Grossman 1979

INT. J. MATH. EDUC. SCI. TECHNOL., 1979, VOL. 10, NO.
4, 525-528
An introduction to non-Newtonian calculus

by MICHAEL GROSSMAN
Department of Mathematics, University of Lowell,
Lowell, Massachusetts, U.S.A.
(Received 1 August 1978)
In this paper we explain the construction of an arbitrary non-Newtonian

calculus from two given complete ordered fields. We then discuss some features
of a few specific non-Newtonian calculi. And finally we speculate as to the
eventual use of the non-Newtonian calculi as alternatives to the classical calculus
of Newton and Leibniz.
1. Introduction
In July of 1967, Robert Katz and I created the first system of non-Newtonian
calculus, which we call the geometric calculus. By August of 1970, we had created an
infinite family of non-Newtonian calculi, each of which differs markedly from the
classical calculus of Newton and Leibniz. Among other things, each non-Newtonian
calculus possesses four operators: a gradient (i.e. an average rate of change), a
derivative, an average, and an integral. For each non-Newtonian calculus, there is a
characteristic class of functions having a constant derivative. In this paper I present a
brief, self-contained description of the construction of an arbitrary non-Newtonian
calculus, and I indicate some properties of a few specific non-Newtonian calculi.
2. Conventions
Let * be an ordered pair of complete ordered fields (A, +, — , x , / , <)and(5, + ,
—, x ,/, <) for which A and B are subsets of the set i? of all (real) numbers. Let a and
P be the isomorphisms from R onto A and from R onto B , respectively; and let i be
the isomorphism from A onto B. Let r and 5 be numbers in A such that r < s. Let / be
the identity function on R. Finally, let us agree that all considered functions have
arguments in A and values in B, unless indicated or implied otherwise.
3. Definitions and results

An en-interval [r,s] is the set of all numbers x in A such that r^x^s, and the
a-extent of [r,s] is s — r.
An a-progression is a finite sequence of numbers alt..., an in A such that
a2-al=a3-a2= ••• =an-an^l
An ^-partition of [r, s] is any ^.-progression whose first and last terms are r and s,
respectively. An a-partition is n-fold if it has exactly n terms.
By using the natural topologies induced by < in A and by < in B, we can define
the concepts of convergence of sequences of numbers in A (tt.-limits), convergence of
sequences of numbers in B (^-limits), and * -limits and * -continuity of functions with
arguments in A and values in B.
The /?-change of a function / over [r, s) is the number /(s) — f (r), which is inB.
0020-739X/79/1004 0525 $02.00 ©1979 Taylor & Francis Ltd 2L 2
526 M. Grossman
A *-uniform function is a *-continuous function on A into B that has the same

jS-change over any two a-intervals of equal a-extent. It turns out that the *-uniform
functions are those expressible in the form i{(m xx) + c}, where m and c are constants
in A and x is unrestricted in A. Clearly, all constant functions on A are *-uniform. In
the special case where a = /? = I, the *-uniform functions are the linear functions, i.e.,
functions of the form mx + c where m and c are constants in R and x is unrestricted
inR.
The following definition is suggested by the well-known fact that the classical
slope of a linear fi .nctiong is equal tog(v) —g(u), where u, v are any numbers such that
V —M = 1.
The *-slope jf a *-uniform function is its jS-change over any a-interval of
a-extent i, where i is the multiplicative identity in A. The *-slope of i{(m x x) + c}
turns out to be i(tt), and the *-slopeof a constant function on A turns out to be 0, the
additive identity in B.
The *-gradie? t of a function/over [r, s] is the *-slope of the *-uniform function
containing (r, f(r)) and (s, f(s)) and turns out to be [f (s) — f (r)]/[i(s) — i(r)].
The *-derivative of / at a number a in A is, if its exists, the *-limit at a of
*
lf( )—f(a)]/U(x)-i(a)] ar>d is denoted by [£)/] (a). The following propositions can
x
be proved: (a) the *-derivative of a *-uniform function equals its *-slope; (b) only *-
*
uniform functions have constant *-derivatives on A; and (c) the operator D is /?-
additive and /^-homogeneous, i.e.,
where bt and b2 are constants in B.

Now it is aapropriate to discuss the *-average, the *-integral, and their
relationships to ;ach other and to the *-derivative.
The p-average of n numbers uu ..., un in B is the number u in B such that
+ ••• +u = u^+ ••• +un

n terms
and is equal to
The *-averaye of a *-continuous function/ on [r,s] is the /?-limit, in B, of the
sequence whose 2th term is the /J-average of /(fli),.. • ,/(«„), where alt..., an is the
w-fold ot-partition of [r, s], and is denoted by Msrf. The operator Msr is jS-additive and
jS-homogeneous.
3.1. The basic theorem of *-calculus

If Dh is *-continuous on [r, s], then its *-average on [r, s] equals the *-gradient of
h over [r, s]; that is,
= [h(s) - h(r)]/b(s) - i(r)]
3.2. The basic problem of *-calculus

Suppose that the value of h is known at an argument r, and suppose that / , the
*-derivative of h , is *-continuous and known at each number in [r, s]. Find h(s).
•An introduction to non-Newtonian calculus 527
Solution
h(s) = h(r) + {[t(s)-i(r)]xMsrf}

The *-integral of a *-continuous function / on [r,s] is defined to be
* * *
[i(s) — i(r)]'xMsrf and is denoted by J*/. The operator J* is j9-additive and /?-
homogeneous. It turns out that jsrf equals the /J-limit, in B, of the sequence whose
nth term is [i(kn) x/Ca^J-f ••• +[i(kn) x f(an_{)\, where ay, •. •, an is the n-fold a-
partition of [r, s] and kn is the common value of a2 — at , a 3 — a 2 > • • • > an ~ an -1 •
3.3. The first fundamental theorem of * -calculus

If / is *-continuous on [r, s], and
^(x) = J^/, for every number x in [r, s]

then
D^ = / , on [r, s].
3.4. T/?e second fundamental theorem of *-calculus

If Dh is *-continuous on [y, s], then
4. Examples
Each choice of specific isomorphisms for a and /? determines a *-calculus. I shall
briefly discuss a few of the many possible choices. First, however, let us make the
following definitions. For each number x:
if x>0
q{x) = 0 if x=0
if x<0
\/x if x # 0
h{x) =
0 if x = 0
The functions /, exp, q, and h are isomorphisms from R onto the complete ordered
fields that we call classical arithmetic, geometric arithmetic, quadratic arithmetic, and
harmonic arithmetic, respectively.
In the geometric calculus, the *-uniform functions are the exponential functions;
the *-gradient is independent of the origin and unit used for measuring the
magnitudes that the arguments and values may represent; the *-derivative is closely
related to the well-known logarithmic derivative; the *-average is the familiar
geometric average; and each of the four operators of the geometric calculus is
multiplicative rather than additive.
In the bigeometric calculus, the *-uniform functions are the power functions; the
•-gradient is independent of the units used for measuring the magnitudes that the
528 An introduction to non-Newtonian calculus
arguments and vilues may represent; the *-derivative is closely related to what
economists refer o as 'elasticity'; and each of the four operators of the bigeometric
calculus is multip licative.
In the anage >metric calculus, the anaquadratic calculus, and some other
*-calculi, the *-ir tegral turns out to be a Stieltjes integral.
In the harmonic calculus, the *-average is essentially the familiar harmonic
average; and in tr e quadratic calculus, the *-average is essentially the familiar root
mean square.
The following table indicates some of the *-calculi that are of special interest.
*-calculus aa fl
Classical / I/
Geometric / exp
exp
Anageometric exp
exp //
Bigeometric exp
exp exp
exp
Quadratic /
i-i
qq
Anaquadratic 1q I
Biquadratic q
q qq
Harmonic / h
>»i
Anaharmonic h
h I
>~*
Biharmonic hh h
5. Conclusion
The non-Newtonian calculi provide scientists with alternatives to the classical
calculus. Presumably, for any given application, the scientist must select an
appropriate calculus (or calculi). Some heuristic guides for making that selection are
presented in Grcssman and Katz ([1], Chap. 9).
At this date we can only speculate as to the fruitfulness of the non-Newtonian
calculi. Perhaps hey can be used to define new scientific concepts, to yield simpler
physical laws, to solve heretofore unsolved problems.
Reference
[1], GROSSMAN, M., and KATZ, R., 1972, Non-Newtonian Calculus (Lee Press).

Grossman 1979

Uploaded by

Copyright:

Available Formats

Grossman 1979

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Grossman 1979

Uploaded by

Copyright:

Available Formats

INT. J. MATH. EDUC. SCI. TECHNOL., 1979, VOL. 10, NO.

An introduction to non-Newtonian calculus

(Received 1 August 1978)

In this paper we explain the construction of an arbitrary non-Newtonian

3. Definitions and results

A -uniform function is a -continuous function on A into B that has the same

where bt and b2 are constants in B.

+ ••• +u = u^+ ••• +un

3.1. The basic theorem of *-calculus

= [h(s) - h(r)]/b(s) - i(r)]

3.2. The basic problem of *-calculus

h(s) = h(r) + {[t(s)-i(r)]xMsrf}

3.3. The first fundamental theorem of * -calculus

^(x) = J^/, for every number x in [r, s]

3.4. T/?e second fundamental theorem of *-calculus

You might also like