|Turned up to eleven: Fair and Balanced|
Wednesday, May 01, 2002
This website gives a very nice overview of the fundamental issues of consciousness, as well as basic neurophysiology. In numerical considerations, I will defer to their information. Many neuroscientists have explicitly endorsed a model of brain function that emphasizes modularity, i.e. that brain functions are localized to different regions of the brain. I will also endorse that view, but I will ignore it for the early part of my discussion. It may come back in later posts as a basis for discussion. I will argue, first at a theoretical, and later based on biological function, for the existence of an "N-dimensional intelligence space", that contains "intellectual vectors" that are interrelated, but distinct. I have never seen this viewpoint in print, but that doesn't mean that it hasn't been presented before. After describing the nature of these intellectual vectors, I will attempt to describe a potential biophysical framework for their existence, and suggest some potential (loosely defined) ways to test this hypothesis. I will also attempt to describe the mind-boggling complexity of brain function, and hopefully shed some light on the way we think.
I asked some questions in a previous post, that were meant to suggest a more global model of intelligence than the currently held popular viewpoint, which essentially endorses abstract problem-solving skills as the core of intelligence. My examples, good or bad, were simply meant to encourage my readers to "think outside the box", to use an overworn phrase, and to consider as facets of intelligence skills that are not normally included in that rubric. Lets try a thought experiment. Lets imagine that we can quantifty every controlled neurological impulse that a human being has, from the most rudimentary non-autonomic skill through the most sophisticated reasoning. By this I simply mean to ignore functions such as breathing and heartbeat, but include emotional responses and what might be called "base instincts". Now, let us further suppose that we can measure the interrelatedness of these functions. For example, we might argue that musical talent is related to mathematical talent, and we might be able to quantify this (this argument has been made many times, although I don't think anything approaching quantitation has been applied to musical talent). On the other hand, algebraic skill might not be at all related to creative writing ability (they might even be inversely correlated!!). We can specify each skill as a vector, and take the total number of unrelated abilities to be a measure of the degree of the space, in other words, the number of dimensions. This could be, as in the "Gardner model" seven different types, but I suspect it is rather a lot more, perhaps millions of different abilities. While this is mathematically daunting, it should be plausible with high speed computing techniques to create a map, given the massive amounts of information needed, that contains all of this information.
This thought experiment, however, leaves out a crucial notion. Some of the things on the massive list of abilities/skills/features of human consciousness are not "intelligence". More to the point, we don't have a vector in this map that is clearly defined as "intelligence". But, you might argue, isn't that because we have chosen the labels, and if we chose to, we could lump some stuff together, and call it intelligence? Well, that is true, but in order for that simplifying action to be valid, we have to minimize the information loss inherent in such an action. Since the axes of our space are orthogonal, we can't collapse them into one another without losing information. We can, however, create a boundary around some subspace, and label all the things within that space "intelligence", and label all the rest "not intelligence". Is this boundary arbitrary? I suspect that it is, at least partially. I will tackle that question in my next post, along with a discussion of a biological basis for discussion.
Food for thought; Is there a good reason why the sheer number of neurons, and more importantly, their connectivity, can make the notion of separating the brain into distinct regions that are not deeply connected irrational? Does the mathematics of connectivity make the notion of large scale discrete separations of brain function of limited utility? I will explain this reasoning as well (Hint: The "small world" hypothesis plays a role here)