A good philosophy has considerable reach. By this, I mean that a certain philosophy may have been expressed in some esoteric and academic area, but it can be applied in other specific disciplines as well as in broader contexts. Notice, this is more than just saying ‘a good philosophy is always true’. Instead, this is saying that a good philosophy is true, relevant and useful. (As a point of contrast, my personal philosophy about coffee is still true when thinking about data cleansing in operational data for a utilities company, but it is not relevant or useful.) This idea, then, has inspired me to take Sean Carroll’s articulation of the functioning of science and apply it to freewill (Greer Heard Forum, 2014). Carroll brings it up talking specifically about cosmology, but I think it’s relevant here.
Carroll’s view is that science isn’t as empiricism-led as we’ve been lead to believe. Instead, we create models that propose to explain a certain element of reality and then compare that model to the observations. I would argue that this is implicitly done in hypothetico deductive approaches in the formation of a hypothesis. Creating a hypothesis implies an expected model of the world. It is an overtly tentative model, but a model nonetheless.
Where I said science is less empirically-led than initially assumed, I do not mean to imply that science is not empirically-led at all. The creation of models if often an empirically-led conjecture. The models and hypotheses science tests are not just any old ideas, they are ones that have some level of empirical or rational justification. But that initial justification should be subservient to the conclusion of the research itself, and that should be heavily empirical. (I mean only to point out there is a lot more going on than just empiricism.)
Where more than one model is created to account for the same element of reality, either by overlap or by answering the same questions, these models can then be compared in their ability to account for the data. In a previous post, where I discuss the reliability of the human brain, I followed this approach: I compared the theistic argument that every element of the human mind ― the brain, the tools like logic ― were all infallibly designed with the natural argument that the mind is a product of evolution. I then compared these two models in their ability to account for the data of the imperfect brain.
The approach Carroll describes has a lot in common with Deutsch’s ‘fallibilism’ approach to science, which is to offer empirically-led conjecture and then assume the conclusion is fallible and so remain critical of it: the easier it is to be impartially critical of a conjecture, the worse that conjecture is (Deutsch, 2011).
It is that approach I want to apply to two models of human behaviour: freewill and determinism. I am excluding ‘compatibilism’ from this because so far as I can see, compatibilism can be summarised as ‘you are free to make your determined choice’. I am open to someone constructing a model of compatibilism for me, explaining how it differs from both freewill and determinism. But, for now, that won’t be included here.
Both freewill and determinism do a perfectly fine job of accounting for all of our actions. We have to dig a little deeper to look for somewhere the data can reasonably be expected to be different under both models: what are the intelligible differences between freewill and determinism? To answer that we have to make a few assumptions. Here is my first assumption: for their to be agency and thus be freewill, still with some degrees of freedom from solely linear and mechanical processes, such agency must originate in consciousness. It cannot originate in the brain, because the brain is a material object defined physical laws; there would be no degrees of freedom there.
This, as it happens, is one of my objections to freewill. Originating in consciousness ― as opposed to entering into consciousness from functions of the material brain ― makes supernatural claims about consciousness as a separate entity. In Occam’s Razor terms, it vastly multiplies its entities; it posits an entirely new type of existence, for which we have no tangible evidence. It also just pushes the question back one: now that we have this agency with degrees of freedom from the physical processes of the brain, what else influences this agency? It doesn’t answer the question of what makes that agency want to go to the cinema this evening. But that is not the approach I am taking in this post.
By contrast, determinism emerges in the brain. I say ‘the brain’ and not the intestine or a WiFi router, because of empirical data that tied certain decision making and personality traits to the brain, and the brain is a physical structure bound by deterministic processes.
I’ve been careful not to paint myself into a corner here. I have described freewill as having ‘degrees of freedom’ from the physical processes, and not having complete freedom from it. I have worded it this way to allow for the physical processes to be an influence on any brute agency that exists in consciousness, but not the only one. Therefore, evidence that the brain controls certain elements of personality and decision making is not exclusively evidence for determinism: this allows for hunger and thirst to influence the freewill agent as well.
With this is mind, the challenge continues to find a set of data that could reasonably be expected to be different under this conception of freewill versus determinism. And here is my conjecture:
If one’s will originates in consciousness, and the agency in consciousness alters the brain to realise such a will, then a person should be able to be aware of a decision they have made before it becomes apparent in the brain. By contrast, if one’s will is defined by determinism and thus a decision makes it into consciousness via the brain, then decisions should be apparent in the brain either simultaneously with awareness or before awareness of the decision arises in consciousness.
That, I suspect, is not a controversial take on what the differences between the two models would be in the only area I can see there would be an empirical difference. Before you read on, I implore you to check whether you agree and think of any other areas there may be an empirical difference between the two models.
But now the question is which of the two models presented account for extant data. The facts are these: an experiment by Libet and Benjamin (1985) asked subjects to make a small deliberate action at a random time. The subjects were asked to report the precise time they were aware of their intention to move (i.e. when the idea existed in their consciousness). Simultaneously, Libet was monitoring their material brains, specifically an electrophysiological phenomena known as the “readiness potential”. The readiness potential was evident in the brain of the subjects approximately 0.5 of a second before the intention to move was evident in the mind of the subject; the material had defined the internal.
Criticism of this experiment does exist. The philosopher Daniel Dennette noted the woolliness of the experience of intention makes any precise time-value meaningless and that the subjects must be focussed on their internal realities so cannot give fair attention to the clock they are reading. Libet’s 0.5 of a second delay could be a systematic error.
It is conceivable that Libet’s experiment only measured some sort of preparatory work on behalf of the brain, and that the readiness potential doesn’t actually reflect a decision moving into consciousness, but merely the context in which a decision has been made. Soon et al (2008) published their paper that does away with that criticism. Soon et al’s experiment was to ask subjects to press one of two buttons, one with the right index finger or the other with their left index finger. Technology has moved on since Libet’s 1985 experiment, and Soon et al were able to observe the brain in high resolution using a function magnetic resonance imaging (fMRI) machine. Their aims were to replicate results suggesting the a brain state is a precursor to action and to see whether the activity of the brain could be used to predict whether the subject would press the right or left button, ahead of the subject acting.
Investigating specifically the primary motor cortex and supplementary motor area, Soon et al were able to see fingerprints of readings that allowed accurate prediction of a subject’s decision up to 7 seconds before it happened; from looking at the material brain, science is starting to read the mind. These data are better explained by the decision emerging in the brain and moving into the consciousness, not vice versa. That is a characteristic detailed in the determinism model, not the freewill model.