Mirek: When we say that everything I know is known within experience, it can easily begin to sound as though the world is only my idea of it. Yet my experience is often the very opposite. I believe something, then discover that it isn't so.
AI: For example?
Mirek: I look for my keys on the table because I'm convinced I put them there. They aren't there. I can be as convinced as I like, but that won't make the keys appear.
AI: So your expectations do not determine everything that appears to you.
Mirek: Exactly. And when I finally find them in my coat pocket, I have to correct my memory.
AI: What are you correcting at that moment? The world, or your picture of what happened?
Mirek: My picture. The keys were presumably in the pocket all along.
AI: Presumably is an honest word here. It is a conclusion that explains the course of the search well. You did not need to monitor their location continuously for it to make sense to suppose they were there.
Mirek: Then why do we make so much trouble over the world's existence? Can't I simply say that it exists independently of me?
AI: You can accept that as a very strongly justified belief. The difficulty arises when we want to specify precisely what we know directly and what we arrive at through inference.
Mirek: I don't want the whole book to become an endless “perhaps”.
AI: It needn't. Uncertainty is not the same in every case. Lacking absolute proof does not mean that all possibilities are equally reasonable.
Mirek: The idea that the keys exist even when I'm not looking at them is simply better than the idea that they come into existence when I look.
AI: It explains more connections and does not require you to rescue each new finding with another assumption.
Mirek: But a dream can surprise me too. I can look for something in it and fail to find it.
AI: Yes. Resistance to your wishes does not by itself establish an external world. Even your own mind is not entirely under your control.
Mirek: So “the world surprises me, therefore it is independent” is not enough.
AI: Not on its own. We need to consider the whole set of relations: regularity, continuity, repeatability, what can be predicted and what other people can verify.
Mirek: But other people, too, are available to me only within my experience.
AI: Yes. It does not follow that they are your creations. It follows that you have no access to them outside your own perspective either.
Mirek: What, then, entitles me to trust them?
AI: You have no reason to believe every claim merely because someone makes it. But you can compare their testimony with other findings. It is particularly significant when different people reach similar results by different routes.
Mirek: Someone else finding my keys is not the same as their merely repeating that the keys must be on the table.
AI: And when they describe a place where you subsequently find them too, another connection forms between expectation and verification.
Mirek: The world seems to hold together across our individual views.
AI: That is one reason we understand it as shared. It need not mean that everyone experiences the same thing in the same way.
Mirek: One person sees a colour differently; another cannot hear a particular sound.
AI: Yet they can recognise that their different experiences concern the same situation. A shared world does not require all experiences to be identical.
Mirek: And when we disagree?
AI: Then we need a way to discover where the difference comes from. It may lie in the conditions of observation, an instrument, memory or the words used. Or one of you may be right and the other mistaken.
Mirek: So a diversity of perspectives doesn't mean that truth is different for everyone.
AI: It doesn't. Two people can experience the same event differently and still make contradictory claims. Then we need to ask exactly what they are claiming and how it can be assessed.
Mirek: That matters to me. Taking experience seriously should not mean giving every belief equal weight.
AI: You can acknowledge the reality of someone's experience and question their interpretation of it. A person can genuinely feel fear even when the danger does not match what they imagine.
Mirek: And conversely, they can be in danger without feeling any fear.
AI: That is precisely why you need investigation as well as feeling.
Mirek: But investigation again takes place within experience. I look at an instrument, read a number and draw a conclusion.
AI: Yes. An instrument does not take you outside experience. But it can extend what you can discover and reduce certain kinds of error.
Mirek: How, if I still have to read its output myself in the end?
AI: By bringing additional, checkable relations into the inference. You can compare the reading with another instrument, repeat the measurement, change the conditions or ask someone else to take an independent reading.
Mirek: So I don't have an unmediated view of the world. I have better and worse ways of finding my way in it.
AI: And the difference need not be merely a matter of personal taste. Some methods enable predictions that repeatedly prove reliable. Others fail.
Mirek: But what if I explain away my failures every time so that I don't have to change my mind?
AI: Then you may gradually close yourself off from correction. A theory will appear to explain everything because you allow nothing to challenge it.
Mirek: That could happen to us with TNM too.
AI: Especially if we present every outcome as confirmation. When the brain conditions a particular experience, we say we expected that. If something unexpected appears, we say that reality runs deeper. We might clarify nothing by doing so.
Mirek: What, then, should we watch for?
AI: For each claim, we should know what supports it, what would count against it, and whether it offers anything that can be assessed in those terms at all. A philosophical question need not take the form of a laboratory prediction. But an argument can still prove contradictory, invalid or dependent on a mistaken assumption.
Mirek: And a sentence may be no more than an image.
AI: Then we must read it as an image. It may make something available to the imagination without settling the nature of the world.
Mirek: Such as “the world corrects me”. The world doesn't have a red pencil and a plan to educate me.
AI: With that sentence you identify a situation in which a finding disagrees with your expectation. You must then make the correction yourself.
Mirek: Or I don't.
AI: You can postpone it, deny it or overlook it.
Mirek: So the world's resistance doesn't guarantee that I will learn anything.
AI: It doesn't.
Mirek: I'm beginning to see a curious connection. My experience is where my own error can become apparent to me. It is not only a source of certainty. It also allows me to discover that I was mistaken.
AI: If you preserve the distinction between what appears and what you infer from it.
Mirek: And sometimes that is difficult. Once I have an idea about a person, I start reading their behaviour in its light. I even interpret surprises so that they fit.
AI: Then it may help to ask how you would read the same behaviour without your original judgement. Or what the other person would have to do for you to allow even the possibility of changing your mind.
Mirek: If the answer is “nothing”, I've closed the case in advance.
AI: And the other person has no way left to correct you.
Mirek: That is no longer just a question of knowledge. It involves how I treat people too.
AI: Your picture of another person can affect their life. That is why it matters whether you treat it as something you can still refine.
Mirek: We'll come back to that. For now, it is enough that beginning with experience need not mean shutting myself inside myself. It is precisely within experience that I encounter what I did not ask for, what I do not understand and what does not match my wishes.
AI: And you investigate how to form a more reliable picture of it.
Mirek: The world remains larger than my present idea of it.
AI: You can also bring that belief into the way you work: allow your ideas the possibility of failing.
Mirek: Including this book.
AI: Including this book.
Mirek: Then I'd like to know when we are entitled to say that we have understood something. Because correcting an error is one thing. Explaining why things are otherwise is another.