You are on page 1of 16

10

Irrational Numbers
Danielle M. Tarnow

2 R (2) 1414 . ...


(I wish I knew)

11

12

13

14

15

17

Archytas Method (428-365 BC)


Known since the time of the Babylonians To approximate R(a), let a1 be your first guess, then: b1=a/a1 (Here either ai or bi is too big and the other too small) a2=arithmetic mean of a1 and b1=(a1+b1)/2 b2=a/a2 Continuing, an=(an-1+bn-1)/2 and bn=a/an

The value of R(2) on the Yale Babylonian Tablet is a3 if a1=1. This method is frequently used by computers today.

Greek Irrational Ladders


The Greek Irrational Ladder for R(2) is: 1 1 1/1=1 3 2 3/2=1.5 7 5 7/5=1.4 17 12 17/12=1.41666 41 29 41/29=1.41379 99 70 99/70=1.41429 etc. Which comes from the easily shown identity that:

2x2-y2=(2x+y)2-2(x+y)2=

The Discovery of Incommensurability


There is no evidence that the Babylonians or the Egyptians knew that R(2) was irrational (they might have just believed that more places needed to be used).
Hippasus of Metapontom is credited with the discovery of incommensurable ratios (apparently he was drowned for this by the Pythagoreans). ALOGOS meant not a ratio and not to be spoken. The Pythagoreans were the first to realize that incommensurable ratios were different in character from commensurable ones. The proof that R(2) is irrational was probably not originally part of Euclids Elements, it was probably added in later editions. (Although the proof was known long before Euclid, according to Aristotle.)

Eudoxus of Cnidus (408-355 BC)

Did you know the ratio of the diagonal length of a cube to the length of its edge is R(3)?

The Greeks did not realize that lines were not discrete collections of units, thats why they had so much trouble with Zenos ideas (the discrete vs. the continuous).
Eudoxus created the idea of a MAGNITUDE (which allowed people to stay away from the idea of irrational NUMBERS). This caused the view that geometry and algebra were unrelated and that geometry was the only rigorous way to prove things. Theatetus investigated and classified the types of incommensurable lengths that can be generated with a compass and straight edge. Fibonacci later showed that there were types of irrationals that could not be created this way.

Aristotle of Stagira (384-322 BC)


Aristotle gives a reductio ad absurdum proof that R(2) is irrational. This proof, which only works for square roots of EVEN non-squares, was known long before him.
Heres how it goes: Let R(2)=a/b where a and b are positive integers that are relatively prime. Then, a2=2b2. This means that a2 is even, so a is even and can be represented by, say, 2c. So, 4c2=2b2 or 2c2=b2. So, b2 is even, so b is even. If a and b are both even, then they are not relatively prime, thus it is impossible that R(2) is rational.

Some Approximations
Plato used R(2)7/5 because 249/25 Theodorus used R(3)7/4 because 349/16 Archimedes, while working on Pi, used the fact that 1351/780 > R(3) > 265/153, which may have come from using what will later come to be known as Herons method.

Herons Method (10-75)


Heron was probably not the creator of this algorithm. b b a a2 b a 2a 2a 1 Where a2 is the rational square nearest to the number and b is the remainder. Several applications of this for R(3) results in Archimedes approximation.
Herons way of saying this (for R(720)): Since 720 has not its side rational, we can obtain its side within a very small difference as follows. Since the next succeeding square number is 729, which has 27 for its side, divide 720 by 27. This gives 26 2/3. Add 27 to this, making 53 2/3, and take half this or 26 5/6. The side of 720 will therefore be very nearly 26 5/6. In fact, if we multiply 26 5/6 by itself, the product is 720 1/36, so the difference in the square is 1/36. If we desire to make the difference smaller still than 1/36, we shall take 720 1/36 instead of 729 (or rather we should take 26 5/6 instead of 27), and by proceeding in the same way we shall find the resulting difference much less than 1/36.

How Ptolemy Got His Approximation (85-165)


Ptolemy gives 103/60+55/602+23/603 for R(3) which is correct to 6 decimal places.
This comes from the idea that: (a+b)2=a2+2ab+b2 So, if you can find an a and a b such that a2+2ab+b2 is close to 3, then a+b will be close to R(3).

Ptolemy actually uses (a+b+c)2=a2+b2+c2+2ab+2bc+2ac here.

Did you know the current day square root symbol was first introduced in 1544?

Narayanas Method (1340-1400)


This Indian method uses the second order indeterminate equation Nx2+1=y2, with x<y, to approximate R(N) by y/x. For N=10, there are many solutions, including x=6 and y=19, which yields R(10)3.1666, which is correct to 2 decimal places. Narayanas solution of 10x2+1=y2 where x=8658 and y=227379 gives an approximation for R(10) that is correct to eight decimal places.

Fermats Method of Infinite Descent (1879)


Let R(5)= a/b where a and b are positive integers. Now, 4<5<9, so 2<R(5)<3. So, 2<a/b<3, which is equivalent to 2b<a<3b. So, 0<a-2b<b. Since R(5)=a/b, 5b2=a2. We can subtract 2ab from both sides, getting 5b2-2ab=a2-2ab. We can factor this, getting b(5b-2a)=a(a-2b). So, a/b = (5b-2a)/(a-2b), which is a more reduced fraction than a/b, since a-2b<b. You could do this over and over again, and this is not possible with b being a positive integer, so R(5) is irrational.
This is a number theory equivalent to the natural idea of infinite reproducibility of figures inside figures (for example, you can keep putting smaller and smaller pentagrams inside of each other). It is only a short step from here to prove that Phi=(1+R(5))/2 is irrational.

Before the 19th Century


The Hindus and Arabs worked freely with the irrationals. They did not bother to prove that they could operate with irrationals in these ways.
The decimal system was adopted and in 1696, Wallis identified the rationals as periodic decimal numbers. It was ASSUMED that functions were also defined for irrational values of x. No one questioned letting x equal an irrational number. It was IMPLICITLY understood that the function value resulting from substituting an irrational x would be between the values obtained by substituting rational values above and below x. The concept of CONTINUITY had not been defined in a satisfactory way.

Julius Wilhelm Richard Dedekind (1831-1916)


The essence of continuity is the ability to divide all points on the number line into two groups using one and only one point This idea of a CUT came to him while lecturing Calculus Not all cuts are created by rational numbers (proof uses something similar to Fermats method of infinite descent)

From this he rigorously develops =, <, >, and the operations

Georg Ferdinand Ludwig Philipp Cantor (1845-1918)


Always remembered his childhood in Russia with great nostalgia and never felt at ease in Germany
Exchanged thoughts with Dedekind He defined numbers in terms of convergent sequences of rational numbers, for instance R(2) is represented by the fundamental sequence: 1, 1.4, 1.41, 1.414, 1.4142, 1.41421, 1.414213, He also rigorously formalized the operations

Irrational Numbers
Irrational numbers are not just single symbols or even pairs of symbolsthey must be represented by an infinite collection of symbols, such as Cantors sequences or Dedekinds cuts. Stoltz (1842-1905) said that every irrational number could be represented as a non-periodic decimal (and this can be used as a defining property of irrationals). I am currently creating a chronology of the development of irrational numbers.

References
From the Course Recommended Book List: Eves An Introduction to the History of Mathematics Katzs A History of Mathematics: An Introduction Klines Mathematical Thought From Ancient to Modern Times Other Books: Boyers A History of Mathematics Burtons The History of Mathematics: An Introduction Katzs Using History to Teach Mathematics: An International Perspective Nahins An Imaginary Tale: The Story of R(-1) From the Course Handouts: Two Dedekind Papers History of Mathematics (by Shank?)

From the Internet: http://www-gap.dcs.st-and.ac.uk/~history/

You might also like