On the Examined Life

An imagined dialogue between Roman, Troy and Socrates.

Roman: I keep finding beliefs that seem secure until I examine what supports them. A memory, for instance. I remember someone saying something, and the memory seems to settle the matter. But I am relying on what I heard, how I understood it, what I retained and what I have since made of it. By the time I say “I know what happened,” several things have already had to go right.

Socrates: And you want to establish that they did?

Roman: As far as I can. Otherwise the confidence belongs to the finished account, while the weaknesses remain hidden in how I arrived at it.

Socrates: Suppose you remember leaving your keys on a table. Must you establish the reliability of your memory before looking there?

Roman: No. Looking costs very little, and it also tests the memory. If the keys are absent, I have learned something. I would be making ordinary life impossible if I demanded a complete account of memory before using it.

Troy: That is a useful kind of decision. Being wrong gives you information without doing much damage. Accusing someone on the strength of a memory is different. You can discover that you were wrong after the accusation has already changed the relationship.

Roman: Which is why the standard has to depend partly on what I am about to do. I can look for keys on a fairly weak recollection. I should want considerably more before treating somebody as dishonest. But I dislike how quickly that sounds as though truth itself changes with the circumstances.

Socrates: Would the cost of looking change where you left them?

Roman: No. What happened remains what happened. My access to it can improve or deteriorate, and the action I can reasonably take depends on that access. The fact and my grounds for believing it are different things.

Socrates: Then why ask the grounds for an action to perform the work of settling the fact beyond all possible doubt?

Roman: Because I know how easily “enough to act on” can become “enough to justify what I wanted to do.” I can admit uncertainty and still use the admission dishonestly. Perhaps especially then, because now I sound careful.

Troy: You want a test that your preferred conclusion cannot quietly pass for itself.

Roman: Yes. If I am suspicious of someone, I can keep gathering details until they look like a pattern. If I trust them, I can explain each detail away. Both accounts may feel coherent. Coherence tells me the parts fit together; it does not tell me whether I have selected the parts fairly.

Socrates: What would you do before deciding?

Roman: Separate what I can establish from what I have inferred. Ask what else would explain it. Look for something that could count against my preferred account, rather than collecting only what supports it. And decide what the situation requires now, because finding the complete explanation and choosing the next action may be different tasks.

Socrates: Give us the difference.

Roman: If a conversation leaves me unsure whether someone has misled me, I can ask them to clarify, check a record or postpone a commitment. I do not have to decide their entire character before deciding what information I need. Sometimes the next reasonable action is small enough that it keeps several explanations open.

Troy: Unless waiting carries a cost of its own. You can preserve uncertainty so carefully that someone else has to live with the consequences.

Roman: Yes. That is where the keys example stops helping. Some decisions cannot be delayed, and some risks become worse while I investigate. I may need to protect a person or set a boundary before I know everything. The question becomes which action is warranted by what I have, including the cost of doing nothing.

Socrates: And has the need to act improved your evidence?

Roman: No. It has changed the cost of waiting. That is an uncomfortable distinction, because once I have acted I want the evidence to have been sufficient in every sense. I want to look back and say I knew. But I may only have had reasonable grounds for a precaution.

Troy: Especially if the precaution itself hurts someone. Certainty would make it easier to carry.

Roman: It would. I could turn a difficult judgement into an obvious necessity and stop feeling responsible for the uncertainty I acted within. That is one reason I want to keep a record of what I actually knew at the time. Otherwise the explanation can improve after the event without the original evidence improving at all.

Socrates: So examination also concerns what you do with a decision once it is yours.

Roman: Probably more than I would like. Before deciding, I can imagine myself willing to be corrected. Afterwards there is a choice to defend, perhaps an apology to make. I have acquired an interest in the account being true.

Troy: Then marking uncertainty is only a beginning. You also need to leave yourself some way of responding when the evidence changes.

Roman: Yes, although I do not want that turned into an elaborate ceremony for every ordinary thought. I would spend the day reviewing whether I had enough reason to make breakfast. Most things work well enough to proceed. The effort belongs where an error would matter, where the evidence is unstable, or where I have a strong reason to prefer one answer.

Socrates: You have described a life in which some beliefs receive more examination than others. Does that trouble you?

Roman: It does, because part of me still wants the whole thing to account for itself. I want what I believe, what happened and what I do to agree without my having to manage the distance between them. I can see why that is too much to demand before acting. I have more difficulty accepting that it may never become a finished condition.

Troy: Would finishing it today settle what you encounter tomorrow?

Roman: No. And I think that is what keeps undoing the project. Even a sound belief has a scope. I can understand one situation accurately and then use that understanding where the conditions have changed. The wish to be finished may itself make me less willing to notice.

Socrates: What, then, would you count as progress?

Roman: Being able to say more precisely what I know, what I am assuming and what would make me reconsider. Choosing an action that fits the evidence and the stakes, instead of quietly enlarging the evidence until it fits the action. And correcting something without having to pretend I was a different person when I believed it.

Socrates: Would that satisfy you?

Roman: It would improve how I live. I am less sure it would satisfy the part of me that wanted certainty in the first place. I can accept that the work remains unfinished without finding unfinished work especially comfortable.