As a subscriber to Scientific American since 1954, “Thinking without Words” on pp. 86-88 of the March 2025 issue struck me as a useful illustration of what GA’s perspective can add to the standard discourse of the human sciences among anthropologists and linguists, by allowing them to rethink their foundations.
No one can contest the value of the scientific method for empirical research, nor is it useful to replay the sterile controversies between Darwinists and believers in Creationism. But the problem the latter are attempting to solve by positing a Creator, however inappropriately in a scientific context, is not thereby proved spurious in itself. Whether or not we need to take seriously a view of the evolution of life as determined by a transcendental will, the great error of positive anthropology is to take for granted that the techniques of empirical natural science in themselves suffice to explain what makes humanity unique. The study of the human is not a simple extension of physics and chemistry, nor even of biology. And the proof is that empirical anthropology can teach us many things about the human, but it has no effective theory to explain its emergence from the non-human, because it lacks the means to define just wherein the difference lies.
In this regard, I have long appreciated Roy Rappoport’s dictum in Ritual and Religion in the Making of Humanity (Cambridge, 1999) that language and religion were “coeval,” emerged at the same moment. But Rappoport poses this primary truth as a given, and never goes into detail in hypothesizing a path from the non-human to the human. In contrast, GA offers an originary hypothesis of the human’s emergence. I hope to show in this Chronicle how it helps clarify the apparent paradoxes of this S.A. article.
The article is, after a brief introduction, a conversation-interview between Gary Stix, a senior editor of the journal, and Evelina Fedorenko, “a neuroscientist who studies language at the McGovern Institute for Brain Research” at MIT. The point of her research, as the author’s first question brings out, is to discover “whether language and thought are separate entities,” a question that she answers in the affirmative, while pointing out that animals too make use of “thinking,” along with people with aphasia who are unable to speak. For indeed the parts of the brain which deal with language on the one hand and thinking on the other are not the same.
There are surely no false statements in Dr Fedorenko’s remarks, including her critique in passing of the Chomskian tradition, “where the dogma has always been that we use language for thinking: to think is why language evolved in our species.” Yet as a way of understanding human language and why it is different from animal modes of communication, Stix’ point that “No words are needed for animals to perform all kinds of problem-solving challenges that demonstrate high-level cognition. Chimpanzees can outplay humans in a strategy game, and New Caledonian Crows make their own tools that enable them to capture prey,” even if arguably true, obscures the very point of what a useful critique of Chomsky’s position would be.
Yes, animals think, but it is useful only from the perspective of a physiologist to concern oneself with the “mechanisms” of thought without considering the very principles by which language operates as being in the first place those of a medium of communication. Since Stix concedes in the following paragraph that “humans perform tasks at a level of sophistication not seen in chimps”—as though this difference of level were somehow comparable to that between a child and an adult rather than involving a wholly different degree of richness and subtlety—he then wonders if “we require words and syntax as scaffolding to construct the things we think about? Or do the brain’s cognitive regions devise fully baked thoughts that we then convey using words as a medium of communication?”
The notion of a “fully baked thought” as expressible without a system of signs begs the question of how this thought could be formulated otherwise even inside the brain. What Stix calls the “high-level cognition” of animals is arguably many degrees less complex than what we would consider ordinary human thought, such as the defense of a thesis in a typical newspaper article; what we know of as animal thinking can typically be presented in the form of a simple diagram.
It is surely useful to point out that animals can indeed think, in the sense of drawing inferences from experience: that the latter is not wholly dependent on language as a shared sign-system. But talking about language and thought in this way distracts us from the real task of anthropology, which is to explain what is particular to humans, who are after all the only animals who have the means to express, and share, such concern about themselves.
It is not useless to know that the brain carries out operations of reasoning in a separate place from its store of vocabulary—just as a computer’s store of data is separate from the modules that process that data. But talking about language without relating it to the fundamental activities of the human beings that it does so much to define implicitly effaces the fundamental difference between humans and other living creatures that, as we should always remember, is the core of anthropology.
In answer to Stix’s question: “So, what is the role of language, if not for thinking?” his interlocutor replies:
I’m . . . sharing some knowledge that I have that you may have only had a partial version of—and once I transmit it to you through language, you can update your knowledge and have that in your mind as well. It’s basically like a shortcut for telepathy. . . . [W]e can use this tool called language, which is a flexible way to communicate our inner states, to transmit information to each other.
Now I don’t want to pretend that Dr. Fedorenko considers this description as an explanation of how language emerged, let alone of why it emerged only among humans. But the comparison with telepathy reveals an indifference to the collective nature of language, let alone to its possible origin. How indeed could we have begun to “telepathize” the words of language to each other? And how, and why, has spoken language given rise to writing and electronic communication that allow humans, also unlike all other creatures, to preserve information in transmissible forms?
What strikes me is not that Dr. Fedorenko does not attempt to answer these questions, but that, at least in the context of this dialogue, she sees them as irrelevant to the matter of how language is connected to thinking. What indeed is it that makes human thinking so much more powerful than that of the chimpanzees and Caledonian Crows that Stix alleges as examples of “thinking without words?” Could it be that a thought, as opposed to “thinking,” can only be conceived as expressed in a system of signs? Ask a logician to give an example of an elementary thought, and it will be in the form of a simple declarative sentence, “all a is b,” or “elephants are bigger than mice.” Our chimpanzee can certainly observe, and presumably “think,” the latter idea; we can say that he “knows” that elephants are larger than mice. What he lacks is a simple means of communicating it to another chimpanzee. But what do we mean by thinking in this case?
Stix’ example of strategy games is a good place to start. When two chess players are playing each other, they are clearly thinking, but it is also clear that expressing their reasoning in language is irrelevant to their action, unless, for example, one is teaching the other and consequently explaining his moves. It is easy enough to see in such cases that language, so useful in communicating our thoughts “telepathically,” is not necessarily required to think them; and this is generally true in the practical matters of daily life. If I decide to put on a raincoat because it looks like rain, I certainly don’t arrive at this conclusion by explaining my decision in words to myself; I draw the inference without expressing it, and this is true for just about all the practical decisions one makes throughout the day. The only time you need to use language to figure something out is, as I am doing now, developing a line of reasoning meant to convince an interlocutor, whether or not actualized in a real Other, of a certain conclusion. Whence the connection between mathematics and logic in the simple and unambiguous architecture of the statements they allow, so that the statement to be demonstrated becomes a theorem to be proved on the basis of a clear set of premises.
And it is precisely when this reasoning process is most rigorous, when “thought” is not just making statements, but deriving true statements from prior ones, that we are farthest from the normal use of language, and that we tend to invent systems of notation to simplify the elements of reasoning by avoiding the use of words. To express a mathematical formula in “language” in the normal sense would have no valid purpose, unless one were dictating the formula to another who would be expected to translate the words back into the symbols that the first party would presumably have been “translating” into everyday speech. But to mention such examples is to demonstrate the weakness of the “telepathy” analogy; the mathematical formula is only conceivable in technical notation, that is, on paper or a blackboard, real or imagined, not as an “idea” in someone’s brain.
And given that Fedorenko no longer accepts Chomsky’s dictum that the purpose of language is thinking, she must assume that its purpose is communication, the “telepathy” that lets us share the contents of our minds—that lets us formulate these contents in communicable signs. But how does this explain why such a system evolved, or how humans, unlike other creatures, came to create it? One would not expect to find in this brief interview a theory of language emergence and evolution, but one might have hoped nonetheless for some acknowledgement that language emerged as a new mode of communication because, as GA insists, it had become necessary.
GA hypothesizes that we needed the “telepathic” mediation of language to separate us from each other’s rivalrous desires. As our hominid ancestors became more intelligent, they found it increasingly difficult to “follow their instincts” in dividing up large food items such as the bodies of animals scavenged or hunted, as had previously been accomplished by a serial distribution system, where the Alpha animal took the first piece and passed the rest of the carcass to the Beta, etc. Challenges in this system took place only one-on-one; if a lower-ranked hominid was unhappy with his rank, he could challenge its holder. Chimps and other apes still use this system on occasion; but it is easy enough to understand that such a system would break down with the growth among the hominids of the sense of rivalry brought about by what René Girard called mimetic desire—something we all know well from our experience when a classmate or colleague gains a thing of value that we feel that we equally or more deserved.
Language, which allows us to share the idea of the desired object by sharing the word designating that idea—the signifier—was the peaceful human solution to the far more difficult problem of sharing the referent of the signifier, which, being a worldly object, could not be multiplied. In the Gospels, one of which (John) begins by telling us “In the beginning was the Word [logos],” Jesus demonstrates his role of Savior by multiplying a few loaves and fishes to feed a crowd—the only miracle (aside from the Resurrection) mentioned in all four Gospels. This figure of divine providence renders real what had already been made possible through the suspension of conflict by means of language—as the course of human history has so plentifully realized.
Language began as a solution to the problem of sharing desirable objects among a group of creatures, who thereupon became a human community. The solution was of necessity a compromise. Jesus was not usually around to multiply the loaves and fishes. But the message conveyed by this miracle is that if the humans truly form a community that puts the welfare of all before that of its individual members, they will devise better ways to increase production of loaves and fishes than if they fought among themselves for the ones they already had.
This understanding of the originary anthropological function of language, of its emergence as a necessary mediation between humanity and its desires, allows us to understand the sense of freedom that Sartre in particular saw as the defining quality of the human mind as opposed to all other objects in the cosmos. The possibility of sharing words/signs when we cannot share things is what allows us to think about them, and subsequently to communicate our thoughts, whether or not the operation of thinking itself is wholly congruent with the sentences we use to communicate its results.
To sum up, we don’t need language to think the simple thoughts of everyday, but it is the thoughts created and preserved in language that have allowed us to accumulate the knowledge of the world by means of which our species, so much more than any other, has been able to shape it to our ends. Speech, and then writing, have permitted the vast accumulation of 44×1021 bytes of information that Google estimates as the current contents of the Internet.
Language, in a word, is in the first place not a means for thinking but for sharing the signs of our desires. It could only have arisen in the context of the need for such sharing, so that the sharing of the sign could subsequently allow us the freedom to conceive, and to bring about, the sharing of the thing itself. For what is most problematic about the human is not our ability to think, which the existence of language and communication so greatly facilitates—since words, signs, are so to speak designed to be easily reproducible, and the history of human communication has been marked by the constant invention of more efficient and effective means of their transmission—but the fact that each improvement in communication adds to the intensification of potential rivalries, not for the signs that are transmitted, but for the objects they designate.
And this brings me back to Rappoport’s dictum about the coevality of language and religion. The gesture that in GA’s originary hypothesis becomes the first sign is simply one of pointing, but the hypothesized origin of the pointing gesture, which is not practiced even by chimps, is an aborted act of appropriation, the result of reaching for the object but aborting one’s attempt in order to avoid conflict with the others reaching for it at the same time—like, in my favorite example, two party guests both reaching for the last canapé. And what we call the sacred, the basis of religion, is in its origin nothing more than this recognition of the priority of the community to the individual—the basis of most of the Ten Commandments and of Hillel’s Golden Rule, and the core of what we call our conscience.
To sum up, thought has existed in some sense from the first living creatures, in the sense of permitting choices necessary to their existence, at first, by purely mechanical means, and with the evolution of nervous systems, through specialized mechanisms by which their environment could be evaluated. But it is only with language and the communication of thought from one human to another that we began to understand ourselves as a community, enduring through the generations and accumulating the “thoughts,” the knowledge that has made us, for better and worse, and however imperfectly, masters of our world.