Principles of Hindu Reckoning

Principles of Hindu Reckoning (Kitab fi usul hisab al-hind) is a mathematics book written by the 10th- and 11th-century Persian mathematician Kushyar ibn Labban. It is the second-oldest book extant in Arabic about Hindu arithmetic using Hindu-Arabic numerals ( ० ۱ ۲  ۳  ۴  ۵  ۶  ۷  ۸  ۹), preceded by Kibab al-Fusul fi al-Hisub al-Hindi by Abul al-Hassan Ahmad ibn Ibrahim al-Uglidis, written in 952.

Although Al-Khwarzimi also wrote a book about Hindu arithmetic in 825, his Arabic original was lost, and only a 12th-century translation is extant. Kushyar ibn Labban did not mention the Indian sources for Hindu Reckoning, and there is no earlier Indian book extant which covers the same topics as discussed in this book. Principles of Hindu Reckoning was one of the foreign sources for Hindu Reckoning in the 10th and 11th century in India. It was translated into English by Martin Levey and Marvin Petruck in 1963 from the only extant Arabic manuscript at that time: Istanbul, Aya Sophya Library, MS 4857 and a Hebrew translation and commentary by Shālôm ben Joseph 'Anābī.

Indian dust board
Hindu arithmetic was conducted on a dust board similar to the Chinese counting board. A dust board is a flat surface with a layer of sand and lined with grids. Very much like the Chinese counting rod numerals, a blank on a sand board grid stood for zero, and zero sign was not necessary. Shifting of digits involves erasing and rewriting, unlike the counting board.

Content
There is only one Arabic copy extant, now kept in the Hagia Sophia Library in Istanbul. There is also a Hebrew translation with commentary, kept in the Bodleian Library of Oxford University. In 1965 University of Wisconsin Press published an English edition of this book translated by Martin Levey and Marvin Petruck, based on both the Arabic and Hebrew editions. This English translation included 31 plates of facsimile of original Arabic text.

Principles of Hindu Reckoning consists of two parts dealing with arithmetics in two numerals system in India at his time. The similarity between decimal Hindu algorithm with Chinese algorithm in Sunzi Suanjing are striking, except the operation halving, as there was no hybrid decimal/sexagesimal calculation in China.
 * Part I mainly dealt with decimal algorithm of subtraction, multiplication, division, extraction of square root and cubic root in place value Hindu-numeral system. However, a section on "halving", was treated differently, i.e., with a hybrid of decimal and sexagesimal numeral.
 * Part II dealt with operation of subtraction, multiplication, division, extraction of square root and cubic root in sexagesimal number system. There was only positional decimal arithmetic in China, never any sexagesimal arithmetic.
 * Unlike Abu'l-Hasan al-Uqlidisi's Kitab al-Fusul fi al-Hisab al-Hindi (The Arithmetics of Al-Uqlidisi) where the basic mathematical operation of addition, subtraction, multiplication and division were described in words, ibn Labban's book provided actual calculation procedures expressed in Hindu-Arabic numerals.

Addition
Kushyar ibn Labban described in detail the addition of two numbers.

The Hindu addition is identical to rod numeral addition in Sunzi Suanjing There was a minor difference in the treatment of second row, in Hindu reckoning, the second row digits drawn on sand board remained in place from beginning to end, while in rod calculus, rods from lower rows were physically removed and add to upper row, digit by digit.

Subtraction
In the 3rd section of his book, Kushyar ibn Labban provided step by step algorithm for subtraction of 839 from 5625. Second row digits remained in place at all time. In rod calculus, digit from second row was removed digit by digit in calculation, leaving only the result in one row.

Multiplication
Kushyar ibn Labban multiplication is a variation of Sunzi multiplication.

Division
Professor Lam Lay Yong discovered that the Hindu division method describe by Kushyar ibn Labban is totally identical to rod calculus division in the 5th-century Sunzi Suanjing. Besides the totally identical format, procedure and remainder fraction, one telltale sign which discloses the origin of this division algorithm is in the missing 0 after 243, which in true Hindu numeral should be written as 2430, not 243blank; blank space is a feature of rod numerals (and abacus).

Divide by 2
Divide by 2 or "halving" in Hindu reckoning was treated with a hybrid of decimal and sexagesimal numerals: It was calculated not from left to right as decimal arithmetics, but from right to left: After halving the first digit 5 to get 2$1/2$, replace the 5 with 2, and write 30 under it:


 * 5622
 * 30

Final result:
 * 2812
 * 30

Extraction of square root
Kushyar ibn Labban described the algorithm for extraction of square root with example of

$$\sqrt(63342)=255\frac{371}{511}$$

Kushyar ibn Labban square root extraction algorithm is basically the same as Sunzi algorithm

The approximation of non perfect square root using Sunzi algorithm yields result slightly higher than the true value in decimal part, the square root approximation of Labban gave slightly lower value, the integer part are the same.

Multiplication
The Hindu sexagesimal multiplication format was completely different from Hindu decimal arithmetics. Kushyar ibn Labban's example of 25 degree 42 minutes multiplied by 18 degrees 36 minutes was written vertically as


 * 18| |25
 * 36| |42

with a blank space in between

Influence
Kushyar ibn Labban's Principles of Hindu Reckoning exerted strong influence on later Arabic algorists. His student al-Nasawi followed his teacher's method. Algorist of the 13th century, Jordanus de Nemore's work was influenced by al-Nasawi. As late as 16th century, ibn Labban's name was still mentioned.