Coronavirus (Covid-19): Latest updates and information
Skip to main content Skip to navigation

The challenge model of free will

The challenge model of free will replicates the free will that humans have. Unlike other approaches it treats free will as a scientific phenomenon, with experimental facts. [Hadley (2018)] The model can be tested against the facts and makes some predictions.

The phenomenon of free will is that humans have a perception that our actions are not predetermined. It is a universal cross cultural phenomenon [Sarkissian (2010)]. Studies to refine the common perception depend strongly on the context [Nichols, S (2011)]. Obviously this applies to a subset actions: having hiccups or a headache would not be normally be included. The phenomenon does not provide any evidence that our actions are technically indeterministic (not predetermined). For complicated objects like our brain, operating in a rich and variable environment there could not possibly evidence of indeterminism.

The philosophical term free will is commonly defined as “Could have done otherwise” [see for example Kane, R ed. (2002)]. While this is a statement about actions, the debate is almost exclusively about human (more generally an agent) decision making. The causal mechanism from a decision in the brain to the subsequent action (typically speech or movement of a limb) is relatively free from paradox.

The challenge model explains the perception that our decisions are not predetermined.

The Construction

The model is based on a utility agent [Russell, SJ and Norvig P (2013)] a goal seeking, decision-making agent that responds to external stimuli, and internal states. It has some analytical capability that does not need to be optimal in any sense. The agent has several goals to satisfy, with different importance or weights; for example satisfying hunger by eating. The word utility refers to internal more abstract goals such as happiness. Such an agent is commonly used in a range of disciplines including economics sociology game theory.


The novel feature of the challenge model is that the agent responds to a challenge, it has a goal that is satisfied when it responds to a challenge. That goal is called independence. The challenges can be external or self-generated. The agent has many goals to satisfy so the response to a challenge is not necessarily predictable, but it does change the pattern of decision making. See Figure 1.

A utility agent with free will

Figure 1. A utility agent with an independence goal, that responds to a challenge.


The challenge model shows free will in a number of ways:

  • When the agent is about to make a decision, or has just made one, we can ask “Could you do otherwise?” this will influence the behaviour and may reverse a decision.
  • The agent itself can generate similar challenges. For any decision it can challenge itself “Could I do otherwise?” and either possibility could result depending on the strength of the challenge and other competing goals.
  • We can apply a highly discerning test where the agent is challenged to do something very unusual, of no obvious value, perhaps even mildly harmful. Such as “Could you write with your left hand?” or “could you put your hand near the flame?”. Given the challenge an action changes from being highly unlikely to probable.
  • The agent itself can apply the highly discerning test.
  • The agent can develop a history of decision making showing clearly that for each decision the challenge revealed alternative possible outcomes.
  • If the agent was capable of abstract reasoning, it would describe its decisions as not predetermined. That it had free will.


The model leads to predictions (which is highly unusual for free will models and theories). An agent that responds to challenges will be perceived to have free will. An agent that does not respond to challenges will be perceived not to have free will. Fictional and cartoon characters can be used to explore the appearance for free will because the visual appearance and the behaviour can be controlled independently. Bender in Futurama is visually a robot, but makes decisions like a human. By contrast Spock in Star Trek, always acts logically. Given a logically superior course of actions he would always take it. Spock would not respond to a challenge test.

There are cases of false attributions of free will, for example a society may attribute personality agency and free will to aspects of the weather or a volcano. They perceive a correlation between their actions and the subsequent behaviour of the weather. Essentially they do a challenge test and judge a positive correlation.



Hadley, Mark J. (2018) A deterministic model of the free will phenomenon. Journal of Consciousness Exploration & Research, 9 (1).

Kane, R ed. (2002) The Oxford Handbook of free will. Oxford University Press, Oxford.

Langer, E J (1975) The illusion of control, Journal of Personality and Social Psychology Vol 32, 311. 

Nichols, S (2011) Experimental Philosophy and the Problem of free will, Science. Vol 331, 1401.

Russell, SJ and Norvig P (2013) Artificial intelligence : A modern approach. Prentice Hall, Englewood Cliffs N.J.

Sarkissian, H et al. (2010) Is Belief in free will a Cultural Universal?, Mind and Language. Vol 25, 346.