NOTE-999: Lenggenhager, Tadi, Metzinger & Blanke — Video ergo sum

Contribution

It moved the rubber hand illusion from a body part to the whole body. Before it, experiments manipulated the ownership and felt location of a hand, which the authors call “a part-to-whole relationship” that “did not experimentally manipulate selfhood per se” (p. 1097). Here a video of one’s own back, projected 2 m ahead in a head-mounted display and stroked in time with one’s real back, was rated as one’s body and pulled felt self-location toward it. It gave the study of bodily self-consciousness a laboratory procedure and two measures, a questionnaire and a walking-back drift, that later full-body illusion work builds on.

Key insight

Where I feel I am, and which body I feel is mine, can be moved by the same temporal match of seen and felt touch that moves ownership of a rubber hand, but only part of the way, and only for something shaped like a body. The self did not leave the body: it was located a little toward a seen body, without any change in the place it seemed to see from.

Assumptions

• Bodily self-consciousness is the route to a model of the self. It is defined as “the nonconceptual and prereflective processing and representation of body-related information” (p. 1096), citing Gallagher, Metzinger, Damasio and Legrand.
• Selfhood is global ownership. The “fundamental sense of selfhood” is “experienced as the transparent content of a single, coherent whole-body representation, rather than as multiple representations of separate body parts”, and “whole-body representations or global ownership are directly associated with the sense of selfhood” (p. 1097). Both are cited to Metzinger’s Being No One (ref. 2); neither is tested.
• The drift measures felt self-location. Blindfolded participants were walked backwards about 20 cm in small steps, then asked to walk back to where they had stood; the error toward the seen body is the measure (SOM pp. 1–2). That this error reports where the self was felt to be, and not something else, is assumed.
• The questionnaire measures self-attribution. Seven items adapted from Botvinick and Cohen’s rubber hand questionnaire, scored −3 to 3 (SOM p. 2).
• Out-of-body experience and heautoscopy are the clinical models. The design is “based on clinical data in neurological patients with out-of-body experiences”, in which “the spatial unity between self and body may be disrupted” (p. 1097).

Key results

• Setup (p. 1097; SOM pp. 1–2). Participants stood with their back to a camera 2 m behind them, saw the live image in stereoscopic 3D in a head-mounted display, heard white noise, and had their back stroked with a pen for 60 s. Study I: n = 14 (9 women, mean age 23.1). Study II: an independent n = 14 (7 women). Asynchrony in study I replayed the video recorded in the training session; in study II it was a constant 200 ms video delay.
• Study I, self-location (p. 1097; SOM p. 3). Drift toward the seen body was 24.1 ± 9.0 cm (mean ± SEM) after synchronous stroking, different from the starting position (p = 0.02, t = 2.67), and 12.5 ± 8.5 cm after asynchronous stroking (p = 0.17). The paired difference was significant (p = 0.01, t = 2.92). There was no left–right drift. (The printed text gives the asynchronous figure as “12. ± 8.5 cm”; the SOM gives 12.5.)
• Study I, self-attribution (p. 1097; SOM p. 3). After synchronous stroking Q1 (touch felt where the seen body was touched), Q2 (the touch was caused by the pen on the seen body) and Q3 (the seen body felt like one’s body) were positive: 2.6, 2.2 and 2.1 on the −3 to 3 scale. Every other item, and every item after asynchronous stroking, was negative. “None of the participants reported sensations of overt disembodiment or a change in visuospatial perspective”; some reported “weirdness” or “strangeness”, and “some found the experiment irritating” (p. 1097).
• Study II, self-location (pp. 1097–1098, Fig. 3; SOM p. 4). Relative to a motor-control condition (displacement with no prior visual input or stroking), synchronous stroking gave drifts of 17.1 ± 6.1 cm toward one’s own filmed body (p = 0.02) and 25.4 ± 8.4 cm toward a mannequin (p = 0.01), but 14.3 ± 7.3 cm toward a white rectangular object, not significant (p = 0.07). An empty room gave 1.12 cm. No asynchronous condition gave a significant drift. The Synchrony × Condition interaction was significant (p = 0.04). Only the mannequin differed between synchronous and asynchronous stroking (p = 0.0004); for one’s own body that difference was not significant. In synchronous stroking only the mannequin and the object differed (p = 0.04).
• Study II, self-attribution (p. 1098; SOM pp. 4–5, table S1). Q1 and Q2 were positive after synchronous stroking for all three, the object included. Q3 was positive for one’s own body (2.1) and the mannequin (0.6 ± 0.6) and negative for the object (−0.4), where synchronous and asynchronous did not differ (t = 1.55). So “Q3 is important to evaluate self-identification …, whereas the first two questions seem more related to the feeling and location of touch” (p. 1098).
• Smaller synchrony effects in study II (p. 1098; SOM p. 5) are put down to the more predictable 200 ms delay, which may have made seen and felt touch seem more related in the asynchronous condition.
• Interpretation (p. 1098). Self-attribution and localization of the entire body “rely, at least partly, on similar visual-somatosensory integrative mechanisms to those of body parts”. The weaker drift and the absent Q3 for the object “speak in favor of bottom-up mechanisms as well as cognitive constraint, rejecting a pure Bayesian account” (the account attributed to Armel and Ramachandran, that correlation suffices). “Bodily self-consciousness and selfhood can be dissociated from one’s physical body position.”
• Not an out-of-body experience (p. 1099). Since there was no disembodiment and no change in perspective, “we have induced only some aspects of out-of-body experiences or rather the closely related experience of heautoscopy”. Full out-of-body experiences need “other mechanisms in addition to conflicting visual-somatosensory information, such as visual-vestibular disintegration”.
• Brain (p. 1099). “We speculate that humans’ daily experience of an embodied self and selfhood, as well as the illusion reported here, relies on brain mechanisms at the temporoparietal junction.” This rests on cited lesion, stimulation and imagery studies; nothing was imaged here.
• The e-letters (June 2008). Meyer argues that “self-localization depends foremost on the visual perception of the application of somatosensory stimuli”, since in this study the self went where the stroke was seen, not to the camera’s viewpoint. Ehrsson and Petkova reply that the rubber hand illusion fails with a hand rotated or more than 30–40 cm away, that they could not induce a whole-body illusion with a mannequin a few metres away “in 10 naïve participants”, and propose that participants were recognising “a video reconstruction of their own body”, as one recognises oneself on a shop’s CCTV monitor. They add that the questions “are somewhat ambiguous when applied to the virtual reality set-up”, that the study does not say whether participants “ever stopped feeling the touch on their on real back”, and that the drift for one’s own body hardly differed from that for the object. Blanke, Metzinger and Lenggenhager reply that under asynchronous stroking participants seemed to locate the self where the touch was felt, that “visual dominance is an important, but not exclusive, mechanism”, and that a condition with no stroking at all, never tested, is needed to know where participants located themselves to begin with.

Claims

id	claim	strength	support
C1	Synchronous visuo-tactile stroking of a body seen 2 m ahead shifts felt self-location toward it	moderate	study I (n = 14, p = 0.02; synchronous against asynchronous p = 0.01); study II own body and mannequin against motor control; effects of 17–25 cm, about a tenth of the distance
C2	It also induces self-identification with the seen body	moderate	Q3 in study I and study II; the mannequin’s Q3 is 0.6 on a −3 to 3 scale
C3	Body-likeness constrains the illusion: a non-bodily object gives a weaker drift and no self-identification	weak	object drift p = 0.07 against mannequin p = 0.01, a significant/non-significant contrast; synchronous mannequin against object p = 0.04; own body against object not shown to differ
C4	The illusion is partial: no disembodiment and no change in visual perspective	moderate as a report	absence of such reports in study I; how it was asked is not stated
C5	Global self-attribution and self-location rely partly on the mechanisms of body-part ownership	weak	inference from the procedure’s analogy with the rubber hand illusion
C6	A pure correlation (Bayesian) account of self-attribution is rejected	weak	rests on C3
C7	Full out-of-body experiences need further mechanisms, such as visual-vestibular conflict	weak; conjecture	absence of disembodiment here; clinical data cited
C8	Embodied selfhood and this illusion rely on the temporo-parietal junction	speculation	cited lesion, stimulation and imagery studies; no imaging here
C9	The sense of selfhood is the transparent content of a whole-body representation, and global ownership is directly associated with it	asserted	Metzinger 2003, cited

Concepts

• Bodily self-consciousness — “the nonconceptual and prereflective processing and representation of body-related information” (p. 1096).
• Selfhood — the “fundamental sense” of being a self, tied to bodily self-consciousness rather than to its cognitive or emotional layers, and experienced as the content of one whole-body representation (p. 1097).
• Body-part ownership against global ownership — the sense that a part is mine, as in the rubber hand illusion, against ownership of the whole body, which the authors associate directly with selfhood (p. 1097).
• Self-attribution and self-localization — taking a seen body (or part) as one’s own, and where one feels oneself to be; measured here by questionnaire and by drift.
• Out-of-body experience — “a disembodiment of the self to an extracorporeal location, an extracorporeal visuospatial perspective, and the sight of one’s own body from this self-location” (p. 1099).
• Heautoscopy — seeing an illusory body and feeling located at it, at one’s physical body or in between, often identifying with it, without the overt disembodiment of an out-of-body experience (p. 1099).

Connections

It is built on the rubber hand illusion (Botvinick and Cohen 1998, ref. 6), its proprioceptive drift and its premotor and insular correlates (Ehrsson et al.; Tsakiris et al.), and on Blanke’s clinical work on out-of-body experience and heautoscopy (refs. 15, 17, 27). Its use of Tsakiris and Haggard (2005) for a cognitive constraint on ownership is the comparison behind the mannequin and object conditions. Its conceptual frame is Metzinger’s, a co-author: selfhood as the transparent content of a whole-body model, and global ownership as its core. Ehrsson’s companion Brevia in the same issue induced a different illusion, self-location at the camera’s viewpoint with chest stroking, and the e-letters set the two side by side. Botvinick and Cohen 1998 and Blanke and Metzinger 2009 are being filed alongside this reading.

Bearing on the record

• NOTE-372, C6, and LIT-450. As NOTE-372 reads it, “Self models” cites this study for full-body illusions that shift felt self-location toward a seen virtual body, and offers global body ownership as “the most simple form of self-awareness”. The first is what the paper reports, and NOTE-372’s grading, moderate for the manipulation, stands; the reading adds that the shift is 17–25 cm over 2 m, and that in study II the own-body condition did not separate synchronous from asynchronous stroking. The second is not shown here: the paper takes it from Metzinger’s book (C9). The paper does not bear on NOTE-372’s C7, on phantoms; it cites Brugger 2002 for disrupted self–body unity, not the congenital-phantom case.
• THEORY-064 and LIT-568. The paper describes selfhood in the self-model theory’s words, “transparent content”. Its result is the kind THEORY-064’s promote_when already says cannot settle it: it shows the bodily self-model can be manipulated, not that a self is nothing beyond it, and nothing in it tests transparency.
• THEORY-458 and NOTE-827. The illusion is induced by passive touch with no voluntary movement, so whatever ownership it produces comes without agency, in line with the split between ownership and agency that THEORY-458 states and Haggard’s review (LIT-1009) draws. The paper does not test agency.
• QUESTION-048. A worked case of what reports show. Q1 and Q2 were answered positively for a box, so the touch items do not discriminate self-identification; Q3 does. The behavioural drift, the measure that needs no report, did not significantly separate one’s own body from the box in synchronous stroking. Report and behaviour diverge in a way that limits each.
• LIT-1186. That study’s participants attempted out-of-body perception. This paper’s distinction between illusory self-location and the disembodiment and shifted perspective that define an out-of-body experience is the line a reading of it would need, and it says the laboratory illusion does not cross it.
• THEORY-768 (filed 2026-10-20) takes this study, with Blanke and Metzinger 2009, as a source for minimal selfhood as identification with the whole body.
• No instruction for machine-learning practice; nothing for the anthology.

Limitations

• Small samples. Two groups of 14, with standard errors of 6 to 9 cm on drifts of 12 to 25 cm.
• Inconsistency between studies. The own-body synchrony effect, significant in study I, was not significant in study II, explained after the fact by the different asynchrony method.
• Slips in the text and supplement. The printed asynchronous drift is “12. ± 8.5 cm”. The SOM says the empty-room and motor-control drifts were negative (“Indeed in these conditions the drift was negative”), but reports the empty-room drift as 1.12 cm. It refers to a “Fig 2C” the paper does not have, and gives “t<3.24” where a lower bound is meant.
• The measures. The drift was measured once per condition after displacement and walking, so motor and memory error enter it; the motor-control subtraction in study II addresses only a constant bias. The critics’ point that the questionnaire cannot tell ownership from self-recognition in a video stands, and the design has no condition that separates them.
• What the authors do not claim. They do not claim to have induced an out-of-body experience, and the temporo-parietal claim is labelled speculation.

Open questions

• Is this an illusion of ownership or recognition of oneself in a video system? Ehrsson and Petkova’s alternative predicts no effect without the camera and display; a direct test within one design, comparing first- and third-person views, is what they ask for.
• Where do participants locate themselves with the display on and no stroking at all? The authors’ reply says this baseline was not tested.
• Does visual-vestibular conflict, added to visuo-tactile conflict, produce disembodiment and a shifted perspective, as the authors predict?