We are Maggie & Anders. A mostly harmless Swedish old-timer couple only now beginning to discover the advanced incompetence that is the proto-science — or “alchemy” — of consciousness research. A few centuries ago a philosopher of chemistry could have claimed with a straight face to be quite certain that a substance with negative mass had to be invoked to explain the phenomenon of combustion. Another could have been equally convinced that the chemistry of life involves a special force of nature absent from all non-living matter. A physicist of today may recognize that the study of consciousness has even less experimental foundation than alchemy did, yet be confident that at least it cannot feel like somethingto be a black hole. Since, obviously, black holes are simple objects and consciousness is a phenomenon which only emerges from “complexity” as high as that of a human brain.
Is there some ultimate substrate, basic to reality and which has properties intrinsic to itself? If so, is elementary sentience one of those properties? Or is it “turtles all the way down” in a long regress where all of reality can be modeled as patterns within patterns within patterns ending in Turing-style “bits”? Or parsimoniously never ending?
Will it turn out to be patterns all the way down, or sentience all the way up? Should people who believe themselves to perhaps be in an ancestor simulation take for granted that consciousness exists for biologically-based people in base-level reality? David Chalmers does. So at least that must be one assumption it is safe to make, isn’t it? And the one about no sentience existing in a black hole. And the one about phlogiston. And the four chemical elements.
This really is good material for silly comedy or artistic satire. To view a modest attempt by us in that direction, please feel encouraged to enjoy this youtube video we made with QRI in mind:
When ignorance is near complete, it is vital to think outside the proverbial box if progress is to be made. However, spontaneous creative speculation is more context-constrained than it feels like, and it rarely correlates all that beautifully with anything useful. Any science has to work via the baby steps of testable predictions. The integrated information theory (IIT) does just that, and has produced encouraging early results. IIT could turn out to be a good starting point for eventually mapping and modeling all of experiential phenomenology. For a perspective, IIT 3.0 may be comparable to how Einstein’s modeling of the photoelectric effect stands in relation to a full-blown theory of quantum gravity. There is a fair bit of ground to cover. We have not been able to find any group more likely than the QRI to speed up the process whereby humanity eventually manages to cover that ground. That is, if they get a whole lot of help in the form of outreach, fundraising and technological development. Early pioneers have big hurdles to overcome, but the difference they can make for the future is enormous.
For those who feel inspired, a nice start is to go through all that is on or linked via the QRI website. Indulge in Principia Qualia. If that leaves you confused on a higher level, you are in good company. With us. We are halfway senile and are not information theorists, neuroscientists or physicists. All we have is a nerdy sense of humor and work experience in areas like marketing and planetary geochemistry. One thing we think we can do is help bridge the gap between “experts” and “lay people”. Instead of “explain it like I am five”, we offer the even greater challenge of explaining it like we are Maggie & Anders. Manage that, and you will definitely be wiser afterwards!
– Maggie & Anders
No qualia or your money back!
Tested for consciousness – zero
Don’t be afraid – she can’t feel anything!
Letter II: State-Space of Matter and State-Space of Consciousness
A core aspect of science is the mapping out of distributions, spectra, and state-spaces of the building blocks of reality. Naturally occurring states of things can be spontaneously discovered. To gain more information about them, one can experimentally alter such states to produce novel ones, and then analyze them in a systematic way.
The full state-space of matter is multidimensional and vast. Zoom in anywhere in it and there will be a number of characteristic physics phenomena appearing there. Within a model of the state-space you can follow independent directions as you move towards regions and points. As an example, you can hold steady at one particular simple chemical configuration. Diamond, say. The stable region of diamond and its emergent properties like high hardness extends certain distances in other parameter directions such as temperature and pressure. The diamond region has neighboring regions with differently structured carbon, such as graphite. Diamond and graphite make for an interesting case since the property of hardness emerges very differently in the two regions. (In the pure carbon state-space the dimensions denoting amounts of all other elements can be said to be there but set to zero). Material properties like hardness can be modeled as static phenomena. According to IIT however, consciousness cannot. It’s still an emergent property of matter though, so just stay in the matter state-space and add a time dimension to it. Then open chains and closed loops of causation emerge as a sort of fundamental level of what matter “does”. Each elementary step of causation may be regarded to produce or intrinsically be some iota of proto-experience. In feedback loops this self-amplifies into states of feeling like something. Many or perhaps most forms of matter can “do” these basic things at various regions of various combinations of parameter settings. Closed causal loops require more delicate fine-tuning in parameter space, so the state-space of nonconscious causation structure is larger than that of conscious structure. The famous “hard problem” has to do with the fact that both an experientially very weak and a very strong state can emerge from the same matter (shown to be the case so far only within brains). A bit like the huge difference in mechanical hardness of diamond and graphite both emerging from the same pure carbon substrate (a word play on “hard” to make it sticky).
By the logic of IIT it should be possible to model (in arbitrarily coarse or fine detail) the state-space of all conscious experience whose substrate is all possible physical states of pure carbon. Or at room temperature in any material. And so on. If future advanced versions of IIT turn out to be a success then we may guess there’ll be a significant overlap to allow for a certain “substrate invariance” for hardware that can support intelligence with human-recognizable consciousness. Outside of that there will be a gargantuan additional novel space to explore. It ought to contain maxima of (intrinsic) attractiveness, none of which need to reside within what a biological nervous system can host. Biological evolution has only been able to search through certain parts of the state-space of matter. One thing it has not worked with on Earth is pure carbon. Diamond tooth enamel or carbon nanotube tendons would be useful but no animal has them. What about conscious states? Has biology come close to hit upon any of the optima in those? If all of human sentience is like planet Earth, and all of Terrestrial biologically-based sentience is like the whole Solar System, that leaves an entire extrasolar galaxy out there to explore. (Boarding call: Space X Flight 42 bound for Nanedi Settlement, Mars. Sentinauts please go to the Neuralink check-in terminal).
Of course we don’t currently know how IIT is going to stand up, but thankfully it does make testable predictions. There is, therefore, a beginning of something to be hoped for with it. In a hopeful scenario IIT turns out to be like special relativity, and what QRI is reaching for is like quantum gravity. It will be a process of taking baby steps, for sure. But each step is likely to bring benefits in many ways.
How do psychedelic drugs produce their characteristic range of acute effects in perception, emotion, cognition, and sense of self? How do these effects relate to the clinical efficacy of psychedelic-assisted therapies? Efforts to understand psychedelic phenomena date back more than a century in Western science. In this article I review theories of psychedelic drug effects and highlight key concepts which have endured over the last 125 years of psychedelic science. First, I describe the subjective phenomenology of acute psychedelic effects using the best available data. Next, I review late 19th-century and early 20th-century theories—model psychoses theory, filtration theory, and psychoanalytic theory—and highlight their shared features. I then briefly review recent findings on the neuropharmacology and neurophysiology of psychedelic drugs in humans. Finally, I describe recent theories of psychedelic drug effects which leverage 21st-century cognitive neuroscience frameworks—entropic brain theory, integrated information theory, and predictive processing—and point out key shared features that link back to earlier theories. I identify an abstract principle which cuts across many theories past and present: psychedelic drugs perturb universal brain processes that normally serve to constrain neural systems central to perception, emotion, cognition, and sense of self. I conclude that making an explicit effort to investigate the principles and mechanisms of psychedelic drug effects is a uniquely powerful way to iteratively develop and test unifying theories of brain function.
Subjective rating scale items selected after psilocybin (blue) and placebo (red) (n = 15) (Muthukumaraswamy et al., 2013). “Items were completed using a visual analog scale format, with a bottom anchor of ‘no, not more than usually’ and a top anchor of ‘yes, much more than usually’ for every item, with the exception of ‘I felt entirely normal,’ which had bottom and top anchors of ‘No, I experienced a different state altogether’ and ‘Yes, I felt just as I normally do,’ respectively. Shown are the mean ratings for 15 participants plus the positive SEMs. All items marked with an asterisk were scored significantly higher after psilocybin than placebo infusion at a Bonferroni-corrected significance level of p < 0.0022 (0.5/23 items)” (Muthukumaraswamy et al., 2013, p. 15176).
Neuropharmacology and Neurophysiological Correlates of Psychedelic Drug Effects
Klee recognized that his above hypotheses, inspired by psychoanalytic theory and LSD effects, required neurophysiological evidence. “As far as I am aware, however, adequate neurophysiological evidence is lacking … The long awaited millennium in which biochemical, physiological, and psychological processes can be freely correlated still seems a great distance off” (Klee, 1963, p. 466, 473). What clues have recent investigations uncovered?
A psychedelic drug molecule impacts a neuron by binding to and altering the conformation of receptors on the surface of the neuron (Nichols, 2016). The receptor interaction most implicated in producing classic psychedelic drug effects is agonist or partial agonist activity at serotonin (5-HT) receptor type 2A (5-HT2A) (Nichols, 2016). A molecule’s propensity for 5-HT2A affinity and agonist activity predicts its potential for (and potency of) subjective psychedelic effects (Glennon et al., 1984; McKenna et al., 1990; Halberstadt, 2015; Nichols, 2016; Rickli et al., 2016). When a psychedelic drug’s 5-HT2A agonist activity is intentionally blocked using 5-HT2Aantagonist drugs (e.g., ketanserin), the subjective effects are blocked or attenuated in humans under psilocybin (Vollenweider et al., 1998; Kometer et al., 2013), LSD (Kraehenmann et al., 2017a,b; Preller et al., 2017), and ayahuasca (Valle et al., 2016). Importantly, while the above evidence makes it clear that 5-HT2A activation is a necessary (if not sufficient) mediator of the hallmark subjective effects of classic psychedelic drugs, this does not entail that 5-HT2A activation is the sole neurochemical cause of all subjective effects. For example, 5-HT2A activation might trigger neurochemical modulations ‘downstream’ (e.g., changes in glutamate transmission) which could also play causal roles in producing psychedelic effects (Nichols, 2016). Moreover, most psychedelic drug molecules activate other receptors in addition to 5-HT2A (e.g., 5-HT1A, 5-HT2C, dopamine, sigma, etc.) and these activations may importantly contribute to the overall profile of subjective effects even if 5-HT2A activation is required for their effects to occur (Ray, 2010, 2016).
How does psychedelic drug-induced 5-HT2A receptor agonism change the behavior of the host neuron? Generally, 5-HT2A activation has a depolarizing effect on the neuron, making it more excitable (more likely to fire) (Andrade, 2011; Nichols, 2016). Importantly, this does not necessarily entail that 5-HT2Aactivation will have an overall excitatory effect throughout the brain, particularly if the excitation occurs in inhibitory neurons (Andrade, 2011). This important consideration (captured by the adage ‘one neuron’s excitation is another neuron’s inhibition’) should be kept in mind when tracing causal links in the pharmaco-neurophysiology of psychedelic drug effects.
In mammalian brains, neurons tend to ‘fire together’ in synchronized rhythms known as temporal oscillations (brain waves). MEG and EEG equipment measure the electromagnetic disturbances produced by the temporal oscillations of large neural populations and these measurements can be quantified according to their amplitude (power) and frequency (timing) (Buzsáki and Draguhn, 2004). Specific combinations of frequency and amplitude can be correlated with distinct brain states, including waking ‘resting’ state, various attentional tasks, anesthesia, REM sleep, and deep sleep (Tononi and Koch, 2008; Atasoy et al., 2017a). In what ways do temporal oscillations change under psychedelic drugs? MEG and EEG studies consistently show reductions in oscillatory power across a broad frequency range under ayahuasca (Riba et al., 2002, 2004; Schenberg et al., 2015; Valle et al., 2016), psilocybin (Muthukumaraswamy et al., 2013; Kometer et al., 2015; Schartner et al., 2017), and LSD (Carhart-Harris et al., 2016c; Schartner et al., 2017). Reductions in the power of alpha-band oscillations, localized mainly to parietal and occipital cortex, have been correlated with intensity of subjective visual effects—e.g., ‘I saw geometric patterns’ or ‘My imagination was extremely vivid’—under psilocybin (Kometer et al., 2013; Muthukumaraswamy et al., 2013; Schartner et al., 2017) and ayahuasca (Riba et al., 2004; Valle et al., 2016). Under LSD, reductions in alpha power still correlated with intensity of subjective visual effects but associated alpha reductions were more widely distributed throughout the brain (Carhart-Harris et al., 2016c). Furthermore, ego-dissolution effects and mystical-type experiences (e.g., ‘I experienced a disintegration of my “self” or “ego”’ or ‘The experience had a supernatural quality’) have been correlated with reductions in alpha power localized to anterior and posterior cingulate cortices and the parahippocampal regions under psilocybin (Muthukumaraswamy et al., 2013; Kometer et al., 2015) and throughout the brain under LSD (Carhart-Harris et al., 2016c).
The concept of functional connectivity rests upon fMRI brain imaging observations that reveal temporal correlations of activity occurring in spatially remote regions of the brain which form highly structured patterns (brain networks) (Buckner et al., 2013). Imaging of brains during perceptual or cognitive task performance reveals patterns of functional connectivity known as functional networks; e.g., control network, dorsal attention network, ventral attention network, visual network, auditory network, and so on. Imaging brains in taskless resting conditions reveals resting-state functional connectivity (RSFC) and structured patterns of RSFC known as resting state networks (RSNs; Deco et al., 2011). One particular RSN, the default mode network (DMN; Buckner et al., 2008), increases activity in the absence of tasks and decreases activity during task performance (Fox and Raichle, 2007). DMN activity is strong during internally directed cognition and a variety of other ‘metacognitive’ functions (Buckner et al., 2008). DMN activation in normal waking states exhibits ‘inverse coupling’ or anticorrelation with the activation of task-positive functional networks, meaning that DMN and functional networks are often mutually exclusive; one deactivates as the other activates and vice versa (Fox and Raichle, 2007).
Taken together, the recently discovered neurophysiological correlates of subjective psychedelic effects present an important puzzle for 21st-century neuroscience. A key clue is that 5-HT2A receptor agonism leads to desynchronization of oscillatory activity, disintegration of intrinsic integrity in the DMN and related brain networks, and an overall brain dynamic characterized by increased between-network global functional connectivity, expanded signal diversity, and a larger repertoire of structured neurophysiological activation patterns. Crucially, these characteristic traits of psychedelic brain activity have been correlated with the phenomenological dynamics and intensity of subjective psychedelic effects.
21st-Century Theories of Psychedelic Drug Effects
Entropic Brain Theory
Entropic Brain Theory (EBT; Carhart-Harris et al., 2014) links the phenomenology and neurophysiology of psychedelic effects by characterizing both in terms of the quantitative notions of entropy and uncertainty. Entropy is a quantitative index of a system’s (physical) disorder or randomness which can simultaneously describe its (informational) uncertainty. EBT “proposes that the quality of any conscious state depends on the system’s entropy measured via key parameters of brain function” (Carhart-Harris et al., 2014, p. 1). Their hypothesis states that hallmark psychedelic effects (e.g., perceptual destabilization, cognitive flexibility, ego dissolution) can be mapped directly onto elevated levels of entropy/uncertainty measured in brain activity, e.g., widened repertoire of functional connectivity patterns, reduced anticorrelation of brain networks, and desynchronization of RSN activity. More specifically, EBT characterizes the difference between psychedelic states and normal waking states in terms of how the underlying brain dynamics are positioned on a scale between the two extremes of order and disorder—a concept known as ‘self-organized criticality’ (Beggs and Plenz, 2003). A system with high order (low entropy) exhibits dynamics that resemble ‘petrification’ and are relatively inflexible but more stable, while a system with low order (high entropy) exhibits dynamics that resemble ‘formlessness’ and are more flexible but less stable. The notion of ‘criticality’ describes the transition zone in which the brain remains poised between order and disorder. Physical systems at criticality exhibit increased transient ‘metastable’ states, increased sensitivity to perturbation, and increased propensity for cascading ‘avalanches’ of metastable activity. Importantly, EBT points out that these characteristics are consistent with psychedelic phenomenology, e.g., hypersensitivity to external stimuli, broadened range of experiences, or rapidly shifting perceptual and mental contents. Furthermore, EBT uses the notion of criticality to characterize the difference between psychedelic states and normal waking states as it “describes cognition in adult modern humans as ‘near critical’ but ‘sub-critical’—meaning that its dynamics are poised in a position between the two extremes of formlessness and petrification where there is an optimal balance between order and flexibility” (Carhart-Harris et al., 2014, p. 12). EBT hypothesizes that psychedelic drugs interfere with ‘entropy-suppression’ brain mechanisms which normally sustain sub-critical brain dynamics, thus bringing the brain “closer to criticality in the psychedelic state” (Carhart-Harris et al., 2014, p. 12).
Integrated Information Theory
Integrated Information Theory (IIT) is a general theoretical framework which describes the relationship between consciousness and its physical substrates (Oizumi et al., 2014; Tononi, 2004, 2008). While EBT is already loosely consistent with the core principles of IIT, Gallimore (2015) demonstrates how EBT’s hypotheses can be operationalized using the technical concepts of the IIT framework. Using EBT and recent neuroimaging data as a foundation, Gallimore develops an IIT-based model of psychedelic effects. Consistent with EBT, this IIT-based model describes the brain’s continual challenge of minimizing entropy while retaining flexibility. Gallimore formally restates this problem using IIT parameters: brains attempt to optimize the give-and-take dynamic between cause-effect information and cognitive flexibility. In IIT, a (neural) system generates cause-effect information when the mechanisms which make up its current state constrain the set of states which could casually precede or follow the current state. In other words, each mechanistic state of the brain: (1) limits the set of past states which could have causally given rise to it, and (2) limits the set of future states which can causally follow from it. Thus, each current state of the mechanisms within a neural system (or subsystem) has an associated cause-effect repertoire which specifies a certain amount of cause-effect information as a function of how stringently it constrains the unconstrained state repertoire of all possible system states. Increasing the entropy within a cause-effect repertoire will in effect constrain the system less stringently as the causal possibilities are expanded in both temporal directions as the system moves closer to its unconstrained repertoire of all possible states. Moreover, increasing the entropy within a cause-effect repertoire equivalently increases the uncertainty associated with its past (and future) causal interactions. Using this IIT-based framework, Gallimore (2015)argues that, compared with normal waking states, psychedelic brain states exhibit higher entropy, higher cognitive flexibility, but lower cause-effect information.
Predictive Processing
The first modern brain imaging measurements in humans under psilocybin yielded somewhat unexpected results: reductions in oscillatory power (MEG) and cerebral blood flow (fMRI) correlated with the intensity of subjective psychedelic effects (Carhart-Harris et al., 2012; Muthukumaraswamy et al., 2013). In their discussion, the authors suggest that their findings, although surprising through the lens of commonly held beliefs about how brain activity maps to subjective phenomenology, may actually be consistent with a theory of brain function known as the free energy principle (FEP; Friston, 2010).
In one model of global brain function based on the free-energy principle (Friston, 2010), activity in deep-layer projection neurons encodes top-down inferences about the world. Speculatively, if deep-layer pyramidal cells were to become hyperexcitable during the psychedelic state, information processing would be biased in the direction of inference—such that implicit models of the world become spontaneously manifest—intruding into consciousness without prior invitation from sensory data. This could explain many of the subjective effects of psychedelics (Muthukumaraswamy et al., 2013, p. 15181).
What is FEP? “In this view, the brain is an inference machine that actively predicts and explains its sensations. Central to this hypothesis is a probabilistic model that can generate predictions, against which sensory samples are tested to update beliefs about their causes” (Friston, 2010). FEP is a formulation of a broader conceptual framework emerging in cognitive neuroscience known as predictive processing (PP; Clark, 2013)10. PP has links to bayesian brain hypothesis (Knill and Pouget, 2004), predictive coding (Rao and Ballard, 1999), and earlier theories of perception and cognition (MacKay, 1956; Neisser, 1967; Gregory, 1968) dating back to Helmholtz (1925) who was inspired by Kant (1996; see Swanson, 2016). At the turn of the 21st century, the ideas of Helmholtz catalyzed innovations in machine learning (Dayan et al., 1995), new understandings of cortical organization (Mumford, 1992; Friston, 2005), and theories of how perception works (Kersten and Yuille, 2003; Lee and Mumford, 2003).
Conclusion
The four key features identified in filtration and psychoanalytic accounts from the late 19th and early 20th century continue to operate in 21st-century cognitive neuroscience: (1) psychedelic drugs produce their characteristic diversity of effects because they perturb adaptive mechanisms which normally constrain perception, emotion, cognition, and self-reference, (2) these adaptive mechanisms can develop pathologies rooted in either too much or too little constraint (3) psychedelic effects appear to share elements with psychotic symptoms because both involve weakened constraints (4) psychedelic drugs are therapeutically useful precisely because they offer a way to temporarily inhibit these adaptive constraints. It is on these four points that EBT, IIT, and PP seem consistent with each other and with earlier filtration and psychoanalytic accounts. EBT and IIT describe psychedelic brain dynamics and link them to phenomenological dynamics, while PP describes informational principles and plausible neural information exchanges which might underlie the larger-scale dynamics described by EBT and IIT. Certain descriptions of neural entropy-suppression mechanisms (EBT), cause-effect information constraints (IIT), or prediction-error minimization strategies (PP, FEP) are loosely consistent with Freud’s ego and Huxley’s cerebral reducing valve.
Qualia Computing comment: As you can see above, 21st century theories of psychedelic action have a lot of interesting commonalities. A one-line summary of what they all agree on could be: Psychedelics increase the available state-space of consciousness by removing constraints that are normally imposed by standard brain functioning. That said, they do not make specific predictions about valence. That is, they leave the question of “which alien states of consciousness will feel good and which ones will feel bad” completely unaddressed. In the following posts about the presentations of members of the Qualia Research Institute at The Science of Consciousness 2018 you will see how, unlike other modern accounts, our Qualia Formalist approach to consciousness can elucidate this matter.
