THE PHYSICS OF INFINITY

Zeno said that, if Achilles was to catch the tortoise, he must first cover half

of the distance that separates them then half of the remainder, half of that and so

and infinitism. That is absurd, said Zeno, for an infinite series is by definition a

series that cannot be completed. Few will follow Zeno in concluding that time is

contractor and must be unreal, but the argument is a half plausible reduction ad

absurdum of the supposition that even finite interval is composed of infinitely many

parts if it were Achilles would indeed have to do the impossible and complete an

infinite task. Zeno’s paradox of the arrow invites a related conclusion, that space

and time cannot be composed of points and instants without extension. At each

instant, a supposedly moving arrow would cover no distance at all and would be

intrinsically as an arrow at rest. But the distance travelled in a time is the sum of

the distances travelled in its ultimate parts and, if no distance is travelled during

any of these parts, none is travelled during the whole interval. Nobody thinks that

arrows cannot fly but did Zeno refute the idea that space and time are composed of

points and instants without extension?

It is widely accepted nowadays that Zeno’s first argument rests on a wrong

definition of infinity: and set is infinite when it has as many members as a proper

part of itself; some infinities cannot be completed and others can as movement

proves. The second argument is answered by the calculus and its formulation

without dubious dependence on the infinitely small. Speed at an instant is the limit

of speeds over intervals that include the instant. Distance travelled tends to zero

as shorter intervals are considered, but speed need not. These are two victories of

THE PHYSICS OF INFINITY 06/01/12 1

many and these victories are now complete. Or so thought the logicists.

Weierstrass analysed puzzling talk of infinitesimals into plain talk real numbers and

limits Cauchy and Dedekind defined real numbers as sets of rationals and the

logicists reduced all numbers to sets. The application of mathematics in natural

science is not a further problem. The scientist postulates that certain relations of

places, times and intensive magnitudes are isomorphic with relations of real

numbers and their sets. He may be right and he may be wrong, but the pure

mathematician is only obliged to furnish the concepts in which conflicting theories

may be written. That is done.

Great chapters of mathematics, certainly, but a complete and definitive

theory of the continuum? There is a difficulty: arithmetic is much less mysterious if

we are constructionists, but constructivism and continuum are not happy together.

Constructivism says that, although we can define ever longer lists of real numbers

by applying the diagonal procedure, there never is, and never could be, a

completed list of all the real numbers. There is only a potential infinity, a rule of

definition that can be iterated as often as we please. However, physical space

does seem to constitute a completed infinity. Each constructive real number is

matched by a point on a line and field theory says that all of these points are

occupied not merely possible places without real existence. There is an element of

physical reality for each of the constructive real numbers. Moreover, further points

stand in the diagonal relation to the totality of constructive real numbers. Cantor

was first led to his transcendental sets by the continuum and the advance seems to

be unavoidable. We cannot make definite descriptions of all these points and real

numbers, but our limitations are irrelevant. The constructive totality is completed in

THE PHYSICS OF INFINITY 06/01/12 2

the material world and its completion requires even larger totalities beyond. Nature

does not wait on what men can describe. The constructivists has a second reason

for concern, even if the physics suits. His philosophy confines mathematics to the

possible doings of Turing machines, which are defined in good part by the banning

of completed infinites: the Turing machine cannot complete an infinity of tasks in a

finite time, either by working at unbounded speeds or by doing infinitely many

things at once, cannot have been working forever, cannot embody an infinity of

information or instructions, either by occupying an infinite volume or by reading

infinite detail from a grain. But it may be said that, though neither man nor his

computes can escape these constraints, a possible being could. Mathematics

should incorporate these possibilities, as Cantor did, not itself to a mere part of the

logically possible.

The constructivist will hope to show that natural science spurns these

supposed possibilities for good reason, even, perhaps, that physics does not really

need completed infinites in the continuum. Can these hopes be realized? The

paradoxes of Zeno do not prove that any of these completed infinites is impossible.

Nor does the empiricism of David Hume, whose writings on space, time and

divisibility move without warrant from the impossibility of our seeing certain things

to their impossibility without qualification. Kant’s contributions, though more

influential, are hardly more compelling.

An infinitely old universe means a completed infinity now, which Kant

thought was impossible for the same reason as Zeno. His argument for the

impossibility of a beginning assumes that, if the history of maths had a beginning,

THE PHYSICS OF INFINITY 06/01/12 3

there was a time before that history began. That is impossible, says Kant, because

there could have been a reason why matter appeared at just that time and not

earlier. Kants’ assumption is surely wrong: times are constructs from the temporal

relations of events and, if there are no events before a time, there are no times

before that time; the beginning of the material universe is the beginning of time

itself. However, that is hardly the end of our difficulties. A beginning of all things

cannot be explained by the previously existing and we wonder how, if we cannot

exploit that resource, the nature or existence of a beginning can be explained at all.

But the understanding is no better suited by a universe that has always been: each

state follows from antecedents, but why should this of all possible series be actual?

There is a mystery if there was a beginning and there is a mystery if there was not.

Kant’s arguments are uncompelling, but a problem remains: finite or infinite, why

this and not another history of the universe?

Kant’s attempted solution is a failure. He thought that descriptions of the

infinite could only be interpreted without absurdity as descriptions of the merely

potential; claims about an infinitely old or infinitely large universe turned into claims

about the future, about the discoveries that people would make on investigating

more and more extensively. The actual infinite disappears. Investigators, may

learn more and more, but their fund of knowledge is never more than finite. There

is at most a potential infinity, an unending growth in the finite as time passes. Is

that credible? Phenominalism says that every description of the material world can

be analysed into a list of conditional truths about experience, a distinctive set of

what would be perceived if.5 The stock objections confuse Phenomenal with

nearby errors, but Kant’s application of the doctrine depends on one of these,

THE PHYSICS OF INFINITY 06/01/12 4

errors, the confusion of explicit and conditional definition. Let us agree that, if we

put aside some quibbling, a bachelor is by explicit definition an unmarried man.

There is not an independent realm of facts about bachelors, floating free of truths

about men and marriage. Consider for contrast a partial analysis of a claim about

solubility, “if this lump of salt were immersed in water, it would dissolve”. What

does that mean? We described the fine structure of water and of salt, describe the

result of putting a thing that satisfies the first description in a thing that satisfies the

second and find that laws of nature entail solution: if properties are continuant

developed invariantly, that is what we come to. This definition does not abolish the

internal structure of salt or water, leaving only the obvious displays of solubility. No

more does the analysis of all the material would into conditionals abolish that

world. To suppose that is to misunderstand entirely the purpose that

Phenomenalism can sensibly be supposed to serve. If this is right, every

description of matter can be analysed into a number of conditional claims about

experience, claims about what would be sensed if states if matter are identified by

their relation to experience, but we do not thereby abolish all but experience. That

is sober Phenomenalism, however, and sober is not to Kant’s purpose. There was

a world before life, logically independent of all categorical truths about experience

and not reducible to a potential infinity insistence on the merely potential character

of infinity must not lapse into Idealism, a temptation that will be hardest to resist

when considering the continuum.

Kant offered no credible reduction of completed to potential infinity and begs

the question, merely assuming that a completed infinity is impossible. Neither Kant

THE PHYSICS OF INFINITY 06/01/12 5

nor Zeno will convert the disciples of Cantor. Can more recent philosophy do

better?

ARGUMENTS FROM MEANING

Putnam and Dummett argue that, in trying to describe the transcendental

infinities, we try to say more than words can possibly mean. Their arguments

depend on a compelling premiss, that people cannot complete certain infinite

tasks, do not quite price all that we need. It is argued that, since we cannot

complete infinite surveys, we cannot even grasp the idea. That is an argument

from an assumption that might be disputed, not a proof absolute: if we could finish,

we would understand. Now, the prevailing philosophy of science says that only

analytical truths are absolutely invariable and available a priori; other propositions

are answerable to experience. People who believe that should also believe that

our inability to complete infinities is a mere contingency, a contingency that may be

abandoned by the science of tomorrow. Russell thought that our inability to

complete infinite tasks was a “merely medical” limitation and, though Dummett was

surely right to disagree, we cannot argue in circles. We must prove the absurdity

of certain completed infinities before Putnam and Dummett can begin, but their

argument is then redundant

The impossibility of these suppositions can only be proved properly if we

consider how very general the impossibility is. Our inability to complete certain

infinities is no more a peculiarity of Mankind than our tendency to fall in

gravitational fields. Neither atoms, galaxies nor all the universe can finish the

THE PHYSICS OF INFINITY 06/01/12 6

infinities that we cannot. We have to do with truths about all temporal things, of

which our limitations are a tiny instance, and the depths of their inevitability can

only be understood in tandem with their generality. The only way to the required

conclusions is proof that certain constraints follow from inalienable properties of

Time and our a priori requirements of explanation, for an argument that really

depended on the peculiarities of Man would have no answer to Russells’ objection.

Let us begin with the most transparent example.

TWO ABSURDITIES

Russell argued that someone might complete an infinity of tasks in a finite

time by working ever faster. He might improve with practice so rapidly that each

task occupied only half the time of its predecessor. If the first needs a minute, an

infinity are completed in two. Thomson tells a story with the same conclusion,

imagining that someone might turn a light on and off infinitely many times in a

minute. The trouble with both of these fantasies is that a body must travel an

infinitely long path in a finite time. The protagonists do not finish at an infinite

distance from their starting places, but their paths are infinitely long nonetheless.

Pure mathematics has not objection to infinitely long paths but Russell’s and

Thomson’s fantasies require more, a continuous path between points at infinite

distance. That is impossible. If distance is measured along the path, not directly,

there are points on the line at an infinite distance from the origin. One part of the

line comprises points at a finite distance from the origin and one part comprises

points at infinite distance. The set of points at finite distance from the origin has an

upper bound in every member at infinite distance, but what is its least upper

THE PHYSICS OF INFINITY 06/01/12 7

bound? Not a point at finite distance from the origin, since every such point is

followed by others at a greater finite distance. Not a point at infinite distance, since

every such point is preceded by others. The line cannot be continuous, since a

bounded and continuous interval invariably has no least upper bound, but

discontinuity excludes analyticity and analyticity is defining characteristic of Time;

Russell and Thomson’s stories cannot, therefore, not descriptions of developments

in persisting things. One symptom of this disconnection is that, although the switch

on Thomson’s lamp must at the end be either up or down, nothing determines

which.

The hands of the people in Russell’s and Thomson’s fantasies are

supposed to cover infinite distances while the hands on their clocks move only

finitely far. We might try, instead, to imagine time infinitely far off, conceived

perhaps as the time at which immortals have the results of their infinite surveys, a

time that stands to ordinary times as omega to the finite ordinals. Our fantasies

exited, we might even try to imagine Times in which soundings of the hour have

the structure of any infinite ordinals. However, these fantasies are absurd for the

same reason as Russell’s and Thomson’s. The mathematician at omega

supposedly comes to a certain judgement after and in consequence of his infinite

survey, but we have imagined away both Time and Cause. Time is by definition

the dimension on which analytical laws unfold, but no such laws join the

mathematician’s supposedly earlier calculations to his state at omega. Causality,

too, is a construct from analytical connections. At best, we are imagining a

universe with no connection to ours, a universe that begins in a miraculously

ordered state. This absurdity is not, of course, to be confused with misleading but

THE PHYSICS OF INFINITY 06/01/12 8

entirely possible descriptions of what happens as magnitudes tend to infinity. We

do sometimes say casually that, if something grows exponentially, it is infinitely

large at infinity. However, this sloppily expressed description is only true when

interpreted as a generalization over finite sizes: however large a finite size you

choose. I can nominate a time at which the thing is larger. The function says

nothing about values at arguments beyond the finite: we might choose anything we

please and define the ordering in which 0, 1, 2, - - - - are followed without

intermediary by that thing. Our conclusion stands: states at infinity cannot develop

analytically from happenings earlier.

These are the first and simplest of physical constraints on mathematics. A

supposition is logically possible, but its impossibility follows from the nature of time,

which is analytical connection. Other constraints follow from the stronger demand

that all is to be explained by the unfolding of continuance properties.

COSMOLOGY

We catch the end of what someone is saying: “ . . . 5, 4, 3, 2, 1, 0. Ah!

Finished at last” Noticing our curiosity, the man tells us that he has just finished

counting the natural numbers backwards. Absurd, but why? Well, there is always

an explanation of ordinary counting – of how the teller come to start counting on

Wednesday, the growth of commerce and accounting the emergence of life.

Moreover, the teller’s rule builds more complex symbols from simpler: as we his

beginning to count, there is a steady diminution of the details that need to be

explained. There can be no explanation of how the backwards counter came to be

THE PHYSICS OF INFINITY 06/01/12 9

counting, however, since he always was. Moreover, simpler symbols are replaced

by more complex as we delve earlier in the hope of explanation. We are doubly

frustrated: the search for explanation takes us earlier and earlier, but an explicable

beginning never comes and, indeed, there are more details to be explained the

earlier we look. All would be well if the future could explain the past, if we could

make a backwards explanation to fit the backwards counter, but that is impossible.

We are affronted less by the completed infinity than by the existence of an

infinity fortuitous thing, by the infinity of details for which there is no explanation: a

chaos of particles that been whirling forever is surely much less offensive. Even

classical mechanics can after a fashion explain the history of such a world. Nearly

all of the most ridiculous fantasies are realized, so long as they are consistent with

the laws of acceleration. There is an eternal shuffling of positions and velocities,

which produces almost every possible distribution of things at some time and,

indeed, at infinitely many times. Some sorts of thing are exemplified by more

patterns o particles, are “more easily made” and appear more frequently, but

almost everything has its day. Almost even this most liberal theory of coming to be

cannot admit quite everything. The emergence of a backwards counter cannot be

explained in this fashion, since he was always there. Moreover, the details to be

explained approaches infinity as we look further backwards. The backwards

counter is, perhaps, the most outrageous example of a consistent but utterly

inexplicable being.

There is another contender for that title. An infinitely large computer would

need an infinity of the right parts in the right relations, but the emergence of such a

THE PHYSICS OF INFINITY 06/01/12 10

thing could never be explained by an accumulation of finitely many parts in a finite

intervals. Either there was always an infinity of parts in right relations or infinitely

many parts were arranged correctly in a finite time. The machine is infinitely

improbable.

The backwards counter and the infinite computer for the rationalist are

merely extreme examples of a universal problem, the problem of why there are

certain sorts of thing and not others, distribution as they are and not otherwise.

Leibniz says: “although we might account for the present book by the book which

was its model - - - we may always wonder - - - why books exist at all and are thus

written. What is true of books is also true of the different states of the world - - to

whatever anterior state you may go back to you will never find there a complete

reason why there is any world at all, and why this world rather than any other.

“The orthodox empiricist thinks that Leibniz asks for more than can possibly be

given. We must put at the foundation of science laws” without our being able to

state a reason why any one of them (. . .was. . .) preferred to the others. It is the

same with possible beginnings, he says: explanation can only derive later from

earlier; there can be no explaining why things were as they were at the origin of all

things, when there was nothing earlier to do the explaining, or why the entire

history of the universe is thus and not so. A backwards counter would be a

disappointingly complicated postulate but the only real objection to such a being

would contradict experience so violently. In its aspirations, if not yet in its

achievements, modern cosmology takes the side of Leibniz. The easiest but least

rewarding conception of propensity moves as little as possible from the

inadequacies of classical mechanics: the continuant properties of things are

THE PHYSICS OF INFINITY 06/01/12 11

propensities, but we must accept why there are just these sorts of things with just

these propensities. That is too sell propensities short, for modern physics offers a

theory of propensity to existence long-lived particles and bodies emerge from the

field. That is only the emergence of continuants from a field, not creation ex nihilo,

but vast new volumes of field are created as the universe expands. The

conservation of energy is not violated – gravitational energy is negative and

conservation requires only that be positive and negative energy always be created

in equal amounts. It may be cosmology must still said that posit something to

grow, some seed from which the laws of propensity make a great world, but that is

wrong too: the universe emerges from nothing at all under the very same laws as

permit the energy of an ordinary field to rise. We might expect that a tiny departure

from zero will stay tiny or be followed by relapse, but that is not so: when there is

only a tiny region of energy, which is the only state compatible with fluctuation from

zero, the result is an expanding universe. I repeat that all of these developments

are subject the same laws as obtain elsewhere including the law that energy is

conserved. No longer need we tolerate a physics in which later states can only be

derived from earlier states no less arbitrary: quantum cosmology describes the

invariant development of propensity from the only possible beginning and thereby

distributes probabilities over all possible over all possible histories of position and

velocity. Logically consistent alternatives can of course be made by putting

arbitrary variations or addenda in the invariant theory superposed other stuff that

had always existed and we might favour certain developments from the very

beginning. An invariant theory can always be wrecked with gratuitous variations

and addenda, but quantum cosmology left alone is the only decent theory of

existence.

THE PHYSICS OF INFINITY 06/01/12 12

 While the modern cosmology can only regard Kant’s discussion of the First

Antinomy as a noble failure, any who read “The Radical Origination of Things”

must surely be started by the prescience of Leibniz, who understood the problem

perfectly and even the outlines of its solution. Cosmology requires a principle that

resolves without arbitrariness the conflict of different propensities to existence: “all

things which are possible tend will equal right towards existence in proportion to

the quantity or reality which they include.” Leibniz even appreciates that outcomes

must be decided by identifying extreme cases, comparing the rules that settle

conflicts of propensity with rules that achieve the greatest on the greatest for least

in geometry and mechanics. Admittedly, this cosmology is set in a theology that

few of us can still believe: propensities are forces that attach to ideas in the mind of

God, conflicts among these resolved by His choosing the best of all possible

worlds – which, is identified, implausibly, with the world in which the greatest

number of possibilities in principle, the principle of sufficient reason which he could

have invoked more plausibly. If all possible developments from the beginning are

accorded the same weight in Feynman’s grand sum of amplitudes, if nine is

favoured arbitrarily, the principle of extreme action dictates a distribution of

probabilities over possible histories that seems to describe the universe observed.

Qualifications notwithstanding, “The Radical Origination of Things” is the finest

anticipation of physics that metaphysics can boost.

These are further arguments for the conception of necessity advocated in

the last chapter. If all necessity is tautology, concept that logicists extended to

incorporate truths of the predicate calculus and set theory, Russell was entirely

THE PHYSICS OF INFINITY 06/01/12 13

right that infinite tasks might be completed in a finite time. Many people are not at

all happy about that; they feel that Russell’s storey, though logically consistent, is

still utterly absurd. They may look to Kant for a more exclusive conception of

necessity, but his cardinal arguments assume that a completed infinity is

impossible. So, too, do more recent arguments from possible behaviour and

meaning. Many, I suspect, think that Russell’s fantasy is ridiculous but are not

quite sure why. A majority, further suspect, continue to accept that geometry is

contingent, that only observation can decide among a variety of theories with the

very same status a priori. Some will say that, although there is the much truth later

in the Critique, Kant’s Transcendental Arithmetic was a failure. Indeed it was, but

there is a better argument for a certain geometry: only a variably curved space can

make an invariant theory of the forces, a theory that does not add unchangeables

with arbitrary powers to the analytical developments of changeables. This better

argument rests on the same premiss as our dismissal of various completed

infinities. I have tried to say why the completion of certain infinities would be very

absurd, not a merely medical impossibilities. The fantasies just dismissed

contradict an unnegotiable requirements of explanation, analyticity and invariance.

Russell’s and Thomson’s tables and times infinitely far off would abolish the

defining condition of a persisting thing. Both contradict the demands of explanation

very blatantly, which is why so many people feel accept not really possible and

accept the constructivist’s constraints. However, we can hardly assign a different

status to the obvious and the hidden consequences of a principle if the completed

infinites are impossible in some sense, alternative to Euclidean geometry are

impossible in the same sense. This conception will come to subsume all logical

and mathematical necessity, but that is for the future. For now, let me say only that

THE PHYSICS OF INFINITY 06/01/12 14

certain principles imply both the constructivists constraints and the right physical

geometry. Some of the former depend on just analyticity, but others depend on

invariance too. We cannot, therefore, both explain the impossibility of completed

infinities and continue to accept the empiricist’s account of geometry. The

necessity forbids both. Science has no need of infinite age or volume, but it does

seem to need completed infinities in the continuum. What can constructivism say

about?

THE CONTINUUM

Modern science and mathematics often embraces without apology what

Zeno thought ridiculous, regions with magnitudes built from points with none. This

flat rejection of Zeno’s assumption carries the authority of the many distinguished

mathematicians who follow Cantor, but strike most laymen and some

mathematicians no less distinguished as preposterous in the stickler’s sense,

putting at the front what properly belongs at the end. Infinite decimals are

unreachable ideals, surely, not properties of real things. And real magnitudes are

not unions of these ideals, unions so huge that nothings are compounded into

somethings.

A construction of points from regions is not always a construction of the

infinitely refined from the tangible. Whiteheads method of extensive abstraction

tried to respect the epistemic order by defining points as bounds of rested regions

<x, y, z>, because the set of spheres with <x, y, z> at their centres defined without

circulating. For all its interest and influence, which included the inspiration of

THE PHYSICS OF INFINITY 06/01/12 15

Russell’s logical constructions, the method of extensive distraction is not rendering

of the infinitely sharp into the palpable. Whitehead’s elements are regions, not

points, but his regions are defined with infinite precision Numbers that identify the

boundaries of Whitehead’s regions are significant to infinity, our measurements of

a boundary only to a handful. There is a second difficulty: if points are to be

constructed from volumes, an infinite series of timer and timer spheres is needed; if

we are confined to the perceptible, that infinite series is not available.

A deeper theory builds the metric from topological relations. Brouwer

conceived space as an “extensive whole”, an initially undifferentiated expanse in

which ever smaller parts can be distinguished by the application of ever more

refined criteria. His scheme is not confined to the measurement of space and time,

but that is its most important application.

Bare right and touch convey some sense of congruence, which is helpful but not so

helpful as we should like. We distinguish places nearer to A from places nearer to

B and, if we push our hunches too far, even the modest predictions begin to fail.

We do better by measuring with a short stick, better still by using a long, finely

marked stick, our decimals are still only significant to so many places and many

THE PHYSICS OF INFINITY 06/01/12 16

desired predictions are still beyond us. The scientist devizes ever more elaborate

of ever sharper categories, but he never comes to a way of verifying that a

magnitude is measured by a certain infinite decimal.

Our knowledge can only be refined in stages and never attain the actually

infinite but what of things in themselves? Does the world become more definite as

we investigate or do previously existing properties determine our measures?

Usually the latter, which seemed for a time to require an actual infinity out there. It

seemed that, since we could in principle measure as sharply as we liked,

previously existence properties must determine the results if we measure to thirty

places, to a billion, to any finite number that we please. Those properties must be

defined unto infinity. There was an infinite series of possible measurements, each

sharper than its predecessor and no merely finite decimal can specify the results at

every stage of the series. Thus classical mechanics, but a splendid example of

Putnam’s shows why classical mechanics is wrong. If you take a large sample of

hydrogen atoms and measure the energy of each electron the results lie in a

narrow range. If you measure the place of each electron, the results entail that

potential energies alone exceed the total energies found on average in the first

experiment. The moral is plain: there are degrees of refinement at which results

are the artefacts of measurement, not evidence of the previously existing. The

point at which revelation turns to creation varies with the context, but is always

finite. Saving a most important extension for a couple of pages, we might say that

quantum theory replaces precise positions and momenta with spreads of

propensity over the various possible results of measuring. There may be no

propensity at all to measure shorter than one nano-meter or longer than two; a

THE PHYSICS OF INFINITY 06/01/12 17

description in meters is significant to twelve places. But there may be a half

chance of measuring shorter than 1.5 nano-meters and half a chance of measuring

longer; there is no significance in the thirteenth place. There turns out to be a

simple dichotomy: if energy at a given frequency can be had in whatever quantities

the physicist pleases, he can measure as sharply as he pleases; if energy at a

frequency is only available indiscrete quantities, there are limits. Shorter

wavelengths cause more disturbance. While long wavelengths give a fuzzier

image and do not determine position so precisely. These interferences cannot be

reduced by reducing energy at a frequency which only comes in multiples of a

smallest possible amount. Quantum theory follows when the same invariances

that classical mechanics demands of absolutely sharp magnitudes are demanded

instead of propensities.

These remarks may seem to exaggerate the indefiniteness admitted by

quantum theory. Putnam says that, while that theory sometimes puts discreteness

where classical physics had continuity, it holds to continuity elsewhere. Energy

only varies discretely at a given frequency, but energy itself, space and time are all

continuous.13 That is true in a sense, but leaves the misleading impression that

quantum theory agrees with classical physics that real things exemplify continuum

of distinctions. Quantum theory calculates probabilities by considering possible

paths through continuous space and each of these paths is defined with perfect

precision by mathematician’s real numbers, but the measurements to which

probabilities are ascribed are always measurements significant to just a few

figures. It is also said that, although quantum theory admits no states in which the

conjunction of position and momentum is defined more sharply than Heisenberg

THE PHYSICS OF INFINITY 06/01/12 18

allows, either by itself can be absolutely definite; either might be measured by an

infinite decimal significance …….? Place. This misconception …………… Position

by itself and momentum by itself can indeed be determined to any finite number of

decimal places in standard units, but neither can be determined to an infinite: since

sharper and sharper measurements of position require waves of higher and higher

frequencies, and infinitely sharp measurement would require waves of infinite

frequency. Perfectly sharp measurement of position requires the opposite

absurdity, waves of no frequency at all. Heisenberg’s is not the only relevant

constraint. Infinitely sharp positrons and momenta, though not forbidden by that

principle, are forbidden by a much more obvious consideration. Thus further

indefiniteness brings another. Positrons are never perfectly sharp and neither are

spreads of propensity: we cannot suppose that although the position of a photon

indefinite, its associated ᴪ - wave has a precise value at all on some points, For

example, we cannot suppose that propensivity show vanishes at π nano-meters

from the nearest electron and not before. Such a precise spread is only possible

when the origin of the photon is equally precise, which it never can be. The psi-

wave ever emanates from an origin whose place is only defined to so many

significant figures.

That is the first of two steps from excessive concern with ourselves. I said

that definiteness is sometimes the artefact of measurement, but that is to

concentrate misleadingly on one sort of case. Measurements are contrived

reductions in the spread of propensity, but most reductions of propensity are

neither designed nor observed. The received laws of quantum theory, without the

aid of any special assumptions about measurement, say when broader spreads of

THE PHYSICS OF INFINITY 06/01/12 19

propensity collapse into narrower. There are unseen collapses all the time, not

only when people measure, but none is described by an infinite decimal.

Collapses of which nobody is aware might contract the span of a magnitude and

infinitism but can never reach infinite sharpness.

The mathematics of this potential continuum” is to Brouwer’s, not Cantor’s.

Brouwer continuum is realized in the world itself and not only in our knowledge.

More refined measurements often reveal the previously existing but there is always

a limit. Whether it is measurement or the collapse of propensities uncontrived and

unobserved, there is always a point at which a sharper magnitude is a new

magnitude; physical magnitude is significant to more than a finite number of

decimal places. And that, submit is the only credible way to make a theory of the

physical continuum from Brouwer’s analysis. The intuitionists do not mention

quantum theory admittedly but they do leave an inevitable question unanswered:

our investigations are as Brouwer describes but what of the world investigated?

Kant’s second Antinomy finds a contradiction in both the possibility and the

impossibility of infinite division and we are offered the same solution as before:

reports of the physical world are reports of a merely potential infinity, a description

of what conscious beings would find on investigating further and further. That is

the incredible version of Phenominalism criticized earlier: the sensible version does

not erase the world before Man and it does not erase the minutiae that Man never

investigates. We tried to answer the questions raised by the First Antinomy in part

by disputing Kant’s assumption and in part by putting our hopes on the aspiration

of modern cosmology, to solve the remaining, real problems in a world that began.

Only physics can solve Kant’s First Antinomy and only physics can solve the hard

THE PHYSICS OF INFINITY 06/01/12 20

questions left by the second. Only quantum theory can make a ‘merely potential’

theory of the continuum without Idealism.

Quantum theory is often thought paradoxical, but Everett and his followers

can solve. The only serious problem, which is to say when spreads of propensity

collapse into greater definiteness. Other charges against the theory are

calumnies. Rather, quantum has all of the advantages over determinism and

definiteness. Its support from experiment is the best of any theory without

qualification and it is demonstrable that no deterministic theory can simulate its

predictions. Moreover, the theory is the superior of its deterministic rivals priori:

quantum mechanics can join with relativity to explain the nature of force and make

a Rational Cosmology. There are further advantages beyond physics, where the

theory belies its reputation and returns to common sense. Quantum mechanics is

happy that organisms should follow autonomous laws, consistent with but

dependent of the laws that govern their constituent particles. It embodies the idea

that earlier produces later and not the other way about, which classical mechanics

can make no sense of. Finally, the theory replaces the actually infinite distinctions

of determinism with Brouwer’s surely far more agreeable conception of the

continuum. Some people think indefinite magnitudes very odd and would be

happier if position and momentum were always defined sharply, but would they still

think the same on reflection? Wrist and elbow were about a foot apart before we

measured certainly, but are we so happy that distances are always significant, not

merely to a few geographic decimal places in millimetres but to infinitely many

places? A limit to these subtleties is surely more agreeable to Reason, not less

Quantum Theory realizes the philosophy of Aristotle, Kant, Gauss and Brouwer,

THE PHYSICS OF INFINITY 06/01/12 21

that infinity is never more than potential. Quantum Theory serves construction by

rebutting the charge of Idealism and transforming a mere philosophy into physics.

Constructivism serves quantum theory insisting, rightly, that and, indefinite

magnitudes are the realization of an commonsensical philosophy and, not at all off.

A physics set in Cantor’s continuum, with its infinities beyond all description, is

really much odder.

THREE ALTERNATIVES

That is why there are no actual infinities in the continuum. Other ways of

abolishing these infinities have been mooted, but one is not viable by itself and the

others are not viable at all.

As the curvature of spacetime varies, so too does definiteness; there is not

a universal degree of curvature, the same everywhere, and there is not a universal

degree of definiteness. Some people would be happier if that were not so, if the

limit of measurement were always the same and the oddities of quantum

mechanics could be forgotten; they would like to abolish the actually infinite by

postulating a definite number of indivisible steps from any one thing to any other.

They may be falsely encouraged by some developments in quantum theory itself,

which try to resolve a conflict with General Relativity by stopping calculations at a

certain bounded. One group of theories say that everything is composed of little

strings, about 10 -36 meters long, and that shorter distances have no objective

meaning.

THE PHYSICS OF INFINITY 06/01/12 22

Sakharou’s theory of gravitation comes to least lengths by a different route. Even if

the advocates of strong theory can keep their promises, which is doubtful, it would

be universe to rest our case against actual infinity on their hopes. The conjecture

elaboration of quantum theory, not an alternative, and offers no escape from

indefinite magnitudes: even if string theory is right, only particles measured with the

utmost precision have sharp sizes in Planck’s unit. Moreover, we must understand

quantum theory before we can understand the possible need of shortest possible

lengths; even if the consistency requires a shortest possible length, there is a less

disputable and logically prior reason for the impossibility of infinite division.

There may well be shortest significant distances in the continuum, but the

most obvious elaboration of this idea conceives travel as a series of discrete steps

from point to immediate neighbour and contradicts all geometry and physics.

THE PHYSICS OF INFINITY 06/01/12 23

The length of a hypotenuse in discrete space is the sum of the length of the sides,

contradicting Euclid far more blatantly than Labachewski or Riemann: equal

treatment of the intrinsically indistinguishable requires that c² = a² + b² in Euclidean

space, that a + b = c when the space is discrete. If steps can only be taken at

discrete times, particles can only travel at different speeds if the slower take longer

rests or the faster skips places. These consequences violate every requirement of

decent explanation. If the difference between two bodies consisted solely in the

difference between two discrete sequences, B is at p 1 at t1, at p2 at t2 1 . . . . ., B is

at p1 at t1, at p1 at t2 1 . . . . . ., there would be no difference in the motion of these

bodies at any time; their relevant difference would consist in their earlier states,

which have ceased to be. Discrete motion requires states to exert an influence

after their demise and contradicts locality. Discreteness also contradicts invariance

by requiring an absolute scale, the distance between neighbouring points, that can

only be constructed from topological concepts by invoking arbitrary objects.

Properties of space and time in themselves would exert an influence; laws would

not be invariant and could only be arbitrary. Worst of all, discreteness abolishes

continuant properties: only continuous developments are analytic.

Although there may be shortest significant length, space is certainly not

composed of discrete places. We may at first be tempted to suppose that one

entails the other: if the shortest distance is d, the possible positions on a line a re . .

. – 2 d, - 1 d, 0, 1 d, 2 d, . . . . . . If we try to measure by counting places, however,

the usual laws of addition fail2: a chain of steps from a to z might be

indistinguishable from a chain that lies in a straight line, for all that measurements

THE PHYSICS OF INFINITY 06/01/12 24

down to 10 -36 meters can tell, and yet the directly measured distance from a to z

might be more or less than the sum of these parts.

These exceptions to the customary laws of addition might be said to reveal that

some of the intervals not immediately distinguishable are shorter than others.

There is no objection, so long as we appreciate that such further divisions report

failures in the usual laws if addition, no more, and are not the beginning of c

system that might be extended ad infinitum.

The perils of excessive division are illustrated in a fascinating but flawed

study by Goodman. Indistinguishable or matching does not ensure identity, says

Goodman: x and y are identical when, for all ȝ, x matches ȝ if and only if y matches

ȝ. Granted that distinction, matching can be used to construct a space with familiar

relations. X is closer to y than to ȝ, when x and y share more matches than x and

ȝ. y is betwixt x and ȝ when x is nearer to y than to ȝ and ȝ is nearer to y than to x.

X is beside y there is nothing betwixt the two. There is even a metric: the distance

of x and y is the smallest numbers of steps from one to the other, all of which pass

from a magnitude to a magnitude beside.

THE PHYSICS OF INFINITY 06/01/12 25

 Goodman’s work suggests a certain theory of uncertainty in perception: the

x before us certainly matches the y before us, but we have no present access to all

of the qualities that either might match and cannot, therefore, tell whether x and y

match the same qualities and are identical. This account is certainly incomplete

and is perhaps not even coherent. Often, we are uncertain whether to declare that

two examples match. We cannot keep to cases we are sure of, because

Goodman’s rules of ordering require us to discriminate even more finely. Can we

always tell, or helpfully stipulate, which qualities stay the same as comparisons are

made? People required to make very fine comparisons will sometimes surely offer

different judgements in identical cases – cases identical in the only respects that

can explain what is happening. A better theory rests on two necessities, the first

being that distinctions in experience are only possible when there is an associated

different in possible behaviour. A psychologist wants to know how finely people

can discriminate among colours and pays well if they do well, judging success by

sharp tests of frequency. His subjects choose with more or less conviction for a

purpose, sometimes succeed, sometimes fail and understand what they are

supposed to be doing. But there comes a point at which they cannot tell whether

two swatches are the same colour and, if they do choose, admit to uncertainty or

even to choosing on a whim. If the psychologist then tells them to describe their

experience, never mind the rewards, his subjects are at a loss. “What seems right

is right which means that we cannot speak of right and wrong here”. The second

necessity is that, if we have to put magnitudes in discrete categories, we must do

so without identifying discontinuities. There is never an abrupt change – at π

nano-meters, say – with identity declared for certain when the distance is shorter

and difference declared for certain at π and beyond. There is rather a spread of

THE PHYSICS OF INFINITY 06/01/12 26

propensity over possible decisions that changes continuously with real magnitude.

That is not a peculiarity of Man. Whenever a continuous array must fall into

discrete alternatives – when some tiny balls are thrown at some slots, for example

– there is always a spread of probability over the marginal trajectories, never

discontinuity. Landé argued directly from the need of continuity in such cases to

the need for a physics of transitional probabilities. I am not sure that continuity is

obviously required in these cases and prefer to argue from the virtues of quantum

theory. However we argue, there are spreads of propensity at the margins. Often,

people are perfectly aware of that. I waver and am aware of my wavering. Can if

smell gas? I’m not sure, which describes my experience as well as I can. It is

sometimes said that although the strange states of quantum theory may be

tolerated among the imperceptibles of modern physics, they cannot possibly be

seen. That is not so: experience is replete with uncertain states.

We try to arrange the colours of swatches on Goodman’s rules, but our

decisions force us to say that some colours have changed. It may well be quite

arbitrary which changes we stipulate and, whatever we decide to say, our decision

may be no help at all in predicting further judgements of matching. At some point

the project presses distinctions too far. There is a spread of propensity over a

continuum, some chance that a match will be declared and some chance that it will

not, but nothing worthwhile can be made from a fall of the chances on one

occasion.

A discrete space as conceived by Zeno or Goodman cannot reconcile the

demands of science with blunt observation. Physics may demand a shortest

THE PHYSICS OF INFINITY 06/01/12 27

length, but the theory making that demand has already reconciled continuity with

the impossibility of infinite division. Quantum theory shows how completed infinites

are to be expelled from the continuum.

INTUITIONISM

Brouwer distinguished two sorts of real number. Some are defined by an

arithmetical rule at every point, but some are defined wholly or in part by “free

choices”. How is free choice related to probability? It is at least disputable

whether the more important of our free choices depend on chance: deciding to

become a doctor or choosing the path of virtue may be entirely determined by our

goals, abilities and characters. If we choose digits haphazardly, however, it is at

least very plausible that choices are a matter of chance. It seems unlikely that our

choices are random – we no doubt favour certain digits over others – but chance

does not demand randomness or equal chances. Now, there is no reason why we

should confine ourselves to chance sequences of this very special and rather

unimportant kind. Rather, let us consider every sequence that chance at least

helps to generate. Since quantum theory teaches that all measuring turns

eventually into chance trials our subject includes every application of real numbers

in natural science. We are considering not just sequences defined by capricious

choices or the fall of a die but a complete theory of the material continuum.

Concepts applicable to determined sequences are no longer appropriate

when sequences are defined in whole or in part by chance. Since we shall only

ever throw the die a few times if any and since we do not believe that some

THE PHYSICS OF INFINITY 06/01/12 28

deterministic mechanism would fix the outcomes if we did, it is obvious that we

cannot sensibly wonder what would have happened at the fourteenth toss or

measurement. Nevertheless, it follows from our rules that not both 5 and 4 will

appear and, of course, that not both 5 will appear and it will not. We must

distinguish a ‘p v ˜ p’ from the corresponding ‘ ˜ (˜ p . ˜ ˜ p). Many further

distinctions are demanded when functions of real numbers are introduced. There

is not a simple dichotomy: either two rules express the same function or they

express different functions. It may be that two rules necessarily define the same

range and it may be that two rules necessarily define different ranges, but it may

also be that two rules permit both identity and difference. “Free Choice” or chance

sequences generate a further distinctions and related theorems. However odd

some of these results may seem to the conservative, there is no conflict with

classical analysis: the different schools are investigating different concepts. That is

true even in the strangest seeming of cases. The intuitionist startles the

conservative by insisting that every total function is continuous, but his claim does

not contradict any theorem of classical analysis and is entirely plausible when

concepts are understood as the context demands. A function that associates

something with a series of chance outcomes, including a series of measurements,

cannot rely on some characterization of an infinite set. I f rule is to define a

function, f (x), it must associate each initial segment of x with an initial segment of f

(x); moreover any initial segment of the latter, however long, must be associated

with some initial segment of x. It is at least plausible that, if a function in this sense

is defined at every point, it must be continuous: no initial segment determines

whether a number is precisely r; if there is a discontinuity at r, approximations to r

THE PHYSICS OF INFINITY 06/01/12 29

do not yield approximations to f( r). And Brouwer does indeed prove that, when a

function are understood thus and everywhere defined, everywhere continuous.

All such principles and theorem are true of the physical continuum, as the

theorems of classical analysis are true in their proper domain. It is sometimes

thought that, although the intuitionist’s doctrine of the continuum is very appealing

observation of its strictures would mutilate the beauties of classical analysis. “ . . . .

the mathematician witches with pain as the greater pore of his towering edifice

which he believed to be built of concrete blocks dissolves into mist before his

eyes”. But there is really no cause for distress. Intuitionism does not force us to

abandon classical analysis and adopt a less pleasing theory of the same subject.

Each mathematics has its place: classical analysis is the mathematics of

determinate rules and intuitionism is the mathematics of chance sequences,

including of measurements. The intuitionists were wrong to context the classical

theory of arithmetically defined real numbers, but they were right about the

application of real numbers in science. There is no need to abandon the

achievements of either school.

Neither philosophy, pure mathematics nor logic can be themselves prove

the absurdity of a completed infinite, of a universe infinitely large or infinitely old, of

infinite tasks completed in plain time or at a time beyond, of infinitely divisible

space or time. Some of the attempted proofs rest on tedious errors, some on

plausible principles that disciples of Cantor deny flatly, plausible principles not quite

plausible enough to serve as axioms. Finitude can only be proved in natural

science, in quantum theory, cosmology and straight from the concept of a

THE PHYSICS OF INFINITY 06/01/12 30

persisting thing. That is not to abandon these truths to contingency: finite age, size

and divisibility are required by the only physics that can explain things.

Some readers will not doubt grumble that logical and mathematical

necessity cannot be founded on physics. I fear the following chapter will give them

a lot more to grumble about.

THE PHYSICS OF INFINITY 06/01/12 31