The Hawkins Frame - 1

In the long run, Hawkins abandons his initial description of "Frame" and (to my mind) wanders into simple Object Oriented Design and Mathematical constructs that lie behind Quantum Mechanics. I'm not sure gets this, but he starts with a frame-based description of consciousness, which is a fine example of putting this phenomenon on the test bench. Many philosophers claim this is not possible. Daniel Dennet is happy to allow what we say about our experience to be evidence of what that experience is. The problem with Dennet is that he was writing decades before Hawkins used the powerful frame language to speak of what it "feels like" to be in the world. Armed with Hawkins' way of speaking, it may be a good idea to loop back and read Dennet again.

With apologies to Hawkins, this is what I think a frame is:

• It is a frame of reference like Cartesian space
• It contains objects, each of which has a location in the space - for example, (x,y,z) coordinates.
• In object-oriented terms, coordinates are attributes of the object within the space. Other attributes include orientation (such as an (x,y,z) vector).
• Although it is not entirely clear, it seems that these objects also have a "velocity" vector - they may be moving through the reference frame.
• The observer occupies a position in the framework and has a complex set of objects that can be used to interrogate objects in the space. In a sense, the observer experiences this set of objects as "self". An example would be fingers exploring his beloved coffee cup. We will call these objects "sensors".
• A key element of Hawkins' account is the interaction between sensors and objects in the space to determine (or guess) the attributes of the objects. This interrogation results in an updated version of the frame itself - as modeled by the brain.
• All this is experienced dynamically - there is time in Hawkin's frame. When we speak of our experience, we call it "now", but "now" is always changing. This is the flow of time. We can be grateful for any model that makes an attempt to explain this most important aspect of the experience.
• Unlike an abstract mathematical "object", a frame is an experience. When we describe it as a mathematical object, we rely on analogy. This is the entire point of the Hawkins "frame". It is claimed to be the fundamental unit of experience, not the actual "God's Eye" view of the world, nor an abstract description of the world "as it is". It is a mental model, presumably implemented with neurons and synapses. As far as I know, this is the only attempt to formalize experience itself.

Some elements here remind me of Einstein's insight - particularly placing the observer in space and making observations relative to the frame of reference. In both cases, there is no "God's eye" view of the space, only the observations of the observer and the conclusions the observer reaches, such as the position and velocity of objects in the frame of reference. If you are familiar with Einstein's view of the world, Hawkins' "frames" are easy to picture.

Along the same line, Einstein was interested in the ability to interrogate objects to determine their properties. It turns out that it is not possible to learn certain pairs of properties exactly, say a and b. Precision in the measurement of a loses precision in b. Worst yet, counterintuitive, certain objects' properties don't exist until they are measured. But that is a different subject.

The point is that Einstein's mental picture of the Universe is a frame. A very powerful one. What it lacks is a sense of "now". Time is just another dimension, like the x,y,z of space.

Hawkins is shy about generalizing this model - sticking to our familiar 3-dimensional world. He could have escaped this a little by referring to the attributes of an object, which can be more or less unlimited. It would have been clearer (to me at least) if his model included the interrogation of the object, returning another frame since the object itself has all the attributes of a reference frame. Certain sensors are limited to certain queries - "touch" returns information that "look at" does not.

Anyone familiar with Object-Oriented design would wish to add (at least) object states which add some complexity to the result coming back from a query sensor. Other things like the concept of "interface" would be helpful to isolate the "experience" of the object from any secret and unobservable properties that may or may not produce the experience. The whole of OOD* seems applicable, and for good reason: OOD* has been developed precisely to describe (at least) the real world.

That said, there are aspects of the world where OOD doesn't seem very helpful, such as chemistry, cosmology, logic, mathematics, and Quantum Mechanics. Hawkins runs into these limitations since he claims his theory covers all we can know. His theory is only a re-invention or application of OOD - a special case. OOD covers a lot more of what we can know. It underlies the design of virtually all modern computer systems, including the Internet. But we must turn to other ways of "knowing" when leaving the familiar 3-dimensional world described in Hawkins' examples.

To be fair, the areas of knowledge where the Hawkins frames start to break down are precisely those that many find difficult to understand. We are also left with the biggest and most urgent things we need to understand: human-to-human interactions.

--------

* OOD = Object-Oriented Design