One sweltering night in the year 628 CE, in the Indian city of Bhillamala, a mathematician-astronomer wrote down a rule that would, centuries later, upend the economies of three continents and make every computer on the planet possible.
🔥 In 628 CE, the Indian mathematician Brahmagupta completed his astronomical treatise Brahmasphutasiddhanta—a compendium of rules for calculating planetary motion, eclipses, and calendar cycles. In Chapter 18, dedicated to the Kuttaka method (solving linear Diophantine equations), he recorded what no mathematician before him had dared to formalize: arithmetic rules for shunya—zero. "A positive minus zero is positive. Zero minus zero is zero. The product of zero and anything is zero." These statements seem trivial to a 21st-century schoolchild, but in the 7th century, they were an act of intellectual audacity: declaring "nothing" a full-fledged number, capable of participating in calculations on equal footing with ones and tens.
⚡ Before Brahmagupta, Indian mathematicians used zero as a positional symbol—an empty placeholder in number notation to distinguish 5 from 50 or 500. A manuscript from Bakhshali (presumably 4th–5th century CE) contains dots denoting the absence of a digit but provides no rules for working with them. Brahmagupta went further: he didn’t just recognize zero as a digit; he gave it an arithmetic life. And he made a mistake that still unsettles mathematicians: he attempted to divide a number by zero, calling the result "khacheda" (a fraction with a zero denominator), and claimed this number remained unchanged. Later mathematicians would prove that division by zero cannot be defined without collapsing the entire arithmetic system, but the very attempt reveals the scale of his ambition—to construct a complete, closed algebra with no forbidden zones.
🌍 Brahmagupta’s treatise remained in Indian astronomical schools for nearly two centuries until, in the 9th century, Arab scholars began systematically importing Indian mathematics. Al-Khwarizmi (c. 780–850 CE), working in Baghdad at the court of Caliph al-Ma'mun, wrote the treatise On the Calculation with Hindu Numerals, where he presented the decimal positional system with zero—"as-sifr" (from the Sanskrit "shunya" via the Arabic "empty"). His contemporary Al-Kindi (801–873 CE) in four volumes of On the Use of Indian Numerals meticulously dissected arithmetic operations, including Brahmagupta’s rules for zero. Arab mathematicians didn’t just copy the Indian system—they reworked it, integrating it into algebra and geometry, and built a bridge between Indian abstraction and Mediterranean practice.
📜 Three hundred years later, these texts seeped into Europe via three routes: through Moorish Spain, Crusaders in the Levant, and traders in Sicily. In the 12th–13th centuries, European translators—Adelard of Bath, Gerard of Cremona, Robert of Chester—translated Arabic treatises into Latin, and for the first time, the word "zephirum" (a Latinized "sifr") appeared in European manuscripts. The Italian mathematician Leonardo Fibonacci in 1202 CE published Liber Abaci (The Book of Calculation), where he systematized Indo-Arabic arithmetic for European merchants. He demonstrated how zero simplified accounting for goods, debts, and interest—tasks for which Roman numerals were hopelessly cumbersome.
🛡️ But Europe resisted. Zero brought not only convenience but also threat: the positional system made forgery trivial—just add a 0 to a number to turn 7 into 70 or 700. In 1299 CE, the city council of Florence issued a decree banning the use of "Arabic numerals" (including zero) in accounting books under penalty of fine. Bankers and notaries were required to keep records in Roman numerals or words. Meanwhile, theologians debated the metaphysics: could "nothing" be a number if God created the world "ex nihilo"—from nothing? Did the mathematical zero contradict the idea of divine creation? Scholastics, including disciples of Thomas Aquinas (1225–1274 CE), engaged in debates whose echoes reverberate in treatises well into the 14th century.
💰 The turning point came in 1494 CE, when the Franciscan monk and mathematician Luca Pacioli published Summa de Arithmetica, Geometria, Proportioni et Proportionalità—an encyclopedic treatise where he first systematized the double-entry bookkeeping method (debit-credit). The system only worked with positional numbers: zero became the key to balance, where the sum of assets minus the sum of liabilities must equal zero. Venetian merchants, who dominated Mediterranean trade, instantly adopted the innovation—it allowed them to track complex chains of transactions, debts, and capital with a precision unattainable with Roman numerals. Within 50 years, all of commercial Europe had switched to Arabic numerals, and the Florentine ban became an anachronism.
🔢 Meanwhile, algebra reached a new level. Gerolamo Cardano (1501–1576 CE) in his treatise Ars Magna (1545 CE) used zero as a reference point for positive and negative numbers, solving cubic equations. His rival Niccolò Tartaglia (1499–1557 CE) applied zero in formulas for roots, and their students engaged in fierce debates over whether an equation’s root could be zero. These discussions laid the foundation for analytic geometry: when René Descartes introduced coordinate axes in 1637 CE, zero became the center of the mathematical universe—the point (0, 0), relative to which all others are defined. Without Brahmagupta’s rules, this construction would have been impossible.
⚙️ In the 17th century, zero became critical to calculus. Isaac Newton (1643–1727 CE) and Gottfried Leibniz (1646–1716 CE), independently developing differential and integral calculus, operated with the concept of an "infinitesimal"—a number approaching zero but not equal to it. The limit of a function as x → 0 became the cornerstone of analysis. Without a clear understanding of zero as a number (rather than mere absence), this concept would have collapsed into logical paradoxes, like those that plagued Greek philosophers when discussing Zeno’s paradoxes.
📐 By the 18th century, zero had become an engineering tool. John Napier (1550–1617 CE) invented logarithms, where log(1) = 0—the anchor point for all calculations. Logarithmic tables, printed in millions of copies, allowed navigators, artillerymen, and engineers to perform complex multiplications and divisions by reducing them to addition and subtraction. Every table began with zero. In 1854 CE, George Boole (1815–1864 CE) published The Laws of Thought, where he created an algebra of logic operating with 0 (false) and 1 (true). It was an abstraction, but 80 years later, it became literal circuit design.
💻 In 1936 CE, Alan Turing (1912–1954 CE) described an abstract "Turing machine"—a theoretical computer operating with a tape of symbols, including 0 and 1. In 1945 CE, John von Neumann (1903–1957 CE) designed the stored-program architecture (EDVAC), where instructions and data were encoded in binary numbers. Every bit is a 0 or 1, every byte a sequence of eight zeros and ones. Without the concept of zero as an independent value—not just the "absence of one"—the binary system would lack the symmetry necessary for arithmetic operations in a processor.
🌐 The first computers—ENIAC (1945 CE), UNIVAC (1951 CE), IBM 701 (1952 CE)—performed millions of operations with zeros: comparison, addition, bit shifting. Every program implicitly uses Brahmagupta’s rules: x + 0 = x, x * 0 = 0, x - x = 0. In 1969 CE, the designers of the C language (Dennis Ritchie and Ken Thompson) made zero a string terminator (\0), an array end marker, and the default value for uninitialized variables. This choice was inherited by C++, Java, Python—billions of lines of code contain zero as a syntactic and semantic unit.
📌 Today, in 2026, zero remains the foundation of digital civilization. Every transaction in the Bitcoin blockchain (invented in 2008 CE) operates with 256-bit hashes—sequences of zeros and ones. Quantum computers (like IBM Quantum System One, launched in 2019 CE) use qubits, which exist in a superposition of states |0⟩ and |1⟩—quantum analogs of classical zero and one. Without a strict definition of base states, quantum logic would collapse.
📌 In machine learning, zero plays a critical role: neural networks initialize weights with random values near zero, and the activation function ReLU (Rectified Linear Unit, which became standard in 2012 CE) clips negative values to zero. Models like GPT-4 (2023 CE) or Claude (2024 CE) process trillions of parameters, many of which, after training, are close to zero (network pruning). Researchers from MIT and Stanford in 2025 CE published a paper on "sparsifying" neural networks, where up to 90% of weights can be zeroed out without losing accuracy—a direct application of the idea of "nothing as a value."
📌 The problem of division by zero, which Brahmagupta attempted to solve in 628 CE, has returned in a new guise: in programming, division by zero throws an exception, halting program execution. Engineers at IEEE in 1985 CE standardized IEEE 754—a floating-point number format where division by zero yields +∞, -∞, or NaN (Not a Number). It’s a compromise: arithmetic continues, but the result is flagged as invalid. 1,398 years after Brahmagupta, humanity is still debating what a fraction with zero in the denominator means—and every new processor, every programming language, must solve this problem anew.