- The primary function of music is to trigger a temporarily altered state of mind in which emotions experienced from daydreaming are intensified.
- The content of music has no intrinsic meaning.
- The criterion for musicality is that for one or more categories of auditory percept, certain perceived values occur and other values don't occur over a certain time-frame (the time-frame being comparable to the durations of musical items). For example, pitch values from the notes in a musical scale occur, and pitch values in between the notes don't occur.
- The occurrence and non-occurrence of perceptual values corresponds to patterns of activity and inactivity in the cortical regions which process the corresponding perceptions.
- The response to music has evolved, possibly from a single mutation, from a pre-existing brain function. This "pre-cursor" to music serves some different function in relation to auditory perception.
- One possible candidate pre-cursor is a response to speech-like sounds, which would be active when an infant is learning its first language, and needs some criterion for distinguishing the types of sounds likely to be part of a spoken language from other types of sounds.
- This pre-cursor would involve a response to activity in cortical regions involved in perceiving speech-like sounds - the only difference between that and the response to music is that the response to music involves a response to both activity and inactivity.
Compulsive ("Maladaptive") Daydreaming
There exists a group of people who consider themselves "addicted" to daydreaming - so-called maladaptive daydreaming. Maladaptive daydreamers daydream much more than "normal" people do, and they seem to enjoy it more strongly.
There is a common association with music and maladaptive daydreaming, although the exact nature of the association varies between different daydreaming addicts.
However a common observation is that music is either a trigger for a daydreaming 'session', or the daydreamer is addicted to a combination of listening to music and daydreaming.
Music and Movies
Music is routinely used to enhance the emotional reponse of audiences to the events happening in a movie.
The use of music in movies is so consistent, that if we starting watching a movie with no music at all in it, it feels a bit strange – we notice its absence.
Movies are somewhat similar to daydreams: they create responses to imagined events which are known by those who experience them to not be actually happening. So it is plausible that if music exists to alter emotional responses to daydreams, then it will have a similar effect on emotional responses to movies.
If This is Correct, What Then is the Value of Music?
Daydreams are a significant component of human experience. Most people spend quite a large proportion of their waking hours daydreaming – at least 30%, (eg see http://news.sciencemag.org/2010/11/daydreaming-downer).
Something that happens so much must be important, otherwise the tendency to do it would long since have evolved away.
And if daydreaming is important, it is likely important that the processing of information within daydreams be carried out correctly, and in particular that the processing of emotional information be carried out correctly.
One thing known about physical representations of thought is that thinking about something activates the same neurons as the thing itself, but not so much.
In as much as thought-related brain activity is a reduced version of reality-related brain activity, the wiring of the brain must be capable of processing the reduced thought-related activity correctly, even though the activity levels are reduced. (This can be compared to how we would process hypothetical information in a computer system that processes live information from an external source – instead of sending "reduced intensity" information through the live system, if such a thing was even possible, we would create a full copy of the live system, and re-wire its inputs and outputs to provide hypothetical inputs and to record and analyse the hypothetical outputs.)
My hypothesis about why daydreaming emotions need to be intensified, at least some of the time, is simply this: the brain's ability to process 'reduced' levels of activity for processing hypothetical information somehow doesn't work so well for processing emotional information, and the only way to better process hypothetical emotions is to intensify them to a level closer to that of "real" emotions, but in such a way that avoids confusing the hypothetical with the real.
And music is the solution that the brain has arrived at (via an evolutionary accident) to solve this problem.
How I Arrived At This Theory
(Added 17 May 2015)
"The purpose of music is to trigger the intensification of emotions in daydreams": this seems like a short and simple idea. Yet it took me decades of observing, analysing, hypothesizing and un-hypothesizing to get there.
What follows is some of the main steps in that journey.
Idea 1: Music is a Thing
It might seem obvious that music is a "thing" – we use a single word to describe it. And in as much as music is composed of smaller things, like notes, or scales, or chords, it seems obvious that those things only exist to serve the larger musical purpose.
One school of thought (or at least one man's idea) is Steven Pinker's "Auditory Cheesecake", which asserts an analogy between the niceness of music and the niceness of cheesecake, in each case assumed to be the result of multiple separate factors each contributing individual components of niceness to the overall niceness, of music or cheesecake in each case.
However, my own feeling is that music is one thing, which is different from other things, and which requires its own particular explanation.
Idea 2: The Niceness of Music requires Biological Explanation
If music is a thing, and it's not quite like any other thing, then, given the subjective and distinctive niceness of music, it must serve some fundamental and unique biological purpose.
Actually, there is one category of "nice" things which have the property of niceness where this niceness is subjectively distinctive, and which do not serve a biological purpose: the category of mind-altering chemicals, ie recreational drugs.
We could consider the hypothesis that music is somehow like a drug. But music and recreational drugs are quite different in their mode of action, even if the results of those actions seem similar, in as much as music has a "mood-altering" effect:
- Recreational drugs are chemicals which must be ingested into the body, and once in the body, they act directly on the brain or nervous system.
- Music, however, only acts through existing perceptual systems.
Because drugs are chemicals that act directly on the mechanisms by which the brain processes information, they effectively provide false information to the brain.
Whereas musical "information" is not false: musical sounds are real sounds, which have to be heard by the ear, and then processed by the same elements of the nervous system and brain that process all other sounds.
Idea 3: Music has Different Aspects, but "Musicality" is Only One Thing
Different aspects of music, such as melody, harmony, rhythm, and structure, are probably processed in different parts of the brain – because that's how the brain works in general. For example, different aspects of object perception, such as colour, shape, size and motion, are known to be processed separately.
But if music is just one thing, then all these different aspects of music must somehow arise from one common criterion.
(Some people ask the question: "Which is primary, pitch or rhythm?". People who ask this question are already assuming that different aspects of music are so different that they can't derive from a single common criterion. People who ask that sort of question just haven't tried hard enough to look for such a criterion.)
Idea 4: Musicality is a Function of Constant Patterns of Activity and Inactivity in Cortical Maps
Following the idea that an understanding of music would require an understanding of cortical "maps" that process different aspects of music, I developed hypotheses about what maps might process those different aspects, assuming that such maps served some function in sound or speech perception when processing information about non-musical sounds.
Without direct access to real-time information about activity in live brains processing music, such analyses can only be hopeful guesswork.
However one pattern did emerge when I considered hypothetical cortical maps for processing melody and rhythm. What I observed in both cases, is the contrast between things that do happen, and things that don't happen, for the duration of a musical item.
In the case of melody, it is the pitch values of notes in a scale that occur, and the pitch values between the notes that don't occur. (And even if there is no direct representation of actual pitch values mapping pitch to position as a simple linear relationship, there is still likely to be some representation of pitch values positionally, such that non-scale pitch values would activate neurons within a cortical map that are otherwise inactive while processing an item of music on that scale.)
In the case of rhythm, we can consider a map that positionally represents regular beats within the rhythm. For a typical time signature, such as 4/4, there might be just 5 regular beats that are persistently perceived, within the range of period from shortest note length to bar length (ie 4 notes, 2 notes, 1 note, 1/2 note, 1/4 note). So the corresponding 5 regions in the cortical map would be active, and the in-between regions would be inactive.
These two examples raised the possibility, in my mind, that maybe all aspects of music correspond to constant patterns of activity and inactivity in corresponding cortical maps, and that these constant patterns are the sole criterion of musicality.
Idea 5: Music is a Super-Stimulus
The idea of constant patterns of activity and inactivity was intriguing, but I could not find any totally convincing explanation for why such patterns would have any meaning that was biologically relevant.
My best idea was that music is somehow a super-stimulus for an otherwise very subtle aspect of speech perception. For this theory to be plausible, this subtle aspect of speech perception would have to be so subtle that it was un-noticeable to the perceiver, perhaps acting very slowly in the long term to alter our perception of those whose speech we listen to, or to alter our response to the content of what they say.
Idea 6: The Emotional Effect of Music is Primarily to Act On Hypothetical Emotions
It is a common observation that music has an emotional effect – often stated as an observation that items of music contain or express emotions, although sometimes it is difficult to express precisely what emotion is expressed by any particular item of music, beyond a general classification of some musical items as happy, sad or angry.
What did occur to me is that music is a very persistent accompaniment to movies.
Which led to the hypothesis that music has its strongest effect on emotions that are about things which are known by the listener not to be real. And to the further hypothesis, that the primary purpose of music is in relation to hypothetical emotions, and not real emotions.
In evolutionary terms, movies as we know them are a recent technological innovation, and are certainly preceded by the creation and enjoyment of music which is known to be at least 9000 years old (and could easily be much older, for example as old as the modern human species, or even older).
Before movies, there were human story-tellers.
However, most of the stories that most people experience are the stories that they make up and experience for themselves inside their own heads. We don't normally call these "stories", we call them daydreams.
So the major target for the emotional effect of music is not necessarily emotions from stories created by other people – it is the emotions of our own daydreams.
The music that we listen to may be created by others, but the emotions that the music affects are the emotions from our own internal daydreams – and probably it has always been this way.
Idea 7: Musical Content Has No Intrinsic Meaning
The difficulty of assigning an actual meaning to the content of music, or to the criterion for musicality according to my hypothesis above (ie constant activity/inactivity patterns), does suggest an alternative: that music actually has no meaning.
In other words, there is a criterion for musicality, but this criterion for musicality is quite arbitrary, and the only thing that matters is the consequence of perceived musicality, which is to trigger an altered stated of mind in which the emotions of daydreams are intensified.
Probably the only thing that matters about the criterion for musicality is that some effort is required to initate it, that the effect disappears as soon as the music disappears, and that the person experiencing the effect of music is somewhat aware of the effect that the music has on the emotions in their own daydreams. In other words, music creates an altered state of mind which can be entered into, which improves the processing of emotional information in daydreams, but which does not lead to any on-going confusion between daydreams and reality.
Idea 8: The Musical Precursor
If the content of music has no intrinsic meaning, then the criterion for musicality is somewhat arbitrary, and is probably the result of an evolutionary accident.
The most plausible evolutionary accident is one that involves just one single simple change to some existing functionality.
This existing functionality must be similar to, but not the same as our response to music.
Music may have replaced the precursor functionality, or, it may somehow have arisen as an altered duplicated of it (which would avoid the cost of losing the precursor, which is presumably of some on-going importance).
Precursor Hypothesis: Interest in "Speech-like" Sounds
The criterion of constant patterns of activity and inactivity in certain cortical maps is a relatively simple criterion in itself. The relevant cortical maps appear to be those involved in speech perception, or at least similar to those involved in speech perception.
A very simple change to this criterion is to remove the requirement for inactivity. This would then correspond to a simple criterion for activity in regions involved in speech perception, which is the same thing, more or less, as a response to speech.
Indeed, such a criterion would more precisely correspond to a reponse to speech-like sounds, and we can imagine that such a response is important in the human brain of an infant learning its first language: there is a special set of sounds to which it should be attending to in a special way, because probably those sounds are speech (even though the specific details of actual languages can vary quite a lot, and there is no 100% accurate criterion that can distinguish speech sounds from non-speech sounds for a previously unknown human language).
This particular pre-cursor hypothesis does happen to explain one particular and peculiar feature of music, which is the relationship between "motherese" and music. Motherese is specially altered language that mothers tend to speak to their infant children, presumably because infants are more inclined to pay attention to such speech. (And lullabies, in the sense of music to be sung by mothers to their young children, do also share some of the characteristics of motherese.)