Beyond Identity: Toward a Logic of Similarity

Take a moment to reflect on what reality would be like without similarity.

Reality would be composed of absolutely unique qualitative identities with nothing at all linking them together into categories or types.

 You would be unable to experience a headache at the internal imponderability of such a reality, because the similarity between distinct low-valence qualities that we label pain simply would not exist to unite distinct pain qualia into a single coherent self-similar sensation.  

Indeed, there would be very little left that's familiar: 

Quale types, valence types, and form types simply would not exist. It would be a universe with no possibility of general terms and concepts, because absolutely everything in existence would have a completely unique and non-overlapping identity. And thus, it would be a universe about which we could have no knowledge, because concepts are contingent upon the existence of similarities between forms and between qualia that they can describe. 

I call the state of affairs that would result from maximally particular identity being more ontologically fundamental than general similarity "The Identity Dust Universe". 

The problem it presents is this:

If non-composite identities were the bottom-up building blocks of reality, we could never come to know similarity at all, because identity is a binary yes/no relation that doesn't admit degrees. No two objects A and B could be more similar than any other pairing (e.g. A and C) in principle, because there would be nothing but fundamental and distinct identities, and thus nothing for similarity to consist in. 

"Identity dust" is not "mind dust." It can still exist given phenomenal binding. If identity dust were temporally bound into a sequence, we would just experience a sequence of 100% distinct identities, where no two elements are more related by similarity to one another than are any other two, and where there are thus no special relationships to describe. 

Why are we not Identity Dust if we live in a bottom-up universe that is built from the ground up by maximally particular identities?

I call this the Hard Problem of Similarity, and I think I have a highly unorthodox solution to it.

But first, let us explore what seems to be the leading "Bottom-up" account of similarity: 

The Blurry Identity Model states that similarities are "blurred" identities. However, an objection immediately leaps to mind: if only maximally particular identities exist as the sole contents of reality, then what is there to constitute "blurry" (i.e. less than maximally particular) identities? Blurring is an instance of the very phenomenon that is in need of explanation. What can it mean for a quality to be a blurry representation of another two, but for it to be indeterminate relative to which it is, simultaneously being both while exclusively being neither?

 Purple, for example, is similar to both red and blue in a way in which it isn't similar to both yellow and green. And red and blue don't magically make purple "by emergence" when we become sufficiently distant from an array of purple and red dots.

Strong qualitative emergence of this sort is a magician's hand gesture intended to distract from the absence of a model, while giving the illusion of explanation.

Rather, the mind uses a single distinct quale type to represent two distinct but closely related colors. And this single distinct quality isn't just an array of spatially separate instances of red and blue. We can imagine a world in which instances of red and blue don't blur (become replaced by purple) with distance or rapid rotation. 

[Side note: given that rotation is just the rapid alternation of two colors at a single point, it's direct empirical evidence of temporal binding. Red one brief moment and then blue the next generates purple by temporal blurring, which could only occur if information isn't confined to so-called time-slices, which takes us into meta-time. But meta-time is sufficiently important to justify its own article, and so I won't get into it in detail just yet.]

What would a top-down account of similarity look like? 

Before we can get into that, we must first clarify exactly what we mean by "similarity" and introduce a few new concepts.

Similarity: from our particular top down view, similarity is constituted by general qualities in which certain distinctions aren't made. Put another way, similarities aren't maximally particular identities, but general identities that undergo change within a bound possibility space. There's no reason in principle why self-presenting, self-knowing qualities have to be maximally particular. And there's no reason why the general can't be more ontologically fundamental than the particular. Indeed, if it were, many mysteries would be resolved - the experience of continuity in spite of change, and the reason for phenomenal binding, most important among them. 

Qualitative Containment: The first thing that we need to know about qualitative containment is that it isn't spatial containment. A particular shade of red doesn't contain both shade and hue like a chest contains a pair of socks; shade and hue interpenetrate within a single formless but extended quality. Sound qualia contain pitch and timbre non-spatially within a single note. Smells can have many simultaneous interpenetrating qualitative dimensions, as can tastes. And smell and taste regularly interpenetrate, as do emotional/touch qualia and visual qualia. 

Qualitative dimensions: They aren't spatial dimensions, but general identities that are more ontologically fundamental than particular identities. Examples include visual qualia, auditory qualia, emotional qualia, and so on. Think of it like a nested hierarchy of being in which meta-temporal awareness (to be explained later) contains spatiotemporal awareness and general qualia and form types, and all of which contain numberless particular shades of qualia.  

Ontological fundamental(ness): The first thing to know about the degree to which a thing is ontologically fundamental is that is not the degree to which a thing is real - for all things that can influence other real things are real in virtue of that influence - but rather it is the direction of logical contingency in qualitative containment relationships.

Qualitative Causality: See my paper on proto-intelligent qualia. 

Generality: For a system to be maximally general, it must contain every other quality as a potential state, being the interpenetration of all qualities collectively, while being none of them exclusively. General identities (similarities) are more ontologically fundamental than more particular identities; they are the spatially and temporally bound basic units of existence in which particular identities arise. The most general system would contain a connectome of similarities that are the various nested and partially overlapping Venn-diagrams that constitute the relational structure of qualia space.

Similarity Connectome: The intrinsic relational structure of reality. 

Observer: The more general, more ontologically fundamental, unitary substrate in which particular qualities arise.

Change: Change is not an instant transition from one identity to another; it is a display of different modes of a single coherent general identity. Because this general identity is logically prior to (and the container of) all of its self-distinctions, it transcends both temporal and spatial self-distinctions, and is thus phenomenally bound -- it naturally bridges spatial and temporal distinctions that it creates within itself, because it's their container and ontological medium. 

And with that we're ready to return to the question of what a top down account of similarity would look like. 

Let's start with a general summary that we'll flesh out as we progress. 

We have never seen a maximally particular quality because, just as with numbers, it is always possible to make a finer distinction, to, for example, find a color that is between any two colors. Every color thus represents a possibility space that is indeterminate relative to all the further qualitative distinctions that could be made within it. This continuum nature of qualia makes it incompatible with a bottom up structure in which maximally particular qualities somehow instantiate more general qualities, for there are no maximally particular qualities for precisely the same reason that there is no smallest number: continuums permit ever finer distinctions to be made without end. 

But what exactly is it to make a qualitative distinction? It is for an ontologically fundamental qualitative gestalt, like meta-temporal awareness - awareness in which the relative past and the relative future are co-represented and interact in a dynamic moment of time that is neither future nor past - to self-observe and individuate aspects of itself.

Observation leads to individuation because even the most seemingly dispassionate observation is goal oriented - one is always looking for something or trying to make some distinction - and this goal oriented searching is a gestalt in awareness, forming a system with the more general space of awareness in which it arises. The nature of the question - its teleology - is written in the very same qualitative language of thought as the more ontologically fundamental (more general) awareness of this sentence. Melodies are modes of your mind, not separate qualitative objects.

Indeed, there are no qualities separate from the general ontological medium of which they are modes, for absent this relational medium, they could not co-exist and relate in the ways needed to create the global similarity connectome -- the general relational structure of reality -- the very similarities that we trace in the course of forming a general model of anything at all. 

Qualitative containment comes in two species: manifest and unmanifest.

Unmanifest containment is the way in which purple contains both shade and hue: we only come to individuate shade from hue by the experience of multiple instances of a single hue that differs in virtue of its similarity relationships to various monochromatic shades. In other words, we have to see multiple shades of purple before we realize purple is a general quality (similarity) that interpenetrates another general quality (shade).

Manifest containment is the manner in which one's mind metatemporally co-represents distinct colors, qualities, and forms. In other words, its the way in which one's mind represents multiple temporally extended and spatially extended qualities in a single bound present duration.

This leads us to the Principle of Qualitative/Teleological Identity:  

To have a certain qualitative property is to have a certain teleology, and the degree of similarity between qualities is the degree of their shared teleology. This is evidenced by the fact that different sensory modalities with different functional roles differ qualitatively from one another, while the qualities recruited to any one functional role are all highly similar. Indeed, we see this even in evolutionarily novel processing, where trip elements made of exotic qualia are qualitatively self-similar through time.

And to have a certain teleology is to have a certain quality, for all teleological processes are constituted by qualities, given a conservative idealist ontology that doesn't multiply entities beyond necessity, and the fact that every goal we've ever had has been instantiated by qualia.

Qualities aren't static entities but goal oriented processes in metatime - i.e. in the present (the duration of intermingling between the relative future and the relative past).

Linear time would emerge from processing occurring both logically and ontologically prior to it - the relative past and the relative future emerge from processing occurring in metatime because it's only in metatime that qualitative causality can occur - i.e. it's only there that sequential qualities can qualitatively intermingle - for metatime is ontologically prior to sequence - as an existentially necessary condition for it - and sequentiality is a self-imposed internal separation within metatime.

The past and future are worked out together in linear-time-orthogonal metatime, where qualitative processing can actually occur. During this metatemporal processing event, there is a sort of pre-temporal feedback that goes on between what will become the relative past and what will become the relative future. When qualitative processing unfolds in metatime, between two teleological/qualitative systems, there is mutual influence and their final product defines the vague past and the even more vague future a bit more than before, setting constraints on the possibilities that can be logically actualized.

This process of ever less general - ever more particular - definition of the past and of the future exists within the present (metatime) can be conceptualized as a set of two partitioned constraint systems: one for the past and one for the future. Past constraints limit what can arise in the future. And what will become the future defines vague areas of the past so as to manifest itself within spacetime. This is how temporal structure arises from metatemporal processing. Time is a system of constraints governing the structures that can arise within metatime.

The rabbit-hole of metatime goes far deeper than this, but this seems like a good place to stop this post. Metatime should have its own article, and everything else I have to say on the holistic idealist ontology that I'm proposing is better said in a dedicated metatime post. So, to be continued in my next post: "All Music Lives in Metatime".