After hearing positive recommendations from several friends, I finally picked up Yuval Noah Harari’s Sapiens. A Brief History of Humankind, and I have to admit that I am somewhat disappointed. I can see why it’s such a popular book: it’s certainly written in a clear, vivid language and brings complex ideas to a layman level. In short, he’s doing a great job in terms of popularizing science. It’s the “science” that he’s popularizing that I find somewhat disappointing, especially when it comes to language.
Harari makes a strong emphasis on fictive language: human ability to speak of entities that do not exist such as the Santa Claus, unicorns, God… and, as Harari points out, nations, states, and money. All these are fictions that large groups of Sapiens (that’s us!) choose to believe in and which allow us as a species to divorce ourselves from the restraints of biology (more on this below). “This ability to speak about fictions,” Harari writes (p. 24) is the most unique feature of Sapiens language”. (Just as a clarification: Harari refers to communication systems of apes, elephants, dolphins and even ants and bees as “language”, whereas most linguists reserve the term to our species’ ability, which Harari agrees is crucially different from the ways other animals communicate. Linguists sometime talk about “human language”, but since Harari uses the term “human” for all Homo species, talking about “human language” in this context would be confusing. The term “natural language”, usually used in contrast to artificial/constructed languages, is no help either since communication systems of other species are natural too. I’ll use “Sapiens language” to refer to the unique ability of our species.)
The ability to talk about fictive entities that don’t exist at all is indeed unique to Sapiens, as far as we know. (We’ll get back to the Neanderthals below.) But it is only a small part of what is truly unique about Sapiens language. Harari defines fictive entities as things “they” (Sapiens who talk about them) “have never seen, touched or smelled” (p. 24). It is not clear to me where he would draw the boundary. Elsewhere, he uses terms such as “fictions”, “social constructs” and “imagined realities” (p. 31), and most of his discussion focuses on such myths as nations, corporations and the like. But what of entities that are not present here and now yet were as real as you and me at some point, like ‘Cleopatra’ or ‘JFK’? What about other abstract entities that cannot be seen, touched or smelled, like ‘love’ or ‘climate’? Are they fictions too? In fact, if we dig slightly deeper, it turns out that even when we talk about “things that really exist, such as rivers, trees and lions” (p. 31), we often talk about abstract/fictive entities. Take, for example, Dogs bark. What does this sentence say? It certainly doesn’t talk about some real entity that is present here and now and can be seen/touched/smelled. It says something about a whole class/kind of entities called ‘dogs’. Notably, this sentence is true even if some entities that can be classified as ‘dogs’ (exceptionally) do not bark. In order to refer to concrete individuals, we can use proper names (e.g. Fido) or combine class-referring terms like dog with a special word (technical term: quantifier) that indicates which/how many “real” entities we have to pick from a fictive class: the dog, some dog, five dogs. By combining a predicate (e.g. bark) with either a kind-denoting (dogs) or individual-denoting (the dogs/Fido) terms, we impart some information (e.g. ‘individuals in the class/kind ‘dogs’ typically bark’ or ‘the individual Fido barks’). Now, our ability to do that is the most unique feature of Sapiens language. Thus, Harari writes (p. 24):
Yet the truly unique feature of our language is not its ability to transmit information about men and lions. Rather, it’s the ability to transmit information about things that do not exist at all.
I disagree: the truly unique feature of Sapiens language is the ability to transmit information, period. That’s what makes Sapiens language different from communication systems of other animal species. For example, Harari discusses the communication system (he calls it “language” but see above) of “green monkeys” (p. 22), which “use calls of various kinds to communicate”. For example, green monkeys have a call (or sign, or sound) that means (or can be “translated” into Sapiens language as) “Careful! An eagle!” and another one that means “Careful! A lion!”. Do these calls transmit information? Not exactly. These are not shared bits of information but calls to action: the first one signals to protect from danger from above and the second one — from the ground. It calls for different reactions (Harari, p. 22):
When researchers played a recording of the first call to a group of monkeys, the monkeys stopped what they were doing and looked upwards in fear. When the same group heard a recording of a second call, the lion warning, they quickly scrambled up a tree.
Such animal calls do not impart information in the sense of truth/falsehood, like Sapiens sentences. A monkey cannot say “that’s not true!”. Sapiens language on occasion uses similarly instructive rather than informative signals, such as “Fire!”. However, the bulk of what we use language for is to transmit information, not to urge our listeners to do this or that action, at least not directly. This is discussed in great detail in Derek Bickerton’s Adam’s Tongue. How Humans Made Language, How Language Made Humans, and I’ll refer the reader to Bickerton’s excellent discussion. In fact, on p. 37 Harari sums up the new Sapiens “Language” ability in three bullet points that all start with “the ability to transmit (larger quantities of) information about…”. That, in and of itself, is the core new ability that “happened in the Cognitive Revolution”: transmitting information.
The second important issue is when that ability first appeared in Sapiens. According to Harari’s “Timeline of History” (a term he also uses rather differently from the commonly accepted wisdom), “fictive language” appeared in Sapiens ca. 70,000 years ago, the event Harari calls “the Cognitive Revolution”. This date seems to be correlated with “symbolic explosion”: the sudden wealth of evidence in the archeological record for Sapiens developing symbolic abilities (such as jewelry and the like), which is more commonly dated as 50,000 years ago. Some scholars have indeed argued that language emerged at the same time as the abundance of more sophisticated (symbolic) artifacts (e.g. Klein and Edgar 2002). The slightly earlier date that Harari picks is necessary in order to peg the emergence of Sapiens language before the species spread from Africa into Eurasia and onwards. Thus, it seems that Harari assumes that Sapiens language emerged only once, before the species spread geographically, an assumption that most linguists share. (As a footnote, I’ll note that in a footnote on p. 21 Harari states: “Apparently, even at the time of the Cognitive Revolution, different Sapiens groups had different dialects”. It is not clear to me what evidence, if any, this is based on, nor how it is possible to have different dialects/languages at the time of the Cognitive Revolution, that is, when language first emerged. Not shortly thereafter, but at the time. I will leave it aside as an example of sloppy presentation.)
Going back to the date when the Cognitive Revolution happened, it is not clear to me what evidence there is that it happened 130,000 years after the emergence of the Homo Sapiens species. Existing genetic evidence suggests that (to the extent that it can be taken as evidence of language at all) the genetic basis for Sapiens language, the H. Sapiens version of the FOXP2 gene, emerged ca. 200,000 years ago. Some scholars have argued that not only did we Sapiens have the genetic underpinning for developing Sapiens language but that Sapiens language allowed us to out-compete other (human and non-human) species despite our lightly built skeleton and overall fragile physique. Harari himself suggests that our ability to use “fictive language” and to cooperate in language groups as a result of our linguistic capacity is what allowed us to out-compete (and ultimately push out of existence) our Neanderthal cousins: “In a one-on-one brawl, a Neanderthal would probably have beaten a Sapiens” (p. 34). How, then, did Sapiens survive language-less for 130,000 years with the same anatomy as their linguistically-endowed descendants in the last 70,000 years ago? And if they had some other evolutionary advantage that allowed them to survive alongside the Neanderthals and other ancient human species, as well as lions, hyenas, gorillas and other competitors, what could have been the impetus for the emergence of language ca. 70,000 years ago? Harari, disappointingly, leaves the whole issue of how Sapiens language emerged open, but it is clear that by his logic it must have been a biological event, one driven by “genes, hormones and organisms” (p. 38). Only the emergence of Sapiens language allowed us to the “bec[o]me exempt from biological laws”. If that’s the case, the emergence of Sapiens language should be correlated with some biological event, such as a mutation in a gene that produced our linguistic ability, whether directly or indirectly. This brings us back to the FOXP2 and its presumed emergence ca. 200,000 years ago, in contradiction to what Harari states.
And that brings us, inevitably, to the question of Neanderthals and their linguistic abilities, an issue that remains rather murky, to say the least. Harari assumes that the Neanderthals had some language (recall that he doesn’t reserve the term for the unique Sapiens ability). They “could share information about the whereabouts of lions, but they probably could not tell — and revise — stories about tribal spirits” (p. 34). In other words, they did not have Sapiens language but they had the ability to transmit information about “real world entities”. Whether such an intermediate stage (i.e. predicative language without “fictive” entities) ever existed, what may have prompted its emergence or its transformation into a fully Sapiens “fictive” language is not at all clear. Which is not to say that it is impossible for such intermediate stage to have existed in our Homo cousins, but at the moment such a possibility remains rather “fictive” (in Harari’s own sense). It appears that the Neanderthals had the genetic basis for Sapiens language (if so, it would need to be renamed “Homo language”); see Krause et al. (2007). Yet, these findings are open to various interpretations. Harari could be right if FOXP2 could be shown to be responsible for predicative but not necessarily “fictive” language, yet I am inclined to agree with Bickerton that predicative language (i.e. language that can be used to transmit information) is impossible without the type of signs that refer to classes/kinds, ergo to abstract entities.
(To be continued, hopefully…)
Klein, R.G. and Edgar, B. (2002) The Dawn of Human Culture, Wiley, New York.
Krause, Johannes et al. (2007) The Derived FOXP2 Variant of Modern Humans Was Shared with Neandertals. Current Biology 17: 1908–1912