Wolfe means--does he mean that individuals invented words as memory devices for their own use or that they were developed collectively for the use of a human community? If the former, then Wolfe needs to explain how words became a means of communication between individuals. If the latter, then Wolfe needs to explain how these words were developed by a community, given the logical problem discussed above.
Fifth, Wolfe doesn't seem to realize that he, Darwin, Chomsky, and Everett all face a similar problem when trying to explain the origin of language. They all seem to be wedded to a naturalistic explanation of the origin of language, but any naturalistic explanation is problematic given the logical problem discussed above, as well as for other reasons (e.g., the lack of any real evidence for a Darwinian evolution of language). In a sense, Wallace was right--a "controlling intelligence" outside of nature must have played a role in the origin of language. And I would name that "controlling intelligence" as the Creator.
In summary, Wolfe's book is a very interesting discussion of the controversy over the origin of human language, but his belief that he has found the answer to the mystery of language is not well-founded. Still, I suppose we should salute him for at least giving us some food for thought.
Image of Daniel Everett from Wikipedia