I’m a philosopher. I work at the University of Technology, so quite often I’m asked, “What do you actually do as a philosopher? Sit around and think?” Basically, yes. That’s what I do. Just when you’re a professor, you also get paid for sitting around and thinking about problems. And on good days, you may even solve a problem. What philosophers think about is formulated very nicely by Bertrand Russel, a famous philosopher and mathematician. He writes, “I believe the only difference between science and philosophy is that science is what you more or less know, and philosophy is what you do not know.” So, uncertainty is where science and philosophy meet. No matter what we do, we always face uncertainty. Will the weather be stable enough for a barbecue tonight? How shall I invest my savings? Which school shall we send our children to? And the decision becomes more complex when we consider society as a whole.
Shall we accept global warming? Or rather go all renewable? Having to face the risk of large-scale blackouts. Or shall we go nuclear? Having to accept accidents in a nuclear power plant, possibly. All these questions are not only about what is worse; the nuclear accident, global warming, the blackout. They’re also essentially about how we can incorporate uncertainty into our decision-making. We’ve had, since the 1950s, professionals in risk and uncertainty analysis. And we’ve made great theoretical and practical advances in how to deal with, how to understand and in how to manage uncertainties. But still, mankind is caught like a deer in the headlights when it comes to dealing with uncertain information.
Take climate change as an example. Since the 1970s, the predicted ballpark figure of global warming has not changed. And still we argue about it. And why? Because some details of climate models may be uncertain, because we’re not sure whether the human-induced climate change is 90 percent certain or 95 percent certain. And a lot of people use this as arguments for inaction. But this is just one example. So we seem to have reached a kind of dead end when it comes to dealing with uncertainty. We know a lot about the future, but we’re paralyzed to take action. I want to show you a way out of this dead end. One way of addressing uncertainty is risk minimization; minimize the risk of your endeavor, and no matter whether you want to start a hedge fund or put on your seat belt, it seems like pretty rational advice. But societal risks are different from individual ones. When we think about the future of our energy supply or when we think about GMOs, then the decision for or against the risky technology affects us all.
Us living today, but also the ones living in the future. So, the question about uncertainty becomes intrinsically intertwined with the questions about what is ethical. So, is risk minimization still a good approach here? Risk is mean expected harm; harm expected on average. But the risks may be distributed very differently among different members of society. Think about a wind farm. When you live in its close vicinity, you have to take all the burden in the form of noise pollution and other, while those living further away only share the advantages of a safe and sustainable energy supply. And focusing on risk as a mean concept, as mean expected harm, completely blurs out this information. So, is this okay? Yes, well, maybe. Maybe we have to accept that. Maybe some have to take the risks in order to do good for the many You may remember this very touching scene from Star Trek, in which Spock dies.
His final words to Captain Kirk are, “The needs of the many outweigh the needs of the one.” Spock is certainly a member of a very rational species, so maybe this is the right approach to risk. But the Star Trek case differs because Spock is pretty certain that with his death he’ll reach the goals he wants to reach. Now, what actually is uncertainty? When we talk about risk as mean expected harm, we make use of the concept of probability, a mathematical concept. And take, again, nuclear power as an example. We may be able to calculate the probability that a certain valve will begin to leak in the course of the coming 10, 25 years. And with this, we can maybe calculate the probability of a nuclear accident of the same time period.
But not always do we have good probability estimates at hand. Consider the nuclear waste and its storage in a geological repository underground. Maybe we can calculate that the rock formation is stable over the coming, let’s say, 100 years, or even a bit longer. But what about really long time scales? Think about 24,000 years, the half-life of a plutonium isotope. Or think about 1 million years, the time span the German government requires nuclear repositories to be safe. We certainly don’t have good probability estimates for such long time scales. Now, in philosophy, we refer to such situations as “decisions under uncertainty.” While for risk, we know all the probabilities, for decisions under uncertainty, some of the possible outcomes cannot be assigned probabilities. You probably learned all about probability in school, and you learned it probably by urns which contained balls in two different colors. Now, we have an urn here, and the equivalent of a decision under risk would be that I ask you, “What is the likelihood, what is the probability of the next ball being black?” What you do in answering it, you begin to perform an experiment, you begin to draw balls from the urn.
And you determine the relative frequency of balls. In our case one third. Then you say, “The probability of the next ball being black is exactly this relative frequency.” This is exactly what we face in a decision under risk. Imagine a similar situation, again, I face you with an urn. I want to know the probability of a black ball. But now I change the rule of the game. You’re not allowed to perform an experiment, you’re not allowed to draw any balls anymore. So, what do you say? What’s the probability? You can make a sophisticated guess, but the problem is, you don’t have … a frequency estimate for your probability at hand. And that’s exactly what we face in a decision under uncertainty. Now consider a third case. Again, like in the uncertainty case, you’re not allowed to draw any balls.
You have to start with a real world, with real events, with the drawing of the balls, in our case, right away. So you draw a few balls, all is fine. All of a sudden, you end up with… not a ball. With a white rubber duck which looks pretty intellectual. Oops, that was kind of a surprise! We call these decisions “decisions under ignorance”. While for risk and uncertainty, you still knew what the possible outcomes were, for a decision under ignorance, you don’t even know all possible decision outcomes. And these are particularly bad, because you only know them with the benefit of hindsight, as we just did. So we have risk, we have uncertainty, we have ignorance. And only for the first, risk minimization would be a good approach. Luckily, we do have other decision rules available, for example, a precautionary approach. The precautionary principle does not look at the harm that occurs on average, it looks at the worst-case scenario only. So the worst imaginable case, and it tries to avoid it at any cost.
Now, quite often, we associate the precautionary principle with politics in the EU, on the European planet, while risk minimization is associated with the US. And indeed, when you look at TTIP or at the different ways people react to GMOs on both sides of the pond, that seems to be the case. While many European countries go for a precautionary approach towards GMOs – they are forbidden until they’re proven to be safe – in the US, they are allowed, as long as they withstand a risk analysis. Precautionary principle and risk analysis provide the two extremes of a whole spectrum of decision rules, but all decision rules share the same problem. There are good reasons for one decision rule, for one decision, and there are also good reasons for another decision rule for the same decision. So, what to do? Our current theories don’t tell us anything about for which decisions we should actually take the precautionary approach, or for which we should take the risk minimization approach.
Why is this? The reason is, all our current theories do is focus on the decision situation, the action itself and all its uncertainties. But we also need to include the actor, with all his characteristics. So far, we focused on knowledge — knowledge about the decision situation, knowledge about decision rules. But what we need is to also include skill; skill of those who are actually taking the action, the person or the group of people who do take the action. Skill differs significantly from knowledge. Knowledge you can learn from a textbook, but skill you have to learn by practicing. You cannot learn how to ride a bicycle only by reading a textbook. The skill we need in order to deal with uncertainty is certainly not a bodily skill. What we need is rather something like a skill, more like a kind of virtue, a character trait.
And now probably for many of you, red flags rise in your head, because training virtues, training character traits? That sounds, at best, impossible. While at worst, it sounds like manipulation. But we do value character traits and virtues even in our society. Think about honesty, think about nonviolent conflict resolution, think about maybe even bravery. And we do train them; we do discourage our kids at school from cheating. And this way, we train them the value, the virtue of honesty. The skill we’re looking for, the virtue we’re looking for, is still much more complex. On the one hand, it needs to resemble honesty; it needs to be an ethical skill that needs to tell us what is worse; a nuclear accident or global warming. But it also needs to be an intellectual skill that needs to sort of tell us whether we actually do face a decision under risk uncertainty or ignorance. Philosophers have also, for a long time, only focused on the action itself. This was completely different in ancient times.
There, the person, the actor, actually stood at the center of our reasoning. Where do we find a solution for our seemingly modern problem of how to deal with uncertainty? Ancient philosophy. The Greek philosopher, Aristotle, has suggested the virtue of phrónesis. And this proves very important for us today. The phrónesis is an intellectual virtue, but it’s not the cunningness of a fox, not a kind of shrewd cleverness; it’s prudence that is always aimed at the ethically good. How does this work in practice? The work of the phrónesis is twofold. First, it’s a kind of problem indicator; it shall single out a situation as ethically relevant. So consider a seemingly ethically neutral situation, like shopping for new clothes. Then the phrónesis will indicate all the things you’re uncertain about. You’re uncertain, maybe, about the production process of the clothes, whether the chemicals are used in a sustainable way, whether there was child labor involved. And the phrónesis hints you to the ethical implications this may have.
The second task of the phrónesis is to mediate between general rules and the specific situation. In our case of buying new clothes, it mediates between the general decision rules we have available under uncertainty – risk minimization, for example, precautionary approach – and it helps you to decide, whether, given your uncertain background information, you should go for the risk approach or for a precautionary approach. You may think that still sounds pretty abstract, and it is. But remember, just like you cannot learn how to ride a bike just by listening to a TEDx talk alone, we are talking about virtues, about character traits, a kind of skill; we need to train them. The fundamental shift we need in order to improve our decision-making under uncertainty is to shift our focus away from knowledge about the action itself, to the character traits, the virtues of the decision maker, of those who are taking the action.
That doesn’t mean that knowledge becomes unimportant. It just means that knowledge does not suffice. And this shift is a very fundamental one. It’s a kind of revival of ancient thinking after it was buried for 2,000 years. With this, I want to give you a taste of how a possible way out of the dead end we, at the moment, are at when dealing with decision-making, could look like. And remember: I’m a philosopher, and paid for thinking and not paid for doing. So, what philosophy does is, it provides the science, the direction of how to reach a good decision-making under uncertainty. And when you practice phrónesis, when you practice your intellectual skill of cycling, you’ll finally get there. And I want to wish you a good journey there. (Applause).