Definition and Some Properties of Elementary

Functions

May 25, 2016

I. Introduction
The term “elementary function” is very often mentioned in many math classes
and in books, e.g. Calculus books. In fact, the vast majority of the functions
that students and scientists come across are elementary functions of a real vari-
able. However, there is a lack of a precise mathematical definition of elementary
functions. Only a few authors in their textbooks, e.g. Stewart in his Calculus
books try to give a description of elementary functions. Unfortunately, these
descriptions are not given properly. For example, from the descriptions of ele-
mentary functions in Stewart’s book, one could conclude that the function
(
sin x , x ≤ 5
f (x) =
ln x , x > 5
is elementary!! This is simply incorrect.
Thus, this note is written to introduce a precise mathematical definition
of elementary functions of a real variable. It is not claimed to be an original
research article, but rather a note that could serve the students to see a proper
mathematical definition of the term Elementary Function of a Real Variable.
After the definition is introduced, it is easy to see that the elementary functions
of a real variable possess properties that could greatly simplify the mathematical
analysis needed to be done on them. Also, many problems in mathematics deal
with elementary functions or even if the functions are non-elementary, very often
the studying of these non-elementary functions leads to elementary functions.
According to the definition of elementary functions given in this note, there are
functions that are usually considered to be non-elementary, but are elementary
functions according to the definition. Since the problems are much simplified
when the functions involved are elementary functions, it is a good idea to have
a precise mathematical definition of the elementary functions. This definition
and some more properties of elementary functions are provided in this note.
The theorems and properties of elementary functions given in this note allow
for problems of continuity of functions, which often arise in calculus, to be
reduced to finding the set of admissible values for a given elementary function.

1
Then the properties of the fundamental elementary functions can be applied to
finding the set of admissible values for any given elementary function, which
becomes the set of points for which the elementary function is continuous. In
addition, more difficult problems, such as those involving piecewise functions
like the function above, may be reduced to considering continuity of several
elementary functions within a restricted domain and checking for continuity
using the definition at certain points.

II. Definitions
Definition 1. The following eight functions are referred to as the fundamental
elementary functions of a real variable,

f1 (x) = c, c ∈ R, with domain D ⊆ R

f2 (x) = x, with domain D ⊆ R

1
f3 (x) = , with domain D ⊆ R {0}
x
√ n n+1
f4 (x) = n
x, n ∈ N, if ∈ N then domain D ⊆ [0, +∞) , and if ∈ N then domain D ⊆ R
2 2
f5 (x) = sin x, with domain D ⊆ R
f6 (x) = ex , with domain D ⊆ R
f7 (x) = ln x, with domain D ⊆ (0, +∞)
f8 (x) = arccos x, with domain D ⊆ [−1, 1]

Definition 2. For any two functions (of a real variable) f (x) and g(x) with
domains Df , Dg and ranges Rf , Rg , respectively, the following operations are
called The Fundamental Elementary
T Operations on Functions:
1. Addition: ∀x ∈ D = Df Dg 6= ∅, f (x) and g(x) are both defined and
have values a and b (a ∈ Rf and b ∈ Rg ) respectively. Thus, ∀x ∈ D there is a
unique corresponding real number c = a + b. Hence, we define a new function

O1 (x) = f (x) + g(x)

T
where the domain of O1 (x) is D = Df Dg , which is called the sum of f (x)
and g(x).
2. Multiplication: In a similar fashion as in 1, we define

O2 (x) = f (x) · g(x)

T
where the domain of O2 (x) is D = Df Dg . O2 (x) is called the product of f (x)
and g(x).

2
 3. Composition of Functions: ∀x ∈ Df and Rf ⊆ Dg , f (x) is defined and
has a value f (x) = a, a ∈ Rf . Since x ∈ Df , a ∈ Rf ⊆ Dg , hence a ∈ Dg . Since
a ∈ Dg , g(a) = g(f (x)) is defined and has value g(a) = g(f (x)) = b. Thus, we
define O3 (x) to be the composite function of f (x) and g(x) iff
(
Rf ⊆ Dg
O3 (x) = g(f (x)), x ∈ DO3 (x) = Df

Definition 3. A function F (x) with domain D is called an elementary function,

if it can be obtained from one and the same set of fundamental elementary
functions using a finite number of fundamental elementary operations in one
and the same way for all x ∈ D.
Remark. If a function is elementary, then it may be written with only one
formula for all x ∈ D.
Definition 4. For a function f (x) defined in a domain D, a point a is called
an isolated point of D, iff
1. a ∈ D and, S
2. ∃ > 0 such that ∀x ∈ (a − , a) (a, a + ), x ∈
/ D.

III. Theorems
Theorem 1. Subtraction could be obtained from applying the fundamental ele-
mentary operations, i.e. subtraction is an elementary operation.
Proof. Let f (x) and g (x) be elementary functions. Then, f (x) − g(x) = f (x) +
(−g(x)) = f (x) + (−1)(g(x)). Thus, the theorem follows.
Theorem 2. Division could be obtained from applying the fundamental elemen-
tary operations, i.e. division is an elementary operation.
Proof. Let f (x) and g (x) be elementary functions. Then, fg(x)
(x) 1
= f (x) · g(x)
which is the multiplication of two elementary functions. Thus, the theorem
follows.
Theorem 3. All polynomial functions are elementary functions.
Proof. A general polynomial function is defined as

f (x) = an xn + an−1 xn−1 + ... + a1 x + a0

where n ∈ N, an 6= 0, aj ∈ R, j = 0, 1, ..., n, and Df ⊆ R. For any j ∈ N such
that 0 ≤ j ≤ n, xj is obtained from j multiplication(s) of x and multiplication is
a fundamental elementary operation. Thus xj is an elementary function. Since
aj ∈ R (j = 0, 1, 2, ..., n), aj is an elementary function. Then multiplication
(a fundamental elementary operation) of two fundamental elementary functions
is an elementary function. Since addition of a finite amount of elementary
functions is an elementary function, the theorem is proven.

3
Theorem 4. All rational functions are elementary functions.
Proof. A rational function could be defined as

f (x)
r(x) =
g(x)
where f (x) and g(x) are polynomial functionsTwith domains Df and Dg respec-
tively. Then the domain of r(x) is Dr = Df Dg \ {x ∈ Dg |g(x) = 0}. Then
division is obtained from applying the fundamental elementary operations by
theorem 2. Since polynomial functions are elementary functions, the theorem is
proven.

Theorem 5. All rational powers of elementary functions are elementary func-

tions.
Proof. Let f (x) be an elementary function with Df . Then, a rational power
m
function may be defined as f (x) n for some m, n ∈ Z (n 6= 0) with domain D. D
non-negative ∀x ∈ D if n is even and m is odd. Then
must be such that f (x) is p
m m
this may be rewritten as n (f (x)) . Then f (x) is multiplication
√ of f (x) m
times so that it is an elementary function. In addition, n x is an elementary
function by definition. Thus, composition of these two is an elementary function,
and the theorem is proven.

Theorem 6. All trigonometric functions are elementary functions.

Proof. We
may define the function cos x with domain D ⊆ R by cos x =
sin x + π2 . Then, x + π2 is addition of two
fundamental elementary operations
π
and thus, is elementary. Then, sin x + 2 is the composition of an elementary
function with an elementary function. Hence, cos x is elementary as was to be
shown.
sin x cos x 1
Then, we may obtain tan x = cos x , cot x = sin x , sec x = cos x , and csc x =
1
sin x .

Theorem 7. All inverse trigonometric functions are elementary functions.

Proof. First, arcsin x = π2 − arccos x, so arcsin x is an elementary function.
To obtain arccotx, let arccotx = t so that x = cot t where x ∈ (−∞, +∞) and
t ∈ (0, π). Then, cos t = sin t·cot t. For t ∈ (0, π), sin t > 0 and 1+cot2 t = sin12 t .
1 x x
Thus, sin t = √1+x 2
. Therefore, cos t = √1+x 2
. Hence, t = arccos √1+x 2
so that
x
we may define arccotx = arccos √1+x2 where x ∈ (−∞, +∞). Since this is the
composition of several elementary functions, it is clearly an elementary function.
Clearly arctan x = π2 − arccotx, so arctan x is an elementary function as well.
Similarly, we may define arcsecx = arccos x1 and arccscx = arcsin x1 where
x ∈ (−∞, −1] ∪ [1, +∞), both of which are also elementary functions.
Theorem 8. All logarithmic function of elementary functions are elementary
functions.

4
 ln v(x) 1
Proof. logu(x) v(x) = ln u(x) = ln v(x) · ln u(x) for values of x such that u(x) 6= 1,
u(x) > 0, and v(x) > 0. Since ln v(x) is an elementary function (composition of
1
ln x and v(x)) and ln u(x) is also an elementary function (composition of x1 , ln x
1
and u(x)), ln v(x) · ln u(x) = logu(x) v(x) is an elementary function.

Theorem 9. Functions of the form u(x)v(x) (u(x) > 0), where u(x) and v(x)
are elementary functions, are also elementary functions..
Proof. For all values of x such that u(x) and v(x) are defined and greater than 0,
u(x)v(x) = ev(x)·ln u(x) . Since the operations involved are either the fundamen-
tal elementary operations or a combination of them, u(x)v(x) is an elementary
function.
The following statements are theorems that are very often proven in text-
books and thus will only be stated without proofs
1. All eight, f1 (x) - f8 (x), fundamental elementary functions are continuous
everywhere in their domains except at the isolated points and are discontinuous
at the isolated points.
2. The sum of two continuous functions is also continuous everywhere in its
domain except at the isolated points and is discontinuous at the isolated points.
3. The product of two continuous functions is also continuous everywhere
in its domain except at the isolated points and is discontinuous at the isolated
points.
4. The composition of two continuous functions is also continuous every-
where in its domain except at the isolated points and is discontinuous at the
isolated points.
Thus, one has the following important theorem concerning elementary func-
tions:
Theorem 10. All elementary functions are continuous in their domains, except
at the isolated points at which they are discontinuous.
Proof. All of the fundamental elementary functions are continuous in their do-
mains, except at any isolated points at which they are discontinuous. Then,
since all elementary functions may be written as one equation consisting of ad-
dition, multiplication, and composition of the fundamental elementary functions
by definition, all elementary functions are continuous in their domains, except
at the isolated points at which they are discontinuous.
Remark. In some real analysis books, continuity is defined in a way that all
functions are continuous at their isolated points.
Theorem 11. The function
(
x ,x ≥ 0
|x| =
−x ,x < 0
is an elementary function.

5
 √ 2
Proof. Consider x2 . Clearly, x2 = |x| for all
q x ∈ R. Thus, since |x| ≥ 0, we
√ 2
may take the square root and obtain x2 = |x| = |x|.
√
Then we may clearly define |x| = x2 . Since this is the composition of two
elementary functions written in one equation, it is an elementary function.

Theorem 12. For a function f (x) defined as

(
g(x) , x < a
f (x) =
h(x) , x > a
where g(x) is an elementary function in Dg = (−∞, a) and h(x) is an elemen-
tary function in Dh = (a, +∞). The function f (x) with domain Df = R \ {a}
is an elementary function.

Proof. Note that

(
|x| + x 0 ,x < 0
=
2x 1 ,x > 0
and
(
−|x| + x 1 ,x < 0
=
2x 0 ,x > 0
Applying a small shifting in x, the function f (x) could be written as

x − a − |x − a| x − a + |x − a|
f (x) = g(x) + h(x)
2(x − a) 2(x − a)
with domain Df = R \ {a}. Since the formula to calculate the value of f (x)
at any point in the domain Df of f (x) consists of only generalized elementary
functions and elementary operations, f (x) is an elementary function.
Remark. Of course, it is very easy to generalize the above definitions and the-
orems for functions of any finite number of real or complex variables. This
generalization may be done by only including the five fundamental elementary
functions f1 (x) , f2 (x) , f3 (x), f5 (x), and f6 (x), and including taking the
inverse of a function as a fundamental elementary operation. However, the
function |z| (modulus) is not an elementary function of a complex variable z,
even if |x| (absolute value) is an elementary function of a real variable.