In the previous article, we introduced the idea that the mind splits the world into abstractions for both for ease of understanding and to facilitate communication. We discussed how some abstractions are more useful than others for building practical models from which we can make assumptions about reality (citing the difference between Chakra Theory and Neurology as an example). In this article we shall expand from this position by delving into the ways in which we can derive and separate useful abstractions from less useful ones. We shall do this by first discussing the what it means to “Know” something.
When you or I say we know something, what do we mean by that? The most basic assumption is that we are sure that the thing we are talking about is true. If I tell you “I know that there is bread in the oven”, I’m really telling you “I believe that it is true that bread exists in the oven at this point in time”. As the astute of you may have noticed from this elaboration however, my knowledge of bread being in the oven has been qualified by the term “believe”. This begs the question, at which point does belief end and knowledge begin? Well, unfortunately for the dear reader, this question doesn’t have a happy ending. Indeed, it is the case that knowledge is a fickle thing.
Linguistically and Socially, knowledge is much less a controversial subject. The thing about language and the social world is that the specifics need only to be good enough to transfer information practically and with adequate pace; so the line between “knowledge” and “true belief” needs not be finely drawn, if drawn at all. Primarily, in the context of the social environment, “knowledge that” is the idea that information can be reliably recalled at will, with the added caveat that such recalled information is assumed to be correct. In such cases, there is often no justification for why a claim on knowledge is correct, we just assume that people know things. The only exceptions to this are in the case of a clear lie, or in the case of dubious or conflicting information presented as knowledge. In our everyday experience, knowledge is almost indistinguishable from memory.
Neurologically, conversations on knowledge primarily regard the mechanisms by which the brain understands and recalls what we label as “knowledge” in everyday life. This concerns the neurological mechanisms which allow us to recall information and commit facts to memory. For example, how do you know your bedroom is behind a certain door in a certain place even though none of your sensory organs can evidence it? There are neurological explanations for this.
Whilst both of the above paragraphs constitute fields of study in their own right, in this article we are concerned with Epistemology, or otherwise, the philosophical justification for knowledge. The question we are seeking to answer is “By what mechanism can we have sureness about the things we claim to know”. For example, we say we know that there are words in front of us, but how are we sure? This is a contrast to the psychological phenomenon of knowledge; where that concerns the neurological computations needed to supply information understood as “knowledge” at a given time, our journey will allow us to draw a line between what can be justified beyond doubt as “knowledge”, compared to what can only be concluded to be belief.
Understanding Falsifiability
You may have heard of the famous “Brain in a vat” thought experiment by Gilbert Harman, an offshoot of Decartses’ “Evil Demon” thought experiment. The thought experiments in their most basic form entail scenarios wherein all external sensations are generated by a either a machine or magical being intent on deceiving the subject. In the case of the “Brain in a vat” thought experiment, events have conspired as such that the subject’s brain is currently (and unbeknownst to itself) in separation from their body and connected to a sophisticated supercomputer which is simulating reality to the minutest detail. The popular conclusion to this thought experiment is to assert that all statements of knowledge by the subject under such conditions must necessarily be “False”. The justification for this is because a simple statement such as “I am looking at a wall”, must be false on all accounts, because the eye which observes the wall as well as the wall itself are both digital (or magical) fabrications. Whilst this thought experiment is not the most philosophically mature investigation into knowledge (see the grey box below), it does do us the favour of highlighting the fact that there is no 100% certain justification for any knowledge. In fact, knowledge as understood outside the context of the linguistic and social practicality of the word, does not exist. That is to say we can never actually know anything for sure.
In answer to the thought experiment, a modest conclusion to its scenario would be to understand it as a hyperbole justifying how the specifics of the universe you are inhabiting are to a large degree irrelevant for you to have knowledge about that universe. Because all knowledge is based on induction (discussed below), uncertainty will always exist to a degree of at least a fraction of a single percent. Therefore, large overarching grand narratives of the nature of the universe and its potential unknowingness constitute naught but a tiny dent in the overall ability to understand the symbiocity of the external world. This combined with the fact that such narratives necessarily cast doubt on all knowledge equally means we can safely ignore them as meaningless in our quest for understanding.
It may help to think of this mathematically. if A = C, and there is some variable X (representing the unknowable grand narrative of the universe) which is applied to all over variables at all times, then A+X=C+X which is equal to A=C+X-X which just brings us back to A=C. Therefore X can be ignored in all cases.
Th thought experiment may also be ignored on the basis that it constitutes an unfalsifiablity. Namely, there is no conceivable and logically possibly system by which we could truly tell whether it is true or false. Unfalsifiable statements cannot be given a value of true or false, as it is impossible to tell, leaving us with nothing but a quaint irrelevancy.
Although we are habitualized into using the word “Knowledge” as part of our common parlance, in actual fact all the things we claim to know are just very well justified beliefs which are currently lacking in evidence to the contrary. To expand on an earlier example, we only know that there is bread in the oven because we had evidence to the fact at an earlier point in time, we may have seen it ourselves, or been informed by a reliable source. The fact remains that we may be wrong about the bread still being in the oven, but whether we are wrong or not is actually not all that important, what is important is that we were satisfactorily justified at the time of communication to claim knowledge of events.
This hits the nail on the head as to what it is we understand as knowledge, well justified belief. Sometimes it is argued that knowledge is “Justified True Belief”, but as we shall see, the idea of anything being true is dubious at best.
Observation and Justification
The only unjustified knowledge comes from sensory observation, although exceptions exist in observational mistakes and optical illusions, generally speaking all knowledge at its most basic level is gained through observation. The physical sensation of sensing a door, whether it be by touch; sight or otherwise is, strictly speaking, the only pure form of knowledge we can ever acquire, the sensory observation itself being the most direct evidence for anything we conceive.
Moving past the pure sensory observation, the idea of an object known as “Door” itself is a conceptualization of the sense data we acquire. We use conceptualization to organize the world and to create anchors from which we can effectively communicate. Conceptualization is a complex topic in of itself and which you can brush up on by reading the previous article.
How about more complicated phenomena? How about statements such as “The door in the building far away is currently open”? or, “Doors made by this man are superior to other doors”? These are statements which require justification beyond personal sensory observation, terms such as “superior” must be adequately defined and reasons and evidence justifying the “truth” of the statements concerning entities currently unobservable must be compiled in order to reach any adequate conclusion.
So if all non-direct-sensory knowedge requires justification, then we need to take a step back and consider what it means to justify something. A justification can be thought of as a series of interlocking pieces of information, which when considered together create a compelling case for the existence of a separate piece of knowledge. No matter how you go about justification, every method falls into one of two categories:
Deductive and Inductive Reasoning
Deductive reasoning is the simpler of the two, it relies on using present knowledge to determine things that such knowledge is dependent on. For example: knowing that the king died on a Wednesday at an age above a single day, I can safely deduct that the king was alive on Tuesday. Because as a person older than a single day, he must have been living on Tuesday to have died on a Wednesday. By reasoning as such I have gained knowledge deductively, but only under the condition that the knowledge of the king being dead and older than a day was indeed true, which unfortunately constitutes knowledge that can only be gained via observation. This highlights that even if you were to primarily use deductive reasoning in your acquisition of new knowledge, someone would at some point need to have gained knowledge observationally; the astute of you may realize this means knowledge is necessarily gained at its most basic level via observation.
Inductive reasoning on the other hand can easily be understood as “The process of creating generalized rules from observed phenomena”, or otherwise using a pool of observations to justify your beliefs. However as you may know, the senses can be devious things. Our brains are also wrought to find patterns in things which may be more complex than we can currently comprehend, a disastrous trait when applied to the pursuit if knowledge.
To give a short rundown on induction, consider the simple example of inferring the probability that the sun will rise tomorrow. Inductively, we may note that the sun has risen every single day for the entirety of our lives (constituting a body of observations), and if we are to reference texts from millennia ago, seemingly this had been the consistent trend for the entirety of recorded human history. Taking this into account, it may seem likely that the sun will rise again tomorrow. This is a very straightforward, albeit superficial assumption to make, and one that we each make every day on a variety of subjects. The problem with this reasoning however, is that past observation does not predict future occurrence. This core problem of induction is highlighted in the story of the farmer and the chicken.
In the story of the farmer and the chicken, a docile chicken is fed by a farmer every day. The chicken, using inductive reasoning, infers that the farmer will come and feed him every day. He seems to be right, until the day the farmer comes to end his life, he was, after all, a farmed chicken.
So what do we do in order to acquire knowledge? For the majority of simple things, (such as the stability of geography and for the most part the fidelity of our relationships), a combination of careless induction and luck are evidently practical enough to get us through life. But how about more complex things, such as detecting liars, developing effective medicines or deducing the laws of physics? In these cases it is not good enough to simply induct; we need as perfect a countermeasure to the flaws of induction as we can muster.
A System of Crystal Induction
For complex phenomena, we must minimize (for it is impossible to complete a full elimination) the confounding effects of the problem of induction, and in doing so we are left with no choice but to attack our assumptions from every direction. Whilst simple everyday phenomena likely don’t need this level of rigorous investigation, complex problems require conscious effort to attack from multiple directions; as an example, had the chicken in the above parable considered the motivation for the farmer’s taking care of him, he may have been led to a direction which caused him to rethink his conclusion. Things that do not seem immediately related could actually have shed evidence on the situation at hand, such as a the infrequent but consistent disappearance of other chicken.
Attacking our assumptions from every direction is, unfortunately, both far easier said than done, and our current methods consistently prove themselves to be far from infallible. For centuries Newton’s universal law of gravitation was believed to be definitive, until the day that astrologists found a single counter example coming in the form of a discrepancy in the orbit of the planet Mercury. After this single observation, it was clear that Newton’s “Law” was in fact not an accurate representation of reality. The truth is, all “knowledge” is but a single counter example away from being reduced to falsity. You may believe the bread is in the oven, but all it takes is to see the oven empty to reveal that you were wrong. This same problem of new evidence potentially wiping out all of our previously established “Knowledge” exists for literally everything – evolution; gravity; everything. This failure of inductive reasoning is formally expressed in the following way:
If A implies B, then Not B implies Not A
Combine the problem of induction with humans ubiquitous predisposition to a variety of cognitive biases, and it becomes clear that knowledge is not an easily acquired treasure. Thus we have had to create an organized, structured and fairly unintuitive method for analyzing our inductions from as many angles as possible in a way which hopefully minimizes our unwanted bias, (it doesn’t always work). This method goes by the common name of the scientific method.
As one might expect from such a duplicitous subject such as knowledge, the scientific method itself is a complex framework which deserves its own article.
A Summary of the Discussion so Far
To summarize: The only direct form of knowledge acquisition is from sensory observation, after which we conceptualize, or abstract, such information for ease of mental organization and social communication.
For more complicated knowledge which can not be justified via single sensory observations, there are two methods of justification: induction (synthesizing rules based on multiple observations), and deduction (determining facts from observation via logical dependency).
Knowledge is primarily understood to be justified true belief, however considering it is literally impossible to conclude that something is true, the best we can do for phenomena too complex to be justified by direct sensory information alone is to use inductive reasoning to synthesize generalized rules based on multiple observations, using said observations to justify such rules to the degree that they can safely be assumed to be consistent in all cases. An example of a generalized rule is “The sun will rise tomorrow, and every day after”. The problem with this approach is that it only takes a single piece of counter evidence to disprove our generalized rules definitively, which in turns means to definitive prove anything amounts to an impossible task.
Deductive reasoning can also be used to determine additional rules on fundamental to previously and satisfactorily established facts, and is often used to deduct complex rules into their simplest forms.
For more complex phenomena, we must abide as best we can by the scientific method in order to gather information in a way which allows us to successfully attack our observations from many angles, and reduce subjective bias as much as possible. For many simple phenomena experienced as part of everyday life the process of using the scientific method is time consuming and practically speaking, unnecessary; we generally run on the assumption that we are mostly correct about simple things, and it is for the most part fine to operate in this way.
Finally, to be considered justifiable, a statement of knowledge must also be falsifiable, to be falsifiable means that there is some conceivable mechanism by which the statement in question may be proven false. If there is no such mechanism, then the statement can not be understood as having a truth-value, leaving it as essentially epistemological nonsense.
I hope you have enjoyed this introduction on why you actually know nothing, for more in-depth analysis on knowledge acquisition, including a discussion on the scientific method and how we categorize our knowledge, here are some more articles in this series:
