The Perpetual Rose and Its Reader

كلما اتسعت الرؤية ضاقت العبارة
Kullamā ittasaʿat al-ruʾya, ḍāqat al-ʿibāra.
“As the vision widens, speech falters.”
- al-Niffārī, Kitāb al-Mawāqif (4th c. AH / 10th c. CE)¹

I was a heretic before I was old enough to drive. The environment that raised me - not merely the people in it, but the whole ecosystem of belief, the assumptions in the air one breathes before one knows air can be questioned - held certain things beyond question, and at fifteen I began questioning them. There was no internet to speak of yet, only its dial-up, pseudonymous infancy; and it was there, in the chatrooms and forums, among strangers who fought their wars of religion behind invented names, that I first met the words: epistemology, ontology, axiology. They reached me as talismans, syllables drawn from a grimoire rather than the names of scientific pursuits. The other heretics were those pseudonyms - and most of them, I came to see, rebelled for the thrill of it, for the cheap electricity of being the dissenter in the room; their unbelief was a posture, and it cost them nothing. Mine was not a posture. While the boys my age went online to meet girls, or to fill their heads with the usual earthly fantasies, I prowled the same wires frenzied, half-possessed, searching for a Magister Ludi² who might show me the way to truth - desperate, even then, for there to be a truth, rather than the bleak indifference I feared was the only thing waiting underneath. One of those words would not let me go. I did not know what a first principle was, but I understood, or perhaps yearned to understand, that nothing could be settled about what exists, or about what any of it is worth, until something was settled about how anyone comes to know. So I started there, methodless and without permission, on no authority but that yearning. Epistemology, reduced to its domestic essentials, is the question of what can be settled from an armchair and what requires leaving the house.

Some truths, I learned from the great philosophers, do require making the trip - an inconvenience for a social agoraphobe such as myself. That a man is courageous is not a fact I could have worked out at my desk. I had to go into the world and ask. Philosophers call such truths a posteriori, and they are, like messiahs in occultation, hidden until sought. The world holds the receipts, and the world must be consulted.

Other truths excuse me from the errand. That all bachelors are unmarried I can certify from my armchair, without consulting a single bachelor, because the answer is baked into the question. Unmarriedness is not news about bachelors; it came woven into the concept. Immanuel Kant called these judgments analytic, and he thought them sterile. They tell you nothing you did not already have.

Then there is the third case - the strange one. I do not need to leave the house to know that two plus three equals five, and yet, Kant insisted, this is not like the bachelors. Stare at two as long as you like; interrogate three in a small cell. Five, he claimed, is not in there. To reach it, the mind must do something - count on its fingers, run the beads, consult what he called intuition. And intuition, for Kant, is not learned; it is the software we arrive with. A newborn has never consulted a geometer, yet she moves her arms and legs through three-dimensional space without hesitation, as if the axes were installed before the tenant moved in. On Kant’s view, they were. Space is not something we learn from the world; it is the form in which any world must be given to us. So arithmetic, for Kant, is knowledge that is genuinely new and yet requires no experience - synthetic a priori - and explaining how such knowledge is possible became the central business of his philosophy. Kant ventured that the mind is a partner in constructing experience; and if the mind is so conjoined a twin to the world it experiences - able to divine real knowledge without leaving the house - then mind and world must be made for each other, fitted like a key to a lock.

But even as a naïve teenager, I could not swallow that articulation. I could not shake off my skepticism. I pressed it on those strangers in the forums, the ones writing under pseudonyms - or noms de guerre, when the fights between the atheists and the theists turned fierce in that early-2000s world of dial-up internet. My skepticism was never partisan; it bit whichever hand was reasoning badly. When the believers said something sloppy, I pressed them. But when the atheists produced a party trick, I pressed them too. - “Can God create a stone so heavy He cannot lift it?” they asked with a smirk, as though heaven had just been checkmated.

The question to me was gibberish, and I said so. Archimedes promised that with a lever and a place to stand he could move the Earth; give me the right lever and there is no stone I cannot raise. Liftable is not a property a stone carries in its bones - it is a relation between a weight and a force, and for any weight there is a force that can overcome it. So “a stone that cannot be lifted” does not name a hard case for omnipotence. It does not name anything. Either the questioner means a stone God chooses not to lift, in which case the answer is yes and divine power was never at issue; or he means an object whose nature is to be liftable and which is also, by stipulation, not liftable - a square circle dressed in scripture. He might as well ask whether God can create a trepundoron. What is a trepundoron? I have no idea - but I have told you it cannot be lifted, and that was the only work the word was ever hired to do. The trick is always the same: smuggle the word cannot into the description of the thing, then feign astonishment when it surfaces in the conclusion.

I dwell on this because I turned the same instinct - the refusal to let a premise be slipped past me, with no more reverence, on Kant. The atheists told me power was missing from God; Kant told me five was missing from two and three. In both cases I suspected the absence had been conjured by sleight of hand rather than found in the thing, and in both cases I wanted to look for myself. I was not convinced that five is missing from two and three. It seemed to me then, and seems to me still, that five is not a new arrival but a package already containing its partitions.

Five fingers do not merely allow a split into two and three; in some sense they are every such split at once. When I bring two fingers together with three, nothing new enters the room - the five was always a five, and if that is right, the equation is analytic after all. I later read that Leibniz had suspected as much, and that Frege built a cathedral on it. Frege’s program, called logicism, set out to prove that Kant was wrong about arithmetic - that every arithmetical truth could be derived from pure logic plus definitions, with no intuition and no preinstalled software; arithmetic, for Frege, was analytic through and through. Numbers themselves would be defined logically: two as the class of all pairs, five as the class of all five-membered collections, addition as a logical operation upon them. If the program worked, 2 + 3 = 5 would be a truth of the same kind as “all bachelors are unmarried” - a mere disclosure of contents, not a new creation. Frege spent decades on the construction. Then, in 1902, as the final volume sat at the printer, Bertrand Russell sent him a letter. Frege’s system permitted the formation of any class by any condition - including, Russell observed, the class of all classes that are not members of themselves. Is that class a member of itself? If it is, then by its own definition it is not; if it is not, then it qualifies, and it is. A contradiction either way; and a system that proves one contradiction proves everything, and a system that proves everything proves nothing. The cathedral came down with the scaffolding still on it. But while Russell wrecked the proof, he could not make my skepticism go away. If my surveying tool turns out to be miscalibrated, my measurement may be void - but the land has not disappeared. The feeling that five conceals two and three sat in my heart exactly where it sat before 1902, as compelling as ever, merely uncertified.

Support for my intuition arrives from an unexpected quarter: the formalists, the cult of cold mathematicians who wanted to rebuild arithmetic as rule-following, with no appeal to meaning or insight at all, aiming to construct it as a pure game of symbols, with every move stamped and decreed by an explicit rule and nothing left to insight. In their reconstruction, the numbers are not assumed; they are manufactured from a single starting point. Begin with zero, written 0. Then introduce one operation, the successor - “the next number after” - written S. One is the successor of zero: S(0). Two is the successor of that: SS(0). Three is SSS(0), five is SSSSS(0), and so on: every number is just zero wearing some number of S’s. Addition gets defined by a simple rule for shuffling S’s from one side to the other. Now write our equation in this notation - SS(0) + SSS(0) = SSSSS(0) - and apply the rules, one careful step at a time. The S’s migrate across the plus sign; the plus sign itself is eventually consumed; and the derivation terminates in SSSSS(0) = SSSSS(0). Look at what that says. Five equals five. The great discovery that two and three make five turns out, at the bottom of the machinery, to be the discovery that a thing is itself, a tautology. The new knowledge Kant offered me has vanished in transit, and the great philosopher’s promise of the synthetic a priori seems to evaporate.

Anguished, I took no pleasure in the demolition. I had gone to epistemology as a boy goes to an apothecary, wanting something I could take that would let me keep living; what I wanted from the synthetic a priori was not a doctrine but permission: permission to believe that the mind could be furnished from somewhere other than the dirt it stands on, that there was a source apart from the world that the world could not account for and that we were built to consult. When five turned out to have been hiding in two and three all along, that sense of grounding crumbled, and what crumbled with the synthetic a priori was the last respectable door I knew of through which anything might come down to us from above; and I was left with precisely the bleak indifference I had gone online at fifteen to escape. It was not an intellectual defeat. It was a moral and spiritual one, and I had no vocabulary for it, and no one to take it to.

So I retreated as a defeated army retreats. I carried it the way one walks with a pebble in the shoe: forgotten for whole seasons, then insistent the moment the mind wandered. I had long since concluded that even to adumbrate a solution was beyond me, and I stopped looking. I studied medicine, which has the great mercy of being wholly occupied with the concrete, and I read prodigiously and selectively - anything that would not remind me of the abyss, nothing that would, amusing myself until death.

There was one writer I could neither administer to myself in that spirit nor do without. The rest of my reading was opium, which takes the mind entire and leaves one waking from the coma having been nowhere; David Berlinski was novocaine, which takes only the pain, leaving other faculties intact. I could be walked to the very edge of the abyss, forced to look down, and find that I was still thinking, still delighted, still in possession of myself. Part of that was ornament, and I make no apology for saying so; his prose is a pleasure, and a man in a stupor will follow a sentence he relishes further than he will follow an argument he fears. But the greater part was a refusal I recognized as a lifeline: Berlinski never conceded that because Kant’s account had failed, the thing Kant was reaching for had failed with it. The demolition of an argument is not the disproof of what it was arguing toward; one does not throw out the baby with the bathwater. That the synthetic a priori could not be sustained in Kant’s form left entirely open whether there is something there of which Kant’s form was the wrong description - and Berlinski, book after book, kept the question open by main force. He gave me the one thing my defeat had taken: permission to suspect that there might be more to it.

Then, through a series of circumstances I have never been able to call accidental, I came to know him. And in the fullness of that unlikely acquaintance an answer of sorts arrived, providentially: Berlinski sent me the proofs of his forthcoming book, The Perpetual Rose, a meditation on mathematics, on naming, and on the ancient suspicion that the world comes inscribed.³

What follows is my record of that reading. It is not a review, and it is not a treatise. It is an account of where the book took one reader who arrived at its first page carrying the baggage I have just described. It will take you somewhere else. A feather does not fall; it negotiates. Released from the same hand, in the same still air, two of them will rise and sidle and hesitate their separate ways down and come to rest yards apart, each having obeyed the same law and arrived nowhere near the other. So it is with a book of this kind, which meets each reader where he stands, and nothing would please me more than for you to set these pages down, find it, and discover where it sets you.

In reading The Perpetual Rose, Berlinski adds the twist that keeps hope alive for the question. The reduction does not extinguish the possibility of the transcendent. At every step, someone must see that the symbols match, grasp the rule, and judge that it was applied correctly. Remove that reckoning mind, and the derivation is no more. Here we arrive at a dilemma: if arithmetic is synthetic, as Kant said, we owe an account of how the mind legislates for a world it has not visited. If arithmetic is analytic, as Frege, the formalists, and my teenage naïveté suspected, then mathematical truths are self-identities - but self-identities that become visible only when a mind performs and certifies them. Either the structure is mysteriously fitted to the world, or it is mysteriously fitted to us. There is a mystery here that migrates rather than disappears - already more than the universe of Richard Dawkins’ imagining, with its blind, pitiless indifference, would seem to permit.

We are between the two horns of the same dilemma: whether the mind supplies the missing content (Kant) or merely certifies an identity that was always empty (the formalists), the verdict is the same: the content was never sitting in the symbols. The marks on the page - the 2, the +, the 5 - are inert. What animates them, what makes a string of strokes mean anything, is the mind that reads them. A theorem is inert without someone to perform it; a word is inert without someone to mean it. So the puzzle about sums has become a puzzle about marks - and the barest mark of all is a name, which has no system behind it, no rules for its use, nothing at all between the letters and the thing but a mind that means them together.

Where the mystery migrates is the subject of this essay, and The Perpetual Rose supplies an elegant map. Naming, Berlinski observes, requires difference. To name something, you must first be able to pick it out: this thing, here, as opposed to that thing, there. Now imagine a universe of perfect symmetry - every point identical to every other point, a featureless fog in which no “here” differs from any “there.” In such a universe nothing could be picked out, so nothing could be noticed, and nothing could be named. Our universe escaped that fate because, in its earliest moments, the symmetry broke: the primordial sameness shattered into differences - this particle and not that one, matter here and not there - and out of those differences came galaxies, anxieties, platypuses, and everything else that can carry a name. Every entry in the world’s vast catalogue exists because symmetry broke, and everything you can point to is a shard of that ancient shattering. But one thing, I will argue, slipped through the metamorphosis untouched - one thing the shattering could not reach, because it had no features to break. It appears in no catalogue, answers to no name, and yet you are more intimately acquainted with it than with anything else in existence. It is the thing reading this sentence.

One caution before stepping into the thicket, offered as a confession rather than a caveat. Much of what follows proceeds by correspondence rather than by proof. I will set structures side by side, drawn from logic and from physics, from the grammar of our sentences and the syntax of our numbers, from esoteric and canonical scriptures, and from the fictions of strange minds, and report that they share a shape. I know what a mind hungry for meaning can manufacture; pareidolia does not confine itself to faces in the clouds, and the intellect is at least as susceptible as the eye. So where a resemblance in these pages seems to harden into an indisputable inference, judge it as a resemblance, and do not close the door behind it; the inferences, if there are inferences to be had, await a mind better endowed than mine. What I vouch for is only this. The shape kept appearing, unbidden, in places that have never heard of one another. It arrived too near the heart to be filed away as coincidence, and it was too beautiful to dismiss, in that immemorial sense in which beauty has sometimes been taken as evidence rather than decoration, the sense in which a physicist trusts an equation because it is elegant and a theologian argues from the splendor of the world to its author. Predisposed to meaning or not, I could not look away.

I. The Wars of the Roses

Berlinski’s meditation is presided over by an image he takes from Borges, who took it from the Arabic sources: a perpetual rose growing somewhere in Hindustan, upon whose petals is inscribed the profession of faith - there is no God but God. If such a rose exists, then language is not something we lay over the world; the world comes inscribed.

In Borges’s story “La busca de Averroes,” the rose is mentioned at a dinner party, and Averroes demurs. He is a strict Aristotelian, the medieval world’s great champion of reason, logic, and natural law, and the miraculous rose makes him profoundly uncomfortable. So he reaches for his tools and tries to dissect the miracle. An older legend, he argues, tells of a tree that grows green birds instead of leaves - and that story, he says, is easier to believe than the rose. Why? Because birds and leaves both belong to nature. A tree confusing one natural product for another is a small error of bookkeeping. But writing is different. Writing is artifice - a human invention, a product of minds. For letters to grow on petals, nature would have to do the one thing that only minds do. Averroes’s demurral is not aversion to miracles in general. It is rather a precise reflection of his worldview: the natural world does not write. And the whole of The Perpetual Rose can be read as a long judgment on that thesis - because if mathematics is genuinely inscribed in nature, then nature writes after all, and Averroes was wrong.

A second rose stands against the first. Umberto Eco closed The Name of the Rose with a medieval hexameter: stat rosa pristina nomine, nomina nuda tenemus⁴ - the rose of old abides in its name; we hold naked names. The flower, and all flowers, will be gone, and only the word survives. But if the name outlasts the flower - if the word endures while every petal disappears - then the line awards the abstract a durability the concrete never had, a Platonism in which the form survives its instances. Averroes, the Aristotelian, could not accept writing growing from nature, because for him the concrete particular comes first and human convention is a latecomer. Eco’s line, read Platonically, says the latecomer is the survivor - that when the rose and the name part ways, it is the name that walks out of the wreckage. The two roses are not merely a war between the realist and the nominalist. They are a dispute about which is more real, the rose or the word for the rose.

Between these two positions hovers Kant, who tells us of the man born wearing rose-colored spectacles, for whom the world arrives forever rose-tinted, and who can never remove the lenses to check what color the world keeps when no one is looking. If Kant is right, the dispute between the roses cannot be settled from inside the greenhouse. The inscription we find on the petals - the mathematics we find written into nature - may be on the world, or it may be on our lenses. The perpetual rose is perpetual precisely because the glasses do not come off.

We are taught to think of mathematics as a language we invented to describe the physical world, the way a map describes a territory: the territory comes first, the map comes after. But again and again, Berlinski’s examples run in the wrong direction. The mathematics arrives before the physics it supposedly describes, predicts things no one has observed, and constrains what the universe is even allowed to do. A map does not behave this way. A map does not tell the territory where its mountains must go. So a suspicion accumulates across the book, never quite stated outright: perhaps mathematics is not the map at all. Perhaps it is closer to the territory’s building code - not a description of the world, but a precondition of it; a fabric that is not written about nature but somehow written into it. Which is, of course, just the perpetual rose in other words.

If that suspicion is right, materialism is in trouble. The standard story says physical stuff comes first and everything else, mathematics included, is our commentary on it. Berlinski’s book argues the reverse, and argues it beautifully, but it conducts the entire argument in what a grammarian would call the third person. Everything discussed is an it: objects, equations, particles, structures, theories, the items on the world’s inventory, the furniture in the room.

Consider, though, what those arguments require in order to work at all. The formalist’s derivation runs only because someone follows it and grasps the rule as a rule, and an inscription no one can read is not an inscription but a blot of ink; in each case the argument presupposes a party who never appears among the things argued about - not in the way an inscription needs ink, which is a fact about how it got onto the page, but as a condition without which the thing under discussion would not be the thing under discussion at all. The reader is not one more item on the inventory, but rather what makes it an inventory.

And here the book does something stranger still. From its first chapter to its last, it interrogates the rose - whether the writing is really there, whether nature can hold an inscription, whether the script sits on the petals or on our lenses - and out of that patient interrogation a fourth thing began, uncannily, to gather at the edge of my reading like an ectoplasm, nowhere argued for and nowhere named, faint at first and then insistent, until it seemed to me that the book had been circling it all along. Before the dispute over the rose can begin, something must already be standing in the garden, bent over the petals; and it is this presence, summoned on every page and never once addressed, that I mean to ask after - what it is, and why, in all gardens, there may be only one of it. I will call it the witness, and I will not be able to tell you much about it, for reasons that will turn out to be the most numinous thing there is to say.

II. The Akashic Catalogue

The Qur’an records that God taught Adam all the names and then challenged the angels to produce them, which they could not. It is a scene about the human vocation, since to name is to sort the world and to hold on to it; but it is also a scene that implicitly assumes its own metaphysics, for the names are older than the man, and Adam does not invent them so much as receive them. Naming, in this telling, precedes the namer: the catalogue was drawn up before any mind grasped a single entry, the categories are real and waiting, and our work is merely to learn them.

One is entitled to contest the fable. It is not obvious that “dog” exists at all, since dogness looks less like a form waiting in heaven than like a region of similarity our minds slice through and our societies agree to call “dog”: the boundary where wolf ends and dog begins is ours rather than nature’s. Yet some names refuse to move under that pressure. Gold, water, and electron seem to reach an essence and to hold it fast, while other kinds - natural, social, and cultural alike - rest on essences shaky enough to be disputed, and their names loosen accordingly.

Whichever reading one’s spirit jibes with, there is solace in finding others who have stood where you stand. Naipaul, traveling in Uganda, was appalled to discover that the valley had no name⁵ - nothing there had a name, he complained, because they do not name things- - and feared that the unnamed world might dissolve into fog. The Babylonians had said as much four thousand years earlier, opening the Enūma Eliš on a heaven above and an earth beneath that do not yet exist because they do not yet bear names: existence itself waits on the baptism.⁶

In ancient Egypt the inverse occurs. As the soul readies itself for the last journey, the Book of the Dead arms it with a remarkable incantation for the next world: “I am he who is not known, the rosy-faced beings are with me. I am the unveiled.⁷” Where Naipaul dreaded the unnamed world, the Egyptian soul yearned to become the unnamed thing, since to pass beyond death, in this telling, is to slip out of classification altogether - to become what no register holds and no mortal taxonomy comprehends, known to itself precisely because it is unknown to everything else. Whether that essence then stands alone or merges into its creator, the incantation does not say; it insists only that the soul is unveiled, in the sense the Sufis of a much later age would give to kashf. And the company it keeps beyond the veil is rosy-faced. Even here, at the far edge of the nameable, the rose attends.

From these cosmic registers one descends, without changing the question, to a lecture hall in New Jersey, where a bearded man with kind eyes, Saul Kripke, speaks without notes and hammers proper names to their objects like nails. He is one of the presiding spirits of The Perpetual Rose. Berlinski follows the rigid designators through chapter after chapter, and I will need them again before the end.

There are two ways a name might work. On the first view, it is shorthand for a bundle of descriptions, so that “Aristotle” means the teacher of Alexander, the author of the Metaphysics, the man from Stagira. On Kripke’s view, it is shorthand for nothing at all: at some point something was dubbed Aristotle, and from that moment the name picks out that very individual, whatever we go on to imagine about him. A friend remains the same friend whether he loses an arm or loses his job, and we have no difficulty placing him in a parallel world with turquoise eyes, breathing through gills, and saying, “That is John, with gills and turquoise eyes”; the sentence is intelligible only because it is John we are imagining so transfigured. The name carried him into the surreal scenario and welcomed him there without consulting his features, which he wears as a man wears clothes; the name grips the wearer and parades him through worlds in which nothing else about him survives.

Medieval philosophy had a word for what the name grips: - haecceity⁸ - “thisness,” the property of being this very individual and no other, irreducible to any list of qualities. With that in hand, we can reconstruct the hierarchy passed to Adam, since not all names dig equally deep. Some - gold, water, electron - reach a hidden structure and hold in every possible world; others - wife, citizen, doctor - organize human conventions and loosen their grip as institutions drift; others still - chivalry, doubt, courage - are thick with judgment and dispute.

However we arrange the catalogue, and whatever thisness we fasten each designator to, one entry will still be missing. It is not missing by oversight. It is missing by necessity.

III. The Disinterested Witness

“Missing by necessity” might strike you as a strange claim, since cataloguing is precisely the sort of work that can always be extended - open more closets, inspect more drawers, hire more clerks - and what I am claiming is not that the catalogue is unfinished. There is one entry that will always elude the clerks because it cannot be catalogued at all, and this is not a mystic’s intuition but one of the great discoveries of the twentieth century, proved several times over in cold logic before anyone thought to look for its counterpart in the mind.

The first formal step belongs to Kurt Gödel, who walks through Berlinski’s pages more than once, at one point defending the perception of mathematical objects by an organ no one has ever located. In 1931, Gödel proved that any consistent system of formal rules rich enough to do arithmetic must contain true statements it cannot prove, and that among the things such a system can never prove is its own consistency. However vast the system, it cannot stand outside itself to vouch for its own coherence; the catalogue cannot certify the catalogue. To see that the rules are sound, you must step out of the rules, for soundness is visible only from a vantage the system itself does not contain.

The same constraint reappears in language. Alfred Tarski asked what would happen to a formal language equipped with its own truth-detector, a predicate inside the language that sorts every one of its sentences into the true and the false, and proved that no such language can define truth for its own sentences without collapsing into the contradiction of “this sentence is false.” Truth for the first language must therefore be defined in a second that talks about it, truth for the second in a third language, and so on forever, the regress halting in practice only because ordinary English eventually steps in and we agree not to ask further. Berlinski states the burden cleanly in his pages on model theory: were a language to interpret itself, it would quickly collapse into inconsistency. The certifying is always done from outside, and whatever does the certifying is not, at that moment, among the things certified.

In both cases, failure occurs at the same instant - the instant a classifier is turned upon itself. The catalogue keeps perfect order so long as the principle doing the cataloguing stays off its own pages. Ask the sorter to sort itself, and logic buckles.

Here I must return to the confession I made at the gate. I am about to set the mind beside these theorems and say that they rhyme, which is not a proof - but rigor often arrives second. Einstein’s relativity began in an armchair, with the thought that a man falling freely does not feel his own weight⁹, and the geometry that made it precise had been built by Riemann more than half a century earlier for no physical purpose and was waiting on a shelf; Darwin published his theory of natural selection with no mechanism of inheritance and was right anyway. A discipline that admits only conclusions never gets unstuck, because whatever unsticks it is, at the moment it arrives, unlicensed.

So let me be exact about what these results do and do not license. Gödel and Tarski proved theorems about formal systems and nothing about minds, and what they establish is a form: whatever certifies cannot, at the moment of certifying, stand among the things certified. That the mind exhibits the same form is not a corollary of any theorem but must be argued on its own ground, and the argument, when it comes, will come from Kant and from a tradition far older than Kant. The theorems play a humbler role, showing that the shape is real, no mystic’s mood but a structural constraint, provable wherever symbols grow rich enough to point at themselves. When the same shape surfaces in the mind’s attempt to catch itself in the act, I report the recurrence; I do not deduce from it.

And within their own territory the theorems speak plainly enough. The mother of all these paradoxes lies in the act of symbolism itself, for the moment you invent a system in which Thing A can stand for Thing B - a word for an object, a thought for a reality - you have opened a door that cannot be closed: sooner or later the system grows rich enough to turn and stand for itself, and that is where the buckling occurs. Mathematics is only a highly rigorous, tightly controlled system of symbols, and for centuries its practitioners believed it immune to the messy paradoxes of ordinary speech, until Gödel showed that a system able to do no more than add and multiply is already able to point at itself, and once it does, Hofstadter’s strange loop arises.

The argument from the mind, which I promised would have to be made on its own ground, begins with Kant, who had already claimed that the mind is no passive bucket receiving impressions but an active structurer of what it receives, and who saw that experience could make no sense at all unless a single unified consciousness were tying it together. He called this the Transcendental Unity of Apperception, and compressed it into a sentence: the “I think” must be able to accompany all my representations. Seeing an apple, feeling the wind, and hearing a dog bark are not three events floating separately in a void; there is an underlying “I” that is simultaneously seeing, feeling, and hearing, and that unified “I” is the ultimate Subject.

Kant then ran into the paradox we have been circling: if the “I” is what does the observing, how is the “I” to observe itself? His solution was to split human identity in two. The first half is the Empirical Ego, the Object - the “you” that can be observed, so that when you introspect on your personality, your memories, your mood, or your body, what you are looking at is a psychological object floating in the space of your mind. The second is the Transcendental Ego, the Subject: the pure, invisible, unified consciousness doing the looking, the condition required for experience to exist at all and yet entirely empty of content. You can never experience or know that second ego, Kant concluded, because the moment you turn your inner gaze backward to catch the “I” in the act of being, you have converted it into an object - into the Empirical Ego - and the true "I" has slipped out the back door, remaining what it always was, the invisible Subject doing the looking. It is the eye that sees everything except itself.

More than two millennia before Kant, the Upaniṣads had compressed all of it into one sentence: you cannot see the seer of seeing; you cannot know the knower of knowing¹⁰. On that sentence the Advaita Vedānta tradition built its doctrine of the sākṣin, the witness-consciousness, a doctrine that sits comfortably beside Kant and the phenomenologists, circling the same vacancy (a comparison Bina Gupta has drawn in scholarly detail¹¹). The sākṣin is consciousness as pure witness, self-luminous in that it needs no second light to be seen by, the way a lamp needs no other lamp, and disinterested in that it does not participate in what it illuminates, any more than the lamp takes sides among the books it lights. Every mental state parades before it, but it joins no parade. Here, then, is the missing entry named at last, by a tradition that believed it was missing by necessity: the witness is to the mind what the metalanguage is to the language and what soundness is to the system, always presupposed, never contained. Any feature I could name as mine is something I can turn and look at - my temperament, my body, my history, the particular grain of my attention are all objects of an inward gaze, and the gaze is not among them. Whatever survives that subtraction has no properties left, not because it is hidden or subtle, but because having a property is what it means to be the sort of thing that can be looked at. The witness is not merely undescribed. It is the residue of every description.

The witness, then, is featureless, and therefore, I want to argue, cannot come in copies: there is not one behind each pair of eyes but one altogether. What I can deliver is not oneness in the way a proof delivers a theorem, but something less ambitious - the failure of twoness, the finding that there is nowhere for a second one to differ. Whether that is unity, or a place where counting simply stops applying, and whether those two conditions are not in the end the same condition, I am not able to say.

The argument begins where Descartes began, with what cannot be doubted. Something is reading this sentence. Whatever else falls away, that does not, since the doubting itself requires a doubter; and so there is at least one. Descartes went on from here to a thinking thing, a substance with a nature, whereas my claim is smaller - that there is a reader, and nothing yet about what sort of thing it is.

Now try to make it two. Counting requires difference, since to say there are two is to say that this one is not that one, and saying so requires something to point at; take away every feature and there is nothing left to point at. The problem is not that the difference is hard to find but that there is nowhere in a featureless thing for a difference to sit. You will object that the difference is obvious - one witness looking out of your life and another out of mine - but consider what you are reaching for when you say so: a body, a history, a position in the world - contents, every one of them, and it was by setting the whole inventory aside that we arrived at the witness in the first place. The objection establishes two lives, which no one disputes, and leaves the witness exactly where it was. At least one, then, and never more than one - not one item selected from a set of similar items, the way a coin is one coin, but the only kind of one that can be said of a thing with nothing to distinguish it.

I like to think that we, marinated as we are in Western thought, can adapt to this hypothesis, and that the precedent for adopting it is already ours. Leibniz made one half of it a principle, the identity of indiscernibles: if two things share every feature, they are not two things but one thing counted twice. Arnold Zuboff drew out the corollary for modern ears¹², observing that everything that distinguishes me is objective and can be handed to a stranger on a sheet of paper - height, history, profession, the particular cast of my interests - while the bare fact that there is someone home, the lit interior as opposed to its furnishings, carries no features at all. So my sense of being is not similar to yours, the way one apple’s taste resembles another’s; if the argument holds, it is the same - the same in you, reading these words, the same in Averroes bent over De Anima in Córdoba, the same in the scribe who cut the Enūma Eliš into clay, the same behind every pair of eyes that has ever opened on the world.

And so we arrive at a staircase: a system cannot certify itself, and something outside must do the certifying; that something, traced through the logicians, through Kant, through the Upaniṣads, is the witness - formless, self-luminous, never among its own contents, and, having no features, incapable of coming in copies. Everything nameable exists because of broken symmetries, and every thisness in the catalogue is a shard of that ancient shattering. The witness offered the shattering nothing to break.

Whatever the shattering could not break, the smaller shatterings should not break either. Consider, then, what happens to a self left alone with time. I walk into my grandfather’s library at six and the room is cavernous, the shelves receding upward past any hope of reaching them; I walk into the same room at thirty and find a pleasant study holding perhaps a few hundred books, though nothing in it has moved. What I called love at six and what I call love now share a word and almost nothing besides. Not one molecule of the boy remains in the man, nor do any of his convictions or sensory convictions remain unamended - all of what made the boy was shattered in the making of the man, and I have never been able to find whatever essence is supposed to run through the two of us. A casuist will say memory: I am the one who remembers being him. But memory is a content, an item in the inventory, a possession rather than a possessor, and it is always of what has already passed; so if the self is memory, the self is entirely a thing of the past, and the present moment contains no self at all. And yet the present is the only place where anything is ever witnessed. Something is here, now, watching these words assemble themselves into this sentence, and it is not made of recollection. The self is behind us. The witnessing is not.

Physics has a name for what survives a transformation intact. A system has a symmetry if you can shift it, rotate it, or move it through time and the underlying laws do not care, so that it behaves the same way before and after; and a century ago, Emmy Noether proved that to every such symmetry there corresponds a conserved quantity, something that persists unchanged while everything around it transforms¹³. Berlinski gives the theorem a chapter of its own, working the derivation out in full until angular momentum falls out of the algebra, and then pausing over what he has just done: a purely mathematical statement, he observes, has overflowed its boundaries and begun leaking into the real world. Berlinski has already noticed that the result will not stay in its container, and having watched it leak once, I find the shape hard to unsee: thoughts shifting, moods rising and falling, the library shrinking, the boy becoming the man, the years going, and through the whole relentless transformation something remaining that the transformation never touches. I offer it as what it is, not a deduction but a correspondence, one more appearance of the shape, this time wearing the uniform of mathematical physics. And if it is telling the truth, then being someone is not a struck match flaring briefly in an infinite dark. It is the standing condition of things.

IV. The Perpetual Rose

So what kind of book, in the end, is The Perpetual Rose? Berlinski is usually described as a skeptic, and the description is accurate but insufficient, since the book’s real method is rarer than skepticism. Each chapter takes a mystery the modern world considers settled, or at least filed away - how mathematics fits the world, what a name fastens onto, whether anything is really conserved, whether anyone is home in the machine - reopens the file, polishes the mystery until it gleams, and then, with the discipline of a man who refuses to sell what he does not own, declines to close it. He walks you to the lip of the abyss, forces you to look down, and then, against all expectation, returns you to the edge with your footing and your suspicion both intact. Even the ending refuses to end, since the book signs off in a Talmudic parable¹⁴, truth wrested from heaven and handed down to the disputants, with the argument left to continue without end. So the book reveals more than it resolves, and the title confesses as much: the rose is perpetual because it is left blooming.

Its every chapter circles a mystery it never names. An inscription presupposes writing, writing presupposes language, and language is the water we swim in so completely that we forget we are wet. Berlinski relays a suggestion Chomsky once made to him, that language and mathematics may share a common origin in the successor function, the same S whose patient migrations across a plus sign we watched earlier reduce arithmetic to identity; and if that is so - Berlinski is not convinced that it is - the lexicon and the number line are branches of a single root, and what holds of one should hold of the other. On Chomsky’s account, beneath the staggering diversity of human tongues, thousands of them, mutually unintelligible and superficially alien to one another, there runs a single hidden architecture, a Universal Grammar built into the biological endowment of every child before the child has heard a single word¹⁵, silently governing which sentences every speaker can and cannot produce. The words are many, but the Word is one.

Beneath thousands of mutually unintelligible languages, a hidden grammar; beneath billions of mutually unacquainted persons, a featureless witness - in both cases the surface multiplies and the depth does not. And the rose gathers the two into a single image, since if the rose is inscribed, it is inscribed in a language, and the only language we know from the inside is, on that account, an invisible unity wearing ten thousand visible costumes, read by an invisible reader who, if the argument here holds, does not come in copies either. The inscription, the language, and the reader: three mysteries, and possibly one.

Or possibly a delusion. I promised at the outset to return to this danger by name, and I am aware of the company I risk keeping. In Umberto Eco’s Foucault’s Pendulum, three editors begin, half as a game, to weave the world’s loose threads into a single secret Plan - the Templars connect to the Kabbalah, the Kabbalah to the pyramids, the pyramids to everything - until the connections seem to glow with necessity and the Plan, which they invented, devours them. I have spent these pages finding roses in the Egyptian Book of the Dead, the soul of the Upaniṣads in a theorem of Emmy Noether’s, and a single Word beneath ten thousand mutually unintelligible tongues: reveries, perhaps, of a delusional man, Casaubon at his desk, mistaking the pattern-hunger of his own tormented mind for the architecture of the world. Or perhaps not. Perhaps these correspondences keep surfacing in scriptures and proofs and equations that have never met because they are all impressions of the same seal, the grain of the fabric showing through wherever the weave is pulled tight. I do not know which, and I have come to suspect that no one standing inside the greenhouse can know which - that is what the rose-colored spectacles were always trying to tell us.

So I will end the way Berlinski ends, by declining to close. I can only report what reading The Perpetual Rose was like from the inside: I would set it down after a chapter and discover that its questions had followed me out of the room. Whether the writing is on the world or on our lenses, I could not say before I opened the book, and I cannot say now that I have closed it; on that question Berlinski left me exactly as free, and exactly as haunted, as he found me. But something did shift in the reading. Somewhere between the rose with the writing on its petals and the rose that survives only as a name, I stopped asking what the inscription says and began asking who has been reading it - and that question, once it arrives, does not leave, because every answer offered is witnessed by the very thing the answer fails to name. The rose, I learned from Berlinski, is perpetual. The reader, I have come to suspect, is too.

Disclaimer: AI was used for grammar, syntax, reference cross-checking, and collaborative editing; the ideas, analysis, and conclusions are my own.