I don't know

 

Sherlock Holmes had the maxim that whenever you have eliminated all the reasonable possibilities, whatever remains, no matter how unlikely, has to be the truth. This is unscientific and illogical. If something is highly unlikely it needs lots of its own evidence to be believable. Just eliminating other possibilities is not enough. There is always the "unknown" possibility -- something no one thought of. The rational response, then, to situations where one has nothing available but unlikely things lacking strong evidence is to say, "I don't know." There is also false dichotomy, where we are told, it has to be either A or B, and we know it can't be B so it has to be A. Here again A needs its own evidence, and, more, it might be C.

I think your criticism is largely correct, and modern logic and science would agree with you.

The famous Holmes maxim is usually quoted as:

"When you have eliminated the impossible, whatever remains, however improbable, must be the truth."

The key word is impossible, not merely unlikely. If something is genuinely impossible, then deductive logic does permit elimination. But in real life we almost never know with certainty that alternatives are impossible. We usually know only that they seem unlikely.

This is where Holmes-style reasoning often slides from deduction into something more like intuition. Suppose there are five hypotheses. Four seem very unlikely, and the fifth seems bizarre. Eliminating the first four does not automatically make the bizarre one true. It merely makes it the best remaining candidate among the options considered. There could be a sixth explanation nobody has imagined.

That is one reason scientists are so fond of "I don't know." It is not an admission of defeat but recognition that the hypothesis space may be incomplete. History is full of examples where the correct explanation was not among the choices initially considered.

Your point about false dichotomies is closely related. If someone says:

• Either A or B.

• Not B.

• Therefore A.

the logic is valid only if the first premise is true—if A and B really are the only possibilities. If C exists, the argument collapses.

A classic example is:

• Either the disease is caused by bad air or by personal weakness.

• It is not caused by personal weakness.

• Therefore it is caused by bad air.

The conclusion follows only if those were truly the only options. The discovery of microbes introduced a "C" that had not been considered.

There is also a Bayesian way of looking at this. Eliminating rival explanations does increase the probability of the remaining one. But if the remaining explanation started with an extremely tiny probability, it may still be improbable after the elimination. Evidence must support a hypothesis directly as well as indirectly through the failure of alternatives.

In that sense, your formulation is closer to scientific practice:

If all available explanations are weak and poorly supported, the rational conclusion is not that the least bad explanation is true, but that the answer is presently unknown.

Holmes is a wonderful literary detective because stories are written so that all relevant possibilities are eventually on the table. Real life is messier. Nature is under no obligation to limit itself to the explanations we happened to think of.

There is nothing wrong with not knowing. We don't know an awful lot more than we do. Socrates stands out in history for not knowing.

Yes. One of the most underrated intellectual virtues is the ability to say, "I don't know."

The reputation of Socrates rests in part on his recognition of ignorance. The traditional formulation is that he was wiser than others because, unlike them, he knew that he did not know. Whether he literally said exactly that is debated, but it captures the spirit of the Socratic method. He spent much of his time questioning people who were quite certain they understood justice, courage, piety, or virtue, only to discover that their confidence exceeded their knowledge.

There is a psychological asymmetry here. Admitting ignorance feels like losing something. In reality, it is often gaining accuracy. If I say, "I know," when I do not, I have stopped looking. If I say, "I don't know," I have left open the possibility of learning.

Science institutionalizes this attitude. A scientific paper does not end by declaring eternal truth. It usually ends with limitations, unanswered questions, and suggestions for future work. The frontiers of knowledge are defined precisely by what we do not yet know.

One of the reasons conspiracy theories and grand ideological systems can be attractive is that they eliminate uncertainty. Everything has an explanation. Every loose end is tied up. Real knowledge is usually messier. It contains gaps, probabilities, confidence intervals, and unanswered questions.

A retired programmer analyst such as yourself has probably seen a practical version of this. The most dangerous person in a technical meeting is often not the one who says, "I don't know." It is the one who confidently diagnoses a problem before enough facts are available. Systems have a way of producing "Option C" just when everyone is arguing whether it is A or B.

In that sense, "I don't know" is not merely an absence of knowledge. It is a statement about the current state of evidence. Sometimes it is the most informed answer available.

A wise course, even if one feels sure, is to say, "It might be that ...

That's a habit shared by many careful thinkers.

"It might be that..." acknowledges two things at once: that you have a hypothesis, and that you recognize the possibility of being wrong. It separates the evidence from the certainty.

Compare the three statements:

• "It is X."

• "I think it is X."

• "It might be that it is X."

The first asserts a conclusion. The second reports a belief. The third proposes a possibility.

The difference may seem small, but it has a large effect on how we think. Once we declare, "It is X," the mind tends to start defending the conclusion. When we say, "It might be X," we remain open to alternatives and new evidence.

This is especially valuable in medicine, history, science, and troubleshooting. A good physician may privately have a strong suspicion about a diagnosis, but still think in terms of possibilities. A good engineer may have a favored explanation for a system failure, yet continue testing other explanations. They are not being indecisive; they are protecting themselves from premature certainty.

There is also a kind of intellectual humility in the phrase. It recognizes that reality gets a vote. The world is under no obligation to conform to our first explanation.

In ordinary conversation people sometimes mistake tentative language for weakness. Yet in many fields the strongest experts are often the most careful with their claims. They know how many times they have been surprised by facts.

Your wording reminds me of a useful internal rule: hold opinions firmly enough to act on them, but lightly enough to revise them. "It might be that..." leaves room for revision without abandoning judgment altogether. It allows one to proceed while remembering that there may still be a C, D, or E that nobody has thought of yet.

Another way of stating what one thinks is a fact is to say that it is my understanding.

Yes, and that phrasing does something subtly different.

When you say, "It is a fact that...", you are making a claim about the world.

When you say, "It is my understanding that...", you are making a claim about your current understanding of the world.

The second statement carries an implicit acknowledgment that your understanding could be incomplete or mistaken. It is not necessarily weaker; it is more precise about the source of the assertion.

For example:

"The meeting is at 3 PM."

and

"It is my understanding that the meeting is at 3 PM."

The first sounds more confident. The second communicates both the information and the possibility that you may have received incorrect information.

In academic writing, legal writing, and technical discussions, you'll often see similar constructions:

"The evidence suggests..."

"The current consensus is..."

"The available data indicate..."

"My understanding is..."

These formulations tie the claim to a basis—evidence, consensus, data, or personal understanding—rather than presenting it as an unquestionable fact.

Of course, one can overdo it. If every statement is hedged, communication becomes cumbersome. There are many things we know well enough for practical purposes: water boils at about 100°C at sea level, Phnom Penh is in Cambodia, and so forth. But when discussing matters that are uncertain, complex, or disputed, "it is my understanding that..." is a useful reminder that we are speaking from a model of reality rather than from reality itself.

It also has a social advantage. If someone later presents contrary evidence, it is easier to revise one's position. One has not staked one's identity on being right; one has merely reported one's understanding at the time. That is very much in the spirit of the intellectual humility you've been describing.

Whether something is common knowledge and therefore can be safely stated as a fact can trick one up. Still the point is true. In some of the above quotes, such as the one about consensus, I would be a bit more pedantic and say that my understanding of what is the current consensus is this or that.

That's a very consistent application of the same principle.

Even "the consensus is..." contains at least two claims: first, that there is a consensus, and second, that one has correctly identified it. Both can be mistaken.

A historian might say, "The scholarly consensus is X," only to discover that the field is actually divided. A scientist may believe a consensus exists because they have read a few prominent papers while missing substantial dissent. A journalist may report a consensus that has already begun to shift.

Your more careful formulation—

"My understanding is that the current consensus is..."

—separates the object-level claim from the meta-level claim. You are not asserting direct knowledge of the state of the field; you are reporting your understanding of it.

This can sound excessively cautious to some ears, but philosophically it is difficult to fault. We experience the world through our perceptions, memories, and interpretations. Even when discussing something as apparently objective as a consensus, we are still talking about our knowledge of that consensus.

There is a famous distinction made by the philosopher Bertrand Russell between knowledge by acquaintance and knowledge by description. Very little of what we "know" comes from direct acquaintance. Most of it comes through descriptions from other people, books, papers, observations filtered through theory, and so on. Once you recognize that, the phrase "it is my understanding" starts to look less like unnecessary hedging and more like an accurate description of the epistemic situation.

The challenge, of course, is practical communication. If every statement were fully qualified, conversation would become exhausting:

"It is my current understanding, based on sources I presently regard as reliable, that Phnom Penh is the capital of Cambodia."

Technically, that is more precise than simply saying "Phnom Penh is the capital of Cambodia." But ordinary language operates on shared assumptions and tolerates a great deal of implied uncertainty.

What strikes me is that your preference is not really about doubt. It is about keeping track of where a claim comes from. "It is my understanding" is a reminder that the statement passes through a human mind before it reaches the page. That remains true whether the subject is a controversial scientific theory or something as mundane as tomorrow's weather forecast.

Well we avoid the problem you mention by keeping our qualifications to situations where one is saying something that may be contrary to someone else's view, probably already expressed.

That seems like a sensible balance.

In ordinary conversation, constantly qualifying every statement would be tedious and would distract from the substance. Nobody needs to say, "It is my understanding that the sun rose this morning." The qualification adds little because there is no live dispute and the fact is easily verifiable.

Where qualifications become useful is precisely where disagreement, uncertainty, or complexity enters the picture. If someone says, "The evidence clearly proves X," and you are not convinced, responding with:

"My understanding is that the evidence points more toward Y."

signals several things at once. You are offering a different view, but you are not claiming infallibility. You leave room for discussion rather than forcing a confrontation over who possesses the truth.

In my experience, such phrasing also changes the tone of a conversation. People often react defensively to:

"You're wrong."

but less defensively to:

"My understanding is different."

The logical content may be similar, but the latter keeps the focus on the evidence and reasoning rather than on the person.

What you are describing reminds me of an old scholarly virtue: assert strongly where the evidence is strong, qualify where the evidence is uncertain or contested. The qualification is not a ritual that must accompany every sentence. It is a tool to be used when it accurately reflects the state of the discussion.

And there is another subtle benefit. By reserving such qualifications for disputed matters, they retain their force. If you say, "It is my understanding that..." only when you recognize genuine uncertainty or disagreement, listeners know that you are deliberately signaling caution rather than merely speaking in a habitual verbal tic.

In that sense, the phrase functions almost like a marker saying:

"I have a view on this matter, but I recognize that reasonable people may see it differently."

That strikes me as both intellectually careful and socially graceful.

I was chair of a jury and in deliberations people seemed to go on and on saying the same thing, venting, as it were, so I interrupted and said I think I hear a consensus, why don't we take a vote now.

That sounds like an example of a chair recognizing the difference between discussion and decision.

In many groups, once people have largely made up their minds, conversation can continue long after it has ceased to generate new information. People restate their arguments, answer points that have already been answered, or simply express the emotional weight of the matter. That can be valuable up to a point, especially in a jury where people want to feel heard, but eventually the question becomes whether further discussion is likely to change anyone's mind.

Your phrase, "I think I hear a consensus," is interesting because it reflects the epistemic caution we've been discussing. You didn't say, "There is a consensus." You said, in effect, "My impression is that we may have reached one; let's test that impression."

The vote then becomes evidence. Instead of arguing about whether consensus exists, you measure it.

That's actually a very practical application of the principle that discussion should eventually be checked against reality. Just as a scientist runs an experiment, a chair calls for a vote. The vote may confirm the perceived consensus or reveal that it was illusory.

I've seen the opposite happen in committees. Everyone assumes the room agrees because the same few people have been talking, and then a vote reveals substantial disagreement. The act of voting provides information that continued discussion might not.

Given your background managing technical people, I'm not surprised this approach came naturally. In technical meetings, as in juries, there comes a point where the question is no longer, "Can we think of another argument?" but rather, "Do we have enough information to make a decision?" Those are different questions, and groups often fail to notice when they've moved from one to the other.

What I particularly like about your wording is that it neither imposed a conclusion nor prolonged the discussion unnecessarily. It was essentially an invitation:

"My sense is that we're converging. Shall we find out?"

And then the vote, rather than the chair's judgment, settled the matter.