What is an internal brain illusion?
An illusion is when the brain perceives something incorrectly.
Almost all illusions are related to the brain incorrectly perceiving something outside of itself.
An internal brain illusion is when the brain perceives itself incorrectly.
I propose that music, or more specifically the intensity of musical emotion, is the result of such an internal illusion.
Why and how does the brain perceive itself?
We have no direct conscious perception of our brains.
We think that we know our own minds, but science tells us all sorts of things about how our brains work that we would never have guessed just by thinking about it.
One thing we know about our brains is that our perceptions are represented by neural activity, ie the activity of neurons, which are a particular type of brain cell.
We know that specific perceptions are represented by the activity of specific sets of neurons, and similar perceptions are usually represented by "maps" of neurons within specific regions of the brain. Many of these maps are called cortical maps, because they happen to exist in the cortex, which is part of the brain.
There are cortical maps that represent all the different kinds of perceptions that we can have about the world outside our bodies, and there are cortical maps that represent our perceptions of things in our own bodies.
It turns out that there aren't any such cortical maps responsible for perceiving the brain itself. There is a sense in which the brain does not perceive itself.
Glial Cells and Glial Perception
But neurons are not the only kind of brain cells.
There are other brain cells, which collectively are called glial cells.
One thing that some of these glial cells do is monitor and regulate the activity of neurons.
So there is a sense in which these glial cells do "perceive" various characteristics of the activity of neurons, and this constitutes a type of perception where the brain is indeed perceiving itself.
We can therefore consider the possibility that there may be situations where regulatory glial cells could incorrectly perceive characteristics of neural activity.
Given that the job of these glial cells is to regulate neural activity, one possible consequence of incorrect perception of neural activity would be the incorrect regulation of neural activity.
"Glial perception" does not directly translate to conscious perception of any kind, but we would likely be consciously aware of the effects of any dysregulation of neural activity.
How to create a glial illusion
So, how would one go about creating an internal brain illusion, ie a glial illusion?
Most illusions are the result of some degree of contrivance.
That is, the perceptual systems in our brains are well adapted to correctly perceive the world around us and our own bodies, if not perfectly, then well enough to survive and prosper most of the time.
So in order to cause a perceptual system to fail, it is necessary to contrive some stimulus that lies outside the normal range of stimuli encountered in daily life, and which exposes the limitations of the mechanisms that make up that perceptual system.
To achieve a glial illusion, it would be necessary to contrive particular patterns of neural activity designed to confound the aspect of glial perception that we are attempting to confound.
But of course there is no easy way to directly manipulate patterns of neural activity. (Short of drilling holes in your skull, and having someone poke very fine electrical wires very precisely into your brain, and doing all that without doing any significant damage, which in practice lies way outside the capabilities of even the most sophisticated modern medical technology.)
To contrive a particular pattern of neural activity, it is necessary to contrive an external stimulus, such that the perception of that stimulus results in the desired pattern of neural activity. Which then results in the false glial perception of that neural activity.
I propose that music is such an external stimulus.
That is, music is a contrived stimulus such that the perception of music creates patterns of neural activity which confound some aspect of the glial perception of that neural activity, which ultimately causes some type of dysregulation of neural activity, which we finally feel as the feeling we get when we listen to music.
And it is important to note that a glial illusion is not an ordinary perceptual illusion.
The perceptions represented by the contrived neural activity are not incorrect – it is only the glial perception of that neural activity which is incorrect.
That is, our perception of the sounds of music, and of the various characteristics of those sounds, such as timbre, pitch and tempo, is not at all false. The only false perception is the internal perception, that is, the brain's own perception of its own activity in response to the external auditory stimulus (ie the music).
A Full Theory
What I have given above is an explanation of the basic concept of an internal brain illusion, or more technically a glial illusion.
By itself, this hypothesis cannot provide a full explanation for the existence of music.
There are two major questions that need to be answered to arrive at a full theory of music based on the concept of a glial illusion.
The first is: What actually is the false perception caused by the glial perception of neural activity when listening to music? (And there are other questions that follow on from that, ie what kind of dysregulation results from that false perception, and how does that correspond to our actual subjective experience of listening to music?)
The second major question has to do with what the glial illusion can explain, and what it can't explain.
It is most likely that glial illusion can only explain the intensity of musical emotion (ie the emotional feelings caused by listening to music).
This leaves completely unanswered the question of what determines the quality of musical emotion, or even why such a thing as musical emotion should occur at all.
In the following sections I give a brief outline of my current hypotheses which attempt to answer those questions. (More detailed explanations of some of these hypotheses can be found in other recent articles in my blog.)
Specific False Glial Perception Hypothesis: Estimated Speech Tempo
With regard to the details of the false glial perception, and what type of neural dysregulation is caused by that false perception:
- Word-based spoken language has the property that the meaning of an utterance is independent of the speed at which the words are spoken. (The speaker's tempo may sometimes provide additional context to what is being said, perhaps telling us something about the speaker's state of mind, but it does not have any significant effect on our understanding of the meaning of what is said.)
- In other words, the perception of the meaning of speech is time-scaling invariant. Faster or slower, our perception of the meaning is the same.
- Time-scaling invariance of perception is unique to speech. This contrasts with the perception of the meaning of perceived motion (be it of things, people or other animals), which is not time-scaling invariant. The speed at which something moves is almost always relevant and significant.
- The human brain has recently (ie within the last one or two million years) evolved an adaptation to globally adjust the speed of downstream neural processing to match the tempo of speech being listened to.
- This global adjustment is managed by glial cells, and as such, it requires glial cells in the auditory cortex to estimate speech tempo.
- The relationship between glial cells and neural cells does not allow glial cells to "know" with any degree of precision the meaning of activity in any specific set of neurons.
- So even though there may be neurons with activity that represents the perception of speech tempo, that information is not available to glial cells, and glial cells must provide their own independent method of estimating speech tempo.
- Glial cells in the auditory cortex estimate speech tempo by measuring the frequency with which different neurons transition from a state of inactivity to a state of on-going activity.
- Music confounds this method of estimation by causing neural activity in the auditory cortex to occur in a manner such that, in each cortical map representing one particular aspect of perception, some neurons are constantly active and most of the other neurons are not active at all. As a result, the transitions from fully inactive to fully active do not occur, and the glial estimate is way lower than the correct value.
- As a result, the glial cells adjust the operating characteristics of "downstream" neurons (ie all those neurons processing the meaning of the speech being listened to) to match a slow rate of speech, even though the actual speech may be coming in at a normal rate.
- The ultimate consequence of this mal-adjustment is over-activation of the neurons representing the emotional meaning of the speech being listened to.
Explaining the existence and quality of musical emotion: Protomusic
The second question is about what determines the quality of musical emotion, ie which particular emotion it is that music expresses, such as happiness, sadness, excitement, etc.
The hypothesis that music causes false glial perception of speech tempo can only explain the 1-dimensional intensification of musical emotion. It cannot explain the existence or multi-dimensional quality of the original unintensified musical emotion.
Indeed that hypothesis only refers to speech, and speech is not quite the same thing as music, even though music can contain a form of speech (ie song lyrics).
To explain the existence and quality of musical emotion, we have to suppose the existence of some other perceptual system which is capable of responding to some category of auditory stimulus which expresses the same emotions as music, only less intensely.
There is no evidence that such auditory stimuli occur, at least not in the modern world.
If they did ever exist, how and why were they produced?
A set of sounds that have prescribed meanings are in effect a type of language. A set of sounds that have emotional meanings constitute a language of emotion.
The most plausible hypothesis that I can think of, to explain our ability to understand a non-musical language of emotion that no one ever actually produces, is to suppose that prehistorically such a language did actually exist.
A good name for such a language, if it did exist, would be protomusic – a prehistorical non-musical ancestor of music.
That is, protomusic was a non-musical language of emotion which our prehistoric ancestors used to communicate emotions to each other, which we, as modern humans, still retain the ability to understand.
This protomusical language is no longer actually produced by anyone, because it has been completely superseded by word-based spoken language.
"Normal" protomusic no longer exists as a thing to ever be heard, but, in the illusory form of music, it still does exist.
(There is an additional twist, which is that although we retain the ability to understand this hypothetical protomusic, we have lost the desire or willingness to respond to protomusical communication in the pragmatic way that spoken speech creates a desire to evaluate and respond to anything that is said to us. That is, even though music can create a very strong feeling that it is communicating something, we never actually use music as a pragmatic form of communication, for example we never introduce music into a normal conversation.)
A final question has to do with the relationship between the glial and neural processing of speech, and the processing of protolanguage.
In particular, the proposed mechanism of glial estimation of tempo exists to serve the optimisation of processing of speech. It can be argued that such a mechanism would not help with the perception of protomusic, because in the case of music, tempo actually affects the perceived emotional quality. For example, sad music is usually slower, and happy music is usually faster. This relationship between tempo and emotional quality would have to apply also to protomusic, which would imply that the perception of protomusic was not invariant under time-scaling, and therefore glial estimation of tempo would not serve any useful purpose in that case.
So how could a glial illusion relating to the glial processing of speech have an effect on the perception of any kind of "illusory" protomusic?
The simplest answer is that the glial mechanism used to optimise the processing of the meaning of speech is not that specific in terms of which neurons are observed and which other neurons are regulated, or in the illusory case, dysregulated.
Regions in the auditory cortex processing speech would be either the same as or very close to the regions that, prehistorically, processed protomusic, and the regions "downstream" of speech would be similar to those "downstream" of protomusic.
Because of the non-specificity, the glial perception of tempo and the consequent regulation or dysregulation would apply equally to speech or music, and equally to the emotional meaning of each of those two.
Exact Repetition
Specific features of music exist because they contribute to the glial illusion.
Examples of such features are:
- Scales
- Harmonic intervals
- Regular beats
Underlying all of these features is some type of exact repetition.
For example:
- If all the pitch values of notes in a melody come from a finite set of pitch values in a scale, then most of the notes will have a pitch value that is an exact repetition of the pitch value of the same note when it occurred previously in the melody.
- If two notes separated by a harmonic interval are played in succession, then some of the harmonics of the second note will be identical to some of the harmonics of the first note, ie those harmonics are exactly repeated.
- Regular beats are intrinsically repetitive. Music not only contains regular beats, it is almost always based on nested regular beats, ie beats with frequencies that are simple multiples of each other. For example, in 4/4 time with 16th notes, there are beat periods of 16 per bar, 8 per bar, 4 per bar, 2 per bar and 1 per bar, and all the notes in any rhythm for that time signature will belong to one of five specific subsets of those beat periods, ie 16, 16 & 8, 16 & 8 & 4, 16 & 8 & 4 & 2 or 16 & 8 & 4 & 2 & 1. So, in that particular example, if notes are identified according to the regular beat periods that they belong to, then only those 5 identities exist, and all others are excluded.
Exact repetition contributes to the glial illusion, because the exact repetitions cause exact repetitions of a small set of perceptual values, and other perceptual values don't occur at all.
In effect we can understand the mechanism of glial tempo estimation as being based on counting the occurrence of perceptual events, but the actual mechanism only counts the occurrence of different perceptual events, and heavy use of exact repetition causes the resulting estimate of tempo to be much lower than it really is.
Another thing we can observe about the features of music that contribute to the glial illusion is that those are also the features of music that require the performing musician to have skill and precision.
Proto-musical did not have those features, and it could not require unusual skill or precision, because it existed as a means of communication. It had to be something where anyone and everyone could produce utterances that were readily intelligible to all listeners, just as is the case with modern word-based spoken language.