Our core research interest is the role of consciousness in decision making and voluntary action. We are also interested in various aspects at the intersection of neuroscience and ethics (sometimes termed neuroethics or neurolaw). Other area of interest include the laws that govern voluntary action and its perception, neural correlates of memory and multi-sensory integration.

The role of consciousness in decision making & voluntary action

To advance towards the long-term goal of deciphering the role of consciousness in decision making and voluntary action, we are working on answering the following interim questions:

  1. How is information accumulated in the brain during deliberate and a random decision making
  2. How do values, biases and random neural fluctuations – both conscious and unconscious – influence random decisions as well as deliberate decisions? 
  3. What steps make up the neural processes underlying decision making? Do these correspond to the stages of intuitive decision making: gathering information and deliberating, deciding, and then acting? 
  4. How can neuroscience contribute to the debates on free will and moral responsibility? 

Research into the role of consciousness in voluntary action selection generally focuses on random decisions. A typical example would be to instruct subjects to raise either their left or right hand with no reason, purpose or consequence. Such research showed that information exists in the brain about upcoming random-action before subject report having decided which action to carry out and when to act (e.g., Libet et al., 1983; Haggard & Eimer, 1999; Soon et al., 2008; Fried et al., 2011). However, such random decisions are not the ones humans typically agonize about, nor are they the types of decisions that "define us", and certainly have no moral consequences. Thus such random decisions seem to directly contribute little to the long-standing debates over free will and moral responsibility. 

Information accumulation for deliberate & random decisions

We have begun to answer the first question using intracortical recordings of local-field potentials and single-neuron activity in humans and monkeys (Maoz et al., 2012; Maoz et al., 2013). We played a matching pennies game with patients, who were implanted with intracortical electrodes for clinical purposes. And we found that we could increasingly well predict their actions (mainly from the supplementary motor area, SMA) while they were deliberating their deliberate decisions and waiting for the go signal to move (Maoz et al., 2013).


Real-time system in action

 A 2.5-minute video explaining a bit about our real-time action-prediction system
 The accuracy of the prediction rises with deliberation time
Most studies that compare information accumulation from the brain with conscious reports of having decided rely on the Libet paradigm. In this paradigm, subjects time the onset of their intention to move using an external clock. The time they report in such paradigms is typically named W time. However, it is still unclear to what extent this W time is a good measure of intention onset. We are therefore now studying W time for deliberate and random decisions

We are also now running an EEG study that specifically compares information accumulation in the brain for deliberate and random decisions (Mudrik, Maoz, et al., in preparation).

Influences on deliberate and random decisions

We showed that information about upcoming random decisions exists in the brain even before the decision alternatives are revealed and rational deliberation begins. We term this predeliberation activity bias activity. We found bias activity correlated with the side of the spatial selection of the animal as well as with the size of the reward. This bias activity was functionally dissociated from the choice activity during deliberation in the prefrontal cortex, and that it was also anatomically dissociated from choice activity in the basal ganglia.



Choice neurons (in blue) are anatomically dissociated from both types of bias neurons (in green and red) in the basal ganglia (right side of left panel), though not in the prefrontal cortex (left side or left panel). All three types of neurons are functionally dissociated (right panel) in the basal ganglia and prefrontal cortex.

We also developed a novel computational, neural circuit-model to describe these findings (Maoz et al., 2013). The model replicated our finding and made additional predictions that were borne out empirically. According to our interpretation of this data, the bias activity constitutes unconscious influences, stemming from slow, random fluctuations in neural activity before deliberation. 

Neuroscience, free will & moral responsibility

There have been claims that neuroscience proved that free will is an illusion (see an early debate on this in Libet and commentary, 1985). I relate to that in detail elsewhere in this siteBut, succinctly, I think that there are critical empirical and conceptual problems with the experiments on which this strong claim lies. And there are strong findings that seem to oppose this claim. So, while there are convincing arguments for and against certain interpretations of free will, they may not directly relate to neuroscientific results. Hence, given what we know at this point, making strong claims against the existence of free will based on neuroscientific findings is hasty.

There were also claims that as neuroscience proved free will to be an illusion, the concepts of moral and criminal responsibilities are also now defunct. In work in collaboration with Gideon Yaffe we discuss to what extent neuroscience is relevant (and irrelevant) to our notion of criminal responsibility.

Neuroscience of ethical decision-making


Image taken from here

Ethical questions, and in particular the trolley problem, have become a topic of interest for neuroscientists and psychologists. However, most of these studies simply ask subjects to imagine facing various ethical scenarios and ask them what they would do. The presentation is typically textual. We are testing the effect of more immersive environments on dilemmas like the trolley problem.

The laws that govern voluntary action

My research looks at modeling two- and three-dimensional voluntary movement as well as noise using analytic and simulation techniques. I showed that a common invariant of bodily movement, further attributed to more-general biological motion, may also result from noise in the motor system (Maoz et al., 2006). I gained theoretical knowledge of differential geometry to generalize invariants of planar motion to spatial motion, based on their affine-geometric interpretation. I also empirically demonstrated that the law of motion that we formulated holds using very accurate recordings and robust analysis of two- and three-dimensional movement (Pollick, Maoz, et al., 2009; Maoz et al., 2009). I also developed a method to accurately find high-order temporal derivatives of noisy signals (Maoz et al., 2009). I further tested the extent to which this invariant holds for the perception of two- and three-dimensional motion (Maoz & Flash, 2014). For this I generated a realistic, stereoscopic, three-dimensional image, that experimental subjects could control online and in real time.

Societal influences of neuroscience

The "my brain made me do it" defense is increasingly used in court now. But what does that sentence even mean? Analysis that Dr. Liad Mudrik and I conducted, tries to explain how this is related to a new type of closet dualism in modern neuroscience (Mudrik & Maoz, submitted)

Has neuroscience shown that free will is an illusion and moral and criminal responsibilities are defunct? I think that it has not at this point. Gideon Yaffe and I discuss to what degree neuroscience has contributed to criminal responsibility in the past and where it might make important contributions in the future (Maoz & Yaffe, forthcoming).

Like other scientists, we are worried about what has been termed the "crisis of confidence" in psychology and possibly in neuroscience. Consequently, we are developing methods to better analyze neuroscientific data. Such methods would yeI am investigating the potential for spurious decoding in fMRI, EEG, intracranial recordings, and so on. In particular, I want to know what are the signal-to-noise ratios that yield rigorous, interpretable, and reproducible data. I have also been developing data-driven methods to create more- reproducible analyses.     

Curriculum Vitae (CV)

A link to my full CV