The “Consequence Argument” is a powerful argument for the conclusion that, if determinism is true, then we have no control over what we do or will do. The argument is straightforward and simple (as given in the Stanford Encyclopedia of Philosophy):
Premise 1: No one has power over the facts of the past and the laws of nature.
Premise 2: No one has power over the fact that the facts of the past and the laws of nature entail every fact of the future (i.e., determinism is true).
Conclusion: Therefore, no one has power over the facts of the future.
Premise 1 seems awfully secure. Authors of history books might change people’s beliefs about the past, but try as they might, they won’t actually change the past. Similarly, scientists may write about the laws of nature however they please, but nothing they write will change those laws. No one can control the facts of the past, or the laws of nature.
Premise 2 looks pretty good too. For at least great big patches of nature, events happen because of the way things are or have been, and because of the continuous governance of the laws of nature. True, there are subatomic phenomena that seem to be indeterministic (Einstein was wrong, and God or nature does seem to roll teensy-weensy dice). But for whatever reason, it also seems that as these subatomic bits are assembled into larger parts of nature, the dice rolling seems to no longer have any effect, and at that point we enter upon a deterministic universe. Certainly by the time we get to big globs of neurons within the skulls of homo sapiens, wired up to eyeballs and limbs, we are in a domain where the fact is that the facts of the past and the laws of nature entail every fact of the future.
And the conclusion follows: we have no power to affect the future. So that’s it. We’re done.
Of course, this conclusion seems false to most of us, since we spend inordinate amounts of time and energy in attempts to shape the future. We buy groceries for dinner, invest in retirement accounts, look both ways before crossing the street, and so on. In fact, some philosophers are so convinced that the conclusion of the Consequence Argument is false that they use it to argue that Premise 2 must be false (that is, that determinism is false).
Now here is another simple and compelling argument, which I will call the “Unpredictability Argument”:
Premise 1: No one has thorough knowledge of the facts of the past and the laws of nature.
Premise 2: Successful prediction of the future requires thorough knowledge of the facts of the past and the laws of nature.
Conclusion: No one can successfully predict the future.
Again, Premise 1 is unassailable. No matter how much history someone knows, they certainly don’t know all the relevant facts of the past that go into making things happen in the way they happen. Even the simple event of pouring cream into coffee entails countless interactions among countless particles that only Laplace’s demon could track. There’s too much information for any real being to master.
Premise 2 is pretty secure also, if determinism is true. If we think of the future as determined by all the relevant facts of the past put together with the laws of nature, then making a successful prediction of the future should require knowing those facts as well as knowing the relevant laws of nature. Making a prediction without that knowledge is simply making a guess, and even if the guess turns out to be true, it would not be honest to call it a prediction. It was just a guess or a hunch.
And, once again, the conclusion follows; though, once again, it sure seems false. Consider the predictions you make while entering a busy traffic intersection. That guy looks like he’s going to turn, even though his blinker isn’t on; that other guy seems to be going straight; the pedestrian is waiting to see what I’m going to do, etc. We make wrong predictions sometimes, with tragic consequences, but it is astonishing how often we are right, and how brilliant we are in general at making accurate predictions of incredibly complex phenomena. And we make these predictions armed with only the slightest knowledge of the most superficial facts. Indeed, none of the data we use in making our predictions is of any use whatsoever for any physicist or neuroscientist who is trying to construct an account of the situation. All our knowledge and predictions are superficial in the extreme. They do not consist of any of the sorts of facts that are at work in the Consequence Argument.
There is a lesson to be learned from these two plausible but ultimately unconvincing arguments. The lesson is that there are many levels of causality. If we insist that the only real causes are those treated by the most fundamental sciences, we end up with the Unpredictability Argument. So, to make sense of our remarkably successful powers of prediction, we need to open the gates a little and allow for causality that isn’t covered by the most fundamental sciences.
We have to allow for entities like beliefs, motivations, and goals, and the sorts of behavior that convey to us the presence of those entities. We have to allow also for further background assumptions, not quite laws of nature, about what people are likely to do, or what makes sense for them to do. We have to let into our account the sorts of things we actually use as we go about making ordinary predictions. Call these auxiliary epistemic resources “squishy stuff”.
It’s not clear that any of the squishy stuff can be identified strictly with neural states, let alone with the more basic physical facts underlying them. Not that there’s no connection, of course; it’s just that the more fundamental account will not help us at all in understanding what those beliefs are or what causal roles they play. It would be like trying to understand the plot of Anna Karenina by doing histograms of the words in the novel, or diagramming all its sentences. Our understanding of what’s going on doesn’t require knowledge of facts at that level. Instead, we require knowledge of states or entities that don’t show up at the basic level. We require psychological and social knowledge, which isn’t unnatural or miraculous, but also isn’t at all augmented by our understanding of the more basic physical facts underlying it.
But once we let in the squishy stuff, we can start talking about people’s beliefs, motivations, and intentions to affect the future. We can identify behavior that arises from an agent’s own beliefs and goals, and behavior that is caused by other factors. In this way we can begin to identify the actions that count as “free” as well as those that seem less free, or forced by circumstance. At that point, we have undermined our confidence in the conclusion of the Consequence Argument. If we try to put our finger on where it goes wrong, we might point to the appeal to “the facts of the past.” For the facts of the past are not the most relevant determinants of what we end up doing. What is more relevant are our goals, our hopes, our motivations, our beliefs, etc., which are not exactly facts of the past, but features of our present states of mind, states in which we make our decisions about what to do. For obviously we do have some power over some facts of the future, even if that power is not evident in an analysis of all the microphysical bits and forces composing us. Once again, if we are skeptical that there is more to us than what the microphysical account offers, we shall have to face the Unpredictability Argument, and we shall have to try to explain our success in prediction when we have no access to the microphysical account.
The moral of this story is that sometimes this is how philosophy works: we learn a bit more about our ways of conceiving the world by noticing the shortcomings of plausible arguments that are based upon intellectual oversimplifications. We learn that the squishy stuff we thought could be safely ignored in one context turns out to be crucial in some other context. Human reality, it turns out, often has two hands: on the one hand is this, and on the other hand is that. The on-going challenge is to assign appropriate weights to the things in those hands.
(The ideas in this essay were caused (at some level) by listening to Sean Carroll’s conversation with Jenann Ismael in episode 80 of his podcast, Mindscape.)