----------

Engrams as Vehicles:

Neural representations are central posits in neuroscience. These representations are individuable patterns of neural activity that perform some function or computation for the brain. In theorizing about representations, and no less for neural representations, it is helpful to distinguish the representational vehicle from its content. Doing so allows us to theorize about what is doing the representing and how. For neural representations, representational content might be some environmental feature that is tracked by an organism, such as the orientation of a line, the trajectory of a moving limb, or fruit in some foraging patch. The vehicle of representation is some re-identifiable physical pattern or structure, such as firing rates in a single neuron, molecular mechanisms of protein modification, or co-activity patterns in a neuronal ensemble. A major goal of neuroscientific inquiry is to individuate vehicles of neural representations to understand how they represent what they do.

Individuating representations from non-representations in practice requires some conception of which properties in the brain carry semantic information (i.e., which properties are the signal (rather than noise)). Hypotheses range from cellular and molecular processes to synaptic structures and from spikes to manifolds. Hypotheses are stronger when it is the case that evidence can weigh both in favor of the hypotheses and against it. For representation discovery, when researchers do not find a representation, the objection is always available that researchers were not using the right tools or looking in the right place. Our project, first, is to develop an account of evidence for vehicles (also called ‘engrams’) such that researchers may reasonably argue that there are no engrams in a particular area or that there are no such representations at all. For example, we might think that there are no engrams in the primary somatosensory cortex (S1) when no such engrams are found using various tools. Strengthening such claims requires evidence that the right tools were used and the right structures were observed.

For engrams, the right structures or processes are hypothetically those that persist for the length of the representation. One challenge for finding engrams is that much of the brain changes over the course of representing. Interestingly, however, stable structures can be found for task conditions in which there is no content. For example, mice can be trained to expect a reward after a pu` of air is sensed via their whiskers. While no stable representation was found for such learning in primary somatosensory cortex, when all of the whiskers were removed, a stable representation formed and persisted for multiple days. Such representations of omission are interesting in that representational content seems to be an absence of sensory stimuli. This raises a number of philosophical questions about whether such phenomena should be considered representations at all.

How do we perceive our body as ours? In the philosophy of mind and cognitive neuroscience, the sense of body ownership (SBO) – the pre-reflective feeling that one’s body belongs to oneself – has emerged as a central topic of investigation. A prominent hypothesis suggests that SBO is not merely a product of multisensory integration but is fundamentally rooted in affectivity (Vignemont, 2018). While research has primarily focused on vision and touch (Chouinard & Stewart 2020; Crucianelli et al. 2018; Della Longa et al. 2022), we propose to expand the investigation to an affectively laden sensory modality: olfaction. Despite growing philosophical and empirical interest in olfaction and embodiment (Cornelio et al. 2020; Young 2023), the specific conceptual and empirical relationship between olfactory experiences and the affective modulation of SBO remains entirely unexplored. 

This project aims to fill this gap by investigating the link between affectively laden olfactory experiences (pleasant or unpleasant ones) and the modulation of SBO through two experiments, one involving the classical paradigm of the Rubber Hand Illusion (RHI; Botvinik & Cohen 1998), and the other a novel virtual reality paradigm (Marucci et al. 2024). Each will involve a 2×3 design with Embodiment (present, absent) and Odor (neutral, pleasant, unpleasant) as within-subject factors, resulting in six conditions per experiment. 

Color Perception Beyond the Human-Visible Spectrum