Thinking Historically about the Future
Part of a Series about the Philosophy of History
A comment on my episode on Ernst Jünger made me aware of a statement by Friedrich Engels made in 1887 that anticipates some of the most surprising characteristics of the First World War, and this from a generation or more prior to the war. Engels’ prediction, which was made after the American Civil War yet before the Russo-Japanese war, appears to show rare insight into the war that was to come. Here’s the quote in question:
“Germany will have allies, but it will leave them in the lurch, and they Germany, at the first opportunity. And, finally, the only war left for Prussia-Germany to wage will be a world war, a world war, moreover, of an extent and violence hitherto unimagined. Eight to ten million soldiers will be at each other’s throats and in the process they will strip Europe barer than a swarm of locusts. The depredations of the Thirty Years’ War compressed into three to four years and extended over the entire continent; famine, disease, the universal lapse into barbarism, both of the armies and the people, in the wake of acute misery; irretrievable dislocation of our artificial system of trade, industry and credit, ending in universal bankruptcy; collapse of the old states and their conventional political wisdom to the point where crowns will roll into the gutters by the dozen, and no one will be around to pick them up; the absolute impossibility of foreseeing how it will all end and who will emerge as victor from the battle. Only one consequence is absolutely certain: universal exhaustion and the creation of the conditions for the ultimate victory of the working class.”
This quote was new to me, but I have subsequently found it in several sources. It can be found in volume 26 of the Collected Works of Marx and Engels and is from an Introduction that Engels wrote for a pamphlet by Sigismund Borkheim titled In Memory of the German Blood-and-Thunder Patriots 1806–1807 (also translated as In memory of the German patriots of murder 1806–1807), originally published in 1887 as Zur Erinnerung für die deutschen Mordspatrioten 1806–07. In the pamphlet, Borkheim was writing about events eighty years’ past. This was the period of the Napoleonic wars known as the War of the Fourth Coalition. There was a war of the third coalition before this, and a war of the fifth coalition after this, as the period of Napoleonic wars was one of nearly continuous warfare.
Why was Borkheim writing a pamphlet about war in 1807 in 1887, eighty years later? A book you might expect eighty years after the fact, but a pamphlet is intended to be timely. To understand what’s going on with Borkheim’s pamphlet, we need some of the backstory. Borkheim was a perpetual revolutionary, a sometime wine salesman, and a friend of both Marx and Engels. He was an active participant in the “Springtime of Nations” in 1848, when revolutions were erupting all over Europe. There was a lot of craziness in 1848, and Borkheim, who was serving in the military, joined revolutionary military units, participated in several campaigns, and eventually served time in a number of prisons for his actions.
During the Napoleonic wars, the revolutionary French forces repeatedly defeated and humiliated any number of armies that it engaged. During the Springtime of Nations in 1848, it looked like revolutionary fervor had been wound up to a higher pitch even than the French Revolution, and many thought that this would be the end of reactionary regimes across Europe. But in the Franco-Prussian war of 1870–1871, the Prussians won, and no one viewed the Prussians as representatives of revolutionary zeal. So, just to be a wet blanket, Borkheim was reaching back eighty years to remind the Germans of a defeat, so they wouldn’t get too cocky over their victory in the Franco-Prussian war. Like Marx and Engels, Borkheim believed that the inevitable revolution was still coming, implying that the German victory in the Franco-Prussian War was a mere flash-in-the-pan.
Engels’ introduction to Borkheim indicated that Engels was not only interested in the past of the War of the Fourth Coalition or the past of the Springtime of Nations, he was also interested in the future. The quoted paragraph sounds like as good a prediction of the First World War as we could find, and as an apparent prediction of something most did not predict, it’s worth looking into. We need to listen to the rare voices that turned out to be right, though we shouldn’t listen uncritically. What we want to know is whether these voices were right for the right reasons, or right for the wrong reasons. In a world of billions of persons, it is statistically almost certain that someone will make a prediction that comes true, but if this accurate prediction is a mere accident, then it is no more of interest to us than white noise.
Of course, when thoughtful individuals take the trouble to write out their predictions and to give reasons for them, they are rarely entirely right or entirely wrong. Usually there is a mixture of prescience and blindness, which is why most predictions are half right, and why most long range predictions are mostly wrong, because the other half of the half right predictions, the half that wasn’t predicted, accumulate and eventually push events in an unprecedented direction.
Another prediction somewhat similar to Engels can be found in H. G. Wells The Shape of Things to Come, which imagines a First World War that doesn’t end, with England broken up into regions ruled by warlords, until a new international force — of course, enlightened and humanitarian — descends upon the warlords and displaces them. For Engels, the enlightened ones are the class conscious proletarians who will eventually triumph by expropriating the expropriators and instituting the dictatorship of the proletariat. For H. G. Wells, the enlightened ones are the engineers of the Air Dictatorship — in the film version they’re called “Wings over the World” — who are the natural technocratic leaders of industrialization. For today, I’m not going to say anything more about H. G. Wells, and instead I’ll focus on Engels.
Engels wrote a bit on military affairs, enough that some actually called him the “Red Clausewitz” So, how right and how wrong was the Red Clausewitz in predicting the war to come? If we take the paragraph I quoted earlier and break it down into its constituent parts we get the following, more or less, though it would be possible to analyze Engels’ prediction into slightly different items:
- A world war
- Unprecedented extent and violence
- 8–10 million soldiers involved
- Europe stripped bare as if by locusts
- Similar to the Thirty Years’ War, but over 3–4 years duration and extended over the entire continent
- Famine, disease, and lapse into barbarism
- Universal bankruptcy
- Collapse of old states
- Universal exhaustion
- Creation of the conditions for the victory of the working class.
Perhaps more importantly than individual events is that, for Engels, all of these events are connected to each other, they follow from the same general cause. All are manifestations of the unfolding failure of bourgeois society. But how good are Engels’ predictions? Let’s consider each of these in turn:
- The First World War was a world war, so Engels is right here.
- As a world war, the extent was unprecedented, but whether the violence was unprecedented is open to question. New weapons or war made new forms of destruction possible, but whether this violence was qualitatively worse than in earlier wars is not clear.
- About 65 million soldiers were mobilized for the First World War, so here Engels actually fell short by an order of magnitude.
- The areas subject to trench warfare were unambiguously stripped bare, but most of the war was fought in the countryside, so that fewer cities were destroyed than during the Second World War, and certainly the whole of Europe was not denuded.
- It is more or less accurate that violence comparable to the Thirty Years’ War was reduced to four years and that it extended over the entire European continent.
- Localized famine, disease, and social unrest occurred, but not over the entire area of the war, and order was eventually restored.
- There were localized financial failures, but not universal.
- Old states did collapse, in particular, Tsarist Russia and Austria-Hungary, but many old states managed to survive.
- Certainly France and Germany were exhausted by the conflict, but I would not call the exhaustion universal. Even the Second World War did not lead to universal exhaustion.
- The Bolshevik Revolution occurred in Russia in 1917, so again we can say that the predicted conditions obtained locally, but not across all of Europe.
A good prediction should involve three elements:
- What will happen
- The date by when it will happen
- The probability with which it will happen
Engels tells us clearly what he thinks will happen, and he seems certain of it happening, so the probability is certain, but he doesn’t give a date or a date range. It could be argued that some kind of date range is implicit in the internal context of the Introduction to Borkheim, from which this paragraph is taken, but there’s nothing sufficiently specific that Engels’ doesn’t have plausible deniability for his prediction. If you never furnish a date for your prediction, you can never be shown to be wrong, because it can always be claimed that the dénouement is still coming. One can always claim that the conditions of the victory of the working class have been created. But leave that aside.
Instead let’s consider the predictions that historians routinely make. Historians make predictions about past events, which we call retrodictions. Most retrodictions made by historians more or less embody the principles I mentioned, of being specific, or as specific as possible, giving a date, and giving a degree of certainty, though in traditional history this wasn’t the language that was used. An hypothesis about the past might later be confirmed or disconfirmed by evidence that comes to light, and so, as with experimentation in empirical science, we expand our knowledge of the past through retrodiction and degree of confirmation. If we could extend this process into the future, we could have knowledge of the future analogous to knowledge of the past obtained through retrodiction, but we can’t because there is no evidence of the future. But predictions are made nevertheless, on the basis of past events and usually some theory of history.
In my episode on Philosophy of History before Augustine I quoted Hayden White on all histories involving a philosophy of history, though this philosophy of history is usually implicit. The same goes for predicting the future: there is always an implicit philosophy of history involved in a prediction. If it makes you feel better, we can call it a theory of history instead of a philosophy of history. And while predictions are made without the possibility of evidence, they can be tested when the present catches up with the predicted date of an event. Philip Tetlock, who wrote a book about prediction, which he calls forecasting, calls those who forecast according to a quantitatively defined methodology “superforecasters” and some of them have a good record. If you predict according to a quantitative methodology, and you regularly return to your predictions to assess your accuracy, you can significantly improve your forecasting over time.
It is to be expected that the intelligence services of all major nation-states are engaged in sophisticated forecasting along these lines — probably much more sophisticated than what we find in open source intelligence, as with Tetlock’s book. Nevertheless, intelligence agencies still manage to be surprised, or at least they pretend to be surprised. We’ll never know for sure.
I dislike predictions. Too much prediction has been a particularly silly futurism. Some years ago, under circumstances I don’t exactly recall, I was corresponding with an editor of Futurist magazine, and I suggested an article on futurism without predictions. If memory serves, my correspondent was lukewarm about the idea, so I didn’t follow up on it, but I did write a blog post about it. I think it is possible to extend the methods of history to the future, and that we need to do this to provide a rational account of the whole of time, and not only the past. In my episodes on Herman Kahn and Robert Heilbroner I discussed the expansion of historical consciousness into the future, which it part of the process of the extension of history. But I don’t think that making predictions about the future is the right way to do this. However, in several episodes of Today in Philosophy of History I have expressed my admiration for those who have demonstrated an intuitive insight into coming events. How can I square that with my dislike for prediction?
Part of my interest is admiration for those who have an intuitive insight into what’s happening and what will happen. What do I mean by intuitive insight? Roughly speaking, “intuition” is used in two ways. Firstly, it is used for a kind of perception, whether empirical or intellectual. Kant, for example, made a distinction between sensible and non-sensible intuition. Both kinds of intuition simply appear to us, though the mechanism of their being present to consciousness is distinct in each case. When philosophers of mathematics and logic talk about intuition, this is the kind of intuition they mean. Gödel, for example, said that we could perceive abstract objects:
“I don’t see any reason why we should have less confidence in this kind of perception, i.e., in mathematical intuition, than in sense perception.”
When we do perceive abstract objects, if we do perceive them, we perceive them through this kind of non-sensible or intellectual intuition.
Secondly, intuition is used to mean the ability to perceive or understand something that is hidden or obscured or not otherwise evident to everyone. In this second sense, intuition is the ability on the part of some people to add up subtle clues, which involves a judgment as to which clues are significant and which are trivial. It is also the ability to fill in the gaps between the clues, and thereby to see what was there to be seen, but which was not obvious. This is the intuition of a Sherlock Holmes or a Miss Marple. The historical intuitive insight that Engels had was of this second kind.
Both kinds of intuition are subject to failure. Kant wrote about what he called transcendental illusory appearance, which is the failure of the first kind of intuition. And it is often the case that we put together the wrong conclusion from a number of fragmentary clues, and this is the failure of the second kind of intuition. We find this failure in Engels. Engels was partly right and partly wrong in his prediction. He put together a partly right, partly wrong prediction from fragmentary clues. His prediction, then, was a partial success and a partial failure. I said earlier that the connections between the events and predictions were as important as the events and predictions themselves for Engels. French historian Fustel de Coulanges formulated something like this such that:
“Supposing a hundred specialists had divided the past of France according to lot, do you think that, in the end, they would have written the history of France? I very much doubt it. At the very least, they should miss the linkage of facts: now, this linkage is itself a historical truth.”
This is, in effect, the principle from which Engels is working, that the linkage of facts is itself an historical truth. But this wasn’t the only principle from which Engels was working. Most predictions that put together a lot of subtle clues into one big picture do so according to a theory. A theory of nature affords us intuitive insight into how the world works and provides us with a framework for experimentation. A theory of human nature affords us intuitive insight into how human beings work and allow us to reconstruct human behaviors even in the absence of records. A theory of history affords us intuitive insight into how history unfolds, and it allows us to reconstruct history even in the absence of evidence. If the theory in question is all wrong, it will lead us astray, so the longer we follow it, the further out in the weeds we will find ourselves. If the theory in question is right, then it will demonstrate its accuracy with a high degree of predictive and explanatory power. If the theory in question is partly right and partly wrong, it will yield mixed results both in reconstructing the past and preconstructing the future, and its predictive and explanatory power will be hit-and-miss.
We know that the basis for Engels’ predictions was the Marxian theory of history, so the mixed success of his predictions is at least partly due to his interpretation of historical materialism. For Engels, Marx’s theory of history provided the linkage of facts of the kind that Fustel de Coulanges called a historical truth in addition to the isolated facts themselves. Of course, convinced Marxists will insist that Marx’s theory of history is just fine, so if it yields mixed results, it does so only because of its faulty interpretation. If Engels got it wrong, or merely partly wrong, it was because the theory of history he was using wasn’t real historical materialism. Obviously, this is an unprofitable path to pursue.
I said earlier that part of what inspires my admiration for those to get a glimpse of the future is their intuitive insight into history yet to come. Another part of my interest is negative: sometimes what’s going on is obvious, and not seeing it is a function of willful blindness. During especially dishonest periods of history, willful blindness can be the rule, and the ability to see what’s actually going on is the exception. When willful blindness is the rule, the few who are not captive to the dominant narrative are the few who see what’s really happening. This is the flip side of using a theory to predict the future. Sometimes it is the ability to stop seeing a theory that makes it possible to see what’s really going on. Being too captive to a theory often means interpreting everything so that it is made to fit the theory. This is precisely what historians criticize most harshly in philosophers whom they believe project a theory onto the past, forcing the facts of history to fit the theory instead of getting the facts right and finding a theory to explain them.
So far, then, part of what interests me is intuitive insight into the historical process, and another part of it is not having your head turned by a sexy theory. We could also say that we want to see the patterns that are there, and which are missed by others, and we want to unsee that patterns that aren’t there, and which delude and distract others. One more part of what’s going on can be understood with reference to the distinction that Fernand Braudel made between the history of the event, and conjuncture, and the longue durée, which are distinct scale of historical time.
For Braudel, the history of the event is the kind fleeting occurrence that flashes like a spark and then disappears. Car crashes, bank robberies, broken water mains, and hurricanes are all instances of the history of the event. The conjuncture is a period of time of about 15–30 years during which some thematically unified historical process unfolds. Boston’s “Big Dig” highway infrastructure project can be understood as a local conjuncture, The Thirty Years’ War constituted a geographically regional conjuncture, and the Cold War Space Race was a planetary scale conjuncture. The longue durée is a period of hundreds of years during over which societies and civilizations slowly evolve. The Early Modern period from 1500 to 1800 is a longue durée, and it is nested within the longer durée of modernity, which includes the present and ourselves in addition to the early modern period.
I think that it’s possible, even if difficult, to grasp what’s going on in the longue durée and to see events and conjunctures as manifestations of the longue durée, which is the deepest strata of history. You could call this prediction if you like, but I would refer to call it extrapolation. We could illustrate this by making a distinction between substantive prediction and formal prediction.
Substantive prediction is the conventional idea of prediction. If I were to make the prediction that the rise of China as a naval power — with two active aircraft carriers and another expected to begin its sea trials soon — will inevitably come into conflict in the Pacific with US naval assets, that would be a substantive prediction. It would be a much better prediction if I specified more about how these navies would come into conflict, and if I were to give a date by which the conflict would become a hot war.
Formal prediction is an extrapolation of a structure existing in the present that will continue to exist in the future and will continue to manifest itself, albeit under changed conditions. If I were to make the prediction that a rising power and a declining power that share an operational frontier often come into conflict, and that this is likely to drive future conflict in any terrestrial theater of operations in which these conditions obtain, that would be a formal prediction. It also bears some resemblance to Carl Hempel’s covering law model of historical explanation.
Obviously from these two examples I’ve given, there is a gray area that separates substantive and formal prediction. Still, there is a separation. A formal prediction is like the Oracle at Delphi telling the representatives of Croesus, “if King Croesus crosses the Halys River, a great empire will be destroyed.” Croesus, believing that this was a prediction that the Persian empire would be destroyed, crossed the Halys river, engaged the Persians at the Battle of Pteria, and in doing so set in motion a series of events that would destroy a great empire — which was his own.
My prediction about conflict between rising and declining naval powers is ambiguous in much the same way as the Delphic Oracle was ambiguous. The ambiguity stems from leaving crucial terms undefined. You get a formal prediction if you substitute variables for values in a substantive prediction, and you get a substantive prediction if you substitute values for variables in a formal prediction. It’s like the difference between arithmetic and algebra. Both have their uses.
This ambiguity is frustrating from a substantive perspective, but the aim isn’t to substitute values for the variables, but to construct a better formal model that is a more accurate reflection of the historical process than our previous model. It’s an iterative process that can be improved in the same way that superforecasters improve their forecasting. I haven’t fully thought this through yet, so it’s still pretty speculative. The distinction between substantive and formal prediction needs a lot more work to be satisfying. And that’s all I will say about it for the time being.
I could have saved all this for Engels’ birthday on the 28th of November, but by that time the connection to the Ernst Jünger episode would have been lost.
