How to Misunderstand History: A review of Yuval Harari's Sapiens, Part 1
I have recently read Yuval Harari's Sapiens, a popular history of humanity. It's a tremendously broad look at human history through a few different periods- the stone ages, from the early evolution of Homo sapiens, all the way through to the development of religion, justice, imperialism, science, money and capitalism. The scope of this work is enormous, but its low-resolution view belies some serious flaws.
In these essays I will start to describe some of these flaws. I will not just minor issues through Harari's prose but try to describe some of the deeply flawed ways he has of describing history. Harari's narratives may seem erudite and convincing, but I contend this is superficial and his work is more an exercise in rhetoric, that leaves the reader less informed than they did before. In this sense Sapiens offers a net loss to our understanding of the world, even if read in the most apologetic way. This is Part 1, which covers Chapter 2 of his work.
Harari's work starts off with a description of Homo sapiens' Cognitive Revolution, an apparent leap in our species abilities that began around 70,000 years ago and ended 30,000 years ago. Note in the previous sentence I use the word apparent. My use of this word is important considering my problems with Harari's history and is to be taken literally- looking at history we only see things that are apparent. We never see things as they are. From appearances we can make interpretations, but to blend the two, and read appearances as the true nature of things is an easy but catastrophic flaw.
Chapter 2 of Sapiens is titled The Tree of Knowledge (a reference to Original Sin, we might assume) considers the Cognitive Revolution which, simply put, is a period when anatomically modern humans started showing modern human behaviours. Evidence for the changes in human behaviour comes in many different forms. Humans before and after this divide, we infer, have some fundamental differences. Complex tools and blades or funerary burials are physical archaeological finds, whereas cooperation beyond kin, language or the ability to do advanced planning must be inferred. Harari does little to describe the fundamental differences between these primary and secondary forms of evidence.
Let’s take a few pages on the development of human language (p. 22-28) and what makes it unique and so revolutionary. Harari makes a number of statements and inferences, somewhat muddled together into a narrative. What makes our language unlike other animals is that: “We can connect an infinite number of sounds and signs to produce an infinite number of sentences, each with a distinct meaning.” Secondly: “...language evolved as a means of sharing information about the world. But the most important information that needed to be conveyed was about humans, not about lions or bison.”
There are a couple of problems with the way this is argued. Harari's first description of human language is a quantitative comparison where we simply add skills together- we can simply make more noises, so can make more arrangements of those noises which allows us to convey more information. As we know now, the information we can convey is as much constrained by each language as it is facilitated. When you define something additively you also run afoul of the Sorites paradox- when you add grains of sand to each other, at what point do the grains stop being just some grains and start to be a pile? Here, how many noises need we make for animal communications to become animal language. Or even- how many particular characteristics of a species' communication are required for it to be a language.
The latter description of human language (“sharing information about the world”) is a qualitative difference and a more fundamental one. Harari uses this aspect to describe a theory of human language that it evolved to a point it allowed us to describe what other humans are doing or feeling (he calls it the Gossip Theory)- apparently a feature missing in other animal communications. Harari jumps here: “Yet the truly unique feature of our language is. the ability to transmit information about things that do not exist at all” (p. 27). And later: “This ability to speak about fictions is the most unique feature of Sapiens language” (p. 27). Well, which is it? It's maybe best to take a step back here, Harari has given a series of somewhat stronger arguments for the uniqueness of human language. The latter arguments, that only humans of all the animals on Earth can tell fictions I find somewhat agreeable, but the narrative consisting of relatively weaker arguments, forming the foundation of this part of the chapter is less convincing, and goes without critical examination. This is a common feature of Harari's writing, and the relevance of these pages comes up again later in this chapter.
We also see problems in his theory as to the cause of this linguistic development. Harari is not a good ætiologist, as we see here: “What caused it? We're not sure. The most commonly believed theory argues that accidental genetic mutations changed the inner wiring of the brains of Sapiens, enabling them to think in unprecedented ways and to communicate using an altogether new type of language”. (p. 23) Here he has stated uncertainty, but then immediately follows it up with a dogmatically certain description of how it occurred, leaning on “most commonly believed” as a kind of appeal either to expertise or popularity. If it is that “commonly believed”, surely, he could drop in a quick reference here? This rather grotesque overstatement, with a veneer of uncertainty is a common yet easy to miss feature of Harari's book. It's an effective bit of rhetorical trickery.
Harari's description of The Cognitive Revolution also gives us an implied boundary between pre- and post- revolution humans. This 'boundary' is frankly lazy scholarship- a cursory glance at people who work on this shows a wide range of views on the nature of the revolution- was it quick, or gradual? Historical boundaries are never boundaries, especially in the development of new and highly complex systems like the beginning of human culture and society. This 'revolution' shares this in common with an area I do know a lot about- the revolutionary Cambrian Explosion of animal life in the oceans, around 550 million years ago. Systems like these can undergo a series of additions that create qualitative changes to a whole system. To give an example, prior to the Cambrian Period, new forms of musculature allowed for animals to start burrowing in the sea floor, changing the properties of the sea and seabed permanently. This in turn created niches for new animals and ultimately changes to seawater chemistry on a much larger scale. This is an example of emergence- a wholly unpredictable system-wide change that comes from apparently small changes which in turn effect a whole new set of interactions. Harari does not appreciate this form of dynamic, rather focussing on lazy assertions such as: “... beginning about 70,000 years ago, Homo sapiens started doing very special things.” He does something again later on page 36: “In 1789 the French population switched almost overnight from believing in the divine right of kings to believing in the myth of the sovereignty of the people.” I hope after the preceding discussion the flaws in this other unsupported claim are clear.
Harari, it is clear, is not a biologist. Yet he keeps making dogmatically certain biological claims. His contrast between the human ancestor species Homo erectus and Homo sapiens is frankly atrocious. You could learn more about human biology by dropping a copy of Sapiens on your head than from reading it. Here, we start with poor inference: “Two million years ago, genetic mutations resulted in the appearance of a new human species called Homo erectus” (p. 37). This sentence is saturated in assumption and implication and is derived from a view of biology from the early 20th Century view of evolution. One major assumption that exclusive and neutral genetic changes lead directly to new speciation which is based in the idea that genes can be directly mapped to morphology. Another is that such speciation can simply be 'seen' in the fossil record. Changes in a genome can cause changes in morphology, but the role of development and morphological plasticity play a much larger role in these changes than we previously thought. Fossil species are also never defined genetically, as they have no (or very rare) genetic material left to study. That leaves us with the unreliable world of morphology to work with to delineate species, making the claim that Homo erectus simply 'appears' 2 million years ago even more spurious.
If the genetic cause of this new speciation is poor, then the next claim is worse: “As long as Homo erectus did not undergo further genetic alterations, its stone tools remained roughly the same – for close to 2 million years!” (p.37). Okey dokey then. Now Harari is not only using the outdated genes-to-morphology mapping discussed above, but the even more problematic genes-to-behaviour mapping, and further applying it to stone tools. Is there a gene for making hammers? No, of course there isn't, but this does not stop Harari from dropping more unfounded and spurious dogma, for example when he makes a similar claim about chimpanzees: “...changes in behaviour would only occur if something changed in the chimpanzees' DNA” (p. 37).
Remember before when we spoke of lazy narrative development in the section on language? Harari made a claim there that our ability to share fictions through language allowed us to build larger cooperative networks. This itself is an inference, one that at least makes sense, even though it is not as empirically verifiable as the ability to make certain tools. He stretches this inference when he claims: “Without the ability to compose fiction, Neanderthals [another human species] were unable to cooperate effectively in large numbers, nor could they adapt their behaviour to rapidly changing challenges.” Harari's assumptions do a lot of work here, again leading to a very weak claim asserted as fact. Any uncertainty he might have discussed previously is conveniently forgotten here as it is of no use to the narrative. Later in this section, he uses evidence used to demonstrate that Homo sapiens had a more complex society. These latter claims are based on the archaeological record but does do not directly support this causal link between language and society. He is mixing up observation and inference.
This essay has gone on a while, so I will end it with a final point, and probably the most important to be made here. It encapsulates Harari's ability to construct a narrative in one part of his work, and completely contradict it later, when it is more convenient to selling a different narrative. This whole chapter covers the uniqueness of Homo sapiens, which is somewhat obvious, through the narrative of an apparent Cognitive Revolution. Problems arise when he tries to describe how and why this happened leading to an absolute mess of observation and theory. There are plenty of examples of where we can see certain skills from the archaeological record associated with Homo sapiens that are not present in other species from the past (eg. Homo erectus) or coeval species (eg. Homo neanderthalis). Harari's interpretation seeks out a large range of specific, intrinsic and generally speculative reasons as to why that is, ignoring the beastly elephant in the room of all historical discussions and causation: Historical Contingency.
Contingency is the idea historical events, biological, or cultural, are contingent on prior conditions. This may seem obvious but is easily overlooked and highly inconvenient to people content on crafting misleading narratives. Historical contingency makes the path of history extremely sensitive to very small events, whose evidence is probably lost to time and utterly inscrutable. We also need to know we have the conceit of hindsight- we forget that when we look back on history and see its events, it does not mean we can see the causes of those events. This time-dependence is very important. Imagine being a human 70,000 years ago at the dawn of the Cognitive Revolution and, by chance, finding a rock that cleaves open a different way. This allows you to cut down a different type of tree. This tree floats better, so you can fish the lakes more effectively, allowing for more protein in your tribe, allowing for a little advance in cognition and a little advance in linguistic abilities. This little story (purely fiction, I assure you, to illustrate my point) is an example of how chance, extrinsic events, can result in wholly unexpected changes. The discoverer of that new type of rock had no idea that it would alter the future of his tribe- changes might have happened through generations. This is a description of societal development through chance events, not an appeal to genetics. A certain progress or direction in development is also no guarantee of long-term success- luck works both ways. In my story the tribe was unexpectedly eaten by crocodiles, and progress had to begin again in a different tribe or in a different time, none of whom had some special intrinsic characteristic. Iterated over generations, small chance events have a much larger chance of creating large and unpredictable changes. You're also more likely, over time, to be eaten by crocodiles.
Harari's treatment of the cognitive revolution and the evolution of humans in general is naïve to this point, which is greatly concerning for a so-called professor of history. In his view Homo sapiens evolved special genes, developed special language and creates special fictions. Homo neanderthalis apparently, did not. He implies constantly that other human species completely lacked the intrinsic hardware to install Homo sapiens special software, a totally spurious claim made totally without evidence. His history of this time is deterministic.
What boggles my mind is there is an entire chapter (13, The Secret of Success) on these sorts of historical pitfalls and a five-page discussion on the dangers of deterministic thinking, the hindsight fallacy, the importance of historical contingency and the chaotic behaviour of complex systems. Here he summarises it better than I have: “Those who have only a superficial knowledge of a certain period tend to focus only on the possibility that was eventually realised. They offer a just-so story to explain with hindsight why that outcome was inevitable. Those more deeply informed about the period are much more cognisant of the roads not taken” (p. 266). On reading this I almost threw my copy of Sapiens out of the window, and I will leave you with the note a scrawled across the top of this page: “DO YOU NOT READ YOUR OWN WORK?!”