## The Method of Exhaustion vs. Calculus

In Lecture 7 of the excellent DVD course *Great Thinkers, Great Theorems* (The Great Courses No. 1471) Professor Dunham of Muhlenberg College discusses Archimedes’ *Measurement of a Circle*. Archimedes’ first Proposition in effect gives us the formula for the area of a circle ─ though this is not quite how Archimedes puts it. He writes

*“The area of any circle is equal to [that of] a right-angled triangin which one of the sides about the right angle is equal to the radius and the other to the circumference of the circle”
* This is ancient Greece, more particularly the ancient Greek ex-colony of Syracuse, and Archimedes had no algebra or modern notation ─ these innovations had to wait for Descartes to be born a millennium and a half later. Historians of mathematics never tire of telling us that the ‘Method of Exhaustion’, which Archimedes inherited from Eudoxus, is an unwieldy and clumsy tool compared to Calculus which up to a point it is. Nonetheless, Archimedes’ approach has several advantages over the contemporary textbook one. For a start, Archimedes’ proof

*makes sense*, is intuitively acceptable and thus far more accessible to the intelligent layman, even to the intelligent child ─ the older Victorian style textbooks invariably used simplified ‘exhaustion’ arguments and that’s how I learned my basic geometry. In comparison, the algebraic Calculus proof, though shorter, is tiresome in the extreme and doesn’t seem to have much to do with real circles and areas ─ it makes one think of an immensely capable stage magician performing a series of tricks.

To Archimedes then. First off, Archimedes does not turn the area of a circle into an abstract algebraic function: he compares the circle to something that we can see and draw, namely a right angled triangle with known area *½ base × height * where the height is the radius of a circle (any circle) and the base is the same circle’s circumference straightened out. Area = *r × C) = ½ r C*.

[Apologies but currently I have problems putting diagrams on this website, see later versions of this post.]

The proof is by double contradiction (as Professor Dunham puts it) and though these kind of proofs are often rather artificial, in this case the reasoning is both impeccable and even entertaining. Archimedes says in effect, “Let us suppose my claim is wrong, and that the Area of a circle is* not *equal to that of my triangle”. It might, for example, be *less *than the area of a circle. We now ‘exhaust’ the circle in the inimitable Greek manner by fitting regular polygons (many sided figures with all sides and matching angles equal) inside the circle. We can start with an equilateral triangle which fits exactly into the circle (how to do this is demonstrated in Euclid Book I), we then double it to produce a hexagon (which, amazingly has six outer sides equal to the radius), double that to make a *12-sider, 24-sider, 48-sider &c. &c. *All these doublings can actually be performed by following instructions given in Euclid Book I using only straight edge and compass.

Clearly ─ because we can actually *see it happening* ─ the gap between the total area of the inscribed polygon and the area of the circle gets less each time. Next, very reasonably, Archimedes asks us to accept that, in principle at least, by constructing polygons with an ever growing number of sides we can *make the difference between the area of the polygon and the area of the circle as small as we please. *(Difficulties with computer graphics stop me giving too many diagrams at this point, see updated versions of this post.)

Now the polygon is made up of identical triangular sections where the ‘base’ is the outer side of each section and the ‘height’ is the length of the perpendicular from the centre to this base, technically known as the *apothem*. Note that the two other sides of the triangular section are radii and, important point, the radius is always *greater *than the apothem since the apothem is the shortest route to the base. Now, every triangular section of the (regular) polygon is the same, so, if the polygon consists of *6 *triangles, the total area will be *6 × ( ½ base × height) *or, to put it another way, *6 bases × ½ h *. The length ‘*6 bases’ *is the length of the *perimeter *of the polygon, the distance you would have to go if you walked all round it. We now increase the number of sides to *12, 24, 48, 96…. *and so on. At each stage, what we end up computing to give the area of the polygon is the perimeter of the inscribed polygon multiplied by half the height, or *½h × P *where *P *is the * *Perimeter. [Apologies but currently I have problems putting diagrams on this website, see later versions of this post.]

What about the original right angled triangle? The area of Archimedes’ right angled triangle is*½ base × height = ½ C × r *─ where *C *is the circumference of the circle and *r *is the radius. Archimedes has suggested, for sake of argument, that the area of the right angled triangle is *less* than the area of the circle. Now, since the area of the inscribed polygon can be made as close to the area of the circle as we wish, it should be a better fit than the area of the right angled triangle. So, according to this supposition, we have the inequality

** Area Triangle < Area n-gon for large n < Area circle
**

**However, this can’t be right because the height of the triangle is**

*r*, the radius of the circle, and the height of

*any*triangular section of

*any*n-gon is the apothem which is always

*less*than the radius. Also, the circumference of the circle, the base of the triangle, is clearly

*greater*than the perimeter of

*any*n-gon ─ because the length of a curve drawn between two points is always greater than that of a straight line.

We thus have a contradiction so we must reject the hypothesis that the area of the triangle < area of the circle.

Suppose, then, the area of the triangle is

*greater*than the area of the circle. This time Archimedes’ constructs a series of regular polygons that fit neatly

*around*the circle, i.e. the bases of each triangular section lie outside, not inside the circle. The number of sides is again increased without limit, making the difference between the area of the circumscribed polygon and the area of the circle as small as we wish. This time we find that the height of each triangular section is indeed the length of the radius but the perimeter is undeniably greater than the circumference of the circle. So this polygonal total area cannot be greater than the area of the right-angled triangle because the base of the triangle is equal to the circumference. Again a contradiction and the second hypothesis must be rejected.

Archimedes now delivers the

*coup de grace*. Since the area of the right angled triangle is

*not*smaller than the area of the circle and is not greater than the area of the circle, the only possibility left is that it is

*exactly equal*to the area of the circle. But this is what he wanted to prove.

But what is the area of the right angled triangle in modern terms?

It is

*½ radius × Circumference*which in our terms (not his) comes to

*½ r (2*

*p*

*r) =*

*p*

*r*(since the

^{2}*½*and

*2*cancel). Archimedes does not write

*π*which though a Greek letter was not given its modern mathematical meaning until the 18

^{th}century of our era (by Euler). But neither does he

*think π.*For the Greeks curved and straight lines (along with certain pairs of straight lines) were

*incommensurable*─ could not strictly be compared because they lacked a basic common ‘measure’. Archimedes restricts himself to saying that the

*ratio*of the diameter of a circle to the circumference (our

*p*) lies between two limits: it was less than (

*3 + 1/7)*and greater than (

*3 + 10/71).*This is not the same thing as

*equating*the circular

*ratio*of two lengths to the irrational number we know today as π ─ since the Greeks did not use irrational numbers which for them were not true numbers. In effect, modern mathematics does (or at any rate

*appears*to do) what the later Greeks knew to be impossible ─ it ‘squares the circle’, a figure that cannot be squared.

The only point in Archimedes’ excellent piece of reasoning that is somewhat questionable is the following. By making the circumference the base of his triangle, Archimedes is in effect

*straightening it out*, something that is in practice impossible to do without falsifying its true length, if only by a little. Even when using a piece of string, the molecules are inevitably going to be more extended when measuring the circumference and more compressed when measuring a straight line. This incidentally is very much the sort of objection that already in ancient Greek times certain sophists and epicureans made against the new-fangled discipline of higher geometry so enthusiastically promoted by the rival school of Plato and his followers.

** ****The Calculus derivation
**The modern derivation of the formula for the area of a circle goes like this. Armed with a co-ordinate system (which Archimedes did not possess) we draw a circle centred at the origin and assess the ‘area under the curve’ ─ or rather the area of the first quadrant which is all we need (because we eventually multiply it by

*4*). By Pythagoras,

*r*where

^{2}= x^{2 }+ y^{2}*r*is the radius of the circle, or

*y = √*

*r*We want the inner area between the right side of the

^{2}– x^{2}.*x*axis and the top part of the

*y*axis, i.e. the limits of the integral are

*x = 0, x = r*. To get out this definite integral you have to reverse the standard derivative of

*arcsin*or

*sin*a formula that is impossible to remember. The integral turns out to be

^{−1},[

*½ x √*

*r*] taken between the limits

^{2}– x^{2}+ ½ r^{2 }sin^{−1}x/r*x = r*and

*x = 0*.

This expression is guaranteed to make the non-mathematical reader give up in disgust and go to the pub instead, and, although I have on occasion taught Calculus to school age kids, I dislike the inverse trigonometrical functions and had to look up the answer in a book and differentiate the result to see if I got back to √

*r*. More to the point, this hodgepodge seems to have nothing at all to do with circles or areas of anything at all though, if you do feed into the expression

^{2}– x^{2}*x = 0*and take this away from what you get with

*x = r*you (eventually) end up with

*¼ π*

*r*which, since this is a quadrant, makes the total area four times this or π

^{2}*r*.

^{2}But how do I know this Calculus result is right? Only by accepting various assumptions which underpin Calculus, assumptions which were, in the days of ‘infinitesimals’, extremely dubious as Bishop Berkeley pointed out at the time and have only been made rigorous by disconnecting Calculus completely from physical reality. The Greeks (correctly) argued that curved lines and straight lines are different species and that areas bounded by curves can be

*approximated*to any desired degree of precision but in general never the twain shall meet. The fact is that the circle cannot be squared and Calculus is being extremely helpful but disingenuous when it gives the area as

*exactly*π

*r*. Like all irrationals, the ‘number’ π is not a real quantity but a numerical

^{2}*limit*that real quantities can approach but never attain. There is no such thing as an object whose length is π, and you cannot perform any action a π number of times.

The ancient

*Method of Exhaustion*rests on very different assumptions about the physical world than

*Calculus*does and on much more sensible ones. Nevertheless, we needed Calculus in the bad old days because the labour of attaining a desired level of precision was horrendous (even after the invention of logarithms which reduced it considerably). Calculus, in its trendy post-Weierstrass version now obligatory in colleges of higher education, is admittedly an ingenious piece of pure mathematics which, by fancy footwork, effectively sidesteps issues about ‘infinitesimals’. But it is a piece of logic rather than mathematics. And anyway the writing is on the wall already for Calculus. Now that we have supercomputers, the trend is not to bother looking for analytic solutions but simply to slog it out numerically. Numerical Analysis, the art of progressive approximation, is rapidly becoming more important than Functional Analysis. In any case, the vast majority of differential equations (equations dealing with change and motion) are

*insoluble*analytically so there is all too often no alternative to approximate methods. Practically speaking, we no longer need all this wretched philosophical lumber of ‘infinitesimals’ and infinite processes: the eminently sensible ancient Greek

*Method of Exhaustion*, which successfully avoided actual infinity altogether, is, against all the odds, making a triumphant come-back . Traditional Calculus is a mighty cultural achievement and a necessary bridge to the present and what lies ahead, but one day it will be probably put in the same class as those impressive looking species that eventually went extinct because they were unable to cope with real-life conditions.

*SH*

## Objects in Space

“*A mental creation that evolved to study objects in the world*”^{1 }─ this sounds to me like a pretty good description of mathematics ─ or at any rate of mathematics prior to the late 19th century. It highlights the two most important features of mathematics:

**(1) **that it is a (human) *creation* and,

**(2) **that it is not a *free *creation ─ it cannot be if its aim is to study ‘objects in the *world’*.

**(2)** in effect means that mathematics is, or rather was, subject to serious constraints. What sort of constraints? If we examine the two earliest branches of mathematics, arithmetic and geometry, we see that the principal constraints are *discreteness *and *distancing*. *Discreteness* because arithmetic sees the world as made up of lots of little bits that *can be counted* ─ if everything was mixed up with everything else like the ingredients in a cake arithmetic would not give sensible results (and would not be needed). Secondly, experience tells us that whatever is happening *here* is not simultaneously happening *over there*, i.e. that objects and events are separated by something, thus geometry, metric spaces, causality, anything that depends on spatial separation.

Viewed thus, the two earliest and (arguably) still the two most important branches of mathematics depend either on *separateness *or *separability *: in the first case, that of arithmetic, the focus is on the *objects *themselves, in the second, the focus is more on what lies *between *and *around *the objects. Practically all pre-Renaissance mathematics was in effect the study

**(1)** *of particular objects separated by empty space *or

**(2)** *of particular objects such as points and curves embedded in empty space.*

No hint of movement so far ─ as we know the Greeks, though they initiated both statics and hydrostatics baulked at the creation of dynamics, the study of objects in motion and, by implication, of the unseen forces that give rise to motion.

^{1} From Lakoff & Núñez, *Where does Mathematics come from?*

## Distinctions by Type vs. Distinctions by Number

Before being represented by a numeral, all entities and collections of entities must be stripped of all characteristics such as shape, size, colour, substance, purpose, origin &c. in imagination, if not in fact. All internal group properties such as rank, proximity, natural affinity and so on must also be destroyed. To become numbers, all groups of entities must be both **DEPERSONALIZED** and **DESTRUCTURED**. In effect,

** DISTINCTION BY NUMBER CAN ONLY BE ACHIEVED BY ABOLISHING DISTINCTION BY TYPE. **

This explains the surprisingly rudimentary nature of number concepts amongst hunting/food-gathering peoples, and their stubborn resistance to the introduction of more advanced number systems and number concepts by missionaries. This resistance is not to be attributed to a lack of intelligence since the complex language structures, imaginative myths, pictorial sense and elaborate rituals of such peoples show that they were capable of first-rate cultural achievements. No, the reason for this resistance to number is deep and essentially well-founded since for such peoples it would have been a rash move to automatically prefer distinction by number to distinction by type.

For classification according to type is absolutely essential to the hunter/food-gatherer: he or she must make quick-fire radical distinctions between plants that are comestible and poisonous, animals that are harmless or dangerous, strangers that are hostile or friendly &c. &c. ─ and errors can easily lead to death of the individual and even extinction of the tribe. But counting objects is of little utility: what is the point of attributing a number sign to objects that are in front of you every day of your life? Do *you *know how many suits of clothes or dresses you own? How many rooms there are in your house or flat? Arithmetic only becomes significant when it is essential to know when to sow or reap, when trade is extensive and, above all, when a state official needs to assess a whole country’s resources. It was the Assyrians and the Babylonians who developed arithmetic just as it was the founder of the short-lived Ch’in Dynasty, Ch’in Shih Huang Ti (he of the terracotta warriors), who imposed the metric system on his citizens nearly two thousand years before Napoleon did and who likewise standardized weights and measures throughout his vast Empire. *SH 17/1/19*

## Assumptions and Abilities Necessary for a Number System

Nature is not deliberately mathematical or even numerate : if certain numbers keep coming up — and few do systematically — there is generally some physical or biological reason for them ^{1}.

In this sense it is perfectly true that numbers, or at any rate number *systems*, are human creations but they are firmly based on features of the natural world that really exist objectively. One might say, to paraphrase Guy Debord, “Number has always existed but not always in its numerical form” ^{2}.

So how do we develop a number system? What are the minimal requirements?

Two, and as far as I can see, *only* two abilities are necessary to develop a number system :

- The ability to distinguish between what is singular and plural, i.e. recognize a ‘one’ when you see it;
- The ability to carry out a one-one correspondence (pairing off).

All the mathematicians who have developed abstract number systems, for example Zermelo and von Neumann, had these two perceptual/cognitive abilities — otherwise they would have been denied access to higher education and would not even have been able to read a maths book. Animals seem to have (1.) but not (2.) which is perhaps the reason why they have not developed symbolic number systems (though a more important reason is that they did not feel the need to). Computers are capable of (1.) and (2.) but only because they have been programmed by human beings.

What is number? One could describe ‘number’ as the ‘property’ that results when we have done away with all other distinctions between sets such as colour, weight, position, shape and so on. This is not much of a definition but it does emphasize the curious fact that number is more of a negative rather than a positive property since it results, as Piaget says, “*from an ignoring of differential qualities”*.

But, notwithstanding the difficulty of saying what exactly number is, practically speaking there is a perfectly simple and universally applicable test which can decide whether two sets of discrete objects are numerically equivalent or not, i.e. can be validly allocated the same number label. If I can pair each of them off with the *same* standard set of objects or marks, the two sets are numerically equivalent, if I can’t they are not. Of course, today if I want to assess the ‘number’ of chairs in a room, say, I associate the collection with a number word, *seven* or *four *or *six *as the case may be, but underlying this is a pairing with a standard set. As a matter of fact I find that, though I use the number words *one, two, three….. *when counting objects, I still find it necessary to use my fingers, either by pointing my finger at the object or pressing it against my side, one press, one object. And the umpire in a cricket match still uses stones or pebbles : one ball bowled, one stone shifted from the right hand to the left. It is not that the finger or stone pairing off is valid because of our ciphered numerals but the reverse : our written or spoken numerals ‘work’ because underlying them is this pairing off of items with those of a standard set.

Now, one could actually derive the Cantor definition of cardinal number — “*that **which results from abstracting from a set the order of appearance of the elements and their specific character”* — from what happens when I apply my test. If I rearrange the objects I am supposed to be counting, does that make any difference to the ‘number’ representing the sum? No. Because if I could pair off the original collection with items from a standard set, such as so many pebbles or marks, I can do the same after rearrangement. Does the actual identity of the objects matter? Apparently not, since if I replace each original item by a completely different item, I can still pair off the resulting set with my standard set (or subset).

We thus arrive, either by reflection or simply by applying the test, at the two basic numerical principles, the **Disordering Principle **and the **Principle of Replacement
Disordering Principle
**

*The numerical status/cardinal number of a collection is not changed by rearrangement so long as no object is created or destroyed.*

** Principle of Replacement
**

*The numerical status/cardinal number of a collection is not changed if each individual object is replaced by a different individual object.*

* * Together these two principles make up a sort of **Number Conservation Principle **since whatever ‘cardinal number’ is, this ‘something’ persists throughout all the drastic changes the set undergoes just as, allegedly, a given amount of mass/energy persists throughout the interactions between molecules within a closed system.

These two principles may either be viewed as *Definitions* i.e. they tell you what we mean by cardinal number, or as *Postulates * since they are the generalisation of actual experiments (pairing off sets with a chosen standard set). They are not ‘logical truths’ and not strictly speaking axioms.

The **Principle of Correspondence **has a somewhat different status and is more like a true Axiom, i.e. something which we have to take for granted to get started at all but which is not directly culled from experience.

**The Principle of Correspondence **

** ***Whatever is found to be numerically the case with respect to a particular set A, will also be numerically the case for any set B that can be put in one-one correspondence with it. *

By ‘numerical’ features I mean such things as divisibility which has nothing to do with colour, size and so forth. We certainly do assume the **Principle of Correspondence** all the time, since otherwise we would not gaily use the same rules of arithmetic when dealing with apples, baboons or stars : indeed, without it there would not be a proper science of arithmetic at all, merely ad hoc rules of thumb. But, though the Principle of Correspondence is justified by experience, I am not so sure that it originates there : it is such a basic and sweeping assertion than it is more appropriate to call it an *Axiom* than anything else. Note that physical science uses a similar principle which is today so familiar that we take it for granted though it is far from ‘obvious’ (and possibly not entirely true), namely that “what is found to be physically the case for a physical body in a particular place and time is the case for a similar body at a completely different place and time”. Newton’s law of gravitation is not just true here on Earth but is assumed to be true everywhere in the universe — a fantastic generalization that many scientists at the time thought unwarranted and arbitrary.

These principles do not by any means exhaust the assumptions we implicitly make when we use or apply a Number System : indeed, if we listed all of them we could probably fill a sizeable volume. For example, we continually assume that there is a physical reality ‘out there’ to number in the first place (which solipsists and some Buddhists deny), that there are such things as discrete objects (which philosophic monists and in some of his writings even Einstein seems to deny) and so on and so forth. But these ‘axioms’ are best left out of the picture : they underlie most of what we believe and are not specific to numbering and mathematics. ** **

**Notes **

^{1} This is (perhaps) not true of the basic constants such as the gravitational constant or the fine structure constant : they seem to be ‘hard-wired’ into the universe as it were and there seems to be no special reason why they should have the values they actually do have, unless one accepts the Strong Anthropic Principle. In theory it should be possible to deduce the values of basic constants from *a priori *principles but to date attempts to do this, such as Eddington’s derivation of the number N, the number of elementary particles in the universe, have not been very successful to say the least. One could argue from ‘logical’ considerations that there must be a limiting value to the transmission of electro-magnetic signals but there is no apparent reason why it should be *3 × 10 ^{8} m/sec *

^{2} The quotation I have in mind is, *“L’histoire a toujours existé mais pas toujours sous sa forme historique” *(‘History has always existed but not always in its historical form’) from *La Société du Spectacle *by Guy Debord. The phrase sounds wonderful but means very little.

*SH *4/03/2018

## DATA RECORDING AND SPACE

Mathematics seems to be not only an exclusively human activity but also a very recent one (in evolutionary terms). Animals, with one or two possible exceptions (bees, whales), do not use symbolic systems for communicating information or as behavioural aids. This is not to imply that our way of doing things is necessarily preferable or even ‘more advanced’: it would be hard to better the extraordinary feats of migrating birds returning to the selfsame nesting spots year after year or the accuracy of predators pursuing rapidly moving prey. Quite how animals achieve such feats is still not completely understood, at any rate with respect to migratory birds, but, certainly, they do not consult ephemerides or solve differential equations. Much the same goes for early societies: they did not develop complex numerical systems because, most of the time, they did not need them. Why did they not need them? For two reasons, firstly because the quantities involved, e.g. number of personal possessions or number of members of the tribe, were small and, secondly, because the sensory apparatus of early humans was extremely acute, more acute than ours. Being able to review objects in their precise locations often does away with the need to count them. Do you know how many chairs there are in your house or flat? How many rooms even? You don’t need to know the numbers involved because you can mentally review such a familiar landscape and even approach it from different sides, go into it, behind it and so on. Moreover, since in earlier times a good visual and auditory memory was necessary for survival, it was intensely cultivated alongside physical fitness. Predominantly oral societies routinely produced individuals whose memory capacities seem scarcely credible to us today: sacred books as long as the Rig-Veda or the Koran were learned off by heart and reputedly still are in some parts of India and Pakistan. Mathematics itself (as we understand the term) only took off with the advent of large, centralised, bureaucratic societies such as Assyria, Babylon and Egypt. In such cases, a good visual memory was inadequate since the scribe or official would not be personally familiar with what he was supposed to be assessing. It was the necessity to record and process data efficiently that gave rise to mathematics in these imperial societies and, conversely, an advanced numbering system only becomes essential when what you are dealing with exceeds the range of your personal experience. “The main condition under which arithmetical operations become useful is economic action at a distance and such conditions do not arise for hunters or for the simpler forms of agricultural society” (Denny, *Cultural Ecology of Mathematics*) (**Note 1**). Unromantic though it sounds, arithmetic seems to have been developed mainly for the purpose of stocktaking, taxation and large-scale warfare while geometry (literally ‘land-measurement’) was, according to Herodotus, invented by the Egyptians in order to survey accurately (and subsequently tax) the irregular plots of peasants bordering the Nile.

Early arithmetic and numbering generally was concerned **(1)** with recording what was already known (at least approximately) and **(2)** finding out and recording what was not known — but which could, hopefully, be extracted from the relevant data. A census carried out in a series of villages would tell a regional official how densely populated the area was, and such a piece of data needed to be recorded in a form that other officials would be able to comprehend. This is **(1)**, recording what is already known — at any rate locally . If we want to work out the food supplies necessary to keep all these people alive in a time of famine, or how many young men the region is likely to be able to provide for the army, we have a primitive kind of equation. This is a case of **(2)**, finding out and then recording what is, prior to the census or other data collection, is not known locally. There is, however, no hard and fast line separating **(1) **and** (2) **since simply combining the separate data about each village does provide new information, i.e. the* total* number of inhabitants in the region which, doubtless, no single villager knew.

An efficient number system is necessary both for *assessing* and *recording* important quantities and in practice this means that *two* systems, or two versions of the same system, are required, a *temporary * system and a more *permanent *system. If quantities are small, we can *assess* a given quantity (how many pigs? how many coconut trees?) using our hands as the temporary recording system but, since we need our hands for other purposes, we also need a separate, much more durable, recording system which could be clusters of shells (Benin Empire in Nigeria), knots in a string (Inca Empire in Peru) or marks on some long-lasting material such as bone, bark or papyrus (Egypt). Even today, numbers are still primarily used simply for recording data — rather than for pure-mathematical purposes. Coping with numerical data has, in fact, been a perennial problem for advanced societies from ancient Egypt right down to the present day.

The early Egyptian ‘hieroglyphic’ number system is perhaps the clearest and simplest number system ever invented. A single item, a datum, was originally represented by a picture of a papyrus leaf which soon just became a stroke. The Egyptians, like most (but by no means all) societies used a base-ten system, i.e. once you have a given collection of strokes, you make it into a ‘first base’ (our *ten*), when you have the same quantity of ‘first bases’ you make it a second base (our *hundred*) and so on. In principle the different bases could be distinguished by size ― if unity is a stroke, ‘ten’ is a longer stroke, ‘hundred’ a longer stroke still &c. &c. The inconvenience of such a number system is that it requires a lot of space if you are dealing with large quantities, which the Egyptian officials often were (it is thought that some Egyptian cities at their height had nearly a million inhabitants). Considerations of space have in fact played a very large part in the development of number systems and recording technology generally. The Egyptians did not distinguish the ‘one-symbol’ from the symbol for first base, the symbol for the first base from the second and so on by comparative size: they had *separate pictograms *for ‘one’, ‘first base’, ‘second base’ and so on. Our *ten *was a bent leaf, our *hundred* a coiled rope, our *thousand *a lotus flower, our *ten thousand *a snake, our *hundred thousand *a tadpole or frog and our *million *a “seated scribe holding up his hands in astonishment”. In this system you only had to learn the meaning of seven hieroglyphs whichis not a very great task. But with these seven symbols repeated when necessary any quantity less than a ‘million million’ (original meaning of ‘billion’) could be represented. “They [the Egyptian officials] could record the number of captives available for slave labour and share them out for public works. They could estimate how much food and drink, how many blocks of stone of different shapes and sizes, how many slaves and overseers would be needed from day to day to build the pyramids” (McLeish, *Number*).

Note that in the Egyptian system, as opposed to the ‘increasing size’ system which hardly any society ever used, a new *single* symbol is needed for each larger base; any given symbol is never repeated more than a certain number of times (nine times in a ten-base system). Each new symbol is thus not just a bigger and better version of the basic ‘one-symbol’ but something quite different. Some of the new symbols seem somewhat arbitrary since one sees no obvious connection between a quantity we call a *hundred* and a coiled rope for example. On the other hand, the Egyptian symbol for our 100, 000, either a frog or a tadpole, may well have been chosen because frog spawn contains a vast number of eggs, as someone recently suggested to me. Since, even today, our brain finds it much easier to store images of real things rather than abstract signs, the Egyptian system was extremely easy to memorise.

This is not really what we mean by a ‘cyphered’ number system, however, since, in the Egyptian system all quantities less than our *ten *are still represented by the one-symbol repeated the appropriate number of times. The Greeks took the ‘different symbol’ principle much further by introducing *single* symbols for *all quantities *greater than one and less than first base, as we ourselves do. Thus our ‘*four*‘ is not represented by a plurality of one-symbols such as * l l l l *but by a

*single*symbol,

*δ*. We are so used to this principle that we do not realize what a significant departure it really was. The Greeks managed this by making their alphabetic letters double up as numerals and so

*β*, the second letter, became ‘

*one-one’*or our ‘

*two*’.The 27 letters of the Greek alphabet, which they took over from the Phoenicians, were divided up into three sets of nine, the first set for quantities up to, and including, our 9, the second set for our 10, 20… 90 and the third for the hundreds. Various artifices such as having a bar above the letter-number enabled one to extend the system beyond 900 but the Greek alphabetic system could not be extended indefinitely in the way that ours can — because the Greeks did not hit on the idea of place value and positional notation. Archimedes himself, the greatest mathematician of the ancient world, felt obliged to write a treatise,

*The Sand Reckoner*, to argue that, in principle at least, all the grains of sand in the world could be numbered ― but even he never hit upon the stratagem of place value.

The great advantage of the Greek alphabetic system was its conciseness. The economy of the Greek, and later the Hindu-Arabic number system was, in its day, as important as the miniaturisation of the components of contemporary computers that has revolutionised the world of communication technology: saving space for the recording of data has been and remains one of the most important of all human concerns.

*SH 23/11/17*

## Extreme and Mean

**‘Extreme and Mean Ratio’**

The Greek geometers never speak of the ‘golden ratio’ and the first recorded use of the term is as late as 1835 ─ when Ohm referred to it as the *goldener Schnitt*. Nor does any ancient Greek give a numerical value for what we now know as phi or *Φ*. What we *do *find in Euclid and other ancient writers is repeated mention of a certain manner of dividing a line segment in “extreme and mean ratio”. **Euclid VI Proposition 30** shows you how to do this. In our terms, this method of division results in “the ratio of the larger to the smaller part of the line segment being equal to the ratio of the whole to the larger part” i.e. *a:b = (a + b) : a *where *a > b *.

← *(a + b) →
*

**———————————————————————**

←

*a →← b →*

Why was this important to the ancient Greeks? Not apparently because of the supposed aesthetic properties of the associated ‘Golden Rectangle’ (formed by making the smaller portion into one of the sides). Although it is sometimes claimed that Phidias used the Golden Section in some of his Parthenon statues this is mere speculation; it was only Renaissance painters and architects who superimposed the proportions of the golden rectangle onto the human figure as in the famous Leonardo da Vinci drawing and claimed there was something especially beautiful about the ‘divine proportion’, as they called it.

Nonetheless, to judge by the number of theorems relating to it in Euclid and numerous references to it in other extant ancient manuscripts, the ‘section’, as Proclus calls it, was famous. So why *did* the ancient Greek mathematicians consider the division of a line in ‘extreme and mean ratio’ significant? Because it was a prerequisite for the ruler and compass construction of a regular pentagon (five-sided figure with all sides and angles equal) and thus for the construction of the pentacle (regular pentagon within a circle) and the starry pentagram (five-pointed star). The pentacle already had a certain history as a ‘magic symbol’, being originally associated with the ‘morning star’ (Venus), and this esoteric reputation has lasted right up to the present day ─ Dr. Faust uses it and so do some contemporary Wicca groups. In ancient Greek times the pentacle had a more respectable, but still somewhat offbeat, reputation since the Pythagoreans, originally a kind of scientific secret society, used it as a sign of recognition amongst the Fraternity ─ compare the Freemason handshake. They sometimes put letters at each point of the five pointed star and these letters spelled out the Greek word for health (*u**g**i**e**i**a*) ─ so it was a sort of “Good Health to you, fellow Pythagorean” message.

But the pentagon had a more serious meaning still for educated Hellenistic Greeks and Romans. Although he did not invent them, Plato was an ardent propagandist for the importance of the regular solids, still called Platonic solids in his honour. For Plato, shape was more fundamental than substance and the supreme shapes were the perfect forms of geometry such as the circle and the regular polyhedral. These ideal Forms were changeless and harmonious whereas everything on the terrestrial physical plane was erratic and unpredictable. The five Platonic solids, which Plato identified with the four elements, Earth, Air, Fire and Water (plus a subtle fifth element Ether), had much the same status as the elements of the Periodic Table have in our eyes today. Indeed, it would hardly be going too far to say that, for Plato, these ideal Forms were cosmic computer programmes while the entire physical world consisted of the fallible execution of such programmes, software compared to hardware, genotype to phenotype. In consequence, it was very important for Platonists to know how to construct these forms, if only in imagination. The five solids are:

- The
**Tetrahedron**(four triangular faces); - The
**Octahedron**(eight triangular faces); - The
**Cube**(six square faces); - The
**Dodecahedron**(twelve pentagonal faces); - The
**Icosahedron**(twenty triangular faces).

Euclid concludes his great work with Book XIII which is entirely devoted to the construction of the five Platonic solids. Although Euclid is generally regarded today as the originator, or at any rate greatest early expositor, of the axiomatic method, this gives the modern reader the wrong impression. Today, the axiomatic treatment of a mathematical topic implies complete disregard of practicalities and ‘realistic’ concerns, but Euclid always has his eye on the actual construction of figures inasmuch as this is feasible. The very first Proposition (Heath calls ‘theorems’ *Propositions*) of Book I is “*On a given finite straight line to construct an equilateral triangle”*. And the penultimate Proposition of his *Elements *(Book XIII. 17) tells you how to *“construct a dodecahedron and comprehend it in a sphere”*. To be sure, this construction is so complicated, likewise that of a icosahedron (20-sided regular polygon), that one is hard put to follow the steps in the argument, let alone produce an actual model in wood or metal. Nonetheless, the mathematical presentation is not abstract in the way that, say, a theorem about Baruch spaces in modern mathematics is.

Such an approach is absolutely in line with the Platonic philosophy. For Plato was not so much an Idealist as a Transcendental Realist: his Ideal Forms were *more*, not less, real than actual artifacts while not being absolutely divorced from material things either. As certain Sophists in Plato’s own time observed, the figures of geometry, when drawn, did *not* have all the properties accorded to them by geometers: points on an actual circumference were not always exactly equidistant from the supposed centre, tangents cut a circumference in more than one point &c. &c. “Yes,” Plato might have replied, “but the drawn circle is not the circle of geometry, only a tolerable imitation of it. The *true *circle and true tangent, of which our human imitations are derivatives, really do have all the properties we ascribe to them, such a tangent really does touch the circumference at a single point only.”

It is interesting to note that Book XIII concludes with the dodecahedron rather than the icosahedron (whose construction is even more complicated) ─ the final Proposition 18 deals with the relations between the entire five Platonic solids and proves as a sort of coda that they are the *only *possible regular solids. The reason for terminating with the dodecahedron is most likely because the dodecahedron was traditionally associated, not with the four *earthly * elements, but with starry matter which was considered to be different from, and superior to, earthly matter. (Tradition has it that the Pythagoreans were especially delighted with their discovery of the dodecahedron and sacrificed a hundred oxen to celebrate the occasion.) And, as stated earlier, the division of a line ‘in extreme and mean ratio’ is essential for the construction of the regular pentagon which is itself essential for the construction of the dodecahedron (since all the faces are regular pentagons).

This may go some way to explaining why the ancients had a particular veneration for the ‘section’. Moreover, Allman makes the interesting suggestion that what we call phi, the golden section, was the very first irrational (the Greeks would have said ‘incommensurable’) to be discovered, rather than √2 as is today usually assumed. This would explain the mystery and slightly sinister glamour attached to figures incorporating the golden section such as the pentacle; for the discovery of incommensurables was, as we know, extremely disturbing for Greek mathematicians and philosophers alike. The Pythagoreans seem to have shifted from an attitude of hostility towards irrationals/incommensurables to one of veneration, at least as far as Phi was concerned since they eventually adopted the pentacle as a sort of logo.

Did Euclid have what we might call a philosophical, almost a quasi-religious, aim in giving the ancient world such a detailed exposition of the *Elements *of geometry? This was certainly the view of Proclus who wrote a commentary on Euclid in which he claimed that Euclid was himself a faithful follower of Plato and that “it was for this reason he set before himself, as the end of the whole Elements, the construction of the so-called Platonic figures”. Heath rejects this out of hand, arguing that Proclus was a biased source since he was himself the leading Neo-Platonist philosopher of his time and keen to claim Euclid as one of his own. Nonetheless, there can be no doubt that philosophical Platonism was inextricably mixed up with late Greek higher mathematics and Heath himself admits that “it is most probable that Euclid received his mathematical training in Athens from the pupils of Plato”. Whether Euclid was himself a Platonist is unknown but he seems to have faithfully transmitted to posterity not only the discoveries of Platonist (or Pythagorean) mathematicians but their overall ‘view of the world’. We do not today consider Book XIII to be the most important part of the *Elements *and usually single out the ingenious treatment of the problem of incommensurables in earlier books because this treatment anticipates the 19^{th} century approach to irrational numbers as pioneered by Weierstrass and Cantor. But the *Elements *was not just an exercise in pure mathematics; at any rate for many later Greek mathematicians, it was a sort of technical preamble to Platonic cosmology as laid out in the *Timaeus*. Kepler, to whom the Alexandrian cultural ambiance of Euclid’s day would have been most congenial, made a persistent attempt to match the orbits of the planets to the outlines of the Platonic solids and, incidentally, singled out the ‘division in extreme and mean ratio’ as the ‘chief jewel of Greek geometry’, on a par with the Pythagorean theorem itself. Although for a long time it was fashionable in scientific circles to look down on interest in the Golden Section as the affair of aesthetes and mystics, it is now known that one version of it, the Golden Angle, does have some importance as a ‘close packing constant’ as Irving Adler relates in his latest book on Phyllotaxis, or Leaf Arrangement. *SH 25/09/17*

## GESTURE NUMBERS

Gestures have been used from time immemorial either as a rudimentary number system, or as a supplement to spoken or recorded systems. Claudia Zaslasky (*Africa Counts*, Lawrence Hill 1999), gives visual examples of a ‘gesture for “six” ’ from Rwanda and a Xhosa woman showing by gesture the number of her children. Zaslasky notes that “in some African societies finger gestures have equal status with the spoken numerals and constitute a proper system of numeration which may or may not agree with the spoken words” (*op. cit. *p. 37).

I would guess the original ‘symbol’ for zero was something like the double open handed gesture that hunters still use to indicate that they have caught nothing that day. This gesture, common amongst country people in the South of France, does not quite signify “nothing” in the absolute sense, but rather “Nothing where something was to be expected” — which is somewhat different.

“In those systems that build by addition to five, counting usually starts with the little finger of one hand and proceeds by the addition of the appropriate fingers in sequence until five is reached. This number is generally denoted by a closed fist. For six, the little finger of the other hand joins in the counting, and the fingers of the second hand are used in the same sequence as those of the first” (Zaslasky, *Africa Counts *p. 49).

That gestures directly gave rise to full-scale finger counting seems unlikely : the sophisticated finger counting systems such as the Venerable Bede describes in his 8th century treatise *De computo vel loquela digitorum *(“On calculating and Speaking with the Fingers”) must surely have developed *after* an advanced spoken number system. No one in their right senses would use finger counting *alone* to represent really large quantities : what generally happened is much more likely to have been a combination of various systems, gestures, spoken words, the use of object numbers alongside recorded numerals and so on. Zaslavsky says that the Arusha Masai of Northern Tanzania “rarely give numbers without the accompaniment of finger signs” (*op. cit.* p. 248). Different ethnic groups had different ‘cut-off points’, most ending with our **50** at most while in the Luo system “there are no gestures for numbers beyond **19**” (*op. cit.* p. 254).

The, at first rather surprising, fact that African languages are predominantly base-five (rather than base-ten) suggests that ‘gestural number systems’ predated written and even spoken ones. Tylor writes: “Word-language not only followed Gesture-language, but actually grew out of it” (Tylor, *Primitive Culture*)

*SH 08/08/17*