DMT micro-phenomenology

The investigators used a single-blind, between-subjects, placebo-controlled design with two testing days separated by at least two weeks. On each day participants completed two IV dosing sessions in counterbalanced order: one with DMT (described in the text as "20mL DMT fumarate dissolved in 10mL sterile saline") and one with placebo (10mL sterile saline). During dosing participants lay blindfolded with eyes closed in an fMRI scanner and concurrent EEG and fMRI were recorded for 8 minutes before and 20 minutes after administration; the present paper reports on the experiential data associated with the second session of each day, for which live experience sampling and subsequent micro-phenomenological interviews were performed. Experience sampling occurred every minute during the 28-minute recording period; recorded prompts asked participants to rate the intensity of the drug effect on a 0–10 scale and to speak their rating aloud for researchers to note. Shortly after returning to baseline, a trained interviewer conducted a micro-phenomenological interview about that session only. Interviews used the experience-sampling timeline to anchor reported phenomena to specific moments and emphasised content-free, iterative prompts to elicit fine-grained, temporally specific descriptions of phenomenal character. For analysis the researchers applied the micro-phenomenological method in three stages: preprocessing, category extraction, and category-dynamics analysis. Transcripts were trimmed of material unrelated to acute phenomenal character, ordered to reflect the subjective timeline, and segmented into descriptive units (descriptemes). These units were coded bottom-up into phenomenological categories, similar categories were merged and hierarchies of superordinate categories were formed. Each interview was further segmented into phenomenological phases using linguistic temporal markers; phase matrices were created to indicate which categories were active in each phase. The investigators compared phase matrices across participants using informal visual searches to detect repeating patterns and performed analytic quantification where patterns emerged. Across the dataset 127 phenomenological categories were extracted. The dynamic analysis excluded two participants for incomplete data, leaving 21 phase matrices for temporal patterning analysis.

The micro-phenomenological analysis yielded a rich taxonomy of structural dimensions that together define an emergent DMT ‘‘landscape.’' Major, repeatedly observed dimensions included multisensory engagement, semantic complexity, amodal feelings (including felt space and felt meanings), spatial characteristics, and multiple configurations of self/world relation. Across the 23 interviews (presences were reported in 15 of these), the visual, auditory and bodily senses were frequently involved; visual descriptions varied on brightness, colour, shade and saturation, and auditory descriptions varied in pitch, timbre, volume and musicality. Bodily phenomena included tactile pressure, changes in position, and temperature sensations, and tactile features from the scanner environment were sometimes incorporated into the emergent landscape. Semantic complexity distinguished semantically simple phenomena (elemental shapes, colours) from semantically complex forms (recognisable objects, architecture, or beings). Amodal feelings encompassed experiences not reducible to a single sensory register, notably a felt sense of space and felt meanings such as the felt reality or familiarity of elements in the landscape. Self/world configuration dimensions included perceptual position (inside, outside, liminal), internalisation (landscape experienced as internal or external to the self), the clarity of self/world distinction (distinct, fused), identification (world elements that resembled self), interactivity (direct or indirect agency over landscape content), and receptive modes (acceptance/letting go versus resistance; attentional engagement versus non-engagement). Phenomena were also described as being in flux: categories captured persistence versus ending, intensifying versus dampening, and modes of encounter (emergence, noticing, revelation). Perceived presences were multifaceted: they could be visually, auditorily, tactilely or amodally perceived (felt presence), vary in semantic complexity from abstract to richly figurative, and engage in social modes such as observing, communicating, affecting, guiding or manipulating the participant. Temporal dynamics were assessed using phase matrices for 21 participants (two excluded). The average number of phenomenological phases per experience was eight (range five to thirteen). Sensory-effects sequencing showed nine distinct trajectories through a three-dimensional state space of bodily, visual and auditory involvement; participants made an average of 2.85 transitions (range 1–3). A general hierarchical tendency emerged: bodily effects most commonly arose in the first transition, visual effects most often in the second, and auditory effects in the third when present. For perceived presences the analysis found that, with one exception, 3D spatial characteristics were not described unless bodily effects plus at least one other sensory modality had already been reported or were concurrent; further, presences did not emerge unless multisensory effects (including bodily) and 3D spatial structure were in place. Mapping these dependencies produced eight distinct trajectories with three possible terminal points; participants made an average of 1.8 transitions (range 1–3) through this state space. Two types of bodily phenomena contributed to multisensory effects: emergent bodily sensations (reported by all 21 participants included in this stage) and complete loss of bodily awareness (reported by nine). The semantic complexity of the wider visual field did not predict the emergence of presences: presences could be semantically complex even when the rest of the visual field was semantically simple, and presences could arise without being visually represented at all.

Sanders and colleagues interpret these results as demonstrating that DMT-induced immersion and perceived presences are structured, multilayered phenomena that can be mapped in terms of distinct phenomenological dimensions and consistent temporal dependencies. They emphasise that a structural, temporally precise approach extends beyond standard constructs like ‘‘ego-dissolution’’ or ‘‘mystical experience’’ by specifying how multisensory engagement, spatial form, and self/world configurations interact to produce immersive states and encounters with presences. One robust pattern was hierarchical: lower-level bodily and multisensory effects commonly preceded the construction of three-dimensional environments and, in turn, the emergence of perceived presences, suggesting higher-order phenomena may require integration of more basic sensory scaffolding. The authors situate their findings in relation to conceptual work on immersion from virtual reality research, proposing that DMT effects align most closely with ‘‘system immersion’’ (the degree to which perceptual inputs are occluded and reorganised), while also implicating narrative and challenge dimensions when encounters take on social or agentive structure. They argue that the observed collapse of self/world boundaries provides specific structural pathways towards non-dual or unity experiences, produced either via synaesthetic fusion or by loss of conventional spatial-relational distinctions. With regard to perceived presences, Sanders notes these experiences are heterogeneous: presences may be sensed through modality-specific channels or via amodal felt presence, vary in semantic depth, and engage the participant in a range of social modes. The authors caution that presence phenomena are not unique to DMT and that comparative neurophenomenological work is needed to identify etiology-specific phenomenological and neural signatures (for example comparing DMT to Parkinsonian presence phenomena, meditative states, or induced illusions). Limitations acknowledged by the investigators include the laboratory context (scanner environment and experience-sampling may have influenced phenomenology), residual reporting biases despite the micro-phenomenological method (selective recall or confabulation cannot be ruled out), and the qualitative nature of the analytic process which may limit replicability. They recommend triangulation with neuroimaging (EEG/fMRI) to test whether recurrent phenomenological categories map onto specific neural dynamics, finer temporal anchoring (for example using button presses) to improve alignment with physiological data, and extension to different populations and compounds to expand and refine the taxonomy. The authors propose the micro-phenomenological approach as a valuable tool for integrating first-person description with third-person measures in neurophenomenology.

The study concludes that a structural, temporally precise micro-phenomenological approach can reveal a detailed taxonomy and temporal organisation of DMT-induced immersive experiences and perceived presences. Sanders and colleagues assert that these categories go beyond traditional psychedelic constructs (such as ego-dissolution or mystical experience) by specifying dimensions like semantic complexity and amodal feeling, and by showing how sensory modalities and spatial properties interact to generate immersive states. They suggest this framework can inform future neurophenomenological work, support comparative studies across altered states, and improve the validity and granularity of phenomenological descriptions in psychedelic research.