Epistemic certainty of objective reality is something that human beings have yet to achieve. This is not to say that epistemic certainty is necessary to make knowledge claims, though. If a claim of knowledge can stand up to scrutiny, then it can be considered true. I call this justifiable knowledge. In this context, justifiable knowledge is defined as a belief that proves to be reliable: these beliefs can be verified, agreed upon, and are not likely to change on a whim. The acquisition of justifiable knowledge is a fairly long and complex process and understanding this process can also shed light onto how this process can be confounded. This makes the identification of faulty claims made by other people much easier. The process will come in stages, starting from complete solipsism and ending with an established epistemology.
To start, my place in the universe must be established. I, like all people, understand myself to ultimately be solipsistic. I know who I am and what goes on inside of myself; I know that I exist; this is the general idea behind Descartes’ statement, “I think, therefore I am.” Additionally, the outside world is filtered into my consciousness through my sensory apparatuses. My understanding of a chair sitting in front of me is represented by sense data from these apparatuses; examples would be my eyes as I gaze at the chair, my fingers as I brush them over the surface of the chair, my nose as I smell the glaze over the wood, my ears as I hear the knock of my knuckles against the chair, or even my tongue if I choose to lick the chair. This provides me with a kernel of doubt over whether what I perceive is in any way accurate. Who is to say that my sensory apparatuses are accurate? It is not like I can step outside of my sensory apparatuses and perceive the chair independently of them. Moreover, who is to say that there is an objective reality at all? In this context, an objective reality is defined as something that exists independent of the mind. Is it not possible that I am just a brain sitting in a vat, where that chair sitting in front of me is simply a projection from my brain? That chair, then, would be dependent on my mind to exist, rather than independent. This is the first major epistemic barrier to my attainment of justifiable knowledge.
My response to this quandary is to define substantive distinctions and unsubstantive distinctions. A substantive distinction is a distinction that has meaning to a perceiver. An unsubstantive distinction is a distinction that has no meaning to a perceiver. The best way to communicate what is meant by these definitions is to give an example. Imagine two possible realities: One reality is the one assumed by every individual from the moment they are born; there is an objective reality that exists and it is being accurately represented by my sensory apparatuses; in short, my sense data can largely be trusted. The second reality is one where everything that my sensory apparatuses are telling me is wrong; for all I know, these sensory apparatuses are making everything up and I am really just a brain floating in space. These two possible realities established, the obvious question should be: is there a substantive distinction between these realities? In both cases, I am subjected to a world of sensory information: my empirical reality. In both cases, this empirical reality has implications for the way I live and understand the world. In both cases, nothing outside of this empirical reality has an impact on me in any discernable way. The only distinction between these two realities is something that I will probably never be able to detect in the first place. Consequently, I consider this to be an unsubstantive distinction.
Moreover, not only is the question of whether my sense data is accurate or not unsubstantive, but I can also quickly establish the existence of things based on their ability to be detected within my empirical reality. Anyone can claim that object X or object Y exists, but simultaneously cannot be detected by sense data; but this is no different from saying that objects X and Y don’t exist at all. The most obvious example of this would be god: god exists, but I can’t see him, touch him, hear him, smell him, or taste him; I can’t sense him in any way. This is no different from saying there is no god. In short, justified knowledge is predicated on my sense data: my sensory experience.
This is empiricism.
Apparitions in the Air
Once empiricism has been established, I can then move on to the next major epistemic barrier to justifiable knowledge. This one has to do with hallucinations and other apparitions that confound my ability to make sense of my empirical reality. First, hallucinations should be defined. Intuitively, people would think hallucinations are images a person sees, but which don’t exist in objective reality. Yet in the last part, I established that there is an epistemic barrier that prevents me from understanding whether my sensory apparatuses are correct in their representation of objective reality. This asks the question of how I even know hallucinations exist in the first place. What makes people think hallucinations are a thing at all?
When I stopped think about this, I recognized that hallucinations are not so much recognizing that my sensory apparatuses are disconsonant with objective reality, but recognizing that my sensory apparatuses are disconsonant with each other. Think about it this way, when I perceive a wooden chair, all of my sense data taken from separate apparatuses line up with one another: I can see a chair made of wood, I can feel the chair made of wood, I can hear the chair made of wood (provided that I knock my knuckles on it), I can smell the chair made of wood, and I can even taste the chair made of wood (if I so choose to do such a thing). Hallucinations are telling because these different apparatuses are disconsonant with one another. Usually, if the chair was a hallucination, I would be able to see the chair, but nothing else. Or, if I were to use another example, perhaps I would be able hear a voice, but nothing else. The nature of hallucinations established, it necessarily follows that justified knowledge is acquired from sensory apparatuses that are consonant with each other. I can see it, touch it, hear it, smell it, and taste it. These five things all match up to one another. When one of these apparatuses falls out of line, then I know something is wrong.
Justified knowledge is then acquired, not only from sense data, but sense data where the different sensory apparatuses are consonant with each other.
Objects and Categories
Like before, there is another epistemic barrier that must be surmounted. This barrier also has to do with how consistent my sense data is, but instead of asking whether the different sensory apparatuses are consonant with one another, I am instead asking whether objects that fall into specific classifications are consonant. Singular objects are characterized by many different traits, pertaining to their appearance and to their use. Objects are then classified by select common traits. This leads to problems, as I begin to question what defines an object at its very core and whether the object can be accurately characterized by the common traits it shares with other objects. As before, the best way to explain this would be to provide an example. I imagined a chair before (specifically a wooden chair). The first thing that comes to mind is that fact that not all chairs are wooden. If I want to understand what a chair is, then I need to broaden it beyond the subcategory of wooden chairs. The second thing that comes to mind is the fact that when talking about a chair, I am not necessarily talking about a specific object, called a chair, but a category of objects that are all labeled as chairs. This category references specific common traits between these objects, rather than every trait for each of these objects: some are made of wood, some are not, and so the material of the object is not one of those specific traits.
So, what are the specific traits that define what a chair is? A few ideas that come to mind would be that a chair is for sitting, a chair has four legs, and a chair has a backrest for relaxation. But there are problems with using these traits. If a chair is used for sitting, then what about chairs used for design, like in an art project? Do they lose their designation as chairs, or are they simply chairs within an exhibit? If a chair has four legs, what about chairs that have five wheels at the bottom, attached to a five-pointed star base? Are they not chairs? If chairs have a backrest for relaxation, what if I snap one off of the chair? Does it no longer classify as a chair, or is it just a broken chair? The initial response to this would be to point out that chairs are made up of multiple traits, not just singular traits. But this asks the questions, how many and which ones? Is there any way to adequately specify all of the common traits that characterize what a chair is? The speculation would be endless.
The solution to this is to recognize the distinction between the concrete and the abstract. Concrete things are the specific objects I perceive through my sensory apparatuses. An abstract is an intangible generalized representation of a concrete thing or a set of concrete things that is conjured by my mind. Earlier, I pointed out that when talking about a chair, I am not necessarily talking about a specific object, but a category of objects. A category of objects is an abstract. Since an abstract is defined as an intangible generalized representation of concrete things, it necessarily brushes over the exceptions to the overall rules. When an abstraction is recognized, it is the amalgamation of all of the generalized traits of a particular set. For example, I recognize that chairs tend to be used for sitting, I recognize that chairs tend to have four legs, I recognize that chairs tend to have a backrest for relaxation, and so forth. By taking all of these recognitions and overlaying them on top of one another, I can amalgamate a cluster of associated traits and then extract a generalized abstraction of what a chair is. This is rather nebulous, but how it is specifically defined is contingent on the purpose. The point is that many of the previous questions are answered in relation to this abstraction. A chair that is an art prop, a chair without four legs, and a chair with the backrest broken off are still chairs because all of the traits that make up these objects fall largely in line with the abstraction. The objects are understood in relation to the abstraction.
At this point, then, justified knowledge is acquired from sensory experience, where the sensory apparatuses are consonant with one another and where an abstraction can be extracted.
Another epistemic barrier to justifiable knowledge would be the fact that everything defined up to this point is rooted in my subjective experience. Abstractions are especially ambiguous and I would be right to question their accuracy. If I want to acquire justifiable knowledge about objective reality, ideally I should step out side of my own subjectivity and relate my understanding with other people. This will make my understanding intersubjective, rather than subjective, taking myself one step closer to objectivity. So the question is how am I supposed to communicate these abstractions with other people?
Abstractions are represented and then communicated with others through the use of models. Models are representations of abstractions that can be broken down into two distinct categories: visual models and linguistic models. Visual models are pictures or images. Linguistic models are mathematical or communicative language (like the English language) describing the abstractions; more specifically, these are equations or propositions. Both of these models allow me to externalize abstractions and make them known to other people, who can then assess them for accuracy. As stated before, this allows me to achieve a degree of intersubjectivity. This is part of the reason why I can put a generic picture of a chair on a screen and have everyone in the room understand the general category being referred to.
Up to this point, justifiable knowledge is understood to be acquired from sensory experience, where the sensory apparatuses are consonant with one another, where abstractions can be extracted, and where abstractions can be represented by models and then communicated to other people.
Systems of Objects
What I have established so far is how to justifiably hold a knowledge claim about objects or sets of objects that fall under a general category, as they are revealed to me by my empirical reality. This knowledge can extend to interactions between multiple objects. In order to do this, models can be expanded upon to describe the patterns in behavior between a few or many objects. Basically, the model describes a system of objects and how they influence one another. An example of this would be the solar system: the motion of the planets around the sun. Originally, Newton’s complex series of calculations helped describe the motion of each planet as it traveled around the sun. This is a linguistic mathematical model being used to represent the system of objects.
Models can also be used to explain why objects are behaving in the way that they are within a particular system. The models represent the underlying mechanisms that account for the behavior within the system, rather than simply describing how the objects are behaving. Going back to the previous example about the planets travelling around the sun, Einstein’s calculations both describe and explain this system. The calculations describe the motion of the planets, bolstered by the explanation that gravitational pull is actually a depression in Spacetime. His calculations, like Newton’s before him, are a linguistic mathematical model that represents a system of objects interacting with one another. Explanatory models are superior to descriptive models, so it should be understood that Einstein’s model superseded Newton’s model.
So justifiable knowledge can be expanded to include knowledge about systems of objects, where many objects interact with one another. The models used to do this are either descriptive or explanatory. This is the logic behind scientific laws and scientific theories, respectively.
The standards described up to this point make justifiable knowledge intersubjective, but intersubjectivity is not objectivity. Just because my peers and I agree with each other about a particular model, does not necessarily indicate that the model is accurate in what it is meant to represent. This is the last epistemic barrier to justifiable knowledge: bridging the gap between intersubjectivity and objectivity. In order to do this, I need to devise a method that will allow my peers and I to agree upon models that are consonant with empirical reality. As it should be obvious, there are many ways in which bias can distort my claims of knowledge—from disconsonant sensory apparatuses to poor abstractions to poor representations of those abstractions to poor representations of systems—so how can these potential biases be filtered out?
The solution to this quandary is the pragmatic testing of linguistic models to see if they can accurately predict future observations in my empirical reality. The idea is that if I rely on a particular model and use it to make falsifiable predictions about future observations, then it’s ability to make true predictions suggests the model’s accuracy. I mentioned falsifiability. If a prediction is falsifiable, then this means it has the potential of being wrong. This allows for honest testing, rather than confirmation bias brought about by non-specific predictions that are correct no matter what. The use of falsifiable predictions is a process that is especially useful when using linguistic models. Both Newton’s and Einstein’s models were accepted as accurate based on their ability to predict the placement of the celestial bodies in the sky. Newton’s predictions were extraordinarily accurate, while Einstein’s were even more accurate. The predictive power of these models were what led people to take them seriously, rather than simply relying on the intersubjective agreement that they seemed accurate.
What can be concluded? Justifiable knowledge is acquired from sensory experience, where the sensory apparatuses are consonant with one another, where abstractions can be extracted, where the abstractions can be represented by models, where the models can potentially be expanded to systems of objects, and where these models can be tested for predictive power by many parties. This is how truth is acquired.
Solipsism is the starting point of every individual. I know that I exist, but whether or not the sense data I receive is accurate is another question. Ultimately, whether sense data is accurate is unsubstantive. This asks the question of where hallucinations come from. Hallucinations are ultimately my sensory apparatuses failing to agree with one another. Once this is understood, I can effectively rule them out as I make empirical observations. In my empirical reality, I can identify objects and group them based on commonalities. From these patterns, I can extract generalized abstractions. These abstractions are represented by either visual or linguistic models for the purposes of communication with others. These models can then be expanded to represent systems of objects, rather than individual objects or sets of objects. The expanded models either describe or explain systems of objects. Finally, these models are verified through testing. Through the use of linguistic models, I can deduce falsifiable hypotheses and test them to see if they can accurately predict future observations. If they can, then this model can be said to be accurate (or true).
The idea behind this epistemological approach is to externalize the search for truth. Instead of relying on my subjective musings of what I think is true, I am tying my search for truth to my actions, as it is reliant on the pragmatic testing of models for accuracy; this way, other people can join in on the search for justifiable knowledge.