As I’ve discussed in my previous posts (here, here and here), in my adulthood I’ve slowly come to understand if not what knowledge and knowing is, then at least what it is not. A naïve model of knowledge could be something like this: an objective reality exists, and the degree of knowledge one has of the world is the degree of overlap between their model of that objective world and the world-in-itself. Facts are the parts of the objective world which can be known – modeled accurately – and serve as measuring sticks against which we lay our models of the world. If our model does not conform to facts or does not logically follow from them, then to that extent it is not knowledge. From our point of view, facts are experiences which can be successfully predicted: if conditions for an experience are created such and such, I can predict that an experience will occur, and eventually that experience will in fact occur.
You may notice that we could just drop the assumption of an objective world from the model right away, and it wouldn’t make a difference as to how we would approach knowledge. I’ll get to that in a moment. Like this, it’s a good enough model to satisfy my immediate curiosity, but it has some caveats.
First of all, people aren’t really that concerned with knowledge. Knowledge is something we are naturally interested in, but only insofar as it has to do with survival and adapting to a changing environment; much of which is social. That relates to the ideas to which we are willing to conform and which we defend or attack. We do not treat knowledge as a value in and as of itself, but instead as a tool to further purposes we value. People are inclined to reject knowledge that would undermine beliefs they value; few people immediately adjust their beliefs and values instead. Likewise that skepticism disappears if we like a piece of knowledge. These biases are called motivated skepticism and confirmation bias, and they are well documented examples of a larger phenomenon of how we create, process and react to knowledge. So we update our model by adding: knowledge is not something immediately obvious, but instead something which we relate to based on its estimated usefulness and the degree to which it corroborates our earlier beliefs.
That has to do with how we react to experiences which might or might not have to do with knowledge. However, even the basic building blocks of knowledge, facts, aren’t easy. What’s essential to knowledge is that it needs to be verifiable, but few things are so easily verifiable that one can expect to be able to verify them all by themselves. If we seek to know anything outside the immediate sphere of our experience, we are required to trust other people’s experiences at least to a degree. Much of knowledge also requires the cooperation of several people to be attained (a simple example would be a group of prehistoric people using scouts to warn of looming dangers). That is, most of what counts as knowledge includes trust between people; for an individual’s purposes, then, our naïve model should also include the notion: facts are also experiences which other people will predictably have, given that such a bond of trust exists that I need not have the experiences myself to ascertain them as facts, and given that I would in fact have the same experiences if I were in their position.
Also the form the knowledge we have takes is dependent on how we experience (the biological component, so to say) and associate experiences, which experiences and aspects of experiences we prioritize and how we formulate those experiences in language, sensory models or emotions. Again we update our naïve model of knowledge with an addendum: As such, the extent to which knowledge can overlap with an objective world or to which it can successfully predict experiences depends on how well the world and eventual experiences can be modeled by means of our psychology, sensory abilities, language and emotions. Or As such, what knowledge means will depend on our psychology, visual abilities… etc.
Knowledge is also context dependent. An easy example is how the laws of physics fit into our experiences: Newton’s laws do not explain everything, but in the context of masses moving at relatively low speeds we can trust to be able to make reliable predictions based on them – that is, we can trust them to produce knowledge. On the other hand, outside that context Newton’s equations cannot be used to produce knowledge in a straightforward manner. So we update our model by saying that knowledge is not universal but instead is defined and limited by its presuppositions (in our example, that masses move at speeds far less than the speed of light). This is practically a variation of the previous addendum; the modes and limitations of our cognition are perhaps the most important presuppositions that exists, as they are extremely hard or impossible to overcome. That is, all (humanly meaningful) knowledge presupposes that it is produced by human cognition.
Now, regarding both the fact that we are prisoners of our biopsychological faculties of cognition – or any other faculties we might have improving upon technology – and the fact that the way which we produce, maintain and process knowledge has social components such as trust and language, and having added the ideas of context specificity and limiting presuppositions, we can safely say that a model of knowledge need not assume an objective reality. Whether or not such a reality exists is a matter of discussion, and I’m already too much out of my depth to engage such a matter here. However, it suffices to say that even if a naïve model of knowledge benefits greatly from such an assumption, our addenda make such an assumption needlessly cumbersome, so we may as well drop it and be content with the idea that knowledge relates to experiences and predicting them.
So, having updated the naïve model this post began with, we have something like this: Knowledge is a model of reality which enables us to predict eventual experiences, depending on context. That model is built upon a more or less predetermined framework of biological necessities, prior experiences, cultural influences, language and prioritized associations. All knowledge presupposes either other experiences or requirements for experiences to occur. Knowledge requires trust to be of use; to trust our own experiences and to trust other people to give us reliable information (whether or not they intend to) precedes knowledge. In the absence of trust, no knowledge is possible. Facts are predictable experiences, which we react to depending on their estimated usefulness and on whether or not we like or dislike them.
It’s by no means a great model, but we can work with it. First, we can completely drop any intuition of “pure knowledge” independent of outside influences; such a thing is by definition impossible, as knowledge is a way to relate experiences to a framework whose purpose is to predict further experiences. We can also drop the idea of knowledge without cultural significance, because culture precedes knowledge (is knowledge, in fact); all knowledge is related to cultural ideas and further influences that culture.
These help us drop the idea of “objective” knowledge. We may assume that an objective world exist (so as not to fall victim to insincere and useless relativism), but our knowledge has little to do with such an assumption. Instead any knowledge we have is always by its very nature subjective and only takes the form our sensory, cognitive, cultural and linguistic frameworks enable. Those pieces of knowledge may or may not relate to an assumed objective reality, but that need not be perceived as an useful aspect of knowledge – all that matters is its ability to help predict further experiences.
That subjectivity, however, is preceded by a notion of trust in one’s own and others’ experiences. Trust is so essential to any knowledge that lack of trust immediately undermines any knowledge related to that trust. I find that part particularly interesting, since it points out that interpersonal trust is essential not only for successful and pleasant interaction, but also for a reliable worldview. Lacking interpersonal trust we regress to questioning basic tenets of reality established by our peers or predecessors and become further unable to form reliable predictions of the future. (If, however, such a lack of trust enables us to better predict future, we are of course right in trusting our own judgment and trusting others’ less). Lacking trust in both ourselves and others will leave us completely helpless, which links our model of knowledge to the psychological concept of learned helplessness.
Such a model also leads us to conclude that knowledge is by no means absolute, and it will vary as a function of our priors, beliefs, emotions and further experiences. To find out things which contradict things we perceive as knowledge should then enable us to relate to our new knowledge in a way surpassing the primitive confirmation and other biases, as we realize that different frameworks of predicting further experiences enable different sorts of knowledge. By learning possibly contradictory things we learn of different vantage points to further predicting our future; those points of view may not and will not be of equal value, but instead of judging them merely as real or unreal we can judge them as knowledge more or less integrated to our other predictions and behavior. Of course, usefulness can also be determined on whether or not we feel content with whatever worldview a piece of knowledge enables, which further distances the concept of knowledge from a naïve notion of objective reality.
That is not to say that knowledge should not be useful to predict eventual experiences or that we should appreciate all claimed knowledge as of equal value. It is to say that whatever knowledge is, it is by its very nature such that merely asking “is it real knowledge or not” will never encapsulate the whole of the processes which create what we call knowledge. Blinding oneself to certain aspects of knowledge (its cultural and biopsychological priors, as some mathematically inclined people seem to do, or its ability to predict further experiences as some relativistic, humanities-oriented people seem to do) will produce a very coherent set of beliefs, which however may be very stiff and awkward when used in contexts alien to its core use (like using logical connectives and numerical probability distributions to predict whether someone would like to be my friend).
What we gain from such addenda to a naïve model of knowledge is dispassion towards matters of fact and knowledge. By now we should be able to appreciate that knowledge which we do not understand or would intuitively reject might have uses or implications we might not immediately be able to understand; and also that whatever reactions we have towards something that might or might not become our knowledge are dependent on our personal worldview and interpersonal trust more than the new information’s perceived “realness”.
I will elaborate my thoughts on knowledge and trust in a further post on the matter.