This series begins with The Machine in the Ghost
Back to Sensing and Perceiving || On to The Brain and Meaning
So far in this series, we have considered the need for a theory of brain, chosen a holistic and biological approach to modeling the brain, considered models for the neuron, and looked briefly at the process of sensing reality. Here, we will discuss what all of that implies about reality and the brain's modeling of reality, and arrive at some fundamental definitions.
Reality
Let's now consider reality, all that is on the other side of the sensory wall from the knowing brain. We tend to talk of the brain and the outside world as two separate things, but let's remind ourselves once again that by environment we mean the sensed environment, that part of reality that is exposed directly to our sensing machines at the lowest level. The brain is a part of reality as much as any rock or cloud or star or electron that we can observe. Given a little technology the brain can even observe itself and thus understand itself.
What does the sensing mechanism as we have described so far in this series imply about the real environment and how can the brain use the very meager sensory input given it to model the highly complex structure of a vast real world that lies far beyond what it senses directly?
Our model of the basic sensory process is the detection of individual events and integration of those events into patterns. Patterns imply stabilities, or states. Events mean that reality is changing; states and patterns of states mean that reality is not changing. Therefore, reality is a structure of states that persist for some extent of time.
Definition: Reality is, at the least, an ontologically primitive structure of static dimensions and time.
We need to begin by grasping clearly the most fundamental meanings of these opposing concepts of stasis and change, because they are the elements of the natural structure of reality as well as the building blocks of the brain's model of reality.
Dimensions, states and events
First consider states and changes. An event is a unit of change, a difference in something than we can measure (detect) from one time to another.
Definition: An event is a significant (detectable) change, or a difference in a relation or state in some dimension from one point in time to another point in time.
Definition: A relation or a state is a static difference in a non-temporal dimension.
Definition: A dimension is a scale for measuring some independent (statistically orthogonal) relation or state.
Another way to think of a dimension is as a frame for differences or other relations between two or more 'things'. (We may have to use the scare quotes a lot to show that we are using a concept very, very loosely or postponing a definition of the concept.) An important implication of these definitions is the essential relativity of events and states (relations). One thing alone in a dimension has no state and cannot change over time; it takes two different things to have a state (relation). It takes two states at two different points in time to have a change. An event is a change in a relation between things in a period of time. Change and time are tightly bound; change implies time and time implies change.
Consider two temperatures, 30 degrees and 31 degrees. We can relate these two measurements in a couple very fundamental ways. They are both of a kind (temperature). They differ from each other (in value). Same-different. There is something that we can say about them. This is structure. This is information. This can be known. Now consider a third temperature, say 100 degrees. There are now more relations among the three temperatures. The difference between 30 degrees and 31 degrees is much smaller than the difference between 31 degrees and 100 degrees. A comparison by relative distance is possible. With just three 'things' in one dimension, we already have the possible relations of same/different and near/far, more/less or higher/lower. Maybe you can think of others. Even reasoning is possible. (If A is less than B and B is less than C, then A must be less than C.) This is already a pretty rich conceptual world for all its simplicity. Now imagine the potential complexity of a world with possibly an infinite number of dimensions and more 'things' than can ever be counted. It is that complexity in which concepts are born, live and die.
Now let's briefly consider the natural dimensions and the human senses, starting with vision. For those of us gifted with complete vision, the eyes provide us with information on three spatial dimensions simultaneously (left-right, up-down and far-near). These dimensions are independent, or orthogonal, which means there can be a change in one without any change in the other two. Two other major visual dimensions are brightness and color (which is actually a composite of three independent dimensions of redness, greenness and blueness for most of us). When we add time to vision, there is potential for an explosive increase in information that the brain can build from the myriad events that occur in a visual scene over time. The brain uses the raw input from the sensors for events in these 5 (or 7) dimensions to construct the subconscious model of phenomenal reality that we will call an ontology as well as the conscious 'picture' of the world that seems so vivid and detailed (real).
Hearing also provides information in multiple dimensions. The brain integrates detections of pitch (vibration frequency) and volume (sound wave amplitude) and their events (changes over time) into secondary dimensions such as timbre that help us know what is in the world around us.
The other sensory modalities likewise are tuned to time-based differences in other specific dimensions that are part of the structure of reality. Several of the other modalities give the brain essential information about the organism itself.
Ontologies
Definition: The natural ontology is the complete set of relations and events that describe the structure of reality.
The natural ontology is the total of what can possibly be known about all of reality up to a point in time. That anything can be 'known' about reality implies that there is a detectable structure in reality, and structure means persistent relations in detectable dimensions and the changes that can occur in them over an interval of time.
Definition: Phenomenal reality, or the reality of the senses, is the time series of the set of raw input signals to the brain from all the organism's sensors.
Definition: An organism's individual ontology is the set of relations and events that constitute the model constructed by the organism through the processing of its own phenomenal reality.
An individual ontology is constructed by processing that associates sensory events to each other through recurrent patterns by the operation of neural networks. The output of one level of processing (one level of association) can be input together with the outputs of other networks to achieve a higher order of pattern recognition. The individual ontology is thus built up from primitive sensory input into a hierarchy of increasingly compounded patterns that model increasingly deeper levels of the natural ontology.
Let's again turn to vision for an example. When we see a red balloon some distance away, for example, the process of recognizing the object begins when the lenses of our eyes focus images of the balloon on the two-dimensional surface of the retinas at the back of our two eyeballs. What eventually comes to represent a complex, compound and dynamic phenomenon in our mind, say the concept of a red balloon situated in a particular way in a particular local environment relative to the organism, has to be constructed from a great many of tiny, individually trivial bits of information that come from individual neurons. That information is integrated and processed in a number of different stages that follow detection by the sensory neurons.
From our definitions of dimensions, events, states and sensory detection, we can see that the signals stimulated by the images of the red balloon are a composite of measurements in multiple dimensions. Those signals are the result of integration that has already been done by the interconnected neurons of the two retinas, also as described above. The two physical dimensions of left-right and up-down are not abstracted at this stage; the spatial relations are mapped directly to the visual cortex, something like a projection of an image by a movie projector onto a screen. The third spatial dimension of near-far is very different, though. The eye does not capture 3-D vision directly. Measuring in the dimension of near-far involves a complex integration of visual features captured by the retina and the positions of the eyes in their sockets and the muscle-controlled shape of the lens that focuses the image. It also relies on higher cognition that gives us the derived dimension of behind (occluded) and in front of (occluding). Our ability to judge depth is very poor compared to the other two spatial dimensions. The processing by the brain identifies objects by compositing measurements in the visual dimensions (bright-dark, red-green-blue, left-right, up-down, and far-near). This layering of dimensions is much like Photoshop or other such graphics programs do to produce a final image.
All of the processing described so far is accomplished by the same simple mechanism of association by neural networks. That such self-organizing networks of essentially similar processing units (neurons) can both model the complex structure of the natural ontology and provide control and coordination for highly complex motor systems is what makes brains efficient and feasible.
The visual experience of a red balloon, or any object, is thus actually a complex reconstruction performed by the brain using the relatively meager signal data produced at the retina. The integration of sensory experiences into concepts requires even further processing for abstraction and memory formation. Perception is a unique interplay of sensory input and the individual ontology. As we will see later, that interplay is modulated even further by the organism's goal states.
Back to Sensing and Perceiving || On to The Brain and Meaning
This series begins with The Machine in the Ghost