Go to other chapters or to other volumes of math
Additional Material for this Chapter
Exercises for this Chapter
Answers to Exercises

Chapter 1
Intuitive and Logical Ideas about Sets

Section 1:   Sets and open statements

Recall that a universe is a collection of all objects to be considered in a discussion. In some discussions the objects are numbers, such as 1,2,3,.... . In geometry the objects are points. The numerical universes in discussions here often are
N5 = {1,2,3,4,5},   N9 = {1,2,3,4,5,6,7,8,9}
and sometimes all the natural numbers   N = {1,2,3,....}. (Natural numbers will be discussed more in volume C.) Most often the geometric universe continues to be is the interior of a drawn rectangle.

[1.1] A set in a universe is a collection of objects that belong to that universe. It is also called a subset of that universe. (The collection may be made by selecting objects from all the objects in the universe. But nothing is removed from the universe.)

(Numerical) Example 1: In the universe N9 make a random collection of numbers, say 6,2,8,5 and then together they form the set {6,2,8,5}. But for convenience the numbers inside the set notation { } are usually listed in increasing order: {2,5,6,8}.
(Numerical) Example 2: In the universe N9 collect the numbers between 3 and 8. This collection becomes the set {4,5,6,7}. But this set is also the truth set of the open statement

x is between 3 and 8
(Geometric) Example 3: In the adjacent figure the universe is all points inside the rectangle. The set is the interior of the circle. If c is the center and ρ (Greek letter rho) is the radius then this interior is also the truth set of the open statement
the distance between x and c < ρ
The point b is inside the circular area because distance between b and c is less than ρ. Point b is assigned a truth value T by the open statement above.

Let subset S be the collection of all selected objects in a universe U. Two extremes are allowed. All objects in U may be selected Then S = U. Therefore, any universe is a subset of iteself. It may also happen that all objects in U are rejected. Then S is empty, S = Φ. Therefore the empty set Φ is a subset of any universe.

To distinguish selected objects from rejected objects in a universe, it may be useful to assign a T to each selected object, and an F to each rejected object. This allows open statements to do selections done in most of the following discussion. Then the truth sets of the open statements form the subsets of the universe. Conversely, for any subset S the open statement

x is in S
has S as its truth set. This open statement is called the trivial open statement that has S as its truth set. Suprisingly, the trivial open set is used many times in following discussions.

[1.2] (Relationship between sets and open statements)
   (a) Every open statement determines a subset of its universe, namely its truth set.
   (b) For every subset of a universe there is an open statement whose truth set equals that subset.

For any set S in a universe U there are many open sets p(x), q(x), r(x), ... whose truth sets equal S. For example:

x > 3 and x < 9,     x + 1 is between 4 and 10,    x is between the roots of x2 - 11x + 24,    x is in {4,5,6,7}
where S = {4,5,6,7}. However, all such open statements must be equivalent to each other, and therefore, equivalent to the trivial open statement   x is in S.

The following "set builder notation" denotes the truth set of an open statement p(x) with universe U:

{x in U | p(x)}
It is the set of all objects in U that are assigned T by the open statement p(x).
For example,   {x in N9 | 3 < x < 8}   =   {4,5,6,7}.



Section 2:   Some operations on sets: intersection, union and complement

In this section certain operations will be discussed. Intuitively speaking, they create "new"sets from sets already existing in a common universe. Special symbols denote these operations. A geometric model in the form of a set or sets of points inside a rectangle often accompany the discussion of each operation. The figure motivates the definition which will state the rule for selection the objects to be in the "new" set. Other examples of the operation will be numerical and often involve the universes N5 and N9 .

Let S be any set in a universe U. Collect all objects outside of S to form a new set ~S. For example, in the universe N9 let   S = {1,3,5,7,9}   then   ~S = {2,4,6,8}.
In the adjacent figure all of the points outside area S form a set ~S.

[2.1] (Complement) The complement of a set in a universe is the collection of all objects in the universe that are not in the given set.
Notation: ~S = {x in U | x is not in S}.
Notation: In a universe U, x is in ~S   if and only if   x is not in S.

If a tilda (~) is "attached" to a set symbol S then the complement ~S of that set is formed. Therefore the tilda (~) is a unary operation on individual sets that forms their complements.
It is obvious that the complement of ~S must be S which contains all objects in the universe not in ~S. Therefore, the complement of the complement of a set is the set itself.   ~ ~S = S.

Another notation for complement is   U\S   (a special form of set subtraction defined in [2.12] below). This notation indicates the need for knowing the universe U in order to determine the complement of S.

If S is empty then ~S must be the entire universe, because no object in the universe can be selected to be in S. Therefore, every object is not in S and must be in ~S.   U\Φ = U.   The complement of the empty set is the entire universe.. Similarly, the complement of the universe is the empty set:   U\U = Φ.

-----

A binary operation accepts two things and produces a third thing. For example, the arithmetic operation of addition accepts 3 and 5 and produces 8. The operation of multiplication accepts 3 and 5 and produces 15. Each operation here accepts two sets, often designated as S1, S2, and produces a set from them. To emphasize coordination with circles in geometric models the sets S1, S2 may be in colors. Then they denote the interiors of red, blue circles. Colors may also be used for sets in N5 and N9.
In any case, there are two trivial open statements

x is in S1,          x is in S2
which together assign any of the following pairs of truth values to each object in the universe:
TT,   TF,   FT,   FF
The selection of all objects that have certain pairs assigned to them will define the binary operation.




The intersection of streets is that portion common to both streets.The adjacent figure shows the intersection of the interiors of a circle and another circle.


[2.2] (Intersection) The intersection of two sets in the same universe consists of all objects that are in both sets.
Notation1: S1 /\ S2 = {x in U | x is in S1 and x is in S2}.
Notation2: x is in S1 /\ S2   if and only if   x is in S1 and x is in S2       (a true statement about universally true equivalence)

The common notation for intersection is an upside down U:
                                             
But for typographical reasons, the notation /\ will be used in all volumes to denote intersection.

Example: In N9 the intersection   {2,3,4,5,6} /\ {4,5,6,7,8} = {4,5,6}.   Simply go through the two sets {2,3,4,5,6}, {4,5,6,7,8} and select those numbers in both sets:

{2,3,4,5,6}, {4,5,6,7,8}.

It is instructive to see the formation of the intersection using truth values. Every number in N9 receives two truth values from the open statements

x is in {2,3,4,5,6},   x is in {4,5,6,7,8}
They assign the following truth values to each number in N9:
1 FF,   2 TF,   3 TF,   4 TT,   5 TT,   6 TT,   7 FT,   8 FT,   9 FF
Select only those numbers assigned TT because only those numbers are in both sets and therefore are in the intersection of the sets. So assign an T to those same selected numbers.) Reject those numbers with the assignments TF, FT, FF because those numbers are outside one or both sets and therefore outside the intersection. So assign an F to those same rejected numbers.) Therefore, for each of the nine numbers, replace double truth values by single truth values according to whether it is selected or rejected:
1 F,   2 F,   3 F,   4 T,   5 T,   6 T,   7 F,   8 F,   9 F,  
The intersection consists of those numbers that have been assigned a T by the open statement
x is in S1   and   x is in S2

From the wording of definition [2.2] it is obvious that the intersection of two sets is the same no matter which set is written first.

[2.3] (Commutativity) The operation of intersection is commutative.
Notation:   S1 /\ S2 = S2 /\ S1.

But what happens to the intersection if one or both sets are empty? For example, what is the intersection {2,3,4,5,6} /\ Φ? The open statements

x is in {2,3,4,5,6},   x is in Φ
make the following assignments of truth values:
1 FF,   2 TF,   3 TF,   4 TF,   5 TF,   6 TF,   7 FF,   8 FF,   9 FF
No number in N9 is assigned TT. Therefore, no number is in the intersection. This means
{2,3,4,5,6} /\ Φ   =   Φ

In general, if either of two sets is empty then their intersection is empty:    S /\ Φ = Φ,      Φ /\ S = Φ,      Φ /\ Φ = Φ
By definition, the open statement    x is in Φ    is universally false. Therefore, any conjunction involving that phrase will have to be universally false.

The intersection of any set and the universe is equal to that set:.    S /\ U = S,      U /\ S = U,      U /\ U = U
Click here to see the an argument supporting this statement.

Intersection of sets has a relationship to nmultiplication of numbers. Click here to see a discussion about this relationship.

The intersection of more than two sets is defined in the obvious way:

[2.4] The intersection of sets in the same universe consists of all objects that are in all the given sets
Notation: x is in S1 /\ S2 /\ S3 /\ ...     if and only if     x is in S1   and   x is in S2   and   x is in S3   and ...

Parentheses surround an operation to be done first. Short arguments can be made to support the two equalities

(S1 /\ S2) /\ S3 = S1 /\ S2 /\ S3,      and      S1 /\ (S2 /\ S3) = S1 /\ S2 /\ S3
Therefore,

[2.5] (Associativity) The operation of intersection is associative.
Notation: (S1 /\ S2) /\ S3 = S1 /\ (S2 /\ S3).

-----------------------


The adjacent figure shows the union of a set and another set. It is the combined interiors of the two sets, including their intersection if they overlap.

[2.6] (union) (In the same universe) the union of two sets is the collection of all objects in either set (or both)
Notation: x is in S1 \/ S2   if and only if   x is in S1 or x is in S2 (or x is in both sets).

The common notation for union is a U:
                                             
But for typographical reasons, the notation \/ will be used in all volumes to denote union.

Example: {2,3,4,5,6} \/ {4,5,6,7,8} = {2,3,4,5,6,7,8}       universe U = {1,2,3,4,5,6,7,8,9} because
  x is in {2,3,4,5,6,7,8}   if and only if   x is in {2,3,4,5,6} or x is in {4,5,6,7,8}
There are three open statements here:

x is in{2,3,4,5,6,7,8}, ,  x is in {2,3,4,5,6},   x is in {4,5,6,7,8}
The last two assign the following truth values to each number in the universe U:
1 FF,   2 TF,   3 TF,   4 TT,   5 TT,   6 TT,   7 FT,   8 FT,   9 FF
But the first open statement assigns the following truth values:
1 F       2 T       3 T       4 T       5 T       6 T       7 T       8 F       9 F
Obviously, only those numbers assigned TT or TF or FT from the last two open statements are in the union. (The first open statement assigns a T to those very same numbers.) The assignments FF exclude numbers from being in the union. (The first open statement assigns an F to those same numbers.)

From the notation following [2.6] it obvious that that an alternation of open statements determines the union of sets. And in doing so, it shows how a union assigns truth values to each object in the universe:

It is sometimes useful to look negatively at a union of two sets. The following is actually equivalent to [2.6] (see [5.6c] in Vollume A, chapter 1):

[2.7] (Not in a union) In a common universe, an object is not in the union of two sets if and only if that object is in neither set.
Notation: x is not in (S1 \/ S2)   if and only if   x is not in S1 and x is not in S2.

Click here for a discussion of duality between intersection and union.


From the wording of definition [2.6] it is obvious that the collection of objects from both unions   S1 \/ S2   and   S2 \/ S1 are the same. In other words, the open statements

x is in S1 \/ S2,        x is in S2 \/ S1
are equivalent. They must determine the same set.

[2.8] (Commutativity) The operation of union is commutative.
Notation:   S1 \/ S2 = S2 \/ S1.


But what is the union {2,3,4,5,6} \/ Φ involving the empty set? The open statements

x is in {2,3,4,5,6},   x is in Φ
make the following assignments of truth values:
1 FF,   2 TF,   3 TF,   4 TF,   5 TF,   6 TF,   7 FF,   8 FF,   9 FF
No number in N9 is assigned FF. Therefore, no number is excluded from the union. this means
{2,3,4,5,6} \/ Φ   =   U
This argument is valid for any set, not just {2,3,4,5,6}, and any universe U.
The union of any set and the empty set is that set..   S \/ Φ = S (Click here to see an indirect proof of this statement.)
The union of any set and the universe is equal to the universe.   S \/ U = U
The proof of this is very similar to that for the union with the empty set. Click here to see the proof.

If 1 represents the universe U and 0 represents Φ then the representations of the four intersections of these two sets produce a strange (Boolean) addition table for 1 and 0. Click here to see this discussion.

The union of more than two sets is defined in the obvious way.

[2.9] The union of sets in the same universe consists of all objects that are in any the sets
Notation: x is in (S1 \/ S2 \/ S3 \/ ...)     if and only if     x is in S1   or   x is in S2   or   x is in S3   or ...

It is sometimeds useful to discuss oobjects not in the union of two sets. It is a form of De Morgan's law.

[2.10] (Outside a union of sets) An object is not in a union of sets all in the same universe if and only if that object is in none of the sets.
Notation:
x is not in S1 \/ S2 \/ S3 \/ ...     if and only if     all the open statements   x is in x is in S1   or   x is in S2   or   x is in S3   or ...

Short arguments can be made to support the two equalities

(S1 \/ S2) \/ S3 = S1 \/ S2 \/ S3,      and      S1 \/ (S2 \/ S3) = S1 \/ S2 \/ S3
Therefore,

[2.11] (Associativity) The operation of union is associative.
Notation: (S1 \/ S2) \/ S3 = S1 \/ (S2 \/ S3).

------

Occasionally useful is a type of subtraction of sets. It is a generalization of the idea of complement [2.1]. A superset replaces the universe. The back slash (\) is used again.

[2.12] (Set difference) The set difference between a second set and a first set is the collection of all objects in the first set that is not in the second set.
Notation: x is in (S1\ S2)   if and only if   x is in S1 and x is not in S2.

Intuitively speaking, the second set takes away objects from the first set. Also, it is the remaining part of the first set after the intersection of the two sets is removed.

Example: Let N9 be the universe.

{2,3,4,5,6}\{4,5,6,7,8} = {2,3}
The numbers 1,7,8,9 play no role in determining this difference {2,3} of sets. Intuitively speaking, the numbers 4,5,6 are removed from the set {2,3,4,5,6}. But {4,5,6} is the intersection of the two sets. Intuitively speaking, the set difference can be obtained by removing the intersection of two sets from the first set:
S1\S2 = S1\(S1 /\ S2)
The set difference {3,5,7}\{,3,4,5,6,7} = Φ } because 3,5,7 are removed from the first set {3,5,7}.

As expected, S\Φ = S because nothing is being taken away from S.



Section 3:   Some relationships between two sets in the same universe

In the geometric model the interior of a rectangle is the universe and sets are the interiors of closed curves, usually circles, inside the rectangle. In each of rectangles in the following figures, except for Fig 2, there are two sets S1 and S2 that are the interiors of red and blue circles.
Statements may be made about relationships between the circles. The statements may be true or false. Therefore, they are logical statements.
   In Fig 4,   S1 is completely inside S2.
   In Fig 1,   S1 and S2 coincide.      But it can still be said that S1 is completely inside S2
   In Fig 3,   S1 and S2 are completely separated.
These expressions "is completely inside", "coincide" and "are completely separated" state relationships between the sets. All three relationships just given are true. However, the following statement
   In Fig 5,   S1 is completely inside S2
is obviously false. Part of S1 lies outside of S2. In that part are points inside S1. And those same points are not in S2.

Any object in the universe that makes a relation between sets false is called a counterexample to the relation.

Sets are determined by the objects in them. The sets S1 and S2 are truth sets of the trivial open statements

(&&)        x is in S1,        x is in S2
They assign a pair of truth values to each object in the universe. All pairs are selected from     TT,   TF,   FT,   FF.
A relation between sets determines what pair of truth values is assigned to each object in a universe.

-----

The most simple relationship is equality. In Fig 1 the two sets coincide:   S1 = S2. As a result, points inside them are assigned TT by (&&). All other points are outside the interiors and are assigned FF. So (&&) assigns a pair of equal truth values to every object in the universe.

[3.1] (Equality of sets) Two sets in the same universe are equal if and only if they contain exactly the same elements.
Notation 1: S1 = S2   if and only if   the open statements x is in S1, x is in S2 assign equal truth values to each object in the universe.
Notation 2: S1 = S2   if and only if   the equivalence   x is in S1 <==> x is in S2   is universally true.

Example: In universe N9, {5,7,6,8,4} = {4,5,6,7,8} because they contain exactly the same numbers. Any number in N9 is in both of them (is assigned TT) or it is in neither of them (is assigned FF). But notice that this shows the listing of the objects in a set can be written in different arrangements of the same objects (permutations).

Click here to see a discussion supporting the fact that all empty sets are equal. In other words, in any universe there is only one empty set. It is unique.

---------

For equality (&&) assignes equal truth values to each object in the universe. The other extreme is that (&&) assigns opposite truth values. But this means that each set is the complement of the other.

[3.2] (Complementary sets) Two sets inside the same universe are complementary if and only if one is the complement of the other.

In Fig 2, S2 contains everything not in S1 and conversely. Every object in the universe is assigned a pair of different truth values: TF or FT from the open sets in (&&).

Inside the universe N9 the two sets {1,3,5,7,9} and {2,4,6,8} are complementary. It is obvious that the complement of the complement of a set is that set. Click here for a discussion supporting this statement.

[3.3] Two sets in the same universe are complementary if and only they satisfy both both of the following conditions:
   (a) their union is the universe;
   (b) their intersection is empty.

Notice that {1,3,5,7,9} /\ {2,4,6,8} = Φ   and that   {1,3,5,7,9} \/ {2,4,6,8} = N9, the universe.

Click here for a discussion supporting [3.3].

------

Relax condition (a) in [3.3] to get a more general situation as shown in Fig 3. The two sets do not overlap. There is nothing in their intersection. No point in the universe can be assigned TT by the open statements (&&). However if their union is not the entire universe then some objects outside the union will be assigned FF.

[3.4] (Disjoint sets) In the same universe a pair of sets are disjoint   if and only if   their intersection is empty
Notation: S1, S2   are disjoint if and only if the open statements

x is in S1,       x is in S2
are incompatible (no object in the universe is assigned TT by the trivial open statements above).

A pair of complementary sets are always disjoint, but not conversely. Fig 3 shows disjoint sets that are not complementary.

-------

Fig 4 shows an important situation: all of one set may be part of another set. This means that all the points in first set are also in the second set.

[3.5] (Inclusion) One set is included in a second set if and only if every object in the first set is also in the second set.
Notation: S1   <   S2   if and only if   x is in S1 => x is in S2.

If one set is included in a second set then the first set is called a subset of the second set. Obviously, every set in a universe is a subset of that universe.

Example: Let the universe N9 = {1,2,3,4,5,6,7,8,9}.
It is obvious that {4,5,6} is included in {3,4,5,6,7,8}. Also the implication

if x is in {4,5,6} then x is in {3,4,5,6,7,8}
is obviously .true.

If two sets are equal then the content of either set is included in the other. If one set is equal to a second set then the first set is included in the second set. The underline in the notation   < allows inclusion to extend to equality:

S1 < S2    means    S1 < S2   or   S1 = S2
read "weak inclusion means strict inclusion or equality."

The reverse of inclusion is containment.
A set S2 contains a set S1 if and only if S1 is included in S2
Notation: S2 >  S1

If one set contains another then the first set is sometimes called a superset of the second set.

If S1 = S2 then trivially everything in S1 is in S2. Therefore every set is included it itself. It is a subset of itself.
The symbol   <   is used in these discussions to mean inclusion but not equality. It is called strict inclusion. The statement

  S1 < S2
is true for the sets in Fig 4 but not true for the sets in Fig 1. The first set is called a proper subset of the second set.
In contrast, the symbol   < allows both strict inclusion and equality. (It is sometimes called a weak inclusion.) Therefore, the statement
  S1 < S2
is true for both Fig 4 and Fig 1.

There is also strict containment, S2 > S1. In this situation S2 contains all of the objects in S1 plus one or more objects not in S1. The statement

S2 > S1
is true for FIg 4 but not true for Fig 1.

------------

[3.6] (Summary of sets and associated open statements)

From sets S1, S2 to associated open statements p(x), q(x): any open statement equivalent to x is in S1, x is in S2 respectively.
From open statements p(x), q(x),\ to sets S1 = truth set of p(x), S2 = truth set of q(x).



Section 4:   More on inclusion, implication and proof

The following three figures are needed in this section:
S1, S2 continue to denote the interiors of the red and blue circles.
Also used are the trivial open statements
(&&)        x is in S1,        x is in S2
which assign a pair of truth values to each point in the universe.

In Fig 4 and Fig 1, set S1 is entirely inside set S2. But in Fig 5, set S1 is not entirely inside set S2. Part of S1 lies outside of S2. There are points in that part. Any of those points is called a counterexample to the inclusion S1 < S2. If point b is one of those counterexamples, then (&&) assigns truth values TF to point b. That point b is also a counterexample to (the universality of) the implication

x is in S1 ==> x is in S2
because b produces a false component
b is in S1 --> b is in S2
(true hypothesis, false conclusion).

The following is seemingly trivial yet sometimes is useful as a basis for arguments supporting inclusions and universally true implications of open statements.

[4.1] (Inclusion) One set is included in another if and only if there are no counterexamples to this inclusion.
Notation1: S1   <   S2   if and only if   there exists no object b such that b is in S1 and b is not in S2.
Notation2: S1   <   S2   if and only if   no object in the universe is assigned TF by the open statements   x is in S1, x is in S2 respectively.

Example: Again let the universe be N9 = {1,2,3,4,5,6,7,8,9}.
A counterexample to {1,2,3,4} < {3,4,5,6,7,8} is 1. So is 2. Each is in {1,2,3,4} but not in {3,4,5,6,7,8}. The remaining numbers 5,6,7,8,9 in U do not qualify as counterexamples, because none of them are in {1,2,3,4}. Therefore, they cannot produce false components. The statement about inclusion {1,2,3,4} < {3,4,5,6,7,8} is false because there are counterexamples.

The following statement provides a useful method to prove two sets are equal.

[4.2] (Equality from reverse inclusions) If two sets are included in each other then the two sets are equal.
Notation1: if   S1   <   S2   and   S2   <   S1    then    S1   =   S2
Notation2: if   p(x) ==> q(x) and q(x) ==> p(x)   then   p(x), q(x) are equivalent.

The inclusions if   S1   <   S2   and   S2   <   S1   prevent any object from being assigned TF or FT. Therefore every object is assigned equal truth values, TT or FF, which is what happens only with equal sets and their trivial equivalent open statements.

The implication q(x) ==> p(x) is called the converse of the implication p(x) ==> q(x). The converse of a (universally) true implication may or may not be (universally) true. For example, in universe N9, x <5 ==> x < 8 is universally true, but its converse x < 8 ==> x < 5 has a counterexample, namely 7. Upon substitution for x, 7 makes the hypothesis x < 8 true, but the conclusion x < 5 false.

Statement [4.1] can be used to prove that the empty set Φ is a subset of any set S in any universe U:

Φ < S.
The basic argument is that it is impossible to find a counterexample to this inclusion. Suppose there is a counterexample b. Then by definition   b is in Φ and b is not in S. But there is no b in Φ because, by definition, Φ is empty. So no counterexample b can exist.

This argument uses a typical approach for denying the existance of a counterexample: assume that there is a counterexample and then arrive at an impossible situation because of that assumption. Such an argument is called an indirect argument.

Both the empty set and the set itself are called the trivial subsets of the set itself. Often they are less interesting than the other subsets.

It is true that 2 < 5. But it is also true that -5 < -2. Negation of numbers interchanges the sides of an inequality. (This can be proven by subtraction of the smaller number from the larger.)
It is obvious that if a first set is included in a second set, then objects outside of the second set must also be outside the first set. But being outside a set is the same as being in the complement of that set. Let the two sets be A and B. Then

if   A < B   then   ~B < ~A
Therefore the operation of complementing interchanges the sides of an inclusion.
Now let B and A equal the complements of S1 and S2 respectively. The implication becomes
if   ~S2 < ~S1   then   S1 < S2
(Recall that the complement of a complement is the original set: ~~S = S.) This statement provides another method for proving the inclusion of one set in another:

[4.3] (Inclusion from complements) A first set is included in a second set   if   the complement of a second set is included in the complement of the first.
Notation1: if   ~S2 < ~S1   then   S1 < S2. Notation2: if   ~q(x) ==> ~p(x)   then   p(x) ==> q(x).

The implication ~q(x) ==> ~p(x) is called the contrapositive of the implication p(x) ==> q(x). It can be proven that an implication and its contrapositive always have the same truth value. Consider the implications

if person x is in Washington then person x is in the USA.    if person x is not in the USA then person x is not in Washington
Both implications are universally true. The second implication might be used by a traveller needing an alibi.

The following is a useful property of inclusions and implications of open statements. It presupposes all sets and open statements have the same universe.

[4.4] (Telescoping property of inclusions) In any universe, if a first set is included in a second set, and the second set is included in a third set, then the first set is included in the third set.
Notation1: if S1 < S2 and S2 < S3 then S1 < S3.
Notation2: if x is in S1 ==> x is in S2 and x is in S2 ==> x is in S3 then x is in S1 ==> x is in S3.
Notation3: for any open statements p(x), q(x), r(x) with a common universe, if the implications p(x) ==> q(x) and q(x) ==> r(x) are both universally true, then the implication p(x) ==> r(x) is universally true.

In notation2 there are three implications involved:

(1) x is in S1 ==> x is in S2,      (2) x is in S2 ==> x is in S3,      (3) x is in S1 ==> x is in S3,     
Implications (1) and (2) are given as universally true. Then neither has any counterexamples. As a result of this, (3) should also be universally true. Now begins the argument that (3) has no counterexamples. Suppose there is a counterexample b. This means that b is in S1, but not in S3. Since (1) has no counterexamples, then b is also in S2. Since (2) has no counterexamples, b is in S3. But this prevents b from being a counterexample to (3). So (3) can have no counterexamples.

In notation3 let S1, S2, S3 be the truth sets of p(x), q(x), r(x) respectively. Then the three implications

p(x) ==> q(x),      q(x) ==> r(x),      p(x) ==> r(x)
are equivalent to implications (1), (2), (3) respectively.



Section 5:   Collections of sets

In the previous sections, one or two set were involved in the previous sections. It is useful to extend ideas developed there to involve more than just two sets. Then it is often useful to collect all the sets into a family. For example, in universe N9 the following are subsets of this universe:
A = {1,5,6}, B = {2,7,8,9}, C = {3,4,5,6}, D = {1,3,5,7,9}
Then {A,B,C,D} is a family of four sets. There is no standard way of writing the name of a family of sets. In discussions here, the name will be underlined:
W = {A,B,C,D}
to avoid any confusion when necessary. It is meaningful (and true) to say that   A is a set in W   or simply,   B is in W. If E = {3,6,9} then E is not in W. Not in W are any of the sumbers 1,2,3,4,5,6,7,8,9, themselves because the family consists of subsets, not numbers.

In the adjacent figure the interiors of all the circles form a family of sets.


Note: In discussions here about any family, all the sets in it are subsets of a common universe.

Since the empty set Φ and the universe U are genuine subsets of U, then Φ and U may be members of some families of subsets of U.

In the universe of the plane of plane geometry straight lines are special sets of points. Therefore, any collection of straight lines is a family of sets. The familiar properties of straight lines can be restated as properties of this family. Click here for more discussion of this geometric family of straight lines.

-----

For any given set, there is a unique family of all subsets of that set. That family is called the power set of the given set. For example, all the subsets of S={1,3,5} are

Φ, {1}, {3}, {5}, {1,3}, {1,5}, {3,5}, {1,3,5}
Therefore,
P(S) = the power set of S is       {Φ, {1], {3}, {5}, {1,3}, {1,5}, {3,5}, {1,3,5}}
The following statement provides a method that helps in checking if all the subsets have been given:

[5.1] (Size of the power set) If a set has n objects then its power set has 2n subsets in it.

Examples: There are 23 = 8 subsets of {1,2,3}. There are 29 = 512 subsets of the set N9.

A discussion supporting this statement will be given in a later volume. The attempt to extend this statement to infinite sets like N leads to an interesting discussion of transfinite numbers.

Let the reader show that [5.1] is true even when the given set is empty.

-----

The idea of two disjoint sets may be extended to a family of sets.

[5.2] (Family of disjoint sets) In a family of disjoint sets, every pair of sets in that family are disjoint.

Example: In N9 if A={1,3,5}, B={2}, C={6,8}, D={4}, then {A,B,C,D} is a family of disjoint sets. The intersection of any pair of sets in this family is empty.

The interiors of all of the circles in the adjacent figure form a family of disjoint sets.
Another example, not shown here, is any family of parallel straight lines.

-----

Consider blankets on a very large bed. to cover the entire bed it may require several blankets that may overlap. The bed is under the combined collection of blankets. A similar situation may exist for some given set in a universe. A set may be "covered" by other sets in that universe.

[5.3] (Family covering of a set) A family of sets covers a given set if and only if the given set is contained in the union of all sets in the family. The family is called a covering of the given set.
Notation: Family W covers set S   if and only if   S   <   union of all sets in W.

Example: In universe N9 let S = {4,5,9}, and let W be the family

W = { {1,2}, {4,5,6}, {8,9} }
The union of all sets in W = {1,2,4,5,6,8,9} which contains S. So family W covers S.
Notice that W does not cover the whole universe N9, because 3 and 7 are not in any of the sets in W. Also the set {1,2} is in W, but it does not even intersect S. In an intuitive sense, {1,2} is not essential to the covering. (But it still is a member of the covering.)

It is quite possible that all the sets in a covering be disjoint. Then the family is called a disjoint covering of the set. The disjoint covering is a special family of sets. Intuitively speaking, the set is broken down into pieces which are subsets in the family.

[5.4] (Partitions) A partition of a set is a disjoint covering of the set by its own subsets.

In other words, a partition separates the entire given set into a family of disjoint subsets. The union of all those subsets is the given set. In the universe N9 an example of a partition of S = {2,3,4,5,6,7,8} is the family

{ {2,8}, {3.4}, {5,6,7} }
More partitions of S are the families:
{ {2,3,4,5}, {6,7,8} }
{ {2}, {3}, {4}, {5}, {6}, {7}, {8} }
{ {2,3,4,5,6,7,8} }
of three, seven and one subsets of S respectively.
The sets in the family { {2,8} {3.4} {5,6,7}, {1,9} } are disjoint, but the family is not a partition of S = {2,3,4,5,6,7,8} because {1,9} is not a subset of S. The family is only a disjoint covering of S.

The adjacent figure shows a partition of the interior of a circle into colored subsets (sub-arias).

Sometimes it is useful to partition a large set into a family of smaller subsets which can be more easily measured. Nothing is lost during the partitioning process. For example, suppose a large number of coins in a pile are to be counted. Partition the coins into several smaller piles. Have each of the smaller piles counted by different people. Add together the number of coins in each smaller pile.

Suppose family

W = {S1, S2, S3, .... Sn}
where S1, S2, S3, .... Sn are n subsets of of the same universe. Then W is a linear family of increasing sets if each set is included in the next set:
S1 < S2,    S2 < S3,    S3 < S4,    ...,    Sn-1 < Sn
This is often shortened and written as a chain of inclusions:    S1 < S2 < S3 < S4 < ... < Sn
Similarly, W is a linear family of decreasing sets if each set contains the next set:
S1 > S2,    S2 > S3,    S3 > S4,    ...,    Sn-1 > Sn
These may also be written as a chain of containments.
A linear family of sets is either a linear family of increasing sets or a linear family of decreasing sets.

Example: if

S1={1}, S2={1,5}, S3={1,2,3,5,7}, S4={1,2,3,4,5,7,8},S5={1,2,3,4,5,6,7,8}
then {S1, S2, S3, S4, S5} is a linear family of increasing sets. There is a a chain
S1 < S2 < S3 < S4 < S5
Similarly, {S5, S4, S3, S2, S1} is a linear family of decreasing sets.
It is customary to list the sets in these families in the order of inclusion or containment.

Let p1(x) be any open statement equivalent to   x is in S1;
Let p2(x): be any open statement equivalent to   x is in S2;
Let p3(x): be any open statement equivalent to   x is in S3;
.............................................................................................................
Let pn(x): be any open statement equivalent to   x is in Sn.


Then the chain of n included sets induces the chain
p1(x) ==> p2(x) ==> p3(x) ==> ... ==> pn(x)
of implications.

[5.5] (Telescoping sets) For a family of n increasing sets, the first set is included in the last set.
Notation1: if S1 < S2 < S3 < S4 < ... < Sn then S1 < Sn.
Notation2: if p1(x) ==> p2(x) ==> p3(x) ==> ... ==> pn(x) then p1(x) ==> pn(x).

A discussion supporting this statement will be given later after a discussion of positive integers, in particular mathematical induction. But it is pattered after the supporting statements for [4.4].