Knols as impressions

Mystics, both ancient and modern, say that all of a person's knowledge consists of "impressions" made upon the person, as impressions are made on a phonographic plate or photographic film. In fact, they say, we do not think "new" thoughts. We only refer to associations of impressions, which can be interconnected in myriad ways, and can have varying degrees of depth to the "engravings". If you ponder this for a while, you may find that you feel this to be true.

When we set out to post a Knol, we are identifying what we believe to be an individual impression and codifying it in our chosen language, making that codified statement about our impression available to some others who may have some affinity with the codified statement, perhaps due to a similar impression experience, and perhaps due to the impressions your selected words happen to bring up for him or her. Still other people may encounter your statement indirectly (e.g. your knol was at the top of a page). Based on Google's architecture, this still could be shown to be an association with your posted knol and the search terms the other person typed, or the website the visited.

The concept of neurons is thought to describe the way information moves around in a brain, and neural nets are a working model of such a concept. This may be (at best) an oversimplification, and possibly misleading, but it is interesting that Google's basic model is similar, and may indeed be closer representing a functional simulation of certain mental and physical mechanics.

Assuming the mystics’ descriptions are correct, it may be that our “knowledge” is indeed associated together more like a hyper-sophisticated variation of a ranking model which includes back-references, frequency of visits, and so on. Consider, for example, how a webpage which is referenced by many other sites can be ranked as more important than a web site which is not referenced by many (ignoring the pay for placement outliers). Again, the mystics have said that an impression can be deep, and can be associated with other impressions. The strength of the associations can depend upon many things, including the emotional experience during the moment of the impression, the level of waking consciousness at the time of the impression, the number of times the impression is revisited, and other things. In addition, if we experience some type of identification with the impression, we may become even less aware of the things that went into the impression, and it may be (for a while, at least) beyond our ability to objectively analyze.

So, when we post knols, we are collectively codifying impressions and gathering them together where others can see our statement about our impressions. A notable gap in the model is that we as people are not generally aware of how the impressions we are attempting to share have been captured, reinforced, and associated. Another way to say this is that we don’t know the ingredients of the impression, and it is therefore an interpretation both the impression itself and an interpretation of the symbols (e.g. words) you are using.

If you feel skeptical, consider how different a three word sentence can be if translated into another language. “I love you” could mean, depending on the next language’s support for pronouns, past & future tense, and multiple meanings for the word love. Even within one language culture, the meanings of statements such as “I want to do this” can vary significantly. In the case of that statement, a person could be referring to a fanciful concept which the person does not mean literally, such as “I want to go to the moon”, or could be referring to something she actually intends to do, such as “I want to go back to school”.

The “gap” we are talking about presently actually has many components. One is the gap between the knol holder and the true nature of the impression, another is the gap between the impression and the words the person chooses, and still more will occur when the reader goes through the process in reverse order as he reads the knol.

Association-based network concepts are certainly interesting and useful. A next step may be to find a way to account for, or even quantify the value or effect of the “gaps”. Consider, for example, how one could compare what is different between two knols which are described using many of the same words. This could produce “gap factors” which could be tagged to those knols and therefore identify a degree of certainty or otherwise that the terms being used within a community have a commonly shared meaning. As well, we can trace back the association chains of the different knols to possibly uncover the “components” which went into a given knol or impression. For example, an analysis of knols in which people describe something frightening (such as global warming) may trace back to a variety of emotional impressions early in life (such as being preoccupied with end-of-the-world scenarios resulting from certain types of childhood implicit “education” (e.g. from intense religious messages) and play).

We consider the things we think, feel, and do to be “us”. However, we do not really know much about the processes which we undergo when impressions are made upon us, and we therefore very likely misinterpret those impressions and further distort them when we attempt to explain them to ourselves or others. Well-developed functional models based on these concepts may help us gain more clarity on the nature of what we consider to be our “knowledge”, and may therefore take us a step closer to understanding who we really are and who we really are not.