�ݺ�ߣ

The Greatest Mathematical Discovery?
David H. Bailey∗
Jonathan M. Borwein†
May 12, 2010
1 Introduction
Question: What mathematical discovery more than 1500 years ago:
• Is one of the greatest, if not the greatest, single discovery in the field
of mathematics?
• Involved three subtle ideas that eluded the greatest minds of antiquity,
even geniuses such as Archimedes?
• Was fiercely resisted in Europe for hundreds of years after its discovery?
• Even today, in historical treatments of mathematics, is often dismissed
with scant mention, or else is ascribed to the wrong source?
Answer: Our modern system of positional decimal notation with zero, to-
gether with the basic arithmetic computational schemes, which were discov-
ered in India about 500 CE.
∗
Bailey: Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA. Email:
dhbailey@lbl.gov. This work was supported by the Director, Office of Computational
and Technology Research, Division of Mathematical, Information, and Computational
Sciences of the U.S. Department of Energy, under contract number DE-AC02-05CH11231.
†
Centre for Computer Assisted Research Mathematics and its Applications
(CARMA), University of Newcastle, Callaghan, NSW 2308, Australia. Email:
jonathan.borwein@newcastle.edu.au.
1

2 Why?
As the 19th century mathematician Pierre-Simon Laplace explained:
It is India that gave us the ingenious method of expressing all numbers
by means of ten symbols, each symbol receiving a value of position as
well as an absolute value; a profound and important idea which appears
so simple to us now that we ignore its true merit. But its very sim-
plicity and the great ease which it has lent to all computations put our
arithmetic in the first rank of useful inventions; and we shall appre-
ciate the grandeur of this achievement the more when we remember
that it escaped the genius of Archimedes and Apollonius, two of the
greatest men produced by antiquity. [4, pg. 527]
As Laplace noted, the scheme is anything but “trivial,” since it eluded
the best minds of the ancient world, even extraordinary geniuses such as
Archimedes. Archimedes saw far beyond the mathematics of his time, even
anticipating numerous key ideas of modern calculus and numerical analy-
sis. He was also very skilled in applying mathematical principles to engi-
neering and astronomy. Nonetheless he used the traditional Greek-Roman
numeral system for performing calculations [9, 10]. It is worth noting that
Archimedes’ computation of π was a tour de force of numerical interval anal-
ysis performed without either positional notation or trigonometry [1, 9].
Perhaps one reason this discovery gets so little attention today is that
it is very hard for us to appreciate the enormous difficulty of using Roman
numerals, counting tables and abacuses. As Tobias Dantzig (father of George
Dantzig, the inventor of linear programming) wrote,
Computations which a child can now perform required then the services
of a specialist, and what is now only a matter of a few minutes [by
hand] meant in the twelfth century days of elaborate work. [3, pg. 27]
Michel de Montaigne, Mayor of Bordeaux and one of the most learned
men of his day, confessed in 1588 (prior to the widespread adoption of decimal
arithmetic in Europe) that in spite of his great education and erudition, “I
cannot yet cast account either with penne or Counters.” That is, he could not
do basic arithmetic [7, pg. 577]. In a similar vein, at about the same time a
wealthy German merchant, consulting a scholar regarding which European
university offered the best education for his son, was told the following:
2

If you only want him to be able to cope with addition and subtraction,
then any French or German university will do. But if you are intent
on your son going on to multiplication and division—assuming that
he has sufficient gifts—then you will have to send him to Italy. [7, pg.
577]
We observe in passing that Claude Shannon (1916–2001) constructed a me-
chanical calculator wryly called Throback 1 at Bell Labs in 1953, which com-
puted in Roman, so as to demonstrate that it was possible, if very difficult,
to compute this way.
To our knowledge, the best source currently available on the history of
our modern number system is by French scholar Georges Ifrah [7]. He chron-
icles in encyclopedic detail the rise of modern numeration from its roots in
primitive hand counting and tally schemes, to the Babylonian, Egyptian,
Greek, Roman, Mayan, Indian and Chinese systems, and finally to the even-
tual discovery of full positional decimal arithmetic with zero in India, and
its belated, kicking-and-screaming adoption in the West. Ifrah describes the
significance of this discovery in these terms:
Now that we can stand back from the story, the birth of our modern
number-system seems a colossal event in the history of humanity, as
momentous as the mastery of fire, the development of agriculture, or
the invention of writing, of the wheel, or of the steam engine. [7, pg.
346–347]
Indeed, the development of this system hinged on three key abstract (and
certainly non-intuitive) principles [7, pg. 346]:
(a) The idea of attaching to each basic figure graphical signs which were
removed from all intuitive associations, and did not visually evoke the
units they represented;
(b) The idea of adopting the principle according to which the basic fig-
ures have a value which depends on the position they occupy in the
representation of a number; and
(c) The idea of a fully operational zero, filling the empty spaces of missing
units and at the same time having the meaning of a null number.
The Mayans came close—before 36 BCE they had devised a place-value sys-
tem that included a zero. However, in their system successive positions
3

represented the mixed sequence (1, 20, 360, 7200, 144000, · · ·) rather than the
purely vigesimal (base-20) sequence (1, 20, 400, 8000, 160000, · · ·), which pre-
cluded the possibility that their numerals could be routinely used as part of
an efficient system for computation.
Ifrah emphasizes more than once that the Indian discovery was by no
means obvious or inevitable:
The measure of the genius of Indian civilization, to which we owe our
modern system, is all the greater in that it was the only one in all
history to have achieved this triumph. ... Some cultures succeeded,
earlier than the Indian, in discovering one or at best two of the char-
acteristics of this intellectual feat. But none of them managed to bring
together into a complete and coherent system the necessary and suffi-
cient conditions for a number-system with the same potential as our
own. [7, pg. 346]
It is astonishing how many years passed before this system finally gained
full acceptance in the rest of the world. There are indications that Indian
numerals reached southern Europe perhaps as early as 500 CE, but with
Europe mired in the Dark Ages, few paid any attention. Similarly, there is
mention in Sui Dynasty (581–618 CE) records of Chinese translations of the
Brahman Arithmetical Classic, although sadly none of these copies seem to
have survived [6].
The Indian system (also known as the Indo-Arabic system) was intro-
duced to Europeans by Gerbert of Aurillac in the tenth century. He traveled
to Spain to learn about the system first-hand from Arab scholars, prior to
being named Pope Sylvester II in 999 CE. However, the system subsequently
encountered stiff resistance, in part from accountants who did not want their
craft rendered obsolete, to clerics who were aghast to hear that the Pope had
traveled to Islamic lands to study the method. It was widely rumored that
he was a sorcerer, and that he had sold his soul to Lucifer during his trav-
els. This accusation persisted until 1648, when papal authorities reopened
Sylvester’s tomb to make sure that his body had not been infested by Satanic
forces [7, pg. 579].
The Indo-Arabic system was reintroduced to Europe by Leonardo of Pisa,
also known as Fibonacci, in his 1202 CE book Liber Abaci. However, usage of
the system remained limited for many years, in part because the scheme con-
tinued to be considered “diabolical,” due in part to the mistaken impression
that it originated in the Arab world (in spite of Fibonacci’s clear descrip-
tions of the “nine Indian figures” plus zero) [7, pg. 361-362]. Indeed, our
4

modern English word “cipher” or “cypher,” which is derived from the Arabic
zephirum for zero, and which alternately means “zero” or “secret code” in
modern usage, is very likely a linguistic memory of the time when using dec-
imal arithmetic was deemed evidence of dabbling in the occult, which was
potentially punishable by death [7, pg. 588-589].
Decimal arithmetic began to be widely used by scientists beginning in the
1400s, and was employed, for instance, by Copernicus, Galileo, Kepler and
Newton, but it was not universally used in European commerce until after
the French Revolution in 1793 [7, pg. 590]. In limited defense of the Roman
system, it is harder to alter Roman entries in an account book or the sum
payable in a cheque, but this does not excuse the continuing practice of using
Roman numerals and counting tables for arithmetic.
The Arabic world, by comparison, was much more accepting of the Indian
system—in fact, as mentioned briefly above, the West owes its knowledge of
the scheme to Arab scholars. One of the first to popularize the method was
al-Khowarizmi, who in the ninth century wrote at length about the Indian
place-value system and also described algebraic methods for the solution of
quadratic equations. In 1424, Al-Kashi of Samarkand, “who could calculate
as eagles can fly” computed 2π in sexagecimal (good to an equivalent of 16
decimal digits) using 3 · 228
-gons and a base-60 variation of Indian positional
arithmetic [1, Appendix on Arab Mathematics]:
2π ≈ 6 +
16
601
+
59
602
+
28
603
+
01
604
+
34
605
+
51
606
+
46
607
+
14
608
+
50
609
.
This is a personal favorite of ours: re-entering it on a computer centuries
later and getting the predicted answer still produces goose-bumps.
So who exactly discovered the Indian system? Sadly, there is no record of
the individual who first discovered the scheme, who, if known, would surely
rank among the greatest mathematicians of all time. As Dantzig notes,
When viewed in this light, the achievement of the unknown Hindu
who some time in the first centuries of our era discovered [positional
decimal arithmetic] assumes the proportions of a world-event. Not
only did this principle constitute a radical departure in method, but we
know now that without it no progress in arithmetic was possible. [3,
pg. 29–30]
The very earliest document clearly exhibiting familiarity with the deci-
mal system is the Indian astronomical work Lokavibhaga (“Parts of the Uni-
verse”). Here, for example, the number 13,107,200,000 is written as
5

panchabhyah khalu shunyebhyah param dve sapta chambaram
ekam trini cha rupam cha
(“five voids, then two and seven, the sky, one and three and the form”), i.e.,
0 0 0 0 0 2 7 0 1 3 1, which, when written in reverse order, is 13,107,200,000.
One section of this same work gives detailed astronomical observations that
confirm to modern scholars that this was written on the date it claimed to
be written: 25 August 458 CE (Julian calendar). As Ifrah points out, this
information not only allows us to date the document with precision, but it
also proves its authenticity. Methods for computation were not explicitly
mentioned in this work, although the frequent appearance of large numbers
suggests that advanced arithmetic schemes were being used.
Fifty-two years later, in 510 CE, the Indian mathematician Aryabhata
explicitly described schemes for various arithmetic operations, even includ-
ing square roots and cube roots, which schemes likely were known earlier
than this date. Aryabhata’s actual algorithm for computing square roots,
as described in greater detail in a 628 CE manuscript by a faithful disciple
named Bhaskara I, is presented in Figure 1 (based on [7, pg. 497–498]).
Additionally, Aryabhata gave a decimal value of π = 3.1416 [7, pg. 465].
From these and other sources there can be no doubt that our modern
system of arithmetic—differing only in variations on the symbols used for
the digits and minor details of computational schemes—originated in India
at least by 510 CE and quite possibly by 458 CE.
3 Modern History
It is disappointing that this seminal development in the history of mathemat-
ics is given such little attention in modern published histories. For example,
in one popular work on the history of mathematics, although the author de-
scribes Arab and Chinese mathematics in significant detail, he mentions the
discovery of positional decimal arithmetic in India only in one two-sentence
passage [2, pg. 253]. Another popular history of mathematics mentions the
discovery of the “Hindu-Arabic Numeral System,” but says only that
Positional value and a zero must have been introduced in India some-
time before A.D. 800, because the Persian mathematician al-Khowarizmi
describes such a completed Hindu system in a book of A.D. 825. [5,
pg. 23]
6

The Aryabhata-Bhaskara I square root algorithm (510–628 CE):
Start with the highest-order digit of the input integer (if it has an odd number
of digits) or with the two highest-order digits (if it has an even number of
digits). Write, as the ﬁrst digit of the result, the square root of the largest
perfect square less than or equal to this one- or two-digit integer; in the next
row of the tableau, subtract that perfect square from the one- or two-digit
integer. Then repeat the following two-step scheme until the input digits are
exhausted: (i) in the tableau, bring down and append the next input digit to
the remainder; divide the new remainder by twice the current running square
root; append the quotient digit (ignoring fraction if any) to the end of the
running square root; subtract the quotient digit times the divisor from the
remainder; (ii) bring down and append the next input digit to the remainder;
subtract the square of the quotient digit produced in (i); if this subtraction
result would be negative, backtrack and reduce the quotient digit obtained
in (i) by one. Note that if the ﬁnal remainder in the tableau is nonzero, then
by continuing the procedure, the fractional digits of the square root can be
generated one by one.
For example, to compute the square root of 45468049 (= 6743):
Tableau Result Notes
4 5 4 6 8 0 4 9 6
√
45 = 6
3 6 62
= 36
9 4 6 7 94/(2 · 6) = 7
8 4 7 · (2 · 6) = 84
1 0 6
4 9 72
= 49
5 7 8 6 7 4 578/(2 · 67) = 4
5 3 6 4 · (2 · 67) = 536
4 2 0
1 6 42
= 16
4 0 4 4 6 7 4 3 4044/(2 · 674) = 3
4 0 4 4 3 · (2 · 674) = 4044
0 9
9 32
= 9
0 Finished; result is exact.
Figure 1: The Aryabhata-Bhaskara I scheme for computing square roots.
7

A third historical work briefly mentions this discovery, but cites a 662
CE Indian manuscript as the earliest known source [8, pg. 221]. A fourth
reference states that the combination of decimal and positional arithmetic
“appears in China and then in India” [12, pg. 67]. None of these authors
devotes more than a few sentences to the subject, and, more importantly,
none suggests that this discovery is regarded as particularly significant.
We entirely agree with Laplace, Dantzig and Ifrah that this discovery is
of the first magnitude. The stark fact that the system is now taught (and
mastered) in grade schools worldwide, and is implemented (in binary) in
every computer ever manufactured, should not detract from its historical
significance. To the contrary, these same facts emphasize the enormous ad-
vance that this system represented over earlier systems, both in simplicity
and efficiency, as well as the huge importance of this discovery in modern
life.
Perhaps some day we will finally learn the identity of this mysterious
Indian mathematician. If we do, we surely must accord him or her the
same accolades that we have granted to Archimedes, Newton, Gauss and
Ramanujan.
References
[1] L. Berggren, J. M. Borwein and P. B. Borwein, Pi: a Source Book,
Springer-Verlag, New York, third edition, 2004.
[2] David M. Burton, The History of Mathematics: An Introduction,
McGraw-Hill, New York, 2003.
[3] Tobias Dantzig and Joseph Mazur, Number: The Language of Science,
Plume, New York, 2007. This is a reprint, with Preface by Mazur, of
Dantzig’s book as originally published by MacMillan in 1930.
[4] Will Durant, Our Oriental Heritage, vol. 1 of The Story of Civilization, 11
vols., Simon and Schuster, New York, 1954 (date of vol. 1).
[5] Howard Eves, An Introduction to the History of Mathematics, Holt,
Rinehart and Winston, New York, 1990.
[6] R. C. Gupta, “Spread and triumph of Indian numerals,” Indian Journal of
Historical Science, vol. 18 (1983), pg. 23–38, available at http://www.new.
dli.ernet.in/rawdataupload/upload/insa/INSA_1/20005af7_23.pdf.
8

[7] Georges Ifrah, The Universal History of Numbers: From Prehistory to the
Invention of the Computer, translated from French by David Vellos, E. F.
Harding, Sophie Wood and Ian Monk, John Wiley and Sons, New York,
2000.
[8] Victor J. Katz, A History of Mathematics: An Introduction, Addison
Wesley, New York, 1998.
[9] Reviel Netz and William Noel, The Archimedes Codex, Da Capo Press, 2007.
[10] Josephine Marchant, Decoding the Heavens: Solving the Mystery of the
World’s First Computer, Arrow Books, New York, 2008.
[11] John Stillwell, Mathematics and Its History, Springer, New York, 2002.
[12] Dirk J. Struik, A Concise History of Mathematics, Dover, New York, 1987.
9

�ݺ�ߣ

Discovery

More Related Content

Discovery