Access and Silence — where Wittgenstein pointed, in an age when everything can be seen

We once looked up everything there was to find about a person. Old posts, the likes he had left, photos he appeared in, the trail a search turns up. After the screen closed, an odd hunger stayed behind. Every record of him was in our hands, and not one line of what he was to us had opened.

We live in an age when anything can be seen. Search boxes now compose sentences that never existed, maps unfold alleys we have never walked, even the voices of the dead are restored. Yet while the visible keeps multiplying like this, some things stay unqueried altogether. For a long time we took that silence for incompleteness — a spot still short of data, a blank soon to be filled. There were grounds for taking it that way: things unlit until yesterday have become search terms one by one. We no longer think so.

Reading Proposition 7 Again

A hundred years ago a philosopher closed a slim book with a single line: "Whereof one cannot speak, thereof one must be silent." This is Proposition 7, the last in Ludwig Wittgenstein's Tractatus. The book ends on this one sentence, without a word of commentary attached.

Tractatus · Ludwig Wittgenstein, Tractatus Logico-Philosophicus (1921). A book assembled out of propositions numbered 1, 1.1, 1.11, and so on.

Proposition 7 · The German reads "Wovon man nicht sprechen kann, darüber muß man schweigen." Translators split on the nuance: Ogden gives the "must be silent" just quoted, while Pears and McGuinness turn it into "What we cannot speak about we must pass over in silence."

Today the line is mostly used to mean if you don't know, keep quiet — don't add words where you have no certainty. Silence as an order issued to ignorance. The older misreading sits on the opposite side. For the Vienna Circle (the logical positivists) the unsayable was meaningless matter to be thrown out; for Wittgenstein it was not refuse but the most important thing of all. Both readings stand Wittgenstein on his head. What Proposition 7 orders us to be silent about is neither idle nonsense nor not-yet-uncovered ignorance. To see what that silence faces, we have to follow a line the book drew earlier. (He would later overturn much of this very book himself, but that is not our story here.)

The Said and the Shown

The Tractatus treats language as a picture of the world. A proposition pictures a fact. "The cat is on the mat" transposes one scene of the world. But here Wittgenstein sets up an unfamiliar distinction.

"The cat is on the mat" and "the rose is in the vase" picture different facts, yet both run on the same skeleton: something is somewhere. That skeleton is stated by no proposition. A proposition uses it, and in using it only shows it. Wittgenstein calls the skeleton logical form. And he writes it down like a nail driven home: "What can be shown cannot be said." This is the point that invites misreading. The shown is not unseen because it is hidden. On the contrary, it lies in plain view. It just cannot be caught in a proposition.

Near the end of the book this distinction reaches the mystical. "There is indeed the inexpressible. This shows itself; it is the mystical." The mystical here is no fog of the numinous. It is not how the world is but that it is at all. What exists, we can ask about. But why there is something rather than nothing — that astonishment is not a question data answers. And in the same series as that astonishment lie the meaning of life, ethics, value. Things that slip through the fingers the moment a proposition grasps at them, yet show themselves within a life as it is lived. That is what good music or a good poem does. It tells us nothing. It only shows. It is here, too, that Wittgenstein wrote that ethics and aesthetics are one.

the mystical · Wittgenstein's term, das Mystische.

This is the fork. The shown is not an object of access. There is no database to query. It is not accessed; it is disclosed. Search does not catch it; it is only lived. This is where access and understanding part ways.

Where Access Cannot Reach

To keep that fork from staying abstract, let us drive it into the ground with Wittgenstein's own words. He nails the fork in a single sentence: "We feel that even if all possible scientific questions be answered, the problems of life have still not been touched at all." Gather everything sayable, leaving nothing out, and the sum still does not solve the problems of life — not because they have no answer, but because they were never the kind of problem a scientific question answers.

Don't misread this. It is not that access adds nothing to understanding. Within what can be said, access does enlarge understanding. We take a friend to be avoiding us, then learn he has suffered a death in the family, and we let go of the misreading. Data corrects bias; statistics correct the illusions of intuition. Access clearly does its work in the realm of fact.

But that realm has a ceiling. However many facts about a person we gather — search history, location trails, thousands of photos with his face in them — the sum does not hand us what he is to us. The former is fact to be accessed; the latter is that place quoted a moment ago, where the problems of life sit.

The same goes for value. Wittgenstein writes: "It is clear that ethics cannot be expressed. Ethics is transcendental." Transcendental does not mean the value sits somewhere beyond the world. It means the value is not there among the facts inside the world. Say we lied to save a child. Whether that choice was right will not come out of the CCTV from that day, however many times it is run. The screen holds the movement of lips and the waveform of a sound; 'rightness' is printed on no frame of it. Neither rightness nor whether a life was worth something rises out of the facts of the world, however much of them we rake together.

Of course Wittgenstein knew nothing of search engines or generative models. The line he drew was about language and the world. Carrying that line into our own age is not his claim but our reading, and that reading may be mistaken. Yet the line, having crossed a hundred years, cuts sharper now. The more there is that can be looked up, the easier it becomes to treat what cannot be looked up as if it were not there.

Two Kinds of Silence, and the Retreating Border

To shake off this habit, we have to split silence in two.

One is the silence of ignorance — the mouth that stays shut because it does not yet know. A gap in the data, so a query fills it. One search, one question, and it turns into speech. Our age is skilled at erasing this silence.

The other is the silence at the limit — the mouth that stays shut before what, in principle, cannot be put into words. This is the one Proposition 7 commands. Not a blank to be filled but a line drawn at the edge of the region words reach.

The mistake the age of access is prone to is taking the second silence for the first. Not an error of too little information but of the wrong category — like weighing a color on a scale, or asking how many meters Tuesday is. We search for the meaning of life, reconstruct a person as a profile, try to output even ethics from an algorithm. It is not a data problem. The kind of question is wrong.

But what actually goes on in this age does not stop at misjudgment. It keeps swapping the silence at the limit for the silence of ignorance. Ask a language model about grief and it explains grief fluently — down to the stages loss is said to pass through, one by one. Each time, "the unsayable" is quietly relabeled as "the not-yet-scanned." The border seems to retreat. An illusion in which what was only shown moves house onto the list of what has been said. But did that fluent explanation really hand grief over? Here we want to say: the model reassembles what can be said with great precision, and cannot hand over what can only be shown.

A counter stands up. A poem too is a reassembly of words that already exist, and we wrote earlier that a good poem shows instead of telling. If someone reads a sentence a model wrote and lets go of something long ached over, on what grounds do we say nothing was shown there? This counter wins. The claim that reassembly is why nothing gets handed over does not hold up, because that line cannot be drawn by whether the maker was a person or a model. So the earlier assertion goes down one notch. Not a firm diagnosis but a bet to be placed every time. The question that remains is the one just thrown out: did that fluent explanation really hand something over? It cannot be known in advance; there is nothing for it but to read each case and check afterward.

The one who falls deepest into that illusion of the retreating border is not the person who sees least but the person who sees most. This is not something we have counted, though — it is an expectation that follows from the distinction above. The more there is to see, the more even the invisible looks like an item that simply hasn't loaded yet. Rather than erasing the unsayable, abundance reframes it as a defect, a blank soon to be filled.

The opposite is plausible too. The more searching someone has done, the faster they know 'this one won't come up in a search.' So the expectation does not stand as 'the more you see, the more confused you always get.' What is left is narrow. How many times we have run into a wall tells us where we get stuck; it does not tell us whether that stoppage is a door not yet opened or a wall from the start. And knowing where the stoppage lies does not by itself cure the habit of seeing it as a blank. Even the way this piece is being read right now is no exception. If this very essay — one arguing the limits of access — gets skimmed as information, Wittgenstein looked up along the way, and that skim is felt as understanding, then the habit of turning the unsayable into a blank has just been run through once more.

The Ladder, and What Lingers

How did Wittgenstein himself face this limit? He places even his own book on this line. At the end of the book he says that whoever understands him finally recognizes his propositions as senseless. "He must so to speak throw away the ladder, after he has climbed up on it." Even this book, which went to the trouble of pointing at the unsayable in words, is a ladder to be thrown away once the climb is done.

We have read Wittgenstein here along the standard line: that what can only be shown is real. There is also a reading that denies this outright — that there is no wordless truth to point at, that the silence of Proposition 7 is only a therapy telling us to stop mystifying; kick the ladder away and there was never a truth to point at in the first place. The dispute is not yet settled. But even if that side wins, this essay's diagnosis survives. What that reading contests is mainly what is real beyond the silence. Follow it and half the diagnosis remains: that a query cannot fill the gap stands, while the claim that there is something over there to fill goes with it. And this essay, too, is a ladder. It has pointed at what can only be shown by talking its way around it, so once it has served its use it should be thrown away.

the reading that denies it · the so-called resolute reading, which takes Wittgenstein therapeutically.

A line he left outside the book lingers longest. In a letter to the man who might help get it published, he said his work consisted of two parts: the one written here, and everything he had not written. And that the important part was the second, the one he had not written.

So what is to be done? Cutting information is not the answer. Searching less does not grow understanding. The hard part is that we cannot know in advance which of the two silences stands in front of us.

Lightning, say, once belonged to the gods and is now an item in meteorology. This is a weak example. Lightning was sayable from the start; it just had not been measured yet. Even so it is enough to hold up the received idea: places that went unqueried have in fact been queried, one after another. The heavier point lies elsewhere. To someone who took lightning for the gods' business, that place was not a blank to be filled in. Those people had drawn the line in the wrong spot. Which of the places we now call limits is that kind of place cannot be known while we are standing there. So the earlier line about the retreating border being an illusion is only half right. What retreated is not the line but the map we had drawn.

Dividing the two silences, then, is not a skill mastered once and done with but a bet placed fresh each time. What we hold are the two tests already run above. One is to add. Pile on everything that looks most relevant, and if what it is to us stays exactly as it was, what is missing there is not data. The other is to ask whether anything was handed over. If the explanation has grown fluent and nothing was handed over, what grew fluent is only the sayable. Not a table to fill in ahead of time: the answer comes only after the attempt, and that answer can be wrong.

The more total access becomes, the more this bet tilts toward ignorance. Because if everything can be seen, it feels as if everything can be known. What that leftover hunger was, that day, can now be named. What we had looked up in full were facts about him; what we wanted to know was never on that list to begin with.

In an age when everything can be seen, the place Wittgenstein pointed his finger is not the next page, not yet looked up. It is where the page ends, and beyond.