Why are the laws of physics the way they are?

Stephen Ames 2
1 Introduction

There are various scientific approaches to answering this question. For

example, arguing to the laws from something physically more fundamental.
String theory is pursuing that kind of argument. Based on the mathematics of
David Finklestein, S.A. Selesnick3 pursues our question in the following form,
where does physics get its Lagrangians? On the other hand V.J. Stenger 4 ,
previously Professor of astronomy and physics at the University of Hawaii,
provides a scientific answer to the question, where do the laws of physics come
from? Remarkably, his elegant and mathematically detailed derivation of the
laws is driven by the requirement that the models physicist develop to describe
objective reality cannot depend on the standpoint of the observer.

At the end

I will comment on Stengers work, partly because of its relevance and partly
because we espouse completely different metaphysical views.
I want to be clear at the start that I assume theoretical physics will
achieve a satisfactory explanation of how our universe comes to be operating
according to the laws of fundamental physics from the earliest moment after
the big-bang.

This paper is an extended version of the paper I will present at the conference
on 5th July 09.
From the History and Philosophy of Science Programme, The University of
Melbourne.
Selesnick, S.A., Quanta, Logic and Spacetime, (World Scientific, New Jersey, second
ed., 2003)
Stenger, V.J., The Comprehensible Cosmos, Where Do The Laws of Physics Come
From?, (Prometheus Books, 2006)

In this paper I will outline the argument for the following purposive answer
to our question. Why are the laws of physics the way they are? In order that

the universe be knowable through empirical inquiry by embodied rational agents.

Given my starting assumption, you should reasonably expect me to highlight
what it is about the way the laws of physics are that justifies such an answer
to our question. I trust this expectation is sufficiently fulfilled in what follows.
Of course there are problems to be confronted. Hume, Kant and Darwin
are just three names that might easily come to mind as representing problems
confronting any possibility of a purposive answer. I will briefly return to them
at the end.
However, I will present an argument to design, not a version of the
traditional argument from design. Also, it is not based on fine-tuning,
anthropic principles, intelligent design, or god of the gaps arguments. The
argument has three parts, the physics, the transition from physics to
metaphysics and the metaphysics, which I call a metaphysics of inquiry.
On the one hand the natural sciences understand the universe as operating
according to processes described by blind natural laws, including the laws of
physics. This is widely taken to undermine the appeal to a purposive
explanation of these laws. It is widely taken as a powerful contradiction of the
belief that the universe is created and ordered to some purpose by God. On
the other hand if a purposive answer exists, it should at least specify some
purpose to which the universe is ordered and show how that explains why the
laws of physics are the way they are. If available the argument would block the
inference from blind to purposeless, for then some larger purpose would be
served by blind natural laws and processes.
The point is that we are surrounded by blind causal processes deployed for
a variety of purposes. One response here might be that while we have plenty of
examples of people using blind causal processes to achieve purposes, nothing
2

about the blind natural processes of the universe suggest they are serving or
expressing or in anyway operating for a purpose. No explanatory task in the
sciences calls for a purposive account of natural laws and natural processes. Of
course a variety of philosophical or theological positions might call for a
purposive account of the universe.

But it is true that no explanatory task in

the sciences calls for a purposive account of natural laws and natural processes
and no mere gap in scientific knowledge justifies raising a purposive answer.
How from a scientific starting point does the alleged possibility of a purposive
answer to our question even arise?
In physics, fundamental laws are commonly presented and used in terms of
what is called extremum principles. Historically, this approach to natural
philosophy was introduced by Leibniz in 17th century and Maupertuis in the 18th
century based on various metaphysical and theological considerations.

It was

called the principle of Least Action.

This approach was taken up by Leonhard Euler and then by Joseph-Louis
compte de Lagrange in a beautiful mathematical formalism for mechanics, while
cutting the metaphysical and theological moorings. A little later William
Hamilton did a similar thing for mechanics and optics. The whole approach was
extended to other areas of science by Hermon von Helmholtz.
Today variational or extremum principles are everywhere in physics, with no
trace of any metaphysical or theological origins. A discussion of extremum
principles in physics is given by Roger Penrose in his, Road to Reality , (Jonathan
Cape, London, 2004).
Given my theological and philosophical commitments I wondered whether
the last word had been spoken about the relation between extremum principles
and metaphysics. Could it be that variational principles still presuppose a
metaphysics? As far as I can see only two physicists in 20th century, took
extremum principles in physics as evidence for the operation of final causes in
3

nature. They were Max Planck and David Bohm. This is because extremum
principles involve integrals which require the starting point and end point of the
integration to be specified. Max Planck and David Bohm took the fact that an
end point had to be specified as evidence for final causes in nature. But this
wont do for the following reason. It is mathematically possible to go from the
integral form to the standard differential form of physical laws. The two forms
are equivalent mathematical descriptions of the same process. The differential
form gives a moment by moment description of the process. No end point needs
to be specified. It is hard to see how a mathematically equivalent statement of
laws in integral form can tell us anymore about the world.
This first objection was made by W. Yourgram and S. Mandelstam in their
important monograph, Variational Principles in Dynamics and Quantum Theory. 5
The belief in a purposive power functioning throughout the universe,
antiquated and nave as this faith may appear, is the inevitable
consequence of the opinion that minimum principles with their
distinctive properties are signposts towards a deeper understanding of
nature and not simply alternative formulations of differential equations
in mechanics. 6

Secondly, Yourgram and Mandelstam appeal to the work of R. Feynman 7 as a way

to use the least action principle while dissolving any ground for proposing any
metaphysical use of this principle. They claim to show,

6
7

Yougrau, W. and Mandelson, S., Variational principles in Dynamics and Quantum

Theory, (Pitman, London, 1968). The importance of the work is indicated not only by

the contents but also by the encouragement, interest, support that the authors
received from L. de Broglie, E. Schrodinger, W. Heisenberg, M. Born, R. Oppenheimer,
M. von Laue and A. Speiser.
Ibid. p.174.
Feynman, R. P., Space-Time approach to Non-Relativistic Quantum Mechanics, Review
of Modern Physics, Vol.20, p.267, 1948

Feynmans principle, although it is actually not a variational principle,

does non the less generate, in a direct and consummate fashion, a
variational principle, when the conditions are such that classical
mechanics provides a valid description of the system under
consideration. 8

In classical mechanics a unique path is picked out for a particle by the

requirement that the path always be such as to extremise the action S
(minimum or maximum) over the whole transition from initial to final state,
represented by S = 0. The odd implication is that the particle somehow knows
what is the unique path to follow. In the Feynmans path integral approach to
quantum mechanics the quantum system traverses all possible paths between
the initial and final states. In the classical limit the least action path is not
picked out, it falls out as the remaining constructive contributions of the
partial transition amplitudes in the region of the classical path after the mutual
cancellations of the amplitudes from the other paths.
I do not follow Yourgram and Mandelstam on Feynman. The reason is
Steven Weinbergs assessment of Feynmans work. 9 Weinberg highlights many
advantages of Feynmans path-integral approach to quantum mechanics

10

but

points out that a better version of Feynmans approach to quantum mechanics

can be derived from the canonical approach to quantum mechanics pursued by
Weinberg 11 using Lagrangians and assuming the extremum condition S = 0. 12
On the other hand Feynmans path-integral approach leads to wrong results in

8
9

10
11
12

Yougrau and Mandelson, (1968), p.137.

Weinberg, S., The Quantum Theory of Fields Vol.1, (Cambridge University Press,
Cambridge, 1995), p.377.
Weinberg (1995), p.377.
Ibid. p.377.
Ibid. pp.292.

certain contexts 13 but these errors do not occur using the Lagrangian approach
pursued by Weinberg. 14 This still leaves the first objection.
My wondering whether extremum principles in physics might be shown to
still presuppose a metaphysics was finally undermined when I considered under
what conditions it might be possible to undermine the first objection. As a first
step there would need to be some deeper physical story from which it would be
possible to derive integral form of the laws of physics and then using the Euler
Lagrange equation derive the laws of physics in the form of differential
equations. This would be another way of doing physics. Of course this might
turn out to be a version of the physical answer to our question here via the
Lagrangian route. There would be no guarantee that a deeper metaphysical
reading of nature would follow from this way of doing physics.
Shortly after accepting this sobering conclusion, I found a physicist who
had pursued precisely this path in physics. It is Prof. Roy Frieden, at the
College of Optical Sciences in the University of Arizona, see his Physics from

Fisher Information, A Unification, (Cambridge, 1998). 15 The deeper physical

story is provided by Fisher information. I stress that Friedens work showed no
interest in nor any grounds for thinking this Fisher information approach to
physics could have anything to do with metaphysics.

13
14
15

Ibid. p. 377.
Ibid. p.377.
For criticism of this work see, Lavis, D. and Streater, R.,(2002), Physics from Fisher
Information, Studies in the History and Philosophy of Modern Physics, 33B(2),
p.327-343.

2 The Physics

Frieden takes Fisher information as the key to explaining the operation of

the laws of fundamental physics and deriving their mathematical forms. For
the list of the laws of fundamental physics he has obtained so far see
appendix1. For the list of publications up to 2007 in well known physics journals
see appendix2. For the list of new predictions he has made see appendix3. I
should say that this approach to physics is not an exercise in pure thought. The
derivations make use of physical knowledge. For example in deriving Maxwells
equations, the existence of electric and magnetic fields are assumed, but not
any relation between them. I need to say something about Fisher information to
help provide a very brief introduction to Friedens work.

(A)

Fisher Information

Fisher Information is a measure of information first introduced by R.A.Fisher,

at Cambridge in the 1920s, who showed that Darwins theory of evolution by
natural selection made sense statistically. Later, Fisher information shows up in
the work of H.L.Cramer and C.R. Rao. They were theorizing about how to
measure a quantity that is subject to noise and so is fluctuating around some
mean value. It is known as classical measurement theory. A celebrated result 16
is the Cramer Rao Inequality, I e2 1, where e2 is the mean square error in
the estimates of the mean value of the parameter being measured and I is

16

Van Trees, H. L., Detection,Estimation, Modulation Theory, Part I, (1968), pp. 66-68.
Trees notes that the inequality was derived by, . Cramer, H. L. , Mathematical
Methods of Statistics, (1946) and, Rao, C. R., Information and Accuracy Attainable
in the Estimation of Statistical Parameters, Bull. Calcutta Math. Soc. 37, 81-91,
(1945). The orignial work was done by,. R. A. Fisher , On the Mathematical
Foundations of Theoretical Statistics, Phil. Trans. Roy. Soc. London, 222, 1922,
pp.309368.

Fisher information.

I 4 ( q /(x) ) 2 dx,

Fisher Information

(one dimension)

(1)

q(x) is the square root of the probability density function describing the
fluctuations x of the parameter around its mean value and is known as the
probability amplitude function. It is assumed that q(x) is shift invariant and
so is independent of the parameter being measured.

It is possible to make

measurements in a way that yields the lower bound, so that I e2 = 1 So, I can
be measured. Notice that I is defined on the probability amplitude functions

q(x) describing the fluctuations, not on measurement and so I is a measure of

the degree of order of the fluctuations. The use of amplitude functions comes
from R.A.Fishers work, quite independently of quantum mechanics, as J.
Wheeler acknowledged. 17
Earlier in his career Prof. Frieden made an international reputation for
being able to turn fuzzy images into clearer images, whether of galaxies of
stolen car license plates. He was able to retrieve the lost information using a
Fisher Information.

One day he noticed that one of the equations he had

obtained using Fisher information had exactly the same form as Schrodingers
equation. He was surprised but thought nothing of it. Some time later he came
across an article by a A.J.Stam 18 , a Dutch mathematician, showing how to go
from the Cramer Rao Inequality to Heisenbergs Uncertainty Principle. Frieden
then seriously started to wonder if there was a deeper connection between the
laws of physics and Fisher information. This became his physics from Fisher
information project, with many of the results indicated in the appendices.
17
18

Wheeler, J.A., (1994), At Home in the Universe, Woodbury, AIP Press, p.304.
Stam, A.J. , Information and Control, 2, 1959, p.101.

Frieden also followed J.Wheelers 19 idea of articulating the whole of physics in

terms of information.

(B) Friedens approach to physics

20

Fisher information arises naturally via the Cramer Rao inequality in the
context of a thought experiment about parameter measurement defined by E
and W (my symbols).

E represents idealised empirical inquiry for parameter measurement

conducted by rational inquirers, (whether human or alien), assuming:

. isolation of the system being measured from the influence of

other physical systems;
. no-noise or other loss of information due to the measuring device;
. unbiased estimation of parameters; on average estimations
equals the actual mean parameter value.

W represents the fluctuations to which a parameter is subject and is

taken over from classical measurement theory. Here the fluctuations are
assumed to be intrinsic to space and time, rather than being the effect of other
forces.
This statement of the premises for the derivation now brings out clearly
how the active role of the inquirer or observer is a premise in the derivation of
the CRI: not just, Ie2 1, but E, W Ie2 1. E and W can be shown to be
premises in all the derivations of the laws of fundamental physics from Fisher
19

20

Wheeler, J., in, Zurek, W.H., ed., Complexity, Entropy and the Physics of
Information, (Adison-Wesley, New York, 1990), p.3.
Frieden, B.R., (1998, 2004); Frieden, B.R., and Gatenby, R.A., eds., Exploratory Data
Analysis Using Fisher Information, (Springer, 2007), chapter 1.

information, including the Lorentz transformation.

The active role refers to

the rational inquirers (whether humans or aliens) that, even in the thought
experiment, are the in-eliminable conductors of empirical inquiry, which aims to
produce epistemically significant results, because the experiments conform to
rationally established epistemic standards.

This emphasizes the rationality of

empirical inquiry, not the consciousness of the inquirers.

Assumption W is motivated by the use of classical measurement theory
with regard to measurement of parameters subject to noise, which is
ubiquitous. Why are the fluctuations in the parameters of a physical system
taken to be due to intrinsic fluctuations in space and time? If the parameter
fluctuations were taken to arise extrinsically from causal interactions from
other physical systems, this would contradict the first point under E. Secondly,
such a causal interaction would presumably be according to one or more laws and
so the idea of physical laws would be smuggled in as part of the premises. This
would contradict one of the aims of the physics from Fisher information
project, viz., to give an account of where physics gets its Lagrangians and
therefore an account of physical laws, without presupposing the operation of
physical laws. The assumption of intrinsic fluctuations in space and time does
not itself presuppose or entail the results to be obtained. The use of space
and time shows that a pre-relativistic context is the presumed starting point.
A first task of the whole project is to derive the Lorentz transformation.
While Frieden illustrates what he means by these fluctuations by reference
to quantum mechanics and to quantum cosmology 21 he does not discuss the
motivation for introducing the assumption here represented by W. The

motivation stemmed from using the classical measurement theory as the

context in which Fisher information naturally arises. The nature of these
intrinsic fluctuations deserves further discussion from the perspective of a
21

Frieden, B.R., (2004), pp. 28,164.

10

Fisher information approach to physics and in the light of other work in physics,
with the work of R. Cahill, R., C.M. Klinger 22 as one example.
It turns out that Fisher information I has many properties relevant to
physics. For example, I is a functional already in the form of an action integral.
The same form as used in the Lagrangian approach to mechanics and as is found
when other areas of physics are put into Lagrangian form. Fisher information I
in eq.(1) may be generalized for parameters n, n = 1, N, in four space-time
dimensions (x,y,z,ict), giving,

I =

23

4 qn(x) . qn(x) dx

dx = dxdydzdct

n=1

I = In
n=1

}
In = in(x) dx

(2)

In (x) = 4 qn(x) . qn(x)

Another property is the I Theorem 24 , dI/dt 0.

A third property is that I

is Lorentz covariant. 25 Gauge invariance ensures observables are unaffected by

choice of reference frame.

Any change of coordinate system that leaves

observable quantities unchanged is called a gauge transformation. The mean

square error, e2 , is an observable, and so should be unaffected by choice of

22

23

24

25

R. Cahill, R., C.M. Klinger, Self-Referential Noise as a Fundamental Aspect of

Reality, in, D. Abbott and L. Kish eds., Proceedings of the 2nd International
Conference on Unsolved Problems of Noise and Fluctuations (UPoN 99), Adelaide;
Australia,11-15th July, Vol. 511, American Institute of Physics, 43, (2000).
Frieden, B.R., Science from Fisher Information, A Unification, (Cambridge
University Press, 2004); pp. 58 - 64.
Plastino, A.R. and Plastino, A., (1996), Symmetries of the Fokker-Planck equation and
the Fisher-Frieden arrow of time, Physical Review E, 54, No. 4, p. 4423.
Frieden (2004), p. 65.

11

gauge for a given physical scenario. From the lower bound of the Cramer Rao
Inequality, I = e -2, and we should expect Fisher information I to be invariant
with respect to choice of gauge.26

A fourth property is that I is an inner

product and so is invariant under a unitary transformation. Mathematically

many unitary transformations hold for complex probability amplitude functions.
Such transformations can be applied to Fisher information I. Frieden shows
how the probability amplitude functions q(x) can be repackaged as complex
amplitudes (x). 27

n N 1/2 (q 2n-1 + iq2n ) , n = 1,2, . N/2,

i = (-1 )1/2

(3)

leading to,

N/2

n n =

n=1

q n2(x) = p(x)

n=1

(4)

Substituting eq. (4) in eq.(2) and using imaginary coordinates, leads to,
N/2

I = 4Nc In dr dt

where ir = i(x, y, z)

n=1

In = [ - ( n )* .

26

27

(5)

+ (1/c2) ( n /t )* n /t ]

Ibid., pp. 106-107. Interestingly, this is supported by the fact that the Fisher
information derivation of Maxwells equations, yields these equations in the form of
an electromagnetic potential A,; ibid., chapter 5.
Ibid. , p. 66. The justification for introducing imaginary numbers is that the Cramer
Rao Inequality holds for real and imaginary parameters; ibid. , pp.441- 444.

12

Frieden thinks of Fisher information I as implicated in the data, (via the

Cramer Rao Inequality) whereas, information J is the level of Fisher
information at the source, i.e. in the effect that is observed. 28 Any
observation is the output of an information-flow process.

messenger
probe particle

Frieden postulates that such information flows are passive so that I can never
exceed J. 29 Thus , I J entails 0 = I J , 0 < 1, a zero principle. The
probe particle disturbs J by J and Frieden takes the information flow as
implying that I is perturbed by I. Furthermore Frieden postulates, I = J
and so (I J) = 0, an extremum principle. 30

Defining K = I J, leads to

K = 0. This is Friedens principle of Extreme Physical Information (EPI)

EPI

K = I - J
K = 0
0 = I J , 0 < 1,

extremum principle }
zero principle

(6)

Frieden uses the EPI in all his derivations of the laws of fundamental physics
save the derivation of the Lorentz transformation, which uses the Cramer Rao
Inequality. 31 This provides some a posteriori justification for the two
postulates of the EPI. On this view the laws of physics are brought into
operation as an effect of the probe particle disturbing the information
28
29
30
31

Frieden and Gatenby, (2007), p. 21.

Ibid. p.21.
Ibid. p.28.
The derivation in Frieden (1998) was criticized by Lavis and Streater (2002). See
the derivation in Frieden (2004).

13

characterizing the system being measured. While the EPI is a general

variational principle, it needs to be supplemented with empirical information for
each derivation to be completed.

(C) My approach

I am much informed by Friedens work and follow many of his results. Two
things particularly mark my interest.

Firstly, I start by highlighting

independent evidence that there is an interesting relationship between physics

and Fisher information. This is found in two papers by W. Wootters 32 and, S.L.
Braunstein and C.M. Caves. 33 Braunstein and Cave extend Wootters work by
showing that the problem of the distinguishability of quantum states can be
mapped onto that of precision parameter estimation, using the Cramer Rao
inequality. They show that the statistical distance between pure quantum
states is attained by maximizing Fisher information via the lower bound of the
Cramer Rao inequality. The conclusion is that maximising the state
distinguishability is equal to the Hilbert space angle between two quantum
states. A similar result is obtained for distinguishability over all quantum
measurements for pure and impure states.

These two papers show an

interesting relationship between Fisher information, via the Cramer Rao

Inequality, and the distinguishability of quantum states and the structure of
Hilbert space.
This provides prior physical justification for further inquiry into the
relationship between physics and Fisher information. Friedens work has
pioneered one way of pursuing that inquiry, which began independently of the
32

33

Wootters, W.K.(1981), Statistical Distance and Hilbert Space, Physical Review D,

23, No.5, p.357.
Braunstein, S.L., and Caves, C.M.,(1994), Statistical Distance and the
Geometry of Quantum States, Physical Review Letters, 72 ,No.22, p.3439.

14

work by Wootters or Braunstein and Caves.

Furthermore all the work being

published in standard physics journals examining other links between Fisher

information and physics points in this direction. One example is the connection
between Fisher information and thermal physics. 34
My second interest is to go beyond the level of a posteriori justification
for Friedens two postulates provided by his derivation of many of the laws of
fundamental physics. I have examined whether these postulates, the
variational principle and the zero principle, may be derived from the same
theoretical starting point as Frieden, viz. classical measurement theory,
concerning a thought experiment about parameter measurement under
conditions defined by E and W.

I will only be able to sketch this work here.

The details of the derivation are set out in a paper 35 presently being scrutinised
by colleagues before being submitted to a physics journal for publication.
Classical measurement theory assumes a measurement interaction between
a probe particle and the system being measured, but it says nothing about what
goes on in that measurement interaction. However to derive the two postulates
requires some idea of what is going on in the interaction.
In this austere theoretical context, the challenge was to start with the
thought experiment about parameter measurement under conditions defined by

E and W, leading to E, W Ie2 1, (with I represented by equations (2) and

(5) and assuming its many properties). On this basis I proceeded to develop
other resources needed so as to be able to elaborate an account of what is going
on in the measurement interaction within time interval t , and then derive the
two postulates.

34

35

See references in, Plastino, A., and Plastino, A.R., Information and Thermal Physics,
in, Frieden, B.R., and Gatenby, R.A. (2007), pp. 119-154.
Ames, S., The derivation of a variational principle for physics, from Fisher
information, pp.1-49.

15

Fisher information I is defined on the probability amplitude functions

describing the intrinsic fluctuations W. It has the same form for all
measurement scenarios. This suggests some other term is needed to distinguish
measurement scenarios, which I introduce as follows. Fisher information I has
the form of an inner product and so is invariant under a unitary transformation.
There are many unitary transformations. I hypothesise that among the many
mathematical possibilities there is at least one (possibly more) physically
significant unitary transformation for Fisher information I, from the spacetime x coordinate to an unknown conjugate coordinate. Let J represent
the unitary transform mate of I. J inherits many of the properties of I.
Since the fluctuations are assumed to be intrinsic, we may think of I as the
measure of the order of these fluctuations, quite independently of whether
measurement is being made. The basic idea is that Fisher information I and its
unitary transform mate J are the measure of the order of the fluctuations

(x) in the x coordinate space and () in the unknown coordinate space

respectively. This holds whether or not measurements are being made.
Parameter measurement proceeds by firing probe particles at successive
copies of the physical system being measured. The probe particles interact
with the system being measured and eventually register on the output-space of
the measuring device.

The system being measured passes out of the

measurement device. Before the measurement interaction the system being

measured is characterised by Fisher information terms Ib , Jb , but after the
interaction by Ia , Ja . The I-Theorem indicates the possibility of loss or noloss of information could take place, even under ideal epistemic conditions.
I assume that within the time interval t of the interaction, the probe
particle disturbs the intrinsic fluctuations W of the parameters of the system
being measured. This disturbs the probability amplitude functions describing
these fluctuations and so disturbs the information terms, Ib and Jb , leading
16

A fascinated group of people at the seminar

enter into a thought experiment
Fig. 1
17

within t, to ID and JD , (D for disturbance ). After t ID becomes Ia and JD

becomes Ja. The measurement scenario is indicated in Fig.1. below.
Any acceptable account of what is going on within t must meet the
following three criteria:

C1

it has to be shown that the equivalent of the two axioms in the EPI
occur in time interval t;

C2

there needs to be a coherent Fisher information account of how

the measurement interaction fulfils C1;

C3

it must be possible to show that any account of a measurement

interaction that meets C1 and C2, also connects to the standard
approach to physics.

The case of no-loss of information, Ib = Ia , Jb = Ja , is not straight forward,

but it is not difficult to meet C1.

My analysis shows that this disturbance has

two components. One is a transposition of the unitary transformation between

I and J, Ib JD , Jb ID. The other is an infinitesimal disturbance of ID by

ID and JD by JD, where, ID = JD = 0.

This is not to misread ID = JD = 0 as

a disturbance rather than a lack of disturbance. Variational calculus shows how

ID = 0 is consistent with a particular kind of disturbance of the probability

amplitude function (x), within t, on which ID depends.

This leads to the

conclusion that the measurement interaction brings about EPID , within the
time interval t.
For the case of loss of information no new theoretical resources are
available to help gain some insight into the measurement interaction in t.
Since loss of information is the opposite of no-loss of information I proceed by
establishing the essential conditions leading to no-loss of information and then
systematically negate them. This generates a range of logically possible forms

18

of the measurement interaction. There follows an argument by exhaustion to

see whether any of these logical possibilities meet the three criteria.
My analysis shows that two of the logical possibilities are acceptable. Each
has two components; one is a loss of information such that ID < JD, and an
infinitesimal disturbance of ID by ID and JD by JD, where ID = JD = 0. The
difference between the two acceptable cases concerns the loss of information.
One suffers a direct loss of information Ib ID , Jb JD , while in the other
case there is a transposition with loss of information, Ib JD , Jb ID . Each
of these possibilities meets C2.
The conclusion is that the measurement interaction brings about EPID ,
within the time interval t. Combining the no-loss and loss of information
cases, we obtain,

EPID

KD = I D - JD
KD = 0
extremum principle
0 = ID JD , 0 < 1, zero principle

(7)

(D) Meeting C3 making the connection to established physics

The disturbance of the system being measured due to the measurement

interaction produces an infinitesimal transformation of the (x) and of the
corresponding amplitude functions in the conjugate coordinate space.

This

infinitesimal transformation leaves the action K = I J stationary, K = 0. It is

therefore a symmetry transformation. The Fisher information approach to
physics can therefore take up all the resources of variational calculus and it can
take up the Lie Group 36, 37 as a theoretical resource for its exploration of

36

Weinberg, S., (1995), The Quantum Theory of Fields Vol.1, Cambridge, Cambridge
University Press, pp. 306 314; (a) pp. 50 - 53.

19

possible physically significant unitary transformations between I and J. For

the same reason this approach to physics can take up Noethers theorem, with
the measurement interaction implying symmetries and thus conservation laws.
The probe particle interaction brings about constraints on the fluctuations of
the parameter(s) being measured, defined by the EPID. These constraints turn
out to be the well known laws of physics.
The analysis shows that the EPID holds for three distinct types of
measurement interaction: (A) no-loss of information, with transposition;
(B) direct loss of information, but no transposition; (C) transposition with loss
of information. Criterion C3 is partly met because (A) corresponds to Friedens
derivation of quantum mechanical results, (e.g. Klein Gordon equation, Dirac
equation, Heisenberg uncertainty principle and others) and (B) corresponds to
Friedens derivation of the equations of General Relativity and Maxwells
equations. The probe particle is the only free variable in this analysis, so it is
assumed that different kinds of probe particles, bring about the different type
of measurement interactions.
It is not yet clear what physical interpretation to offer for (C). Comparing
and contrasting interaction types suggests (A) (transposition and no-loss of
information) and (B) (no transposition and loss of information) could be special
cases of (C) (transposition with loss of information). Furthermore, whereas (A)
leads to quantum mechanics and no-loss of information, and (B) has no quantum
mechanics and loss of information, we should possibly consider (C) as leading to
quantum phenomena with loss of information. Since (A) and (B) lead to physical
laws that are time reversible, it is worth exploring whether (C) leads to physical

37

Olver, P.J., (1993), Applications of Lie Groups to Differential Equations, (SpringerVerlag, New York, Cambridge, Cambridge University Press) Chapter 4.

20

laws that are time irreversible. This suggestion is based on a paper by Ilya
Prigogine. 38

(E) Conclusion
Dropping subscripts and gathering all the premises used to derive the

EPID , we obtain the following result,

ith measurement scenario

E , W, I, OA

EPI

EPI , OAi

Li IDI

For all scenarios,

E, W, I, OAi IDi
R1:
E, W, I, AOA

Li

Li

Li
L

OA

Other assumptions besides E and W

OAi

Other assumptions for the ith measurement scenario

Li

An equation with exactly the same form as a well known law of

fundamental physics

IDi

Identity postulate for unknown constants. This is supported by

empirical evidence made relevant by the details of the
derivation.

Li

The well known law of fundamental physics

AOA

38

Represents all the other assumptions for all the measurement

scenarios.

Prigogine, I., Why Irreversibility? The formulation of classical and quantum

mechanics for non-integrable systems, International Journal for Quantum
Chemistry, 1995, Vol. 53, Issue 1, pp. 105 118.

21

represents all the laws of fundamental physics so derived.

Given E, W, I and all other assumptions, AOA, the operation of the laws of
physics L may be explained and their mathematical forms derived.
I call R1 the rational tuning of the laws of physics to the idea of idealised
empirical inquiry conducted by embodied rational agents, somewhere in the
universe.

This is quite different from the well known physical fine-tuning of

universe for carbon based life.

R1 holds within a thought experiment or, equivalently, it holds within a

fiction. Does it hold in the actual univese? An affirmative answer is supported
by five reasons.

The laws obtained by the Fisher information approach to physics are

the laws of physics that operate in the actual world.

The measurement interaction can take place outside of a measurement

context. It does not depend on the consciousness of inquirers.

All the terms in the result R1 are well known in the actual world.

New results obtained by Frieden et. al. open to testing.

Independent support for the Physics from Fisher Information project

from the work of W.Wooters and, S.Braunstein and C.Caves noted
above.

The function of the fiction is to bring to light the relationship between the
terms shown in R1 that is otherwise unnoticed.
What follows from R1 is that from the earliest moment after the big bang,
when the universe is operating according to the laws of fundamental physics,
these laws are already rationally tuned to idealised empirical inquiry, even
though actual inquirers dont show up until billions of years later. This is what
22

it is about the way the laws of physics are, that needs to be explained.
Commonly, the initial reaction is that R1 can be explained within the resources
of the natural sciences, by an extension of evolutionary cosmology. This leads
to discussing the move from physics to metaphysics.

3 The move from physics to metaphysics

This move confronts at least two challenges. We need to ask whether it is

logically possible for a scientific theory to explain how it is that R1 holds from
the earliest moment of the universe after the big bang. If so, end of story
no further seeking for an explanation is justified. If not there is another
challenge. We need to ask whether it is reasonable to treat R1 as a brute fact.
If so, again, end of story.

If not, well then, but only then, would we be

justified in further seeking an explanation.

It turns out that both challenges can be met successfully. Logically, no
scientific theory (as construed) can explain R1. Understandably this claim has
occasioned some surprise. The immediate thought of more than one person is
that evolutionary cosmology can do the job. That is the scientific account of
the 13.7 billion year old universe from the big bang until now, including the
evolution of life on planet earth, can account for R1. This deserves some
attention.

(A) Evolutionary Cosmology

Evolutionary cosmology (EC) tells the story of the emergence of carbon

based life on planet earth. Contemporary evolutionary biology tells a complex
story aiming to show, among many other matters, how human beings have evolved
on planet earth equipped with rationality and acting accordingly. It is well
established that we can only tell the story of the evolution of carbon-based life
23

on planet earth by telling the story of the evolution of the universe from the
moment of the Big Bang.
Broadly speaking we can summarize EC as follows: assume the laws of
fundamental physics L, also fluctuations W, Fisher information I and all other
assumptions AOA, some initial conditions and whatever else is currently used to
tell an accurate and adequate account of the development of the cosmos, even
multiple universes, and so call all these COSMO; assume the key processes of
chemistry, call these CHEM ; assume evolutionary biology, call this BIO; and
assume a whole raft of contingencies, call these XX for X factors. Given
these assumptions it is possible to explain how human inquirers have evolved
equipped with rationality and with various historical and cultural processes, call
these CUL , it is possible to explain in particular, how there came into
existence human inquirers who hold to and carry out experiments in accord with

E, who therefore instantiate E.

Thus EC may be summarised as follows:

EC:

explains
E

COSMO including W, I, OA, L, + CHEM + BIO + CUL + XX

Thus EC entails the L, W, I, OA and E and therefore EC entails R1, since,

R1:

E , W, I, OA

explains
entails

Since EC entail R1 and R1 defines the rational tuning of the L, EC entails

the rational tuning of the L.

24

The EC explanation of R1 proposes an account of the evolution of intelligent

life on planet earth, claiming thereby to explain R1. This is mistaken. To see
the problem, we need to take stock of the fact that E is obtained initially quite
independently of EC, from some general beliefs about rationality and the world
in which inquiry is assumed to take place. It is true that we know about
rationality because human beings instantiate rationality, whereby they think and
act for reasons, and this is known independently of how the origins of that
instantiation might be explained. It is on this basis that EC can be assessed as
a successful explanation or not of E, where E is known independently and
epistemically implicated in generating EC.
The EC explanation of R1 terminates with an entailment from EC to R1.
Now an entailment is not necessarily an explanation and this is the case here.
What EC proposes is an explanation, not of R1, but of how, via evolution on
planet earth, E is instantiated in human inquirers. The EC proposes an
explanation of how, eventually, (borrowing the arguments for R1), we come to
know R1. It is not an explanation of R1 because it does not explain how it comes
about that the L were already rationally tuned to E from the earliest time
after the big-bang. The EC explanation of R1 fails.
My claim that no scientific theory can explain R1 does not depend on
rebutting one counter argument. Here is a more general argument.

(B) More general difficulties for a scientific theory to explain R1

It is worth reiterating that I assume theoretical physics will be successful

in its aim of providing an account of the universe that will include how it comes
to operate according to laws L. But will it be able to explain how R1 holds from
the earliest moment of the universe? Assume some physical theory TPhys can
explain R1. What will TPhys look like? As a physical theory TPhys ,
25

is a blind theory ( causal/mechanical; final causality not

included),

concerns physical conditions, processes, leading to the Big Bang

and thereafter to the L ( let it include string theory, M-theory,
multiple universes, ),

presents a valid argument, to deduce the result R1, viz. :

TPhys R1.

Three problems confront Tphys as an explanation of R1.

#1 For TPhys to offer a valid argument, the conclusion cannot contain terms
that are not in the premises proposed by TPhys. Therefore, TPhys must include or
lead to E. Recall that E helps define a thought experiment concerning idealised
empirical inquiry by embodied rational agents for valued epistemic ends, in the
future of (at least) one of the multiple universes to be produced by the physical
processes proposed by TPhys. But this means TPhys is not blind and so cannot be
a physical theory.
#2 At the point where TPhys must include E, in the conditions prior to the
Big Bang of our universe, E represents a thought experiment about empirical
inquiry, conducted by rational agents according to ideal epistemic standards, in
principle anywhere in the universe. What provides the thought experiment?
Even naturalists think it requires all of evolutionary cosmology to account for
the evolution of E. A physical theory TPhys alone cannot provide E.
#3 TPhys has to explain R1. But R1 is not a causal physical process but a
rational inference (explanation/logical entailment). It is not the kind of thing
explained by physical theories. See W. Sellars 39 and more recently the

39

Sellars, W., Empiricism and the Philosophy of Mind, in Feigl, H., and, Scriven, M., eds.,

26

discussion by J. McDowell, 40 which locates the relations of explanation and

entailment in the logical space of reasons and this is logically different from the
space of subsumption under natural law.

The result R2 follows:

R2: On the proposed account of a physical theory,
if TPhys is a physical theory it logically cannot explain R1.

(C)

Can R1 reasonably be treated as a brute fact?

Consider the following argument.

(1) R1
(2) If no scientific or non-scientific explanation of R1 is possible,

R1 is a brute fact.
(3) No scientific theory can explain R1.
(4) No non-scientific explanation of R1 is possible.
(5) Therefore R1 is a brute fact.

The argument is valid. If we reject (5), which premises will we reject?

(1) OK
(2) Stipulates what is meant by a brute fact.
(3) OK
(4) Says that there is nothing outside or beyond what the natural
sciences can tell us, that can explain R1.

40

Minnesota Studies in the Philosophy of Science, vol. 1, (1956).

McDowell, J., Naturalism in the Philosophy of Mind, in, De Caro, M. and Macarthur, D., eds.,
Naturalism in Question, (2004).

27

Are there any arguments against (4)? I argue that E is obtained initially
quite independently of EC. It is obtained by rational inquirers, with certain aims
and some general beliefs about rationality and about how the world operates,
deciding what should be done to ensure inquiry is effective. We know about
rationality because human beings instantiate rationality, whereby they think and
act for reasons, but this is known independently of how the origins of that
instantiation might be explained. Of course particular applications of these
standards of inquiry may invoke specific scientific knowledge, for example
controlled random, double blind trials in testing drug efficacy.
This is one argument for thinking of E as something beyond the theories of
natural science and that E is non-trivially involved in explaining the laws of
fundamental physics L, as shown in R1. This provides rational grounds for
wondering if something beyond the natural sciences might explain R1.

But (4)

would lead us to expect any such explanation to be impossible. Hence (4) should
be set aside as unreasonable. Therefore (5) does not follow and we reasonably
set aside the claim that R1 is a brute fact.
Those holding to scientific naturalism as an ontological thesis would reply
that epistemology ultimately has a natural ground in natural processes described
by the natural sciences as argued in the EC explanation of E. Hence, they would
say that E is not something beyond the theories of natural sciences and so
does not create a problem for (4).
But this does not take account of two points. The first is the fact that
evolutionary cosmology, EC, or any Tphys , cannot explain how the laws of
fundamental physics, L, are already rationally tuned to E in the earliest moment
after the big-bang. This is another argument for thinking of E as something
beyond the theories of natural science.

28

The second point is the way R1 is a new problem for scientific naturalism.
Physicalism is the strongest form of scientific naturalism. According to
physicalism, the ultimate ontological constituents of what exists are physical
constituents. 41 For both reductive and non-reductive physicalism, what exists
is either an instance of these ultimate ontological constituents or complex
configurations of the same. These ultimate ontological constituents are not
explicable in terms of something more fundamental. By definition there are no
deeper non-physical constituents of these ultimate ontological constituents.
For physicalism the laws of fundamental physics L are either among the
fundamental ontological constituents of what there is or are themselves
brought about by what is physically more fundamental.

Either way, physicalism

denies the L are explicable in terms of non-physical constituents. 42

But R1 says that the operation of L may be explained and their
mathematical form derived from premises that non-trivially include E, where E
helps define a thought experiment about rational inquirers, pursuing the valued
ends of empirical inquiry, according to idealised epistemic standards. R1 holds in
the actual universe. The operation of the L is explained and their mathematical
forms derived from what is non-physical, since E manifestly includes the
mental. Furthermore, R1 cannot be explained by any scientific theory,
including any physical theory. These points contradict the basic tenant of
physicalism and so create a new problem for physicalism.
true this situation would not be possible.

If physicalism were

Thus R1 provides evidence against

physicalism and so against scientific naturalism. It is unreasonable to appeal to

41

42

For example, Stoljar, D., Physicalism, Stanford Encyclopedia of Philosophy,

http://plato.standford.edu/entries/physicalism/ , 2001, pp.2; Papineau, D.,
Philosophical Naturalism, (1993), p. 9.
The possibility that the mental may be incorporated in what counts as physical is
recognized as a challenge to phyicalism by Stoljar (2001), p.12; Poland, J.,
Physicalism, (Clarendon, Oxford, 1994), p. 287; and would appear to be a problem for
Melnyks account of physicalism, see Melnyk, A., A Physicalist Manifesto: A
Thoroughly Modern Materialism, (Cambridge University Press, New York, 2003), p.27.

29

scientific naturalism to support (4). 43 Again, it is unreasonable to claim that R1

is a brute fact.
A weaker objection has also been offered to this conclusion: while it is not
possible to prove that R1 is a brute fact, it cannot be ruled out. Perhaps the
universe just turns out to be such that R1 is a brute fact.

My response is that

I neither want to nor need to prove that R1 is not a brute fact. Something like
the principle of sufficient reason would be needed to prove that point and that
would be to import a metaphysical position, which I have aimed to avoid. All I
need to do is to show it is unreasonable to treat R1 as a brute fact. I think the
weaker position is also unreasonable. Under what conditions could one have
reasonable grounds for thinking that R1 is a brute fact? Since (3) still holds,
the weaker position would be reasonable if there were some grounds for holding
(4). But I have already shown why (4) is unreasonable.

I conclude,
R3:

It is unreasonable to treat R1 as a brute fact.

R4:

Given R2 and R3 it is reasonable to seek further for an

explanation of R1.

4 A metaphysical explanation of R1

What must minimally be assumed to hold in order to explain R1 ? Any

explanation of R1 must include the thought experiment E.

Something capable

of rational thought, could envisage E. This something should be thought of as

some kind of rational agent, envisaging rational agents (inquirers) pursuing
valued epistemic ends in some universe. A rational agent must also be assumed

43

This is a fragment of a much larger discussion of scientific naturalism.

30

because thought alone is not enough to explain the existence of our universe in
which R1 holds.
In addition, we must assume, minimally, some kind of basic stuff, which this
rational agent orders in the light of envisaging E.

(The minimum needed to

explain R1 does not require God creating ex nihilo as Kant made clear. A
separate argument would be needed to identify the rational agent of this
argument with the God of theism or of various religious traditions.) This
correctly treats this rational agent as envisaging E as a purpose to be enacted.
For our universe, R1, the rational tuning of L to E , may be explained as
follows. Assume some kind of rational agent with a purpose that at least
includes forming a universe that is knowable by embodied rational inquirers
conducting empirical inquiry under ideal epistemic conditions in that universe.
Assume there is also some basic stuff and that in the light of this purpose,
this rational agent orders this basic stuff so that it eventually brings about a
universe (possibly one among many others), of which our best physics gives an
account, which among other things is characterized by W , I and AOA. Given
this purpose and these characteristics, the universe operates according to
laws, whose operations is explained and mathematical form entailed by E, W, I
and AOA, and these laws are L.

Call this result R5, which I summarise as

follows:

R5

R1 can be explained by minimally assuming that some kind of rational

agent has structured some basic stuff to form this universe, (one
among many) of which our best physics gives an account, and
characterised by W, I, AOA, and laws L, in order that the universe
can be known through empirical inquiry by embodied rational agents.

Notice this is not a god of the gaps argument. The argument assumes the
(eventual) success of theoretical physics in giving an account of the physics of

31

the universe, including multiple universes. It is unaffected by a scientific

elaboration of COSMO, the cosmology in EC , the evolutionary cosmology story
represented above.
This is an argument from physics to metaphysics.44 It is a metaphysics
because it goes beyond physics to what physics does not inquire into. It is
however a metaphysics of inquiry. As such it logically cannot be in conflict with
empirical inquiry. Most importantly, this answer to our question is not a science
stopper! It logically cannot inhibit either empirical or theoretical inquiry in
physics or any other science. On the contrary, it encourages the continuing
exploration of both physics and metaphysics.
From R5 we should expect other aspects of the universe besides L to
provide evidence that the universe has been ordered to the end that the
universe is knowable through empirical inquiry by embodied rational inquirers.
This expectation is fulfilled as follows:

it is now well established that the physical universe is fine-tuned

across many physical constants and the laws of physics, for
producing carbon based life, including intelligent carbon based
life;

the universe it is also fine tuned to be widely characterised by

locality and linearity, which are the physical bases for being
able to inquire into particular processes and phenomena, without
having to study everything else as well;

44

Contrast the move from metaphysics to physics by J. Leslie, see his, Infinite minds:
A Philosophical Cosmology, (Clarendon Press, Oxford, 2001). Leslie argues that our
life producing universe exists because it is ethically required.

32

evolutionary cosmology shows how carbon based intelligent life

has in fact evolved in the universe.

Here I also want to briefly comment on the work of V.J. Stenger mentioned
earlier. Stenger addresses the question, where do the laws of physics come
from?
The laws of physics are not handed down from above. Neither are
they rules somehow built into the structure of the universe. They
are ingredients of the models that physicists invent to describe
observations. If the models of physics are to describe
observations based on an objective reality, then those models
cannot depend on the point of view of the observer. This suggests
a principle of point-of-view invariance that is equivalent to the
principle of covariance (or cosmological principle or Copernican principle)
when applied to space-time. As Noether showed, this leads to the principles
of energy, linear momentum, and angular momentum conservation and
essentially all of classical mechanics. It also leads to Lorenzt invariance and
special relativity. When generalised to the abstract space of functions such
as the quantum state vector, point-of-view invariance is identified with gauge
invariance. Quantum mechanics is then just the mathematics of gauge
transformations with no additional assumptions needed to obtain its rules,
including the superposition and uncertainty principles. Conservation of
electric charge, isospin, and other quantities follow from global gauge
invariance. The forces in the standard model of elementary particles are
fields introduced to preserve local gauge invariance. Gravity can also be
viewed as such a field. Thus practically all of fundamental physics as we
know it follows directly from the single principle of point-of-view
invariance. 45

I take Stengers main point that the model of physics is developed on the
main assumption of point-of-view invariance. This is the assumption that in this
universe there is no preferred reference frame for observation. The physics
45

Stenger, V.J., Where Do the Laws of Phyiscs Come From? p.1.

This paper is based on his book and may be found at,
http://www.colorado.edu/philosophy/vstenger/Nothing/Laws.pdf .

33

of the actual universe turns out to conform to this model as indicated by a

myriad of precise observations. This is further independent support for my
conclusion that the universe is structured in order to be knowable by embodied
rational inquirers, in principle anywhere in the universe. Furthermore the pointof-view invariance or covariance which is a presupposition of Stengers approach
is a natural consequence of the Fisher information approach to physics.
Another question I wish to explore about Stengers argument is what
happened to his point-of-view invariance presupposition. If the argument was
laid out in full would this presupposition appear in the premises? A comparable
illustration is the derivation of the Cramer Rao Inequality. It is easy to simply
use the conclusion Ie2 1. However the argument is, E, W Ie2 1. Likewise it
is possible to derive many of the laws of fundamental physics using Friedens

EPI principle. But the fully stated argument is represented by R1. Is it

possible that Stengers argument might entail another version of the rational
tuning of the laws of physics to the requirement of observation?
Lastly, Stenger regards the laws as reflecting underlying symmetries and
these symmetries are taken to be symmetries of the void, since on his view,
nothing is more symmetrical than nothing. But symmetry is unstable and so
there spontaneously occurs symmetry breaking that eventually leads to our
universe. Accepting this part of Stengers view for the moment, it may be
positioned in relation to the discussion of evolutionary cosmology, above, as the
origin of all that is in Cosmo. It extends EC. It may help explain L, but this
will still not explain R1. These points set up an interesting discussion with
Stengers work.

34

5. A brief note on Kant, Hume and Darwin

Kant says of the physico-theological proof for the existence of God,

usually called the argument from design or the teleological argument, that it,
deserves to be mentioned with respect. It is the oldest, the
clearest, and the most accordant with the common reason of mankind.
It enlivens the study of nature, just as it itself derives its existence
and gains ever new vigour from that source. 46

That argument rests on analogy with human art. A watch points to a

watchmaker, a house to an architect. In these examples the designer
works with raw materials. Nothing is said about the origins of these
materials. Kant concludes,
[t]he utmost therefore this argument can prove is an architect of the
world, who is always very much hampered by the adaptability of the
material in which he works,not a creator of the world, to whose idea
all things are subject. 47

This paper conforms to that limit since it claims only that the world is
designed, but leaves open the question of whether the existence of the
universe is metaphysically contingent and therefore largely leave open the
question of whether the designer is also the creator of the world. 48

46
47
48

Kant, I., (trans. Smith, N.K.),The Critique of Pure Reason, (1933), B561 2.
Ibid. B656, (italics original).
The question is addressed by Craig, W.L., retrieval of a cosmological argument for
God, in, Naturalism and Cosmology, Craig, W.L. and Moreland, J.P., eds., Naturalism,
a Critical Analysis, (2000), Chapter 9. My preferred approach is represented by
Lonergan, B. , Insight, (Dartman Longman and Todd, London, 1958), Chapter XX, and
by Ward.K., Rational Theology and the Creativity of God, (Basil Blackwell, Oxford).
In different ways each authors approach the argument for God from an inquiry into
inquiry.

35

At the risk of oversimplification I summarise one major strand of

Kants thought as follows: whether as a presupposition of inquiry that
aims to attain a whole body of knowledge concerning nature 49 , or in the
distinction between organisms and aggregates in nature 50 , reason finds
indispensable, as a presupposition, the idea of nature as the product of a
highest intelligent cause acting according to purpose 51 , utterly different
from though doubtless thought on analogy with human art. 52 Natural
phenomena are treated as if they are all purposively interrelated. For
example, Kant asserts that we must,

consider from a teleological point of view not only merely certain

parts of nature, such as the distribution of land, , but make this
systematic unity of nature completely universal, in relation to the idea
of a supreme intelligence. For when we treat nature as resting on
purposiveness, in accordance with universal laws, from which no special
arrangement is exempt, however difficult this may be to establish in
any given case. We then have a regulative principle of the systematic

49

50

51

52

This unity of reason always presupposes an idea, namely, that of the form of a whole
of knowledge a whole which is prior to the determinate knowledge of the parts and
which contains the conditions that determine a priori for every part its position and
relation to the other parts. This idea accordingly postulates a complete unity in the
knowledge obtained by the understanding, by which this knowledge is to be not mere
contingent aggregate, but a system connected according to necessary laws. Critique
of Pure Reason, op.cit., B673.
Kant, I, Critique of Teleological Judgement, Ak., xx, 217, 219); cited by McFarland,
J.D. , Kants Concept of Teleology (University of Edinburgh Press, Edinburgh, 1970),
pp.93 96.
This highest formal unity, which rests solely on the concept of reason, is the
purposive unity of things. The speculative interest of reason makes it necessary to
regard all order in the world as if it had originated in the purpose of a supreme
reason. Such a principle opens out to our reason, as applied to the field of
experience, altogether new views as to how the things of the world may be
connected according to teleological laws, and so enables it to arrive at their
greatest systematic unity. The assumption of a supreme intelligence, as the one and
only cause of the universe, though in the idea alone, can therefore always benefit
reason and can never injure it. Critique of Pure Reason, op.cit., B 714 715.
Critique of Judgment, Ak., v: 182; Critique of Teleological Judgment, op. cit., 64,
370.

36

unity of teleological connection.. 53

These purposive connections in nature are to be considered

supplementary to mechanical connections, not a substitute for them. 54
Thus according to Kant,

The worst that could happen would be that where we expected a

teleological connection (nexus finalis), we find only mechanical or
physical connection (nexus effictivus). In such a case we merely fail to
find the additional unity; we do not destroy the unity upon which
reason insists in its empirical employment. 55

What of the relation between mechanism and teleology?

It is true that in teleology we speak of nature as if its purposiveness

was designed; but to avoid all suspicion of presuming in the slightest to
mix up with ours sources of knowledge something that has no place in
physics at all, namely a supernatural cause, we refer to design in such
a way that this design is ascribed to nature, that is, to matter. In this
way there can be no misunderstanding , since no design in the proper
meaning of the word can be ascribed to lifeless matter. We thus give
notice that this word here only expresses a principle of the reflective,
not of the determinant judgement, and consequently is not meant to
introduce a special ground of causality, but only to assist the
employment of reason by supplementing investigation of mechanical
laws by the addition of another method of investigation. 56

According to Kant, the teleological view point helps us seek for purposive
connections in nature, that we might not otherwise have noticed, but

53
54
55

56

Critique of Pure Reason, op.cit., B719.

McFarland (1970), op. cit., p.33.

Ibid., p.33, cited by McFarland from, The Critique of Pure Reason, trans. N.Smith,
1933, B715 716.
Critique of Teleological Judgment, 68: 383; see also, ibid., 66: 376.

37

which might be otherwise explained. One such connection, of interest in

view of my use of extremum principles, is Kants discussion of Maupertuis
principle of least action, discovered in 1746. Kant noted,

Maupertuis proved that the most universal laws according to

which matter in general operates are themselves subject to an
overriding rule according to which the greatest economoy is
observed in every action. 57

Such discoveries, or the natural purposiveness of organisms, provide

examples of the expected teleological connection in accord with viewing
the nature as designed by a higher intelligence. But even so we have no
basis for saying that there is such an intelligent being. For to argue so,
we would have to establish the prior proposition, Organisms are not
possible except through the operation of an intelligent cause. 58 This,
says Kant, we cannot possibly establish, since,

strictly speaking we do not observe the purpose in nature as

designed, but only read this conception into the facts as a guide to
judgment in its reflection upon the products of nature. 59

McFarland points out a parallel between Kant and Hume on this point.
Hume argues that, to say that all this order in animals and vegetables
proceeds ultimately from design is begging the question 60 because we

57

58
59
60

Kant, I., Only Possible Argument for a Demonstration of the Existence of God, (Ak.,
II, 98 99).
Critique of Teleological Judgment, op. cit., 75: 399.
Ibid., 75: 399.
Ibid. p.126, McFarland cites, Humes, Dialogues Concerning Natural Religion, p.179.
Hume also accepts the principle the nature does nothing in vain one of the roots of
the least action principle ibid. p.214.

38

would first have to prove that they are impossible in any other way. 61 All
the appearance of purposiveness, final causality, teleology or design in
nature doesnt allow us to argue that nature is in fact purposive or
designed, as Darwin vividly demonstrated.
There are other important comparisons between Kants argument and
my argument above. According to Kant, teleology is a regulative idea
introduced as a presupposition by which the highest form of systematic
unity of knowledge may be envisaged if not attained. No such
presupposition is at work in the argument above. Of course, we readily
recognize the growing systemic unity of theoretical and empirical
knowledge that may well be compared with a complex cross-word
puzzle, 62 though with no presumption of an intelligent agent designing
the cosmic puzzle. This recognition, however, has no role in the above
argument.
On the other hand I agree with Kants view that mere appearances of
design could be explained in other ways. As noted above Maupertuis
principle of least action is entirely consistent with the view that laws of
physics in standard form are no more than second order differential
equations, which, as it just turns out, may be conveniently and elegantly
recast in Lagrangian or Hamiltonian form. To read this recast form of
physical laws as evidence for final causes in nature begs the question as
to why such a reading should be preferred to a reading entirely in terms
of efficient causality. At the very least we would need independent
evidence for preferring the former reading rather than the latter. This,
however, is precisely what my argument provides, starting with Friedens
Fisher information approach to physics.
61
62

McFarland, (1970), op. cit., p.126.

See Haack, (1995), op. cit., especially Chp. 4.

39

I agree with Kant that we cannot argue from the mere appearances
of purposiveness or design in nature to an intelligent being who is the
designer, because we cannot be sure there is no other way to explain the
appearances.

But the argument above has a different structure from

Kants argument. It does not start from a presupposition of teleology as

a view point for organizing our knowledge, nor does it point to the
appearances of design in nature. Rather, it shows that the result R1
logically cannot be explained within the resources of the natural sciences.
Specifically this means physics, cosmology, and evolutionary theory. The
latter is part of the reason Darwinism is not a problem for this
teleological argument. The conclusion holds for any development of the
natural sciences in which physical causality remains blind.
This differs from Kants view that there are aspects of nature that
appear to fit the teleological stand point but are explainable in terms of
mechanical causality. Contrary to Kant, in my argument it is the empirical
laws of fundamental physics, which underwrite causal explanations, that
are themselves explained, and this by first explaining how the extremum
principle, Maupertuis principle of least action, physically comes about,
from premises that non-trivially include Fisher information I and
idealised empirical inquiry E.
Lastly, the rational agent to which my argument to design leads, does
not, in turn, immediately incur the burden of a call for its explanation. A call
for an explanation of this agent must be argued, just as I have justified
further seeking an explanation of R1.

40

6 Conclusion

The paper has argument has obtained the following main results.

R1:

E, W, I, AOA

If we assume the idea of idealise empirical inquiry conducted by

embodied rational inquirers, E , in principle anywhere in the universe,
and some other assumptions, then operation of the laws of
fundamental physics L may be explained and their mathematical forms
derived.
R2:

On the proposed account of a physical theory, if TPhys is a physical

theory it logically cannot explain R1.

R3:

It is unreasonable to treat R1 as a brute fact.

R4:

Given R2 and R3 it is reasonable to seek further for an

explanation of R1.

R5

R1 can be explained by minimally assuming that some kind of

rational agent has structured some basic stuff to form this

universe, (one among many) of which our best physics gives an
account, and characterised by W, I, AOA, and laws L, in order that
the universe can be known through empirical inquiry by embodied
rational agents.

Why are the laws of fundamental physics the way they are?.
In order for the universe to be knowable through idealised empirical inquiry
by embodied rational agents.

41

Appendix 1

Laws of physics that Frieden has derived

L1
L2
L3
L4
L5
L6
L7
L8
L9
L10
L11
L12
L13
L14
L15

Lorentz Transformation.
Relativistic mechanics, including the energy equation.
Klein-Gordon Equations
Time independent Schrodinger Equation non-relativistic limit
of L3.
Dirac Equation.
Heisenbergs Uncertainty Principle.
Maxwells Equations.
The Field Equations of General Relativity.
Newtons Laws, by approximation from L2 and
L8 including the Virial Theorem.
The Maxwell-Boltzman velocity law of classical statistics.
The power spectral 1/f noise.
The Wheeler-Dewitt equation for a pure radiation universe.
The wave equation for quantum chromodynamics.
The EPR-Bohm effect describing entangled states.
The Higgs mass effect.

Frieden, B.R., and, Gatenby, A.R., eds., Exploratory Data Analysis Using
Fisher Information, (Springer, London, 2007), pp. 28-29.

42

Appendix 2
B.R. Friedens main publications 1988- 2007
1988, Frieden, B. R., Applications to optics and wave mechanics of the criterion of maximum CramerRao bound, Journal Of Modern Optics, 35, No.8, p.1300;
1989a, Frieden, B. R., Fisher Information as the basis for the Schrodinger wave equation, American
Journal of Physics, 57, (11), pp.1004 1008;
1989b, Frieden, B. R., Fisher Information as the basis for diffraction optics, Optics Letters, 14, No.4,
pp.199 201;
Frieden, B. R., Fisher Information, disorder, and the equilibrium distributions of physics,
Physical Review A, 41, No.8, pp.4265 4276;
1991, Frieden, B. R., Fisher Information and the Complex Nature of the Schrodinger Wave Equation,
Foundations in Physics, 21, No.7, pp.757 771;
1992, Frieden, B. R. Fisher Information and uncertainty complementarity, Physics Letters A, 169,
pp.123 130;
1994, Frieden, B. R. and Hughes, R. J. , Spectral 1/f noise derived from extremised physical
information, Physical Review E, 49, No.1, pp.2644 2649;
1995a,Frieden, B. R.,Physical Information and the Derivation of Electron Physics, in Hawekes P.E.,
(ed.), Advances in Image and Electron Physics, Volume 90 ,( Academic Press, San Diego,
pp.123 204;
1995b Frieden, B. R. and Soffer, B. H., Lagrangians of physics and the game of Fisher-information
transfer, Physical Review E, 52, No.3, 1995b, pp.2274 2286;
1996, Frieden, B. R. and Cocke, W. J. , Foundation for Fisher Information-based derivations of
physical laws, Physical Review E, 54, No.1, 1996, pp.257 260;
1997, Cocke, W. J. and Frieden, B. R. , Information and Gravitation , Foundations of Physics, 27,
No.10, pp.1397 1412;
1998, Frieden, B. R., Physics From Fisher Information: A Unification, (Cambridge University Press,
Cambridge);
1999a Frieden, B. R., Plastino, A., Plastino, A. R., and Soffer, B. H., Fisher-based thermodynamics:
Its Legendre transform and concavity properties, Physical Review E, 60, No. 1, pp. 48 53;
1999b Frieden, B. R. , F-Information, A Unitless Variant of Fisher Information, Foundations of
Physics, 29, No.10, pp.1521 1541;
2000a Frieden, B. R. and Soffer, B. H., A Critical Comparison Of Three Information Based
Approaches To Physics, Physics Letters A, 304, pp.1 7;
2000b Frieden, B. R. and Plastino, A.,Composite fermion particles and Fisher information, Physics
Letters A, 272, pp.326 332;
2001a Frieden, B. R., Plastino, A. and Soffer, B. H.,Population Genetics from a an information
perspective, Journal of Theoretical Biology, 208 ,pp.49 64;
2001b Frieden, B. R. and Plastino, A., Higgs mass generation from the standpoint of information,
Physics Letters A, 278, pp.299 306;
2002a Frieden, B. R., Relations between parameters of a decoherent system and Fisher information,
Physical Review A, 66, Article No. 022107;
2002b Frieden, B.R.and Soffer, B. H., Black holes and optimum coding, Physics Letters A, 304,
pp.1 7;
2002c Frieden, B. R., Plastino, A. and Plastino, A. R., Schrodinger link between non-equilibrium
thermodynamics and Fisher information, Physical Review E, 66(4), Article No. 046128 Part2;
2002d Frieden, B. R., Plastino, A. and Plastino, A. R.,Non-equilibrium thermodynamics and Fisher
information: An illustrative example, Physics Letters A, 304, pp.73 78;
2003 Flego, S. P., Frieden, B. R. and Plastino, A., et al., Non-equilibrium thermodynamics and
Fisher information: Sound wave propagation in a dilute gas, Physical Review E, 68, Article
No. 016105;
2004a Hawkins, R. J. and Frieden, B. R., Fisher information and equilibrium distributions in
econophysics, Physics Letters A, 332, pp. 126 130.
2004b Frieden, B. R. Science from Fisher Information, A Unification, (Cambridge University
Press, 2004).
2007, Frieden, B.R. and, Gatenby, R.A., eds., Exploratory Data Analysis Using Fisher Information,
(Springer, London).

43

Appendix 3
New results predicted by Frieden and colleagues
. The predictions regarding free quarks and the combination of
quarks. Frieden, B. R., Plastino, A. (2000)
. On population genetics. Frieden, B. R., Plastino, A. and Soffer,
B. H.(2001)
. The prediction of an upper bound to the mass MH of the Higgs
particle H0 and additional bosons to H0, Z0 and W. Frieden, B.
R. and Plastino, A. (2001b)
. Prediction of enhanced number of black holes with large mass,
resulting in an increased occurrence of gravitational lensing, and an
enhanced ability to entrap relatively distant stellar objects.
Frieden, B. R. and Soffer, B. H. (2002).
. A relation between the Weinberg angle W and the Cabbibo angle
C. The standard model of elementary particles does not predict
a relation between the two angles, whereas Frieden (2004) does.
. In biophysics: the prediction that the power of the grow process
for cancer is the Fibonacci constant. Frieden, (2004).

44