Everyone agrees that consciousness needs coherence. The question of coherence in Gwendolin Kirchhoff’s reading goes beyond neuroscience — she carries the thought further by asking whether consciousness is not the condition for coherence rather than the reverse. The wonderful and at the same time treacherous thing about this agreement is that it obscures the actual question: does coherence also need consciousness? Or more precisely: does coherence suffice to explain consciousness? The answer of natural philosophy is a clear no, and this no has consequences that reach far beyond an academic distinction.
#The blockchain argument
In the Everlast AI debate (2026) Joscha Bach represented the position that consciousness is an emergent control process arising from the coherence between communicating cells. Cells send signals to one another until they become coherent, and from this pattern of coherence the model of the organism and its environment arises. Consciousness, on Bach, populates the space of coherence (cf. Gwendolin Kirchhoff, Everlast AI Debate, 2026, 40:26–40:28; 57:54).
Gwendolin Kirchhoff cut through this thought with a single counter-example: one can produce coherence, for instance in a blockchain. But the blockchain is therefore not conscious. One can produce second-order perception, a camera observed by a secondary sensor as it records. That too would not be consciousness. Coherence in itself is not yet consciousness (cf. Gwendolin Kirchhoff, Everlast AI Debate, 2026, 45:10–45:18).
The point is not trivial. It exposes the logical error that runs through the entire computationalist programme: the confusion of a necessary condition with a sufficient one. If every coherent system were conscious, then a heating thermostat would be conscious, a tide-algorithm conscious, a well-synchronised traffic-light circuit conscious. That no one seriously claims this shows: coherence itself is not what constitutes consciousness. Something is missing, and what is missing is precisely what the functional description systematically leaves out. When you next hear the term coherence in the context of the AI debate, it is worth looking closely: is coherence described as accompanying feature, or already passed off as explanation?
#What is missing: the first person
Thomas Nagel formulated in 1974 in What Is It Like to Be a Bat? the question that any coherence theory of consciousness would have to answer: what does it feel like to be this system? No functional description, no coherence metric, no information processing of the third order can answer this question, because it asks about something accessible only from inside — subjective experience.
Kirchhoff drew the arc in the debate: all these definitions — coherence, second-order perception, substrate independence — do not grasp the actual content of consciousness, which can only be apprehended from the first person. Consciousness is first of all evident in experience (cf. Gwendolin Kirchhoff, Everlast AI Debate, 2026, 45:34–45:41). To the first person belong phenomenology, the gathered knowledge of meditative practice, and the entire access to the world of indigenous peoples. What David Chalmers formulated in 1996 in The Conscious Mind as the Hard Problem — the explanatory gap between physical description and subjective experience — reappears in the coherence debate as a specific finding: you can describe the coherence of a system in finest detail and stand at the end before exactly the same question as at the beginning, namely whether anything feels like something for this system. This is no gap of knowledge that more data could close. It is a category difference: coherence belongs to the description of structures, experience belongs to the reality of the one who describes.
#The reversed direction of explanation
Bach formulates the relation between coherence and consciousness in one direction: coherence produces consciousness. Natural philosophy reverses the direction of explanation. Schelling developed in the Ideas for a Philosophy of Nature (1797) the thought that the living cannot be assembled from the dead: the inorganic is only the negated organism, the dead only the suppressed life (cf. Schelling, 1797). There is nothing absolutely dead.
In this perspective coherence is no principle that produces consciousness, but a manifestation of the already living. The living organism brings forth its inner order from itself, from inside out. The cell does not communicate coherently with other cells because an algorithm synchronises them, but because it is alive. Jochen Kirchhoff (1944–2025) described the difference between the organic and the mechanical as ontological difference: the mechanical is steered from outside in, toward the purposes of the human understanding. The organic organises itself (cf. Kirchhoff, J., Was die Erde will, 1998).
If you bring the consequence to mind, the scope of the difference becomes visible. For Bach, coherence is what consciousness emerges from. For natural philosophy, coherence is what consciousness leaves as order in the material world. Not the coherence of neurons produces spirit; spirit produces the coherence of neurons. The direction of explanation decides which questions can be asked and which cannot.
#What metaphysics has to do with the problem
Kirchhoff made the decisive move in the debate: the question is what metaphysics produces a boot problem and a Hard Problem at all. The mechanistic produces both, that of the living cosmos does not (cf. Gwendolin Kirchhoff, Everlast AI Debate, 2026, 75:53–76:34). The coherence problem inherits the same structure. Whoever thinks consciousness as product of coherence must explain why some coherent systems are conscious and others not. They cannot give this explanation, because their metaphysics does not contain the criterion of distinction. If coherence were the essential, every coherent system would have to be conscious. That this is not so shows that something essential is missing from the description.
Computationalism and functionalism operate with the same basic error on different levels of abstraction: they describe features that accompany consciousness and declare the accompaniment to be identity. Coherence accompanies consciousness. Functional organisation accompanies consciousness. Information processing accompanies consciousness. But accompaniment is not production. The philosophical diagnosis: an ontological levelling in which the functional description of a phenomenon is confused with the phenomenon itself (cf. Gwendolin Kirchhoff, Die Abschaffung des Menschen, 2024, 02:35).
In the philosophy of consciousness the coherence question marks the boundary between two metaphysics: one that starts from the dead and must explain how the living arises from it (and fails at this), and one that starts from the living and recognises in the dead what was suppressed. The decision you encounter here is not technical. It concerns the image of the human being underlying the question, and first-person phenomenology offers the access that no coherence metric can replace.