A five-year-old recently asked me, "Who invented words?" It is sobering how impotent we linguists are in the face of that question. Because language is ever in a state of transformation, today's languages cannot tell us what the words were like in the first language humans spoke, likely in Africa between 150,000 and 200,000 years ago. And because those humans lacked writing and we humans lack time machines, we cannot know how the first words were created. Was a cat called a "meow"? Possibly. But that just barely scratches the surface. What decided the word for arm? Or maybe, or or? Like the origin of the universe, the death of Richard II, and Orson Welles' cut of The Magnificent Ambersons, the answer to that five-year-old's question will lie forever tantalizingly beyond our reach.

The fashion nowadays is to insist that animals use words as well, such that human words are less of a miracle than we might think. However, between human language and even the most sophisticated animal communication systems lies an awesome gulf. Be it primates mastering some sign language, or bees dancing in the hive to communicate where food is to be found, parrots' "speaking," and assorted animals' alarm calls, the animal has yet to be found whose language serves beyond the primal functions of avoiding danger, showing aggression, seeking mates, and seeking food. Certain signs for a small set of physically urgent goals, always within the immediate context—human language is vastly more than this. The ape using sign language, for example, rarely initiates conversation just for its own sake, and just as rarely comments upon something to share an impression.

The commenting aspect is key. Human language is not merely about labelling, but combining a subject with an observation about it—a predicate—to yield a statement. The statement, as opposed to the one-word yawp, lends itself to another cognitive leap: reference to the past, the future, or the hypothetical rather than the here and now. This renders what seem the simplest of statements into something unknown in any other animal. Who'd even know? one might ask in any number of situations. The very activity of posing a question, as opposed to making a statement, goes beyond anything known in the most charmingly sophisticated of apes or parrots. Referring to a mental state such as knowing and being understood is, again, a miracle compared to vervet monkeys having a call that means "jump up in a tree to avoid a leopard" and another one that means "run under a bush to avoid the eagle." Then the even in Who'd even know? adds the nuance of specifying that the proposed action is, first, naughty, and second, would nevertheless not be detected and therefore would occasion no punishment. To be able to effortlessly produce and process language of this kind is unique to humans, no matter how much you surmise that your cat "understands" you. One wishes to know how it got started.

Many humanists feel that the very question is misposed. They prefer a scenario under which humans developed higher intelligence—a process easy to imagine as a typical instantiation of Darwinian natural selection, they argue—and that language merely emerged as a system of vocalizing thought processes. Here, the question as to how the first word labels were created remains—how do you come up with a sequence of sounds that means run?—but how humans strung the words together into thoughts becomes less of a head-scratcher. The thoughts were already there; the words just tracked along with them.

In The Language Animal, Charles Taylor argues against this reconstruction, which he traces to foundational work by philosophers Thomas Hobbes, John Locke, and Étienne Bonnot de Condillac, who saw language as simply expression of thought that already existed. Taylor elaborates upon a contrary hypothesis, traceable to Wilhelm von Humboldt, Johann Herder, and Johann Hamann, that it is language that makes sophisticated human thought even possible. In the terms Taylor uses, language was constitutive of thought rather than merely descriptive of thought already thriving.

Given that we can think without words, the idea that language creates thought rather than merely reflecting it is inherently stimulating in its counterintuitiveness. Taylor's book, however, will only partially satisfy most readers interested in how language makes us different from animals. Taylor's prose is careful, and even accessible. Yet sprightly is not one of the adjectives one might choose to describe his writing style, nor is brisk—the book could easily lose a good 75 pages. Taylor seems most accustomed to an audience of academic philosophers; for those outside that group, treatments such as Derek Bickerton's Adam's Tongue and Thom Scott-Phillips' Speaking Our Minds are more engaging explorations of why humans are the only animals with language.

However, there are precious insights in Taylor's text. For example, on the very substrate of language in contrast to animal communication, Taylor zeroes in on the difference between recognizing something and having a word for it. This involves more of a leap than we readily suppose. An animal might recognize a triangle as the shape of a hole to jump through, or point to, to get a treat, such that the triangle is the conduit to a concrete reward. However, only humans have a label for the triangle whose usage, via utterance, yields the reward simply of corresponding to the thing we know as a triangle. That is, we humans can regard a triangle and see not only a way to get something done but also the embodiment of something to which there corresponds a name "triangle."

This means that the biblical tale of Adam naming the animals already makes a jump, in the assumption that he would seek to come up with a "name" for anything. A strain in the philosophy of language, espoused in passing by Taylor and classically by Rousseau, even claims that without language one can only conceive of individual entities, rather than perceiving categories, such that everything is, as it were, a proper noun. Bickerton, for example, imagines a prelinguistic hominid encountering every pink, fragrant flower as a new instance of such, incapable of grouping all those traits as "roses" without the ability to apply to them a word with that meaning. Now, if philosophers have ever provided a conclusive argument for this reconstruction I have yet to encounter it. Allowing that modern human thought represents a phase shift beyond any animal language doesn't require assuming that non-linguistic cognition is incapable of brute generalizations. Yet, Bickerton's point is that, at a minimum, a word makes the neurons perceiving each aspect of an object fire together more powerfully than encountering the object without using a word for it, entrenching it in a newly powerful way as a unitary concept. A nice analogy is contemporary American slang's idea of a practice or assumption having become "a thing," such as looking at one's smartphone while taking care of children: a fusion of two activities once considered unrelated and perhaps antithetical.

Less controversial would seem another of Taylor's central points, that only with language are evaluative emotions such as admiration, pride, remorse, and respect possible. To wit, for all of our tendency to anthropomorphize animals, and how obvious it is that they feel more primal emotions such as joy, sadness, and anger, the notion of the non-linguistic creature swelling with pride, or regretting something as having grievously contravened a code of ethics, is nonsensical. This is because this very kind of code of ethics applies, by definition, to a community, and the nature of ethics is such that if all members of the community are to live by such a code, language is necessary to convey and maintain it.

Taylor focuses on the example of respect. One could without language impart basic rituals of deference such as sharing food, not showing aggression, physical postures of submission. However, only with language could we impart to a child the general concept that all of these actions are intended to indicate, as in respect due to certain members of the community because of benefits they brought to the community in the past, because those who are older have knowledge useful to the younger, or even because of a more ritual idea that one's parents, or elders, are to be treated with this brand of esteem. The baboon can be taught to defer to one higher in a brute power hierarchy, but not that certain individuals, outside the bounds of who can physically abuse who with impunity, merit esteem. Remorse requires similarly broad cognitive horizons, impossible to communicate, police, or preserve without language in the human sense.

It is here that the flaw becomes clear in the objection that we can think without words. Taylor specifies that this is an epiphenomenon of a basic linguistic ability that must, in the logical sense, have developed as a communal tool, not as one for solitary reflection. Language emerged as speech, not writing, and the heart of speech is conversation, as in joint attention. Language is fundamentally conversational, not monologic. Of course, equipped with the level of thought language allows, we now can turn language off and think without it. However, the temptation to therefore imagine cats having the same richness of thought that humans do but simply not vocalizing it is as illogical as it is charming. Surely, the exigencies of natural selection would lead at least a few creatures thinking on this level to leverage their capabilities of forethought and reflection into cooperation, leading to elaborate invention and very complex societies, just as humans did. Okay, cats are solitary and "self-satisfied"—but that's just them. Why not a single more social and extroverted primate, or breed of parrot or canine? The hypothesis beckons that language makes the difference here, rather than simply piggybacking on thought.

Taylor builds on his basic formulation with further ways that language renders us cognitively unique. Hobbes' The Leviathan opens with an almost curiously lengthy disquisition about the exact meaning of assorted terms, which charity dictates excusing any undergraduate class from in a philosophy survey. This prologue makes sense only within Hobbes' conviction that metaphor is antithetical to effective communication: he wanted to keep words within the cages of their core meanings. Taylor objects that metaphor is actually part of the heart of what makes human language special, ever shedding light on new facets of the basic referent. To bat one out of the park, to go all the way, to go the whole nine yards, to go for the gold—all of these expressions expand the mind beyond the confines of the mere triumph or complete, and sometimes even contract the mind as well (such as in these expressions' implication that achievement is inherently masculine, competitive, and/or rapid). Whatever communication is going on among baboons, the closest thing to a metaphor would seem to be mimed bites and swats of the arm. Also, language allows us not only to feel abstract emotions such as respect and remorse, but to discuss them, elaborate upon them, the highest rendition of this being narrative and other arts.

A prime takeaway from Taylor is that a model of human language as a mere logical system along the lines of a computer program hopelessly undershoots and distorts what language actually is. Statements, questions, and commands—which lend themselves most readily to symbolic representations of language as a variation on mathematics—are only the surface. Hence how resistant actual language usage is to calls to use language as if it were, indeed, a sequence of logical formulas under which two negatives make a positive and a word like literally cannot "coherently" be used to mean "figuratively" (I was literally dying of laughter). The language never "goes to the dogs" as grammar hounds fear, because language is so much more than classically diagrammable sentences like The boy bounced a ball.

Rather, the miracle of language is, as I write this, that a week ago my daughter, one year and a few months old, pointed at the door on a night when her mother was working late, and I said "No Mommy!" and she started crying. The sheer fact that she understood that I was communicating information to her—naming a thing (Mommy) and saying something about it (as in indicating that it was not imminent)—with this information neither directing her to sustenance, informing her of bodily danger, nor indicating aggression, made this a uniquely human moment. That morning my four-year-old had trotted up to a friend and asked, "What are you up to?" Seemingly so uninteresting, this one phrase involved using words in distinctly non-literal meaning (up to what?) and was also used without expectation of a literal answer, in that the person in question was just standing waiting for the school bus like everybody else, not "up to" anything. What are you up to? was a question only in form; in essence it was a greeting, as in a ritual acknowledgment of the other person's presence, indicating a basic respect.

Taylor's The Language Animal, although it needed more stringent editing, is about a kind of animal in all of us—an animal whose language makes us smarter than we are often taught to suppose.

John H. McWhorter teaches linguistics, English, American studies, comparative literature, philosophy, and music history at Columbia University. His book The Language Hoax: Why the World Looks the Same in Any Language (Oxford Univ. Press) was just reissued in paperback.

Copyright © 2016 by the author or Christianity Today/Books & Culture magazine.
Click here for reprint information on Books & Culture.