Analogies of Human Consciousness

This article discusses three analogies that QRI believes to illustrate properties of human consciousness. The overarching model of the brain as a hybrid computer is discussed in the first section.

The Hybrid Computer Model

The conventional view within neuroscience is that computation in the brain can be analyzed in terms of the operations of individual neurons. While this view admits significant differences between the brain and a digital computer (the brain has millions of concurrently active units and incorporates analog logic), it still models the brain as performing standard computation, meaning that it can be understood in terms of simple logical operations performed on discrete informational units.

In QRI's view, the above at least partially describes the unconscious parts of the brain. However, whereas the conventional view holds that consciousness is an emergent phenomenon from the same architecture, QRI believes that consciousness arises due to a separate component of the brain, which is shaped primarily by a topologically segmented pocket of the electromagnetic field. Unlike the unconscious component, the conscious component performs nonstandard computation, thus obeying a widely different set of principles. The remaining section will go into more detail on the properties of both components.

Anatomy

In the hybrid model, the unconscious component comprises a majority of brain mass, including every area that intersects the brain's surface (such as the cerebral cortex). Conversely, the relevant part of the electromagnetic field is likely contained entirely within subcortical regions (possibly the thalamaus). As such, the electromagnetic pocket is both spatially confined by the physical skull (which is why people cannot merge their consciousness by moving closely together) and isolated from electromagnetic interference by the surrounding brain (which is why electromagnetic radiation has no noticeable effect on consciousness).

Function

As mentioned above, QRI's views of the brain's unconscious component largely align with the conventional view (with the most significant additions relating to communication, which is discussed in the subsequent section). As such, there is a vast literature on reverse-engineering the function of this component within academia, most prominently within computational neuroscience.

Conversely, the purpose of the analogies discussed in this article is to provide a starting point for thinking about the functions of the conscious component. As a preliminary comment, note that while such "field computations" are alien relative to our conventional understanding of computation, they still operate entirely within the laws of physics. QRI generally advocates for broadening our conception of computation to include processes that rely on complex physical mechanisms.

Communication

Since the brain's conscious component – the segmented part of the electromagnetic field – is proposed to spatially overlap sub-cortical brain regions, those regions can "read off" properties of the field and communicate them to the unconscious brain through regular synaptic connections. (Note that the thalamus has a rich set of such connections to many points in the cortex.) Thus, there is a straightforward mechanism for "outward" communication, allowing the unconscious component to receive and execute consciously created motor commands.

Conversely, "inward" communication (i.e., from the unconscious to the conscious component) is nontrivial because the voltage levels of individual neurons typically vary by only around 0.1V, which is too small to meaningfully affect the electromagnetic field at distances on the order of centimeters. However, synchronized neuronal activity (also known as brain waves) can greatly amplify the strength of the electromagnetic signal, thereby creating a disturbance strong enough to reach the conscious component. Thus, contrary to the conventional view, brain waves have a major functional role within this model, serving as the primary vehicle of inward communication. Moreover, the model permits the unconscious component to carry out internal computations that do not synchronize neuronal activity without interfering with the conscious component.

Note that since the conscious component has no direct access to the external world, all conscious percepts are of this kind, i.e., all qualia are signals broadcast by the unconscious component. This is why QRI endorses indirect realism: what we perceive is a world simulation created by our brain, not the external world itself.

Consciousness as a Sheet of Paper

An illustration of how the visual world appears to us, by psychonaut Steven Lehar. Notice the spherical geometry, the changing spatial resolution, and the inclusion of a self-model at the center.

Recall from the preceding section that human consciousness does not directly interface with the external world. The analogy to a sheet of paper is one where the brain continuously paints new qualia, thus creating a private world simulation. Importantly, this view asserts that consciousness is inherently spatial. Whereas regular computers must encode spatial information as numerical coordinates, our consciousness can represent spatial information as spatial information, i.e., the spatial relationships of external objects are replicated within our world simulation.

Another property relates to the continuity of consciousness. Whereas classical systems represent information as bit patterns, which are discrete bundles of information, the spatial nature of consciousness means there is no decomposition into modular units: any region of the sheet is an equally valid subset. Note that the same continuity applies not only to the qualia we actually experience but also to the state-space of qualia, i.e., the set of all things we could experience. There is no clean mapping from input types to qualia – there's no one qualia of red, tiredness, or a B-minor piano chord. Instead, qualia types are more accurately viewed as painting styles or archetypes, and the same input may be drawn in subtly different ways. In particular, meditative techniques like equanimity can be applied to arbitrary qualia to make them more fulfilling and/or less painful without changing their information content; this may be analogized to an art style with smoother outlines and fewer rough edges.

We also appear to have a near-universal ability to introspect on elements of our consciousness, which can be likened to the ability to make "photographs" or "screenshots" of arbitrary regions of the sheet, with the important caveat that the resolution is necessarily degraded (e.g., we can't remember every detail of a visual scene). Note that this introspective ability is crucial for the viability of psychonautics.

The final and most exotic property is that the geometry of consciousness is malleable, analogous to how a sheet of paper can be deformed. (Since it is roughly spherical by default and generally forms a closed manifold, a better analogy may be to the surface of a balloon; however, note that it is really three-dimensional, so no simple analogy is perfect here.) Psychedelic substances and intensive meditative states may alter this geometry, thus explaining a category of eccentric phenomenal reports. Furthermore, while qualia are spatially distributed, it is also the case that we often perceive them as unified objects – e.g., a bottle standing on a desk is perceived as a singular object with a degree of separation from the environment. This effect, which is referred to as local binding, can similarly be analogized as local deformations of the paper that impose a partial boundary around the relevant regions.

Consciousness as a Musical Instrument

The similarity of human consciousness to an instrument relates to the consonance and dissonance between musical tones. The analogy asserts that any two qualia in the same field of consciousness may be consonant or dissonant with each other (or anything in between), leading to an increase or decrease in valence, respectively.

To a first approximation, two musical notes will sound consonant if the ratio between their frequencies is close to a rational number with a small denominator, though this idea is difficult to quantify. The emerging paradigm of brain eigenmodes suggests that a similar thing may be true for qualia, although published research measures spatial rather than temporal frequencies.

The implications of this analogy are most easily observable on the macro level, where they relate to moods and their effect on the valence associated with activities. It is not uncommon for the same task (e.g., cleaning a room or running an errand) to be pleasant at one time but unpleasant at another. In the analogy, this happens because the overall field of consciousness has a certain "tone" (sometimes called its noise signature), which may be consonant or dissonant with the qualia exhibited during the activity.

The analogy also applies on smaller scales, where it may be less apparent and more amendable to intervention. Attention can locally bind different qualia together, which tends to amplify their mutual consonance or dissonance. Thus, paying attention to, e.g., two sources of pain may mitigate their combined negative valence. Note that it is also possible for attention to modulate the tone of a quale over time, which is why sustained attention on a mental object tends to be pleasant even if the object itself is intrinsically neutral.

Another application is actual music, which, for many people, has the potential to evoke a wide spectrum of qualia tones. (Note that since music often has a strong emotional component, there is not necessarily a close link between qualia tones and literal musical tones or notes.) Consequently, the enjoyment derived from a given piece of music can vary greatly between different contexts.

Consciousness as a Resonance Hierarchy

The resonance hierarchy is a proposal for how the consonance and dissonance between qualia discussed in the preceding section can be utilized for computation. This section will first explain the concept and then sketch possible implications. Out of the three analogies discussed in this article, the resonance hierarchy is the only one that might reasonably be called a computational model, and it may even apply to parts of the unconscious brain, although with a different physical mechanism implementing resonance.

Resonance

An example of the matching process in the jagged lines toy setting. Note that the first template is an imperfect match (even after resizing) but would still be the template of choice since both alternatives are inferior. Conversely, if a larger set of templates were available, the best match might account for the fourth spike in the input.

Suppose we are given the following:

• a class of objects
• a fixed relationship determining how well any two of them fit together, or "resonate"
• a set of such objects functioning as templates
• a mechanism that, given a newly encountered object, finds the template with the best fit (i.e., the one that has the highest resonance with the new object)

As a toy example, you may think of the set of jagged lines, and the relationship as how well their spikes align, possibly after one of the lines is squeezed or stretched. The graphic to the right illustrates an example with one input and three available templates.

In a formalization of this setting, we have a set ${\mathcal {O}}$, a resonance function $r:{\mathcal {O}}^{2}\rightarrow [0,1]$, a set of templates $T\subset {\mathcal {O}}$, and a mechanism that, given any $i\in {\mathcal {O}}$, finds ${\text{argmax}}_{t\in T}r(t,i)$.

Given this setting, suppose our goal is to classify novel objects into categories. In our line example, one such category could be "triangle spikes", i.e., lines in which a tall spike is surrounded by two smaller ones. To do this, we can create a template that has high resonance with objects within that category and low resonance with objects outside it. Then, our mechanism will determine the template as the best fit whenever an object in the category is encountered but not for objects outside the category, thus classifying the object accurately.

In the application to consciousness, the objects are the qualia of sensory inputs, and the templates are stored qualia patterns that represent features. Thus, whenever a novel qualia pattern enters consciousness, the brain will try to match it with an existing pattern (that serves as a template). If you look at a spoon, you may notice that the gestalt in your consciousness has an additional component above and beyond the raw pixel information (or even the information of lines and surfaces). This additional "spoon-ness" is the qualia template that resonated best with the input. The analogous process occurs for any other object, assuming we recognize it as something familiar.

Recall from the instrument analogy that consonance results in high valence states. Because of this, the goal of finding the qualia pattern with the highest resonance for a given input (i.e., the mechanism mentioned above) is synonymous with the maximization of valence. Therefore, if the only thing consciousness intrinsically does is to maximize valence, then the resonance hierarchy is a principled way in which consciousness can be utilized for feature recognition. Note that this is a form of nonstandard computation.

Hierarchical Application

The "hierarchy" part of the model refers to the idea that the mechanism sketched above is applied recursively. In other words, there may be qualia templates chosen in response to patterns that are themselves templates (and thus resonate with lower-level patterns). If you look at a house, for example, you can imagine qualia templates for windows or the front door, but also a high-level template for the entire gestalt. This extension is analogous to hierarchical feature recognition in artificial intelligence, which is present in virtually all modern models, including all neural networks with at least one hidden layer.

Phenomenal Character vs. Narrative Content

When it comes to the study of qualia, especially when consciousness is influenced by meditation or psychedelics, the model implies that reports of low-level properties (shapes, energy flow, flickering, strobing effects, etc.) tend to be significantly more useful than reports of high-level properties (hallucinating creatures, feeling a sense of awe, etc.) because

1. high-level properties likely correspond to templates high up on the hierarchy, which tend to be different for different people; and
2. even for a single person, the same template may be matched to several possible patterns.

Thus, a report of seeing a machine elf under LSD will shed little insight on the effects of the substance, whereas an observation that visual content flickers with 17Hz may reveal a general property.

QRI tends to emphasize the above point because reports of psychedelic experiences (known as "trip reports") can provide valuable data for any theory of consciousness, similar to how particle accelerators provide essential data for physical theories. Due to the numerous difficulties in providing such reports, it is particularly important to follow these guidelines to increase the chance that the resulting document is useful.

Resources

• Cartoon Epistemology by Steven Lehar – An article outlining the basics of Steven Lehar's model of vision, including its geometry.
• Guided Meditation: Textures of Valence - Consonance, Dissonance, and Noise – a guided meditation by Andrés exploring aspects of the instrument analogy.
• Guide to Writing Rigorous Reports of Exotic States of Consciousness – an article by Andrés relating to the focus on phenomenal character over narrative content.
• Harmonic Gestalt – a presentation by Steven Lehar for his model of visual processing, which heavily relies on resonance.
• Conscious Mind, Resonant Brain: How Each Brain Makes a Mind – a book by renowned neuroscientist Stephen Grossberg laying out a family of computational models for the brain that rely on resonance.