The Middle Is Not Compromise

There is a prejudice so deep in modern thought that most people don’t even know they carry it: the idea that the middle is weak. That to stand between two extremes is to have failed to choose. That yes and no are positions of conviction, and anything between them is cowardice dressed up as nuance.

I want to argue the exact opposite. The middle is not compromise. It is the most powerful position there is. And I want to argue it from two directions — one ancient, one mathematical — because I think they say the same thing, from different ends of time.

The Third God

The Hindu Trinity — Brahma, Vishnu, Shiva — is often translated for modern audiences as Creator, Preserver, Destroyer. The first two names are straightforward enough. But “Destroyer” is a mistranslation that has done enormous damage to how people understand Shiva.

Shiva does not destroy as an act of malice or termination. Shiva transforms. The Sanskrit root is closer to dissolution — the returning of form to formlessness, so that new form can arise. It is not an ending; it is a clearing. The forest fire that looks like catastrophe but is, in fact, the condition for new growth. Shiva is the third because without the third, creation and preservation calcify into stagnation. The two poles need the third to remain alive.

Look at the Trishul — Shiva’s trident. Three prongs. Not two. The central prong does not cancel the outer ones. It gives them meaning. Without the center, you have two isolated extremes with nothing between them, no connection, no resolution, no path from one to the other. The center is the spine.

I grew up in Nashik, a city on the Godavari. Every twelve years, Nashik hosts one part of the Kumbh Mela — millions of people, one river, one reason. The river itself is Triveni in some traditions: three waters meeting. Ganga, Yamuna, and Saraswati — the third river is said to be invisible, underground, the hidden one. The one you cannot see is the one that holds the other two together.

Three appears everywhere in Vedic thought not because anyone decided it was a holy number, but because reality keeps arriving in threes. Day, night, and twilight. Creation, preservation, dissolution. Tamas, Rajas, Sattva. Waking, dreaming, deep sleep. Satya, Chitta, Ananda. The middle term is never the weak one. It is usually the most subtle and therefore the most essential.

The Sunyata

Zero is not nothing.

I want to stay with this for a moment because the confusion here runs very deep. When Brahmagupta formalized the arithmetic of zero in the 7th century, he was not inventing a placeholder for “the absence of something.” He was recognizing that zero is a number in its own right — a quantity, a position, a starting point from which all other quantities are measured. You can add zero. You can multiply by zero. You can be zero. And in my musings, I always tend to keep “dividing” by zero, to arrive at that mathematically impossible singularity… ;)

The ancient Buddhist concept of Sunyata — often translated as “emptiness” — carries a similar misunderstanding when it reaches modern ears. Sunyata does not mean the world is empty in the sense of hollow or meaningless. It means that phenomena do not carry inherent, fixed, independent existence. Everything arises in relation to everything else. The “empty” cup is not without value — it is the emptiness that makes it a cup. The empty space inside a wheel’s hub is what allows the axle to turn.

Zero, in balanced ternary arithmetic, works exactly this way. It is not the absence of a trit. It is a trit. It encodes information — the information that this position carries neither +1 nor −1. That is a specific, definite, measurable fact. When a balanced ternary circuit outputs zero, it is not failing to output something. It is actively transmitting the middle state, and the receiver knows exactly what to do with it.

In binary, zero is absence — the light is off, the voltage is low, the switch is open. The zero channel is silent. In balanced ternary, zero is a message. The zero channel is speaking. This is a small distinction that changes almost everything about how information flows through a system.

The Mathematics of Three

Now let me come at this from the other end — cold, formal, numerical.

In information theory, there is a quantity called radix economy. It measures how efficiently a number base encodes a given number: you take the base, multiply it by the number of digits required to represent your value, and the base with the smallest product wins. If you let this quantity run continuously — if you ask, for every real number r, how economical is base r — you get a function with a single, sharp minimum.

The minimum is at e. Euler’s number. 2.71828…

Not 2. Not 3. Not 10. e. The same constant that appears in compound interest, population growth, radioactive decay, the shape of a hanging chain, the distribution of prime numbers. The base of the natural logarithm. The most natural constant mathematics knows.

e is not an integer. You cannot build a physical logic gate in base e. (I am trying my best to do so, though.) So the question becomes: which integer is closest to e?

Three.
Not two. Three.

Binary — base two — is on the wrong side of optimal. It is less efficient than e by a small but real and consistent margin. Ternary — base three — is on the other side of optimal by an even smaller margin. Between the two, ternary is fractionally closer to e, and it is the only integer radix that is.

This is not an approximation or an engineering tradeoff. It is the mathematics of information itself saying: if you must choose an integer, choose three. The universe, if it were designing a number system, would choose ternary.

I find this extraordinary. The two traditions — one arriving from ancient Sanskrit cosmology, the other from 20th-century information theory — point to the same answer without knowing each other. Three is optimal. The middle state is not a compromise between the extremes; it is what makes the extremes coherent. It is the reason the system works at all.

The Rest in Music

There is one more way I think about zero — from music.

The oldest music tradition I know of is in the Sama Veda — the Veda of melody. The Rig Veda contains the knowledge; the Sama Veda sings it. But what makes the Sama Veda remarkable, from a technical standpoint, is how it encodes pitch. The chanted syllables of the Sama Veda are marked with three tonal accents: Udatta (raised), Anudatta (lowered), and Svarita (the combined, falling tone that carries the memory of both). Three. Not two. Three, thousands of years before information theory, three, before anyone knew what a trit was. The Svarita is the middle tone — not a compromise between the raised and the lowered, but a distinct phoneme that carries its own semantic weight. Remove the Svarita from Vedic recitation and the chant is wrong. Not slightly wrong. Wrong in a way that changes what is being said. The middle tone is load-bearing.

In any composition, there are notes and there are rests. The rest is not a failure to play a note. The rest is a note of silence. It has duration, position, and meaning. Remove the rests from a Raga and you don’t have more music; you have noise. The silences are what allow the notes to be heard as distinct. The pause before the sam is what makes the sam land.

A Raga in Kharaja — the deep lower octave — uses silences that a student doesn’t understand for years. Why hold there? Why not fill that space with a note? Because the held space is doing the work. The listener’s ear is completing what the musician has left open. The silence is the invitation.

Zero in balanced ternary is that rest. Not silence as absence, but silence as active participation. The trit 0 in a sequence +1, 0, −1, 0, +1 is as necessary as any other member of the sequence. It is doing something: it is holding the phase, maintaining the transmission line, passing through the inflection point that is mathematically required whenever a signed quantity transitions from positive to negative. You cannot go from +1 to −1 without passing through zero. Zero is not a gap in the sequence; it is the hinge.

AC electricity — the kind that powers every home, the kind I use in my device — obeys the same rule. The sinusoid passes through zero every half-cycle. That zero-crossing is not a failure of the wave. It is the mechanism by which the wave changes direction. Every motor, every transformer, every transmission line on earth depends on that zero.

My device inherits this. Trits are not voltage levels. Trits are current polarities driven by an AC signal. +1 is positive phase, −1 is negative phase, and 0 is the photon absent, the signal quiet, the channel at rest. The three states are not a hierarchy with zero as the weakest. They are a cycle, and the cycle depends on all three.

Why I Build in Ternary

I am sometimes asked why I chose ternary. The question usually implies that binary is the natural choice — the historical standard, the established industry, the thing that works — and that ternary is therefore a detour requiring justification.

I think the question has it backwards.

Binary is the choice made under constraint. Early engineers needed a logic that was robust against noise, cheap to build, and easy to reason about. Two states — on and off — satisfied those requirements admirably. Binary is not wrong; it is a very good solution to a very specific set of early constraints. But it carries costs that have been accumulating for seventy years: the overhead of two’s complement arithmetic to handle signed numbers, the asymmetry between positive and negative, the need for increasingly elaborate countermeasures in cryptographic hardware because power consumption correlates with bit transitions in ways that leak secrets, the sheer size of the instruction encodings required to express operations that ternary handles more compactly.

Balanced ternary doesn’t add states. It removes distortions. The signed number is native — you don’t need a sign bit or a convention about two’s complement because every trit is already signed. The negative of a number is its trit-by-trit negation; no special cases, no integer overflow, no minimum-negative-number that has no positive counterpart. Multiplication is symmetric in a way that binary never achieves without tricks.

And the power symmetry — I find this as the most beautiful, most elegant part. When my device outputs +1, it draws the same current as when it outputs −1. The magnitudes are equal by physics, not by engineering effort. The MWCNT modulates the SWCNT conductance identically in both directions because the AC terminal voltage is symmetric and the quantum transport is symmetric. The Hamiltonian commutes with time-reversal. This means that a statistical attack on power consumption — the class of attacks that broke smart cards and AES implementations and SSL accelerators all through the 2000s and 2010s — cannot work, not because I added shielding or randomized the timing, but because there is nothing to measure. The device tells you nothing. The physics withholds the information before any engineering decision is required.

I did not design that symmetry in. It was already there, waiting to be recognized.

The Middle Path

The Buddha’s Middle Way — Majjhimā Paṭipadā — is one of the most misunderstood teachings in Asian philosophy. It is usually explained as a moderate path between extreme asceticism and extreme indulgence. This is true, but it is the small version of the idea.

The larger version is that the Middle Way is not a position on a spectrum between two extremes. It is a different axis entirely. When you stand in the middle, you are not equidistant from both poles; you are orthogonal to them. You are no longer playing the game that the two extremes are playing. You have stepped off the line.

Three-valued logic does something similar to two-valued logic. It is not binary logic with an extra state added for comfort. It is a different system in which the questions themselves change. In binary, every proposition is true or false. In balanced ternary, a proposition can be true, false, or unknown — and unknown is not a failure to decide; it is a precise characterization of the epistemic state. Quantum mechanics uses exactly this: a particle in superposition is not secretly one thing that we haven’t measured yet. It is genuinely, physically, in a state that binary logic has no word for.

The middle state is not weakness. It is precision. It is the willingness to say: this is the thing that is neither +1 nor −1, and that fact is information, and that information matters.

Closing

I am an engineer. I build things. The reason I think about Shiva’s trident and Brahmagupta’s zero and the rests in a Raga is not that I think mythology is physics. I think they are different languages for the same structural insight that keeps appearing wherever humans pay careful attention to how the world is organized.

The world is organized in threes more often than twos. The information-theoretically optimal radix is 2.718. The most natural constant in mathematics is e. The nearest integer to e is 3. The Vedic trinity is three. The Trishul has three prongs. The Triveni is three rivers. The three gunas. The three states of consciousness. The Sama Veda’s three tonal accents. The three-valued truth of quantum superposition.

I do not think this is coincidence. I think it is the signature of a deep structural truth that different traditions have found by different routes.