The Language Hoax: Why the World Looks the Same in Any Language
by John H. McWhorter.
Oxford University Press, 2016.
Hardcover, 208 pages, $20.

Reviewed by Gene Callahan

John H. McWhorter is a linguist at Columbia University, and a fascinating and sometimes controversial figure. For instance, although he describes himself as “a cranky liberal Democrat,” he also wrote a book, Losing the Race, which claims that African American culture is partly to blame for the relatively poor performance of African Americans in school and work life. That book made him popular with conservatives, and led the Manhattan Institute to name him a contributing editor to its publication City Journal. On the other hand, he is a vigorous defender of “Black English” as a perfectly valid form of English, just as grammatical as “standard” American English. (That is a position that traditionalist conservatives ought to embrace, first of all because it is correct, and secondly because it recognizes the priority of concrete ways of life over abstract rules. But alas, too many conservatives have mistaken the position of nineteenth-century grammar rationalists as being conservative.)

In addition to these positions that have brought him into the public eye, McWhorter is a serious linguist, specializing in language change and Creole languages, and conversant in Spanish, French, German, Italian, Portuguese, Russian, Dutch, Swedish, and Hebrew. In The Language Hoax, McWhorter takes on an important and contentious view that has been kicking around linguistics and the social sciences since at least the eighteenth century, a theory which has sometimes been called the “Sapir-Whorf hypothesis” (itself a controversial term). The theory is popular enough that it played a central role in the major motion picture Arrival, where it was explicitly mentioned by name. The essence of the theory is that the worldview of us humans is … well, here things get murky: determined by? strongly influenced by? somewhat influenced by? … the language we speak.

Evaluating the case for and the case against “Whorfianism” (the term McWhorter prefers to “Sapir-Whorf hypothesis”) is confounded by the fact that various proponents and opponents of Whorfianism have used their support for or opposition to it in all sorts of polemical and politically charged ways. For instance, in the early nineteenth century, Wilhelm von Humboldt argued that “inflectional” languages—those with lots of prefixes and suffixes that modify a word’s grammatical role—are superior to other languages as aids in understanding the world, and that is why their speakers have come to dominate the globe. But that idea cannot survive even moderate scrutiny: English is far less inflected than German or Italian, so why is so much more of the world Anglophone rather than Germanophone or Italophone? And since some of the worlds most inflected languages were spoken by Native Americans, why didn’t the groups that spoke those languages wind up dominating the globe?

Once I got a sense of where McWhorter was going with this work, I began to suspect that the book’s title was the creation of the publisher’s marketing department rather than the author’s choice: nowhere between the front and back covers does McWhorter actually suggest that anyone is deliberately tricking people into thinking Whorfianism is true. From the evidence he brings forward, the worst that could be said of enthusiastic Whorfians is that they are guilty of wishful thinking: they embrace Whorfianism because they like the idea that the languages of various “primitive” people actually demonstrate that they have insights into reality that “Westerners” lack. A famous example is the notion that Eskimos have “one hundred words for snow.” Despite this legend apparently being false, many people take a fancy to it because they think it shows how “in tune with nature” non-Westerners can be.

Among the pieces of counter-evidence to this thesis that McWhorter offers is the example of the Amazonian Tuyuca, who use “evidential markers” extensively in their language. Their grammar requires speakers to make their evidence for some claim explicit. For instance, if the claim they are making is based on their seeing the event in question—in English we might say, “I see it is raining”—in Tuyuca the suffix í would be attached to some word in that sentence. (I haven’t learned enough Tuyuca from McWhorter’s description to be sure of which word.) If, instead, the assertion is based on hearing (“I hear the rain falling”) a different suffix is used. If they are just surmising it is raining (“From the wet streets, I’d guess it is raining”), there is another suffix, and if it is just the word of others (“They tell me it is raining”), there is yet another. Whorfians tend to like stories like this: “See how these people you might think of as ‘savages’ are actually scrupulous evalutors of evidence!”

However, McWhorter argues, this kind of argument relies on cherry-picking one’s examples. He notes that Bulgarian and Turkish have evidential markers, while Greek and Persian do not. Are Bulgarians and Turks really more sensitive to the backing for claims than are Greeks and Persians? McWhorter presents a wealth of evidence suggesting it is unlikely that we will find many macro-level explanations like that for the enormous diversity of phonetic and grammatical features in human language.

But McWhorter errs in assigning another cause of this profligate variety: “In fact, there is a coherent explanation.… That explanation is, quite simply, chance.” But “Chance” is not an explanation for anything: chance is the word we use for happenings we can’t explain. In cases such as the presence of evidential markers in the Tuyuca language, McWhorter is fighting against the Whorfian view that there is typically a cultural/environmental explanation for the features of a language. As mentioned above, he makes a strong case that things are not so.

Yet there is some explanation for how a language’s features arose, even if it not at the level the Whorfians would like it to be: in the case of evidential markers in Tuyuca, perhaps (and I am merely offering a purely imaginary example here!) a long-lived chief was an especially suspicious fellow, and some little tic led him to put small sounds of doubt at the end of otherwise factual statements when he questioned their factualness. Now, this would be a fact of history that is long lost, and that we probably have no hope of ever recovering. And to that extent, McWhorter would be accurate if he had said, “From a linguistic point-of-view, it might as well have been the flip of a coin that determined how evidential markers arose in Tuyuca.” But, of course, coin flips themselves are not chance events! Or, rather, saying they are a chance event is an expression of our ignorance of, and, indeed, despair at ever detecting, what really caused a coin to come up heads or tails on some particular flip.

Consider the opening toss before a football game, which determines who kicks off. Such tosses are governed by “chance” from the point of view of the players and referees involved. But whether the coin comes up heads or tails is, in fact, determined by the force and angle of the ref’s flip, the air resistance offered to the coin, the gusts of wind that impact it, the participants’ movements, and so on. If anyone involved could exactly determine the influence of all of those factors and could rapidly calculate them, they could say for sure whether the coin would land on heads or tails. It is the participants’ inability to do such calculations that makes the coin flip, from their point of view, chance, and a fair way to decide who kicks off. But “chance” is certainly not an explanation for why the coin landed as it did: it is an admission that the participants can’t explain it.

McWhorter goes on to what he sees as the chance nature of these language changes, “There is a comfort in this reality.” Well, as Rocky Rococo would say, “Maybe yes, maybe no.” He brings similar considerations into play a bit later, when he writes, “Whorfianism, here, seems dangerous” (in making Chinese speakers seem less clear-headed than English ones). But if Whorfianism is true, then even if it is “dangerous” in the way McWhorter suggests, well … what? We should suppress it? I thought scientists were proud iconoclasts, who followed their theories however many idols were crushed along the way.

To be fair to McWhorter, he is attempting to discourage those who are inclined to believe in Whorfianism because of what they think are its sociopolitical entailments. And McWhorter is attempting to alert them to the fact that those entailments don’t always entail what Whorfians want them to. Fair enough, but it may have been a more accurate rejoinder simply to tell the Whorfians that it is invalid to judge scientific theories based upon our opinion of their social implications.

McWhorter also attacks the idea that a language, like English, shapes one’s worldview. His arguments include: “if English conditions a worldview, then it that has to be a worldview that encompasses the frames of reference of that Jersey City boy, Mary Tyler Moore, Margaret Cho, William Jennings Bryan, and Sting … Clearly, that is a worldview so general as to be equivalent, essentially, to simply being human.” McWhorter would appear here to have conflated shaping a worldview with completely determining a worldview, and he ridicules Whorfian claims that language might do the former on the basis of the fact that it certainly does not do the latter.

However, things are not that simple, for towards the end of this book, McWhorter insists that it is a misinterpretation of his stance to think one can defease it by noting “that no one has claimed that language absolutely determines thought.” And he admits that “no Whorfians make such simplistic claims.” But once McWhorter admits those points, then what are we to make of his list of English speakers? If speaking English, like being Catholic (or being Marxist, or being a jazz musician, or being a racial or religious minority in one’s society, or being gay), is only supposed to be (per the Whorfians or their equivalents in other fields) a factor making one “significantly more ‘likely’ to think in certain ways,” then the truth of Whorfianism would seem to turn on what we deem to call “significantly.”

McWhorter says this more restrained Whorfian claim is “highly fraught,” but that is not what this work has demonstrated: instead, he has shown that more extreme Whorfian claims about the significance of language’s impact on thought don’t bear close scrutiny. (And this reviewer wonders why McWhorter puts “likely” in scare quotes: does he believe the notion of likelihood is also “fraught”?) Since McWhorter admits that at least some moderate Whorfian claims have been borne out in rigorous testing, it would seem the sensible conclusion would be, “Extreme Whorfianism is false; less extreme, particular Whorfian claims may or may not be true: we’d have to test them out to see.” But that’s a claim no Whorfian would dispute.

Although McWhorter has taught me much about linguistics (I have gone through several of his digital courses) and, as a fan of his work in general, I wanted to see what he had to say about Safir-Whorf, nevertheless, I fear he has muddled a number of propositions together in his discussion of Sapir-Whorf, or Whorfianism, a muddle that becomes apparent on the last page of this book, where he asserts:

  1. “Rejecting” Sapir-Whorf is “preferable” to accepting it, for various political/ethical reasons.
  2. “Science does not support” Sapir-Whorf as it is popularly interpreted.
  3. Sapir-Whorf “just isn’t necessary,” since “languages are fascinating in their own right.”
  4. Rejecting Sapir-Whorf is “more truly progressive” than is accepting it.

But points 1, 3, and 4 are entirely irrelevant to deciding if Whorfianism is scientifically true. At most, they might bear on why people might embrace Sapir-Whorf despite a lack of evidence. And, as McWhorter accurately notes, the pop embrace of Whorfianism is more enthusiastic than the evidence warrants … but how much more enthusiastic than warranted is surely a matter of rigorously testing one claim after another, rather than asserting how “more truly progressive” is rejecting Sapir-Whorf.

In short, while this work contains a wealth of interesting linguistic findings, and certainly serves to debunk the more extreme Whorfian claims that we sometimes see in the popular press, it would have been a more coherent work had it clearly separated its scientific claims, its psychological analysis of the motives for strong Whorfian positions, and its political advocacy of “why the world looks the same in any language.” Instead, since these issues are mingled together throughout this book, it leaves the reader wondering to what extent McWhorter has allowed his position on one or more of these questions to influence his stance on the others.


Gene Callahan is a Lecturer in Computer Science and Economics at St. Joseph’s College and a Research Fellow at the Collingwood and British Idealism Centre at Cardiff University, Wales. He is the author of Economics for Real People and Oakeshott on Rome and America.