Temporal constraints on intentional binding and the perception of causality

Marieke Rohde, Meike Scheller and Marc O. Ernst

The perceived timing of both an action and its visual feedback are altered if humans assume that they are causally linked [Haggard et al, 2002, Nature Neuroscience, 5, 382–385]. However, it is unknown how visual and motor events have to be timed relative to one another to be perceived as causally linked. Two possible criteria apply: (1) The visual and motor event should be maximally close in time. (2) The visual event should occur after the motor event. We tested which of these criteria dominates the human perception of causality. Participants had to press a button and two visual stimuli were presented at variable times shortly before and after the press event. In order to present a visual flash at a given time before the voluntary button press, we predicted the moment of the press in real time from online measurements of finger position. Participants decided (forced choice) whether their button press had caused the first or the second flash. We found a trade-off between the two criteria: There is a window of intentional binding of ca. 100 ms where closeness in time dominates causality judgments, even if the visual stimulus leads. For larger visual leads, participants adhere to the criterion that cause precedes effect.

The time course of feature integration in plaid gratings identified by metacontrast masking

Maximilian Bruchmann

Using oriented gratings in a metacontrast paradigm we recently showed that masking functions – i.e. the dependence of target visibility on the stimulus onset asynchrony (SOA) between target and mask – are determined by inhibitory processes within and between spatial frequency selective channels. By contrasting collinear and orthogonal target-mask sequences in an EEG/MEG-study we find that within-channel inhibition peaks around 120 ms after target onset. In a behavioral study we present simple grating annuli as masks preceded by either collinear, orthogonal, or plaid targets, or no targets, asking the participants whether a target had been presented, and if so, which one it was. The results for plaid targets show that at short SOAs (10-30 ms) the mask selectively removes the collinear component from the plaid stimulus, causing subjects to perceive it as orthogonal instead. At longer SOAs (50-60 ms) the mask reduces the visibility of the plaid pattern as a whole. We conclude that masks interfere with isolated features at short SOAs and with integrated objects at longer SOAs. In combination with the results mentioned above we infer that the beginning of feature integration precedes within-channel inhibition by around 30 ms, so about 90 ms after target onset.

Priming the recognition of one's own and other faces: perceptual shifts and reaction times

Wolfgang G. Röhrich and Hanspeter Mallot

The recognition of the own and other faces was investigated with morph sequences creating smooth transitions between a portrait of the participant and portraits of three other people [cf. Kircher et al, 2001, Cognition, 78, B1-B15]. In a time-constraint two-alternative forced choice paradigm, participants were asked to reply "yes" if they thought a portrait depicted themselves, and "no" if they did not. The recognition process was manipulated by presenting periliminal (30 ms) primes [Pannese and Hirsch, 2010, Consciousness and Cognition, 19, 962-968] that either showed the participant's own face or the face of the other person appearing in the current morph sequence. Percentages of answers "yes" were plotted as a function of morph level (ranging from 0 to 100%), yielding a psychometric function for self-recognition. Priming with "self" or "other" increased the probability of recognizing the respective face, shifting the point of subjective equivalence (PSE, 50% level of fitted curves) by about 7% away from the primed face. Reaction times indicated faster response in concruent trials (target picture consists of more than 50% of person shown in prime) of about 40ms as compared to inconcruent trials. Taken together, priming-induced PSE-shift and reaction times provide a stable probe for measuring the influence of contextual stimuli on self-recognition.

How many roads are there to act upon visual stimuli?

Izabela Szumska, Rob Ven der Lubbe and Michael Herzog

It has regularly been proposed that there are different routes of visually based actions. The dorsal visual processing route is thought to be involved when manual actions have to be carried out whereas the ventral route is crucial for judgments about object identity. These findings seem in line with views on consciousness such as the multiple drafts model of consciousness [Dennett, 1991, Consciousness Explained, Boston, Little, Brown and Co.], and point against a global workspace that guides our actions. In the current study, we examined whether this multiple drafts models is supported by focusing on different effector systems. A diamond was presented either to the left or the right from a fixation dot accompanied by a distracting square on the other side. Participants had to indicate the side of the diamond either by making a corresponding button press, a corresponding saccade, or by giving a verbal left or right response. Meta-contrast masking was employed at various SOAs to manipulate the visibility of the target and distractor. Response accuracy increased as a function of SOA, but importantly, the pattern for the different effectors was indistinguishable. These results favor global workspace models in which incoming visual information is first analyzed before triggering the required action.

Object-centered binocular suppression

Mark Vergeer, Marco Boi, Haluk Öğmen and Michael H. Herzog

In binocular rivalry, one image is perceived consciously whereas the other image is suppressed. It is often thought that binocular suppression results from inhibition between (monocular) neurons coding for the same visual field location and, hence, suppression occurs within retinotopic coordinates. In a binocular rivalry experiment, we presented three disks to both eyes. In the central disk, a low-contrast horizontal or vertical grating was presented to one eye and a high-contrast bull’s eye to the other eye. The bull’s eye fully suppressed the grating. Next, the three disks were shifted in both eyes to the right by one position, creating apparent motion and, consequently, a non-retinotopic reference frame. In the central disk of frame two, again a horizontal or vertical grating was presented. No bull’s eye was presented. Detection of the grating with the previously suppressed orientation was worse than detection of the orthogonal one, even though the gratings in frame one and two were presented at different retinotopic locations. Hence, feature suppression occurred in non-retinotopic, object-centered coordinates. We suggest that binocular rivalry in non-retinotopic coordinates allows the visual system to create binocular perceptual stability despite continuous changes in the retinal images due to eye, object, and head movements.

Chromatic evoked potentials modulated by hypnosis

Michael Crognale, Chad Duncan, Christopher Jones and Molly Finnegan

Prior research has demonstrated that the major negative component of the chromatic visual evoked potential is resistant to modulations from attention, suggesting an early origin for this component. There have been other reports that brain responses in early visual cortex can be selectively suppressed in “multiple personality disorder”. In one report, the brain responses were suppressed for blind personalities but nor for sighted personalities within a single subject. This result suggests a feedback contribution to early visual areas from regions involved in higher functions (e.g.) consciousness. We hypothesized that perhaps early visual areas can also be modulated in other manipulations of consciousness such as hypnosis. We tested chromatic and achromatic VEP responses in subjects with and without the hypnotic suggestion of blindness to the visual display of the stimulus. Results showed that the responses of many subjects were suppressed during the suggestion compared to responses before and after hypnotic suggestion. Other subjects’ responses were unaffected, consistent with the observation of high variability in suggestibility to hypnosis within the general population. Preliminary controls have ruled out some other explanations of this phenomenon. We continue to test alternative explanations for these results.