Conference Agenda

In the context of conscious decision-making, the distinction between intention and foresight shapes how we judge actions in ethics and law. Consider a doctor administering a high-dose painkiller to a terminally ill patient, who dies as a consequence of the medication. Whether the doctor is criminally liable depends on whether they intended for the patient to die or just foresaw this consequence as a side effect of pain relief. To uncover whether the critical distinction between the mental states of intention and foresight is associated with distinct neural signatures, we recorded EEG while participants performed a task inspired by the game Frogger, where their actions varied in intentionality and expectedness (foresight).

We found that the feedback-related negativity (FRN) was stronger for intended outcomes, while the P3 increased for unexpected results—suggesting these signals track intention and foresight, respectively. However, neither ERP component uniquely reflected only one of those mental states. Nevertheless, effect-matched spatial filtering identified a distinct neural pattern that may better separate intention from foresight.

Hence, intending a specific consequence of an action and foreseeing that consequence may rely on different neural mechanisms. This lends empirical weight to the “doctrine of double effect”—which argues that foreseen but unintended harm may be morally permissible. We propose a follow-up study that explores the neural signature of causing harm to another person, comparing foresight of harm to intention to harm. This line of research demonstrates again how basic research in the neuroscience of consciousness has practical implications in ethics and law.

Is motor-cortical neuronal activity during movement preparation hard-wired or under conscious control? An intracranial study in humans.

After more than a century and a half of research, much about the functional role of motor cortex remains unknown. In particular, it is not well understood how much the neuronal activity leading up to movement is hard-wired versus under conscious, volitional control. Working with four human participants implanted with intracranial electrodes in motor and somatosensory cortex, our machine-learning classifier found spiking patterns that were highly predictive of upcoming movement in real time (~80% accuracy). We then investigated two research questions. First, to what extent can humans alter their motor-preparatory neuronal activity before a specific movement and still carry out that movement? Second, how similar is the pre-movement neuronal pattern for the same movement under different cognitive contexts?

Participants played a strategic game, where their objective was to flip as much slime as possible from a slow-filling bucket on a virtual “robot” opponent before it flipped the bucket on them. The robot was activated by the real-time classifier that indicated the participant’s intention to move. Hence, to win, participants needed to alter their preparatory neuronal activity. However, in multiple sessions over days and weeks, participants lost to the robot on more than 80% of the trials. Moreover, the classifier predicted movement just as well in a different setting, when participants scrolled through visual content in a self-paced manner. Our study thus provides evidence for hard-wired motor-preparatory neuronal activity in motor cortex rather than conscious control.

Understanding the temporal dynamics of intention formation is crucial for uncovering the neural basis of decision-making. Here, we leveraged the suggestibility of subjective reports to implicitly track the time course of intention formation. Forty-two participants were led to believe that their brain activity was monitored by a sophisticated AI which could predict their upcoming actions. In reality, random predictions were presented during their deliberation process, and participants reported whether these predictions were correct. Notably, predictions shown early in deliberation were reported as accurate more often than chance, with this effect diminishing as the decision moment approached, β=-0.084, SE=0.040, t(43.818)=-2.130, p=0.039. This pattern suggests that suggestibility decreases as the intention strengthens over time. The onset of intention, estimated using this method, aligned more closely with neural data compared to traditional explicit reporting methods. These findings highlight the utility of suggestibility as a tool for investigating the formation and time course of mental events and offer a promising avenue for studying the interplay between subjective experience and neural activity.

Mind wandering (MW) is a mental state process which has been correlated to prefrontal executive function and local theta synchronisation. Here we investigated whether transcranial magnetic stimulation (TMS) at a theta-frequency over the left dorsolateral prefrontal cortex (DLPFC) may causally enhance attention and executive control during a MW task. Using a block design, a cohort of human participants received rhythmic (7Hz), arrhythmic, and sham TMS while performing a finger-tapping task (FTRSGT) engaging executive control. In two counterbalanced TMS-EEG sessions, participants reported their attentional state via thought probes and objective measures of approximate entropy (AE) and behavioural variability (BV), indexing executive control and temporal attention, respectively.

Behavioural measures supported task validity: AE increased during on-task states while BV showed the opposite pattern, thereby linking subjective reports to objective measures of attention. Results show that rhythmic TMS induced a global increase in scalp theta power and enhanced intra-trial coherence (ITC) of theta, alpha and beta rhythms. EEG electrodes near the DLPFC exhibited significantly higher theta power and ITC during and immediately following stimulation, suggesting successful local theta-entrainment. Group-level analysis revealed that at the group level participants reported higher levels of attention (though-probes) during rhythmic theta-TMS blocks compared to arrhythmic, suggesting enhanced attention under theta entrainment. At this stage, no significant differences emerged for BV, AE or the remaining thought-probes. Future analysis will determine the correlations between subjective and objective MW reports across conditions, as well as dynamic changes in BV and AE in relation to subjective attention and EEG-modulated activity.

Mind-wandering (MW) is a mental state occupying 30–50% of waking time, where attention drifts away from the task at hand. This phenomenon has been linked to sustained attention and executive function (EF), but this relationship remains unclear. Here we hypothesized that modulating neural oscillations in the left dorsolateral prefrontal cortex (DLPFC) with theta-frequency transcranial alternating current stimulation (tACS, 7 Hz) would enhance sustained attention and reduce MW. To this end, we delivered tACS to the DLPFC and assessed temporal attention and cognitive control using a finger-tapping task (FT-RSGT). We expected increases in approximate entropy (AE, a proxy for cognitive control) and reductions in behavioral variability (BV, a proxy for attention) during tACS compared to SHAM and transcranial random noise stimulation (tRNS) conditions. Additionally, we expected participants to report being more "on-task" (via thought probes) and to exhibit larger pupil sizes, indicating increased arousal.

Our findings show significant BV reductions and increased pupil size when participants were on-task, with opposite patterns when off-task. However, AE did not differ significantly between on-task and off-task states, nor did AE or BV vary across stimulation conditions. Unexpectedly, tACS reduced attention scores compared to tRNS and SHAM. Pupil diameter increased during tACS and decreased during tRNS, suggesting theta stimulation uniquely influences arousal. Theta and alpha power increased post-tACS, and higher theta power was linked to off-task states. These results suggest that while theta DLPFC stimulation affects arousal, its impact on EF and MW requires further investigation, focusing on interindividual variability in attentional responses.

Accurate assessment of residual auditory function in patients with severe acute brain injuries, from acute coma to prolonged states, is crucial for clinical decision-making. Recent studies have demonstrated the feasibility of simultaneous recording of auditory evoked responses at multiple levels for comprehensive assessment of the auditory hierarchy. However, an approach that achieves both high sensitivity and reliability in detecting individual-level responses has yet to be established. We propose a frequency-tagged roving paradigm employing two amplitude-modulated tones (carrier frequencies 220/440 Hz modulated at 40/80 Hz) as both standard and deviant stimuli. This design enables concurrent recording of brainstem frequency-following responses (FFRs), auditory steady-state responses (ASSRs), and cortical event-related potentials (ERPs). Using an optimized two-channel recording setup (Fz and Cz), we tested this paradigm on 30 healthy participants using a roving design to minimize adaptation effects while maintaining efficient assessment of change detection. Our results demonstrate 100% sensitivity in eliciting high-frequency FFRs and gamma-band ASSRs, as verified through spectral analysis with permutation tests. Machine learning classification successfully distinguished stimulus conditions from resting state in the N1-P2 complex across all participants. For mismatch pairs, ascending transitions predominantly elicited N1-P2 responses (30/30 participants), while descending transitions evoked MMN (28/30 participants). Systematic analysis of recording duration revealed that with 30 minutes of recording, the paradigm achieved 93.33% sensitivity across all components. These findings validate our frequency-tagged roving paradigm as an efficient approach for comprehensive assessment of the auditory hierarchy, offering potential advantages in both research and clinical applications where rapid yet reliable evaluation is desired.