Neurons and free will

In another post in this blog I have described the four theories used by philosophers to try and solve the problem of human mind: What is intelligence? What is consciousness? What is free will? Are we actually free, or are we determined, just like meat machines?

At the end of last year, Javier Pérez Castells published a book where he addresses some of these issues from a scientific and philosophical point of view. Its title (in Spanish) is the same as the title of this post. In particular, chapter 8 of the book describes some of the models with which various scientists and philosophers have tried to explain how we make decisions more complex than those studied by the experiments performed by Libet, Fried and Haynes, which don’t go much further that pressing a button or raising a hand. These models are called two-stage, because they try to explain our decisions assuming that they are made in two phases: the first, more or less random, in which the brain generates the available alternatives, followed by a second phase, when we actually make a decision, after weighing those alternatives.

Two-stage models date back from the late nineteenth century. The first one was proposed by the American philosopher William James in 1884. In 1906, the French mathematician Henri Poincaré proposed another. Soon after, quantum mechanics and its intrinsically random interpretations introduced possible mechanisms that could act during the first phase of the process. In the last decades, several of these models have been developed. Let us look at a few.

• John Carew Eccles, an Australian neurophysiologist who in 1963 won the Nobel Prize in Physiology and Medicine for his work on synapses, takes a clearly dualistic position and thinks that certain brain neurons are subject to the influence of a spiritual principle that controls their joint activation.
• David Hodgson, Australian philosopher and judge, uses non-local quantum indeterminacy in the first phase of his two-stage model, but for the second phase he resorts to another causal source (the conscious mind) that would choose among the available alternatives without being determined by the preceding states.
• Daniel Dennett, American philosopher, takes a compatibilist determinist stance that redefines freedom and takes away most of the meaning of this term. In his model, proposed to explain complex decisions, in the first stage a set of alternatives is generated unconsciously and a few are selected, among which the agent will take his final decision “freely.”
• Peter Tse, American neuroscientist, argues that our current decisions are influenced by previous ones, for the brain modifies the weights of synaptic connections. His model (criterial causality) is an intermediate between determinism and pure indeterminism, but like Dennett’s it does not seem to leave much room for true freedom.
• Brigitte Falkenburg, German philosopher, proposes a dualistic model like Eccles, based on the increase in entropy inherent to irreversible processes.
• John Searle, the American philosopher who proposed the mental experiment of the Chinese room against the Turing test, is working on a two-stage model of consciousness and free will, in which the path from the neurobiological level to the conscious level is not continuous, but contains gaps where a free decision-making capacity can take place. According to Searle, if free will were not real, there are no evolutionary reasons why so many resources, energy and effort would be used to simulate it.

John Searle

Note that most of the previous theories are due to philosophers, rather than neuroscientists. This is understandable, because this question is more philosophical than scientific, and contradicts the idea, spread by the media and some popular books, that neuroscience has shown, or is about to prove, that we are meat machines, determined by our neurons, and that human freedom is just a delusion. Note also that the authors of the models distribute evenly among the underlying philosophical theories alluded to, at the beginning of this post.


The same post in Spanish


Thematic thread on Natural and Artificial Intelligence: Preceding Next

Manuel Alfonseca