Sunday, May 07, 2006
Models Within Models
This is only a half-baked idea, which needs more thought (a small disclaimer). Political Reasoning and Cognition by Rosenberg at al. categorizes cognitive levels as sequential, linear, and systematic. These are increasingly sophisticated views of the world that they apply to political reasoning. One of their hypotheses is that people operate at different cognitive levels and that this affects their political reasoning. This was (and is, I think) apparently controversial, and in conflict with many social scientists who reject the idea that one person's thinking is more sophisticated than another's. Of course, I am with Rosenberg on that. The egalitarian's odd view, that we are all the same, is the sort of thinking that makes non-academics roll their eyes and gripe about political correctness. It's not science to take a position because it is "nice" and reject reality because it is "unfair".
But, back to the model. The levels are part of the "structural developmental" school of developmental psychology. They are well-developed concepts, with related behaviors. But, to me, they resemble a model that still too high level to be fundamental. I think there is an underlying model that interacts with other capabilities. My hypothesis is that what is really at work here is the level of abstraction of a cognitive modeling capability.
Humans observe patterns in the world and construct a model based on their observations. Their models help them predict the outcome of actions and allow them to make choices and intervene to have a more favorable outcome to themselves. This is not to say that every observation and every instance of model construction is deliberate and consciously intended to improve prediction and help you to survive. This modeling capability is a big part of our cognition and we use it all the time. Creatures that model all the time, building more and more accurate and predictive models, survive better. We are the descendents of those survivors. We model, model, model.
As babies, we start out observing and modeling the physical world. Over time, our models get better at working with longer time frames. People become adept at modeling and become aware of the models themselves - and are able to think about them as if they were physical objects - these are abstract concepts. As we advance in capability, we can observe and model about abstract items, and the processes of abstraction. This is basically a recursive process, builing level on level. On this blog, I am observing and modeling the process of modeling itself. More and more sophisticated thinking require more levels of abstraction.
Put in another way, first people understand the behavior of objects, then they understand rules about them, then the understand how rules are made and how to make them, then they understand which rules are better than others and what constitutes an effective rule making process - they become experts on rulemaking itself.
An example can be made about religion. As babies, religion has no meaning - it is too abstract. A schoolage child can learn the rules of religion and understand how to apply them. A more sophisticated person can understand the morality and value behind the rules and apply them more effectively to special cases. Another level of sophistication beyond that, and a person can change the rules to fit evolving circumstances. Beyond that, a person can create a new religion, defining the rules that its followers would adhere to. There are even more levels of sophistication - as someone could become an expert at religion defning, with a model of how religions work, in general.
This is only a half-baked idea, which needs more thought (a small disclaimer). Political Reasoning and Cognition by Rosenberg at al. categorizes cognitive levels as sequential, linear, and systematic. These are increasingly sophisticated views of the world that they apply to political reasoning. One of their hypotheses is that people operate at different cognitive levels and that this affects their political reasoning. This was (and is, I think) apparently controversial, and in conflict with many social scientists who reject the idea that one person's thinking is more sophisticated than another's. Of course, I am with Rosenberg on that. The egalitarian's odd view, that we are all the same, is the sort of thinking that makes non-academics roll their eyes and gripe about political correctness. It's not science to take a position because it is "nice" and reject reality because it is "unfair".
But, back to the model. The levels are part of the "structural developmental" school of developmental psychology. They are well-developed concepts, with related behaviors. But, to me, they resemble a model that still too high level to be fundamental. I think there is an underlying model that interacts with other capabilities. My hypothesis is that what is really at work here is the level of abstraction of a cognitive modeling capability.
Humans observe patterns in the world and construct a model based on their observations. Their models help them predict the outcome of actions and allow them to make choices and intervene to have a more favorable outcome to themselves. This is not to say that every observation and every instance of model construction is deliberate and consciously intended to improve prediction and help you to survive. This modeling capability is a big part of our cognition and we use it all the time. Creatures that model all the time, building more and more accurate and predictive models, survive better. We are the descendents of those survivors. We model, model, model.
As babies, we start out observing and modeling the physical world. Over time, our models get better at working with longer time frames. People become adept at modeling and become aware of the models themselves - and are able to think about them as if they were physical objects - these are abstract concepts. As we advance in capability, we can observe and model about abstract items, and the processes of abstraction. This is basically a recursive process, builing level on level. On this blog, I am observing and modeling the process of modeling itself. More and more sophisticated thinking require more levels of abstraction.
Put in another way, first people understand the behavior of objects, then they understand rules about them, then the understand how rules are made and how to make them, then they understand which rules are better than others and what constitutes an effective rule making process - they become experts on rulemaking itself.
An example can be made about religion. As babies, religion has no meaning - it is too abstract. A schoolage child can learn the rules of religion and understand how to apply them. A more sophisticated person can understand the morality and value behind the rules and apply them more effectively to special cases. Another level of sophistication beyond that, and a person can change the rules to fit evolving circumstances. Beyond that, a person can create a new religion, defining the rules that its followers would adhere to. There are even more levels of sophistication - as someone could become an expert at religion defning, with a model of how religions work, in general.
Subscribe to Posts [Atom]