Choice in a Determined Universe

by Luke Muehlhauser on October 1, 2010 in Science

drescher good and real smallI’m blogging through Good and Real by Gary Drescher, perhaps the best book on naturalism I’ve read yet. (See the series index.)

So far I’ve summarized Drescher’s views on the mind, consiousness, the nature of time, and quantum mechanics.

Now, Drescher turns his attention to

the paradox of choice: how can it be that we make choices for ourselves if we are just machines, completely constrained by the physics of our constituent parts? What difference could choice even make, if… the universe is deterministic, with everything (including the entire future) already just sitting statically in spacetime?

Drescher’s first goal is to argue that inalterability does not imply futility: Though reality is unchangeable, this does not mean our choices are without use or purpose.

First, consider the consequences of fatalism:

…a false fatalist theory (for example) can inspire suboptimal goal pursuit. You probably know someone who would undoubtedly act to dodge an imminent collision, but who, on the other hand, declines to wear a seatbelt on the grounds that when your time is up, there’s nothing you can do about it… [Thus] a fatalist theory may indeed impede an appropriate action that a better theory would instead promote.

We are, after all, prediction-value machines rather than just situation-action machines. Many simple creatures seem to be entirely made of situation-action rules. Given certain stimuli, a particular action is performed. Given another stimuli, another action is performed. But humans (and probably some other animals) are prediction-value machines. There are certain things that hold value to us, and we can cognitively predict which actions will best make true states of affairs we value, and thus act prudentially to realize that which we value. We contemplate and predict which actions will realize what we value, and we act in consideration of such contemplations. We areĀ choice machines.

In response to the actions of a choice machine, it makes sense to ask all of the following questions:

Why did it take that action? In pursuit of what goal was the action selected? Was the goal achieved? Would the goal have been achieved if the machine had taken this other action instead? … Would it have better achieved its goals [if it had possessed other information]?

And all of these are sensible questions even if the choice machine’s actions are entirely determined. With this in mind, let’s consider some challenges from the fatalist:

Why (in some cases) does it make sense for us to take an action for the sake of a [goal that will necessarily be realized]?

It makes sense for us to take that action because if we didn’t, the goal wouldn’t obtain. But didn’t the goal have to obtain? Yes, but so did the action!

But why, then, does it make sense to contemplate alternative actions, and their results, and select from among them, if it is already determined which sole action [will be] chosen?

The contemplation makes sense because without it, the preferred action might not have been taken. This is true even though the contemplation had to occur in the way that it did.

The choices of the choice machine are also compatible with our intuition that

when you do X, you believe you did so only because, on the whole, at that moment, given the circumstances, you wanted to. If instead you had on the whole preferred at the time to do Y, you would have done that instead.

Of course, a choice machine’s desires are fully determined, but this is compatible with the above intuition. Drescher concludes:

Thus choice, in the sense just delineated, is a mechanical process compatible with determinism: choice is a process of examining assertions about what would be the case if this or that action were taken, and then selecting an action according to a preference about what would be the case. The objection “The agent didn’t really make a choice, because the outcome was already predetermined” is as much a non sequitur as the objection “the motor didn’t really exert force, because the outcome was predetermined.” …Both choice making and motor spinning are particular kinds of mechanical processes. In neither case does the predetermination of the outcome imply that the process didn’t really take place.

In the book, Drescher goes into much more detail, and applies this reasoning about choice machines to Newcomb’s Paradox. In the next post, I’ll discuss chapter 7 of Drescher’s book: “Deriving Ought from Is.”

Previous post:

Next post:

{ 8 comments… read them below or add one }

Reginald Selkirk October 1, 2010 at 7:03 am

Determinism != fatalism.

I’m surprised by how often this comes up in discussions of free will.


lukeprog October 1, 2010 at 7:37 am



G'DIsraeli October 1, 2010 at 8:33 am

I liked very much the last paragraph, it was sitting on the edge of my tongue.

I just started reading “Reason and Practice: A modern introduction to philosophy” by the Kai Nielsen.
He puts it rather nicely: “A fatalist believes that all events are fated…events are bound to happen no mater what he choices he makes or actions he takes…Any effort to make a better world is benighted and pointless. It is out of our hands whether the future will be better or worst”

on the other hand a determinist: “But human actions sometimes do effect what will happen…(this does not mean) the government’s stubbornness or folly on this issue is uncaused.”


Thomas October 1, 2010 at 3:20 pm

I’m looking forward to your post on Drescher’s views on ethics and how they compare to desirism.


Joel October 1, 2010 at 5:54 pm

All depends on whether choice in a meaningful sense can be defined only in mechanical terms.

In any case, I am always worried that we who accept determinism, and yet want there to be free will, may be biased towards compatibilism. I think we ought to be really careful with our cognitive biases here, lest they get in the way.


svenjamin October 1, 2010 at 9:49 pm

“The experience of choosing is not a fiction, regardless of how the brain works. It is a real neural process, with the obvious function of selecting behavior according to its foreseeable consequences. It responds to information from the sense, including the exhortations of other people. You cannot step outside it or let it go on without you because it is you.”
Steven Pinker, The Blank Slate


lukeprog October 1, 2010 at 10:07 pm




Revyloution October 3, 2010 at 6:50 am

The ability to predict the future is certainly the cog in the wheel of fatalistic determinism. I have to disagree that certain organisms are either situation-action or prediction-value machines. I think the entire range of animals shows an wide range of prediction and reaction in their actions.

With humans, and other primates, we seem to be spending most of our brain power on predicting the near future, yet we also have portions of our brain dedicated to situation reaction. Stick a pin in someones finger, or take a doctors mallet and strike that nerve under your kneecap, and you see humans simply reacting to a stimulus. Likewise, the simple worm can learn to associate a lightbulb with the pain of electric shock. When it senses the electric shock, it reacts as though it is about to be electrocuted. That is, in a sense, predicting the near future.

Like everything else we thought was unique to humans, cognitive ability is really just on a sliding scale.


Leave a Comment