Laboratory of Psychophysics LPSY
Consciousness is one of the most enigmatic phenomena of neuroscience and humankind in general. What are the neural processes that give rise to consciousness? Which processes are related to unconsciousness? Whereas consciousness was considered a non-scientific question for centuries, the last two decades have faced a wealth of computational and neuroscientific models proposed to explain consciousness. How good are these models and theories? A joint cooperation with Michael Esfeld (philosophy) and Wulfram Gerstner (neural computation) showed that all models fail to explain consciousness by what we call the small network argument. We showed that each model can be realized by a model with less than ten neurons. For example, Tononi and colleagues proposed that a complex firing pattern of neurons is necessary and sufficient for consciousness. We showed that a small network of only nine neurons meets all the criteria and hence needs to be considered conscious. One may accept this conclusion but then one also needs to accept that the human brain creates at each moment millions of consciousness, which we think is not a tenable move.
Our main research in this area is about understanding unconscious processes. To this end, we developed paradigms where we can fully mask a target and prove it is indeed unconscious, which no previous paradigm achieved (only features of a target but not the target itself are unconscious in mask priming for example; Otto et al., 2006). We were able to show that the features of these elements are still visible at conscious elements and are, similar to the TMS experiment described in Decision making, stored and “influential” for a substantial period of time (Scharnowski et al., 2009).
Usually, we assume that consciousness is a smooth stream of percepts. However, continuous perception cannot explain most temporal phenomena, such as apparent motion, favoring discrete models. What is the duration of a discrete percept? We can perceive apparent motion with 3m differences? Is perception hence that quick? Other paradigms show much worse temporal resolution. We proposed a 2 stage model of discrete perception and unconsciousness, where quasi-continuous unconscious processing precedes slow conscious perception (Herzog et al., 2016).
- Herzog MH, Esfeld M, Gerstner W (2007).
Consciousness & the Small Network Argument. Neural Networks, 20(9), p1054-1056. [⇒ pdf]
For commentaries see Taylor 2007a & 2007b
- Herzog MH, Kammer T, Scharnowski F (2016). Time Slices: What Is the Duration of a Percept? PLoS Biology, 14(4):e1002433, p1-12.
- Grainger JE, Scharnowski F, Schmidt T, Herzog MH (2013). Two primes priming: Does feature integration occur before response activation? Journal of Vision, 13(8):19, p1-10.
- Hermens F, Scharnowski F, Herzog MH (2010). Automatic grouping of regular structures. Journal of Vision, 10(8):5, p1-16.
- Scharnowski F, Rüter J, Jolij J, Hermens F, Kammer T, Herzog MH (2009). Long-lasting modulation of feature integration by transcranial magnetic stimulation. Journal of Vision, 9(6):1, p1-10.
- Otto TU, Öğmen H, Herzog MH (2006). The flight path of the phoenix – the visible trace of invisible elements in human vision. Journal of Vision, 6(10), p1079-86.