Epistemology offers lots of theories of knowledge as well as their refutations. These theories practically all focus on knowledge as private. This far, definitions of knowledge as residing in private minds have been rather disappointing. One of the reasons is that knowledge is seen as a qualified belief that requires assessment, stating it to be ‘justified’ or ‘true’. Many epistemologists think justification to be a necessary condition for knowledge, although part of them – the internalists – admit justification to be subjective. If a certain knowledge is also true, this justification boils down to some level of consciousness of a belief. The opposite position – of the externalists – seeks to objectivate justification by stating that it needs to be grounded in events that cause beliefs. As knowledge of these events in their turn needs to be grounded in earlier events, they enter on a path that can only end in spontaneous observations not in need of justification. But the issue is not if our senses may trick us. They may, but illusions may be overcome by rational justification. The issue is: from where does this justification process start? Do we see reality as it is, or do we see it only as it is to us humans? And if we see it as humans only, how can we be sure that we as humans share the same reality?
A purely internal condition of knowledge is that we are conscious of knowing it. Tacit beliefs, even if true, cannot count as knowledge. Most epistemologists agree to this, but they need a solution for distinguishing between consciously held and tacit knowledge, which is far from easy. Difficulties related to an agent’s expertise, power of memory, observational qualities and so on disappear if we stop seeing the human mind as the prime bearer of knowledge. Recently, more attention has been given to knowledge as a collective asset, a theory known as social constructivism. We share and store knowledge ever since oral tradition started, meaning that it exists independently of individual people.
Bedtime, fairytale-time. The brave knight defeats the ferocious dragon, frees the beautiful princess and they live happily ever after. But my little son worries that a dragon may still come after him while he’s asleep. “Don’t worry,” his elder sister says, “dragons don’t exist.” “Yes, they do,” the boy says, “they are there, or else, how did this one get into the story?”
“Through phantasy,” I answer, “not because dragons are real.” But the fact that nobody ever saw one does not prove that dragons don’t exist. Moreover, there are parts of reality to which we have no access yet, for example, the precise functioning of minds. The famous mind-body problem is about facts that may be “beyond the reach of human concepts”, as Thomas Nagel famously wrote in his 1974 article What is it like to be a bat? The human mind is also part of reality, so, the dragon did exist in the mind of the author of the fairytale book, she then wrote down the word “dragon”. My son had no idea what a dragon was, but he got some idea in the course of the story. It mentioned concepts that he understood: big tail, claws, sharp teeth, fire, smoke, etc., the dragon’s properties.
We now have three dragons. One in the mind of the author, one in the mind of the reader, and one in the book. Knowledge about this last dragon can be objectivated by bringing together all properties of the dragon the story mentions. Let us assume that the dragon the author had in mind was green, but that she didn’t mention its color. This would allow for dragons of any color in the mind of the reader, but for no color at all in the description of the dragon based on the story. The dragon in the story would be colorless, and presumably, the story leaves open quite a lot more: the number of the dragon’s legs, the length of its tail, if it has spines or not, etc. Moreover, our thoughts must be expressed in language, impressing still another limitation on knowledge.
The shared dragon-knowledge is necessarily incomplete. The next day, my son makes a drawing, His dragon is red, has ten legs and two tails with spines. “Can dragons have two tails, like lizards?” he asks. The nature of dragon-knowledge is the same with all knowledge. Let us assume that the author of the fairytale book wrote “reality” instead of “dragon”. This would imply that we have the author’s concept of reality, the reader’s concept, and the concept knowable from her – enormously voluminous – book of reality. But the book of reality is not written by one single author. We are all writers, editors, and readers of that book.
Knowledge is essentially a written corpus, its own never-ending story of hypotheses, mistakes, correction, renewal, etc. Humanity as a whole is far better capable of hosting knowledge than any individual; therefore, discussions about how knowledge resides in individuals are rather trivial. Any justification of private knowledge in individuals or groups is futile if it is not shown to be coherent with the rest of human knowledge.
The story of knowledge is not only contingently incomplete because we still need to make discoveries or because we don’t know the future, but also necessarily incomplete because some things can never be known or cannot be expressed in language, and because of knowledge’s generalizing character: it can never mention all singular cases of general principles, see Truth and identity, nor can it predict the possible outcomes of all processes, see Coincidence, determinism and free will.
In his famous regression-argument, the skeptic philosopher Sextus Empiricus argued that knowledge is impossible because all knowledge has to be justified by other knowledge. This leads to either endless regression, circular reasoning or axiomatic belief.
Foundationalists answer that some fundamental beliefs are justified by direct impressions en therefore are not axiomatic. But how can we be sure these offer a view of reality? We already saw Wilfrid Sellars speaking of the “myth of the given”, see Knowledge and Truth. Most foundationalists however adhere to the correspondence theory of truth, claiming that impressions of the mind-independent world can be directly expressed in propositions or sentences. But they cannot justify this claim: of course not, if they try to do so, their ‘fundamental’ impressions are no longer fundamental. Some solace is offered by the observation that human sense perception works. If not, we would have died out long ago. The drawback is that foundationalists need a pragmatic lifebuoy to save their theory. The coherentists’ answer to the regression-argument is that within the scope of reality as a whole, circular reasoning is no problem. But what, their critics say, if we are brains-in-a-vat or misled by Descartes’ evil demon?
I argue that our evil demon is rather our craving for fundamental truths.. We just cannot know things mind-independently. Knowledge is a mental representation of aspects of reality, see Knowledge and facts.
We need to start from the metaphysical beginning: does something exist? The question itself exists, even if not yet worded. Questions then need minds to pose them, and humans have minds capable of asking that question. Maybe other minds exist besides the minds of humans, but is senseless to have oneself disproving that we are brains-in-a-vat, puppets on a string or messed around with by whatever supernatural power, because first, proof of non-existence is impossible, and second, these theories violate the principle of parsimony – to reject theories introducing new existent, like for example vat-maintainers.
A different case is the other minds problem, also known as solipsism: how do I know that I am not alone in the universe and surrounded by zombie-like figures who behave like humans, but maybe do not have minds? This idea violates the principle of the uniformity of nature. My existence was caused by some chain of events. These events repeat in a uniform way, causing other minds to exist as well. But this observation does not take away the difficulty to perceive other minds directly. Indirectly, however, through our limited language-lens, we notice that other minds are capable to produce beliefs, such as red dragons with two tails.
How do minds relate to the material world? First of all, minds need matter to be able to exist, whether they reside in living beings or are cases of artificial intelligence. Mind and matter are capable of causal interaction. They belong to a monistic system of being.
There is no such thing as the perception of the world by the mind as if all input were bundled into one channel. All theories claiming a connection between the mind as a perceptive entity and the outside world are remote echoes of René Descartes’ dualism.
We perceive the world not as it is, but as it is to us. This position is known as indirect realism or representationalism. Strong support for this position was offered by the German biologist Jakob Johann von Uexküll, who stated that each animal species has its own Umwelt or representation of the world, dependent on its specific sensory organs necessary for survival. This view endorses enactivism, the theory that cognition is created by the interaction between an organism as a whole and its environment. Since we are also each other’s environment, building up knowledge is essentially interactive.
Views like Von Uexküll’s were central in Daniel Dennett’s 1991 book Consciousness Explained, but Dennett was unable to account for consciousness’ subjective aspects and the relation between consciousness and personality. The human mind remains largely a secret to us.
For shared knowledge, however, the situation is more favorable. Shared knowledge is explicit and therefore conscious. Its reliability is also made explicit, allowing it to be judged externally.
We cannot directly observe the world as it is, but only as it is to us humans: foundational human knowledge can never be objectively true, only intersubjectively acknowledged. The observation that human knowledge is apparently functional for our survival confirms Von Uexküll’s theory, but it does not prove that foundational knowledge is mind-independent. Humans are capable to establish knowledge beyond direct perception, but for such knowledge, internalist justification cannot be taken seriously. Knowledge must relate to facts and come from sources whose reliability is checked. For higher-order knowledge, explicit externalist justification is required, showing why we are justified to believe something.