Interview by Richard Marshall.
(Art: Julio Larrez)
'The theory of mind that story-telling relies on for its explanatory power is false. It’s not just that we can't get into the heads of historical agents to figure out exactly the beliefs and wants that paired up to determine their choices, decisions and actions. My argument is not that narrative history is underdetermined by the evidence.'
'There are lots of false but useful theories. Ptolemaic astronomy was around for a 1000 years or so, getting better all the time, as the epicycles were piled on. The theory of mind is just another one.'
'The neurologically agnostic research program of cognitive science is in effect the search for cognitive processes at much higher levels than the neural circuitry, or even vast assemblages of it. It’s safe to say that so far this search hasn’t produced anything much more impressive than a jargon-ridden predictively impoverished formalization of the theory of mind.'
Alex Rosenbergis the R. Taylor Cole Professor of Philosophy (with secondary appointments in the biology and political science departments). He has just published a new book that argues that narrative history is bunk because it uses a theory of mind that is false. Consequently, relying on such history is pointless and has lead to catastrophic decisions. Here he discusses some of the salient points.
3:AM: Academic history is rarely narrative and so you’re not attacking that. But what you are attacking is a history that we use to explain human actions and the lives they constitute and so forth. So before we look at your reasons, can you sketch for us what kinds of history are ok and which aren’t, what you take history to be and why you think matters that we get the narrative history stuff wrong? Where’s the harm in it?
Alex Rosenberg: My “issue” is with story-telling narratives, a mode of explanation that relies on the theory of mind, ‘mentalizing’ in some idiolects, ‘mind-reading’ in others (though I prefer using that label for the ability to track prey and predators we share with other mammals and birds), or what 20thcentury philosophers used to call “folk psychology.”
Historical explanation free from theory-of-mind driven story telling is innocent of the defects I allege. That includes a lot of macroeconomic history, history that exploits game theory, history driven by Darwinian cultural evolution, cliometric history. Among the major contributors to these kinds of history are Jared Diamond, Robert Fogel, Douglas North, Peter Turchin, and Gerald Mackie.
3:AM: Why should we be skeptical from the get go about narrative history being a path to understanding?
AR: The theory of mind that story-telling relies on for its explanatory power is false. It’s not just that we can't get into the heads of historical agents to figure out exactly the beliefs and wants that paired up to determine their choices, decisions and actions. My argument is not that narrative history is underdetermined by the evidence. It’s that the explanations get the causes of human action completely wrong. The theory of mind turns out to have all the virtues of Phlogiston theory and substantially fewer virtues than Ptolemaic astronomy had.
3:AM: Why do you say that because there is no settled account of what started the First World War we have a telltale sign that narrative history can’t give us the required answer – and never will?
AR: The long-term disagreement (a.k.a. “revisionist history) about even such well-trodden topics as what caused the First World War is a symptom of the fact that historians are on the wrong track completely when they try to fathom what the major players were “thinking” in August 1914. Notice that the trajectory of explanations in natural science never looks like the endless loop we get from generation to generation in the explanation of the same historical events. Epistemically credible explanations successively narrow down their disagreements with their predecessors. Historical explanations don't.
A century after the outbreak of the First World War there is no more agreement on the matter than there was when Winston Churchill published “The World Crisis” in 1923. (A wag said, “Winston has published his autobiography, cleverly disguised as a history of the world.” My critique of narrative story telling in history goes for biography too—history one life at a time.)
3:AM: And what’s the reason for you saying that if we want to win the next war we shouldn’t study military history? Surely the point of military historians is to help with that job?
AR: Well, winning the next war (or at least not losing it) is the aim of military history. But as pointed out in chapter two How History Gets Things Wrong,“How many times can the German army play the same trick?”, it’s never succeeded. The historians, like the generals, are always preparing for the last war. You want a handle on the future, don't try to get into Napoleon’s head, or Guderian’s.
Of course lots of military history is fun to read, and pumps patriotic emotions, not to mention revanchism. But we shouldn’t mistake entertainment and patriotic fervor for knowledge or wisdom.
3:AM: So your reason for these conclusions is that we have evolved a wired-in theory of mind that is false. First then can you say what this theory of mind is, why it evolved as it did and what it claims? Presumably at some point it must have been very useful even if it was false?
AR: As I suggested above, the theory of mind is a relatively inchoate set of platitudes (David Lewis’ word) about beliefs and desires, their causes and effects, that we employ to explain our actions. Hypotheses like “seeing is believing” and “Other things being equal, people do what they want to do.” It tells us that people make choices and act in light of, as effects of, resulting from pairs of beliefs (about means) and desires (which give the ends).
The theory evolved from the mindreading ability we share with other predator and prey species, once gestures and grunts emerged as ways of shaping the behavior of others. The languages these pidgins initiated eventually colonized our conscious experience with silent speech. Once words like ‘belief’ and ‘desire’ were invented we had the resources to express this theory and start using it.
Though false, the theory was useful in the Pleistocene and for long afterwards.
There are lots of false but useful theories. Ptolemaic astronomy was around for a 1000 years or so, getting better all the time, as the epicycles were piled on. The theory of mind is just another one. It emerged as a quick and dirty Darwinian solution to an evolutionary “design problem” we faced when we found ourselves at the bottom of the food chain on the African savannah. It was pretty good at predicting the behavior of a small number of people in our immediate vicinity over short time periods. Just what we needed to gang up on the mega fauna (and other Homininsfor that matter).
Trouble is we started employing the theory to explain the actions of large numbers of people, far beyond our immediate observable range, and over long times, far beyond the domain in which it is even roughly right in its predictions. Those explanations are what narrative history (and biography) consists in.
3:AM: So what does neuroscience have to say about this? Is there nothing like the theory of mind working in our brains?
AR: Neuroscience has identified the process and the locations where information is electrochemically encoded, stored, retrieved and deployed in complex “voluntary” behavior. Or at least the Nobel Committee has awarded a set of prizes to neuroscientists who appear to have done so, Eric Kandel, John O’Keefe, and May-Britt and Edvard Moser. Their work has of course been replicated and developed in further detail by researchers like Loren Frank. A lot of my book reports the work of these scientists (while trying hard not to package their achievements as ‘stories).
What that research shows is exactly how large assemblages of neural circuitry do the work of encoding, storing, and deploying information, but do it ways that are not even remotely like what the theory of mind would lead us to expect. In particular, there is no level of organization in the brain where there is anything like “content.” The neural states that guide behavior don't represent the way the world is arranged in the way beliefs are supposed to, nor do they represent possible outcomes of action in the way desires are supposed to. Neuroscientists can construct a map of the rat’s environment by reading it off the neurons firing in its midbrain. But the rat’s brain doesn't construct a map and so doesn’t consult it either. And the electrochemical oscillations that carry the information, what we quaintly call its “beliefs,” don't vary in some code we or the rats for that matter can decipher. Neuroscientists know this because they’ve detected the oscillation patters of distinct “memories.” The way these work to direct behavior is just the boring old conditioning processes behaviorists discovered, only operating on the neuronal circuits.
“Belief; and ‘Desire’ have about the same grip on the way the brain works as “Phlogiston” has on the way combustion works. One task we face is to explain how conscious experience forces the illusion of content on us, and makes us willing slaves to the theory of mind long after it passed its “sell-by” date.
3:AM: Much of the work has been with rats – what is the pushback against those who think we’re much, much more sophisticated than rats and therefore what you can see happening with rats is not applicable?
AR: You are absolutely right; much of this work has been undertaken with rats. Not all. There has been some with bats, some with primates, some even with humans, especially in the course of medical diagnosis and treatment. The whole research program began with a human, the most famous of all experimental subjects in neuroscience, “HM” whose hippocampus and entorhinal cortex were destroyed in an attempt to alleviate his epilepsy. All the research on all the brains points in exactly the same direction. That’s no surprise to neuroscientists. Otherwise they wouldn't have worked so hard on rats to begin with. After all, the relevant parts of the human, monkey, bat, and rat brains are “homologous” right down to the brain regions, tissues, cell types, and somatic genes that control their electrochemical circuitry.
3:AM: What’s wrong with thinking that although we can’t see the theory of mind at brain state level we would find it at a different level of explanation of what we’re doing? Doesn’t this view of the mind actually throw out beliefs and desires and representation and so forth – and aren’t they things we need to communicate and think about the world? Or is your claim that desires are just illusions – perhaps post hoc rationalizations?
AR:The neurologically agnostic research program of cognitive science is in effect the search for cognitive processes at much higher levels than the neural circuitry, or even vast assemblages of it. It’s safe to say that so far this search hasn’t produced anything much more impressive than a jargon-ridden predictively impoverished formalization of the theory of mind.
Trying to predict the future successes or failures of a scientific research program is a mug’s game. But the fact that the theory of mind we employ is the same one Homer employed when he told the tales of the Iliad and the Odyssey, strongly suggest that this research program is (to use some jargon from the philosophy of science) degenerating.
What’s more, though its somewhat embarrassing for mainstream philosophy of mind to say so publically, ever since Descartes 400 years ago, it has not been able to refute seductive arguments that brains couldn't possibly be the seat of anything like desires and beliefs, no matter how its parts are organized. This is why ‘physicalist antireductionism’ is the default position in the discipline.
Our conscious introspective experience is of course screaming at us all the time that we have beliefs and desires and they pair up to drive our choices and actions. That’s a useful illusion, like so many others consciousness tricks us into, like thinking of things in the world as colored and solid when they’re really mainly empty space full of leptons and bosons.
3:AM: Why does computer science and the ability of a computer to out- perform humans at Jeopardy show that the Theory of Mind is completely wrong – and why is that game a better example to show this than Chess?
AR: Watson, the IBM computer who beats human champions in trivia contests, seems to have a lot of beliefs and desires, or at least the theory of mind provides accurate predictions of its correct answers. But we know it doesn’t have any. It lacks what John Searle calls “original intentionality.” We have to interpret the distribution of charges on its microprocessors to actually attribute all the content it seems to have. There is much less temptation to attribute strategic beliefs to Deep Blue, the IBM chess champion computer. But if Watson’s circuitry gets its content from our interpretation, where do we (or rather our brains) get our content from? Someone else interpreting our brain circuitry? That’s the start of a regress of course. And the only way to get off the merry-go-round is to see that original intentionality—the content, representation, aboutness of thought, whatever you call it, is a mistake.
3:AM: You draw on characters from history and narrative history to illustrate your points – so why is it pointless trying to figure out what the Kaiser was thinking in the First World War?
AR: It’s pointless because the Kaiser wasn't thinking about anything at all when he gave the “blank-check” to the Austrians. He didn't have any desires about how matters should turn out, or any beliefs about how to organize things to make them turn out that way. He didn't because no one has such thoughts with content. Of course people have thoughts, mostly non-conscious and not available to introspection, and these thoughts are neural processes that drive behavior (without having any content themselves of course), including what the Kaiser did. But trying to locate what he believed and wanted, even roughly and inexactly, in contentless neural circuitry firing, is a fool’s errand.
3:AM: And what does the example of Talleyrand show us?
AR: In the 170 years since Talleyrand died, people have been trying to figure out that devious guy’s machinations in the first 40 years of 19thcentury diplomatic history. They’ve been arguing about what he really believed and wanted. When Talleyrand died, the Austrian nobleman Metternich is famously and falsely reputed to have said, “I wonder what he meant by that?” Well, he didn’t mean anything by it, nor by anything else he ever did. There are no meanings in the head.
3:AM: You also write about the more recent historical (and still living) figure of Henry Kissinger reading his way through the Congress of Vienna. What does he help us grasp?
AR: Kissinger doesn't help us grasp anything, because he thought he could get into the heads of the Talleyrand, Metternich, Castlereagh, and the Tsar, to understand what went on at the Congress of Vienna, and use the knowledge to craft American foreign policy for Richard Nixon and his successors. The results were predictably awful: he got most things wrong and the result was untold human suffering. Chalk it up at least as much to the theory of mind as to Henry Kissinger’s hubris.
3:AM: So are books like Guns, Germs and Steeland the Gulag Archipelagouseless? Shouldn’t we heed anything in them?
AR: Well, the former is my poster-child for the way history should be undertaken. It’s entirely free from the theory of mind in its explanatory apparatus. And the latter is my poster child for the real power of theory-of-mind-driven story-telling: its emotional power, its ability to move people by stirring their emotions. Alas, there are more such story-books that drive humanity to moral catastrophe than there are ones that do good, like Solzhenitsyn’s three volumes. Think of Mein Kampf, another story.
3:AM: It’s not just history that seems to be a problem if all this is right – just things like the evening news is going to be false isn’t it – and what about the law where criminal law, for example, seems to work by mobilizing narratives explaining motivations and so forth.
AR: Well, our whole culture and every civilization that we have any record of is constructed on the rickety foundations of the theory of mind. It has given us free will, moral responsibly, praise and blame, moral norms and political institutions, things we can’t dispense with in the normal course of life. My book is a plea that when we try to mitigate the worst features of human interaction, to design better institutions, control an uncertain future, we try to use theories that have a chance of being on the right track instead of the theory of mind.
3:AM: And finally, are there five books you can recommend that will take us further into this philosophical world?
AR: None of the five books I’ll mention agree with me about the theory of mind. But they are cognizant at least of the issues that we have to grapple with in figuring out what really makes the mind the brain and its implications for human culture and so stand a chance of moving psychological science in the right direction.
Dennett, Daniel, From Bacteria to Bach
Pinker, Steven, The Stuff of Thought
Neander, Karen, The Mark of the Mental
Sterelny, Kim, The Evolved Apprentice: How Evolution Made Humans Unique
ABOUT THE INTERVIEWER
Richard Marshall is biding his time.