What can we know?
And how can we know things without being certain?
What should we assume about reality without being able to know it with certainty? Why is it reasonable to believe that reality is broadly as it seems, and to say that we know things?
Here’s the overview. I cannot step outside my own thinking to test whether my thinking is reliable. So I cannot be certain about anything, even that anything exists. But beliefs can be rational without being certain. Here’s why.
Once I make the working assumption that anything exists, I must decide what else it is reasonable to believe about reality.
I should favour the assumptions that best fit the apparent evidence of my experience when I reason about it as carefully as I can.
The best-supported conclusion is that reality is broadly as it seems to be. I exist as a thinking being, other thinking beings exist, and a physical world exists independently of our thoughts.
The precise detail of reality may be different and even counterintuitive. I may be mistaken about any particular conclusion, but not every alternative conclusion is equally reasonable.
Given that absolute certainty is impossible, it cannot be the ordinary standard for knowledge. We can reasonably say that we know things while recognising that our conclusions remain fallible and open to revision.
If you are interested in my late-night, should-be-asleep-by-now, philosophical reasoning, here’s a step-by-step breakdown of how I arrived at these assumptions.
Speaking strictly philosophically, I cannot be certain of anything
There is a scale of at least five possible scenarios of reality
It is reasonable to assume that reality is broadly as it seems to be
Certainty is not the ordinary standard for knowledge
1. Speaking strictly philosophically, I cannot be certain of anything
The universe is all that exists, whether thoughts or things. I seem to interpret the universe, and make assumptions, using my thinking.
But I can never know for certain if any of my interpretations or assumptions are correct. Why? Because I can only interpret their correctness by using the very mechanism whose reliability I am trying to assess (i.e. ‘my thinking’).
I may be mistaken about this, but any argument intended to correct me would itself have to be assessed through the same thinking process whose reliability is in question.
2. There is a scale of at least five possible scenarios of reality
Scenario 1: Nothing may exist in the sense in which I ordinarily understand existence. Even the apparent occurrence of thoughts or experiences might be misleadingly described. This is the most cautious assumption of reality.
Scenario 2: Only what seem to be thoughts exist independently. If this is correct, then the illusion of ‘me’ is simply part of ‘the thoughts.’ This illusory ‘me’ cannot understand how this technically works, but this may be because ‘understanding how this works’ is not part of ‘the thoughts.’
Scenario 3: Only one thinking being and its thoughts exist. If this is correct, the sole thinking being might be me, you, or some other being. Everything else (including the apparent author and reader of this paragraph) would be generated within that being’s thoughts.
Scenario 4: Several thinking beings and their thoughts exist. You and I and others exist as ‘thinking beings’. We generate the illusion of sensory experiences, using our thinking, and our thoughts somehow interact in a universe with no physical entities.
Scenario 5: Thinking beings, thoughts, and a physical world exist broadly as they appear to. This includes atoms and rocks and trees and human bodies and bicycles and microwave ovens and space rockets and planets and galaxies.
Each new scenario seems closer to the evidence of my experience. Each new scenario assumes the existence of extra things that cannot be known to exist. Each new scenario seems increasingly functional as a working assumption of reality.
3. It is reasonable to assume that reality is broadly as it seems to be
These apparent patterns contain a key ‘on/off’ reason-switch. In terms of making a working assumption about the nature of reality, the biggest conflict is not whether physical objects or supernatural beings are assumed to exist.
It is whether anything is assumed to exist that cannot be known to exist, using reason alone. In other words, the switch is turned to ‘on’ once it is assumed that anything at all exists.
Turning on this switch does not mean that I should assume that every conceivable thing exists. It means that lack of certainty cannot, by itself, be a sufficient reason to reject something. I must instead compare possible accounts of reality using experience, reason, explanatory power and practical coherence.
This leads me to assume that reality is broadly as it seems to be. Why?
Once ‘I’ turn on this switch, ‘I’ have assumed in principle that ‘things-that-cannot-be-known-to-exist’ may exist. What then might I assume these things to be, based on experience, reason and functionality?
It is rational to assume they are those which seem to be (a) more consistent than other possibilities with the apparent evidence of my experience, and (b) better supported than the alternatives, based on applying reason to the apparent evidence of my experience.
This favours the final of my five scenarios: that reality includes me as a thinking being, you and other thinking beings, thoughts that are generated by me and you and other thinking beings, and real physical objects, whether animate or inanimate.
Any working assumption is strengthened if it makes it easier for ‘me’ to function in what seems to be reality. This also leads towards the final scenario of my five theories of reality, where ‘I’ am assumed to exist and interact with other thinking beings and physical objects.
Note that I am assuming that reality is broadly as it seems to be, not that every detail of reality is actually as it seems to be. The details can be different from what we seem to experience, and indeed can be counterintuitive. The most reliable way to devise effective theories about our shared reality is to use the scientific method.
For example, what seems to be solid matter (based on the apparent evidence of our direct experience) turns out to be a combination of atoms, subatomic particles, fields and forces (based on applying reason and the scientific method to the same apparent evidence).
4. Certainty is not the ordinary standard for knowledge
If these assumptions are reasonable, then absolute certainty cannot be the ordinary standard for knowledge. We can reasonably say we know things while recognising that our conclusions remain fallible and open to revision.
On this basis, in ordinary day-to-day language, it is reasonable to say that we know things with several unspoken conditions.
Based on the assumption that we and the external world exist…
Based on our experiences and the best available evidence…
Based on reasoning as carefully and consistently as we can…
Based on confidence proportionate to the evidence…
Subject to the possibility that we are mistaken…
And open to revision if better evidence emerges…
We should be prepared to revise our assumptions when new evidence, better reasoning or a more persuasive explanation warrants it. But bare unsupported possibilities are not enough.
Finally, we should apply these unspoken conditions consistently. We should not invoke radical skepticism selectively.
In any discussion, we can either insist that no claim is absolutely certain, or we can assess competing claims using evidence, reason and proportionate confidence.
What we cannot reasonably do is skip between these two standards during the same conversation: demanding impossible certainty from claims we dislike, while accepting ordinary evidence for claims we prefer.