Everything has exceptions. Look at how balloons fall up! I do not understand why you think "increased cognitive load lowers willpower" is something that can't be objectively measured and shown to be true or false.
Cognitive load is hard to measure, but you don't have to measure that part, only induce it. There are plenty of ways to get one group performing easy tasks without much thought, and another performing tasks with a lot of thought.
With willpower you can be more clever. You don't actually care about an abstract willpower meter, you care about the application of willpower to make good choices. Application is downright trivial to measure in a narrow way: just provide a tempting choice. Do the same thing with as many different choices as you can think of and you've shown that willpower is impacted.
Your examples are indirect methods. For the claim "increased cognitive load lowers willpower" to be falsifiable you actually need to have meters with numbers on them that let you measure these things directly from the brain. How do we know the brain even has a willpower construct?
I think you could probably substitute brain activity for cognitive load and measure it with MRI or something.
edit: I realize that if all you are interested in is the phenomena that you've decided are proxies for our concepts of cognitive load and willpower, that this claim is falsifiable; the experiment did demonstrate causation there. But it's not falsifiable if you're saying that cognitive load in the brain causes willpower in the brain to go down. My primary concern is that the new hypothesis that is presented as the conclusion of the study is too far extrapolated from the experimental results, and untestable.
> But it's not falsifiable if you're saying that cognitive load in the brain causes willpower in the brain to go down.
That depends on how you define 'in the brain', which seems like a waste of time. If I run for a while I can say I've 'run out of energy', and I can do show it objectively, but the question of whether the body has an 'energy construct' is semantic at its root.
Another way of putting it is: treat willpower as a measurement, not as something to be measured. Sorry that I can't force that thought into English better. I hope you figure out what I meant.
But your body does have an energy construct, in the form of fats and sugars, and we can measure them, and we can relate their levels to the amount of force you're able to exert. We know that these things provide energy because we can measure the heat produced when we oxidize them. Compare with willpower - where is the physical thing that we can measure? It might be there, but presently we have no way of knowing.
The body has thousands of kinds of 'energy' levels we can measure, but combining them into an overall value is a matter of preferences, numerical weights, and semantics. I see willpower very similarly.
Perhaps you're right, but if so, what is a single 'willpower' level the body has that we can measure? The comparison you made with energy is very appropriate.
As an aside, and not related to the broader point here about this kind of psychology experiment, scientific fatalism rejects the notion of willpower altogether.
It doesn't matter what it 'truly' is. The colloquial definition is pretty easy to measure by proxy, and in reality every measurement every taken is a proxy of some sort.
And as far as I know fatalism is only concerned with the abstract philosophical definition of willpower, not the practical definition that relates to desire and objective behavior.
Science is only concerned with what things 'truly' are. Anything that is presently an untestable claim - such as the brain having explicit circuitry devoted to our concept of willpower - is not a scientific hypothesis.
Measurements tell you about the thing you are measuring and you can do science with them. However they do not tell you the mechanism by which the thing you are measuring works internally. For example, when Newton discovered gravity he did not know how gravity worked, so gravity was a black box that he simply measured. It's now a much smaller black box thanks to better experimental methods. The problem is not making scientific claims about the behavior of the black box, the problem is extrapolating to the point where you're claiming that your model of what's in the box is the right model, with no way of checking. Hopefully as neuroscience makes progress we will be able to test our claims about the inside of the brain.
I don't know how to make it clearer. As far as fatalism is concerned, some scientists believe that everything since the big bang is predetermined, and choice is an illusion. It's unfalsifiable, but the point is that's just as valid as the claim that we actually have willpower.
Science is based on cause and effect. Science doesn't care the slightest bit about philosophy. If you hold a sphere of silicon in your hand, is it one object, or is it 10^25 objects? Science doesn't care, it depends on context, it's a semantic issue.
The mechanism of willpower-actions, the actual choices, is based on neurons firing and chemical shifts and such. We don't fully understand these mechanisms, but we have a rough idea of how the brain works and any specific hypotheses are completely falsifiable.
On top of that, it doesn't matter if there is explicit willpower circuitry. I think by most definitions my muscles don't have explicit energy circuitry, but it'll still become harder to balance if I've been running or lifting weights extensively. And the reason is that they're low on energy. For any particular definition of an aspect of 'energy' I can show you the precise biological mechanisms and give you objective measurements. It doesn't matter that the term 'energy' is overarching.
So, in terms of objective science and provable mechanisms: Bringing up an argument about whether things are predetermined has nothing to do with whether people make willpower-actions. It's a metaphysical question about whether the actions have meaning. Science does not care about meaning.
Basically, I understand you when you say that metaphysical willpower in the sense of free will is not a falsifiable thing. But that's not what these experiments are about! These experiments are about the objective side, the measurement side, the willpower-as-defined-by-people-with-heads-outside-of-their-navels, about whether people make the hard choice or the bad choice. As far as I can tell you're claiming the entire thing is trash just because they used the word willpower. These experiments are talking about a real thing, with partially-understood mechanisms, that are falsifiable in every way except for word choice.
The word 'willpower' is not a specific mechanism that's claimed to be in the black box. It's just a description of something the box does.
I agree that fatalism was an unnecessary distraction.
The point about energy vs. willpower is that you cannot point to any precise biological mechanisms for willpower, but the claim that willpower competes with cognitive processing for processing resources suggests that there is a precise biological mechanism, in the same way that the coexistence of aerobic and anaerobic energy processing suggests a precise biological mechanism.
The box doesn't yield measurements of willpower and cognitive processing. It yields measurements of memorizing a phone number and choosing to eat cake. We bind those things to our terms in order to generalize and create a model. Models are fine. But the problem is taking the model and saying the model actually maps to reality inside the brain. Even the authors of the paper in question hedge on this mapping with the words "seems like".
Models are unfalsifiable because they are models and we made them up. The only thing that is falsifiable are hypotheses about behavior derived from the model. My irritation is when models of the brain that we do not possess the tools to validate in the brain are taken as neurological truth. I can suggest other models that fit the experimental results, and there's no way to say whether mine are worse or better.
As far as I can tell, we agree about the utility of things outside of the box. But I'm not sure if we agree that the way the box works on the outside doesn't generate testable claims about how it works on the inside. As an example, if you write me a program that converts one list of numbers into another list of numbers and give me a binary (without the GPL...), it is actually not possible for me to tell what algorithm your program uses. I would have to disassemble it to figure things out.
Well, I don't think we're going to come to a consensus here. You see it as a problem that a model might not represent actual processes, with no way to prove one way or another. I see it as no problem because it's a non-mechanical label for whatever the real process is, and you can of course disprove it if it's inaccurate.
I'm replying to this part in a separate comment to avoid it distracting from the real discussion.
>Anything that is presently an untestable claim - ... - is not a scientific hypothesis.
Screw that. With the miniscule size of transistors in today's chips, it's untestable whether an i7 has an NSA backdoor. That doesn't make the theory unscientific. It's unscientific if I can't propose a buildable test at all, but it doesn't matter if the test can be built in 2013.
(If you don't like that example, replace it with something that involves reaching a Voyager probe. Completely impossible to do in 2013, completely possible to be scientific.)
Strangely enough, for me this is the real discussion. I think it speaks to the extent we are miscommunicating here.
> It's unscientific if I can't propose a buildable test at all, but it doesn't matter if the test can be built in 2013.
I agree that this is a better definition. I am unable to propose a buildable test that determines whether our brains have neurons dedicated specifically to cognitive processing and willpower and moreover that these functions draw separately from the same pool of processing resources. Perhaps neuroscience will get us there in the future.
