Philosophers have always enjoyed making sweeping pronouncements on epistemological questions (see, e.g., last week’s post on Socrates). Plato’s contribution to the field of epistemology follows logic just as counterintuitive as that which supports his Republic, and with perhaps a more radical conclusion. And, as usual, Plato gives voice to a theory that more closely matches reality than his own, but it is voiced through a minor character and ends up rejected.
Plato’s epistemology is, in essence, that nothing we come to understand from the senses (“empirical” knowledge) counts as knowledge. Only that which is derived from pure reason can be said to be known. Russell puts it this way: “In this view, ‘2+2=4’ is genuine knowledge, but such a statement as ‘snow is white’ is so full of ambiguity and uncertainty that it cannot find a place in the philosopher’s corpus of truths.”
Understanding Plato’s “rationalist” argument is prerequisite to understanding Western philosophy, especially folks like Descartes, Hume, Berkeley etc.—and frankly, to understanding how to read legal and academic writing. But living with the conclusion of this argument, that genuine, non-math-related knowledge is impossible, is intolerable. Not only that, the argument fails to clear a very basic bar: distinguishing between the quality of various propositions. Some of our perceptions are more reliable than others. Rejecting perception as a way of knowing outright forecloses the genuinely useful, if (definitionally) less than 100% certain, “knowledge” we get from perception all the time.
Plato’s Argument
In severely abridged fashion:
The notion that “knowledge is nothing but perception” is put forward by one of Plato’s interlocutors, but identified as the position of the late great Protagoras. Also phrased as: “Man is the measure of all things.”
Socrates adds to Protagoras’ argument that “Perception, then, is always something that is, and, as being knowledge, it is infallible.” He tweaks this to add to it Heraclitus’ doctrine that everything is always changing, i.e. that “all the things we are pleased to say ‘are’ really are in the process of becoming.”
For Plato, this move is enough to dismiss perception as a source of knowledge. Knowledge is eternal. Objects of the senses are ever changing. We can’t finish the sentence “Snow is white” without it changing its meaning.
(Bertrand Russell goes through the argument in much more detail in his chapter “Knowledge and Perception in Plato.” I will spare you, but it is quite interesting. Among the gems: “Plato’s arguments as regards all other knowledge are fallacious. This does not, of course, prove that his conclusion is false; it proves only that he has given no valid reason for supposing it true.”)
We can agree, I think, that everything in the sensible world is constantly changing. The law of entropy works on all of us over time. Everything is made of atoms which are in a state of perpetual motion. Different neurons are firing in my head right now than were five minutes ago. Snow is clear when you look at it under a microscope! Also among the things that change are the meanings of words, albeit slowly.
And we can agree that our perceptions are often mistaken. In fact, we experience the world a split second after it happens, as it takes our brains time to process what we see and hear. This is a global error in perception; more local errors occur when we think we are awake but are in fact dreaming, or when an illness makes food taste different.
By setting the bar for knowledge from perception as infallibility, as Socrates does in this dialogue, he makes it all too easy to dismiss perception writ large. Equating knowledge with infallibility is, of course, limiting to the point of absurdity. In fact, as Karl Popper pointed out, for empirical knowledge to be at all useful or “scientific,” it must be disprovable. Scientific knowledge relies on falsifying past beliefs with empirical data.
Some perceptions are better than others, in the sense that they more closely approximate the true nature of the world. Plato gives voice to this idea in the dialogue. An objection to Protagoras’ doctrine of perception is that, if correct, no person knows more than any other; each knows what they know from perception, and no more. But consider the doctor who examines me when I am sick; she knows more of my future than I do. Given her experience, wisdom, and memory of past medical lessons and patients, my doctor’s perception is worth more than mine.
Again, Plato uses this as a counterargument to the notion that we can gain knowledge from perception at all. But of course it is perfectly cogent to say that some perceptions are closer to the truth than others! This is the notion behind more closely aligning your map with your territory. An expert’s forecast is worth more than a fool’s. An understanding of patterns and causal relationships within a field can move our map from being, say, a 60% accurate representation of the territory to something more like 80-90%. And that matters a great deal!
As with many humans, Plato is uncomfortable with ideas like probability and contingency. They are too messy, too uncertain. Anything between 0% and 100% is undeserving of the name “knowledge.” And that rules out everything beyond abstract mathematical concepts.
Assorted (tenuous) connections
Bayesian Reasoning: The idea that propositions about the world have a given probability of being true, based on a combination of our prior beliefs/experiences/theories and any new evidence we encounter. We can update our beliefs about the world in response to new evidence, assigning probabilities in the messy range between 0 and 1. We can even revise what constitutes “evidence” in light of, well, new evidence. Plato would not have been a good Bayesian.
Pragmatism: William James’ (correct) epistemological worldview which posits, in effect, that “truth is ‘a species of the good,’ like health. Truths are goods because we can ‘ride’ on them into the future without being unpleasantly surprised” (Stanford Encyclopedia of Philosophy). Bayesian reasoning is simply a mathematical construct meant to allow us to better describe the truth; pragmatism is the normative doctrine that truth is good, and we seek it, because of its good effects. (James prefigures Ludwig Wittgenstein’s later doctrine of “meaning as use”: the meaning of a word is its use in the language.) Plato was not a pragmatist.
Was Plato depressed? Scott Alexander’s recent blog post, “The Precision of Sensory Evidence,” analyzes a new theory of depression as an underweighting of sensory evidence, relative to priors. Scott writes, “All perception and cognition is the combination of evidence and priors. But in depression and otherwise neurotic people, the evidence is only a weak signal and the priors are a much stronger one.” It is, of course, entirely speculative to assign Plato’s apparent undervaluing of sensory evidence to depression. (Incidentally, this dissertation makes the case that Plato discusses clinical symptoms of depression, and recommends philosophical thought as an intervention! (Okay now this is officially a rabbit hole…))
I’m going to move on from Plato, at least for now. Thinking about him too much has made me crochety, and besides, Aristotle is due for some consideration in this blog. What can I say, except…it’s bad to think in all or nothing terms! It’s debilitating to see knowledge as a rarefied, elusive non-entity. Aristotle will be much more down-to-earth.