An Introduction to Symbolic Logic

Guram Bezhanishvili and Wesley Fussner∗

1 Introduction
This project is dedicated to the study of the basics of propositional and predicate logic. We
will study it based on Russell and Whitehead’s epoch making treatise Principia Mathemat-
ica [9]. Published in three volumes between 1910 and 1913, Principia was a culmination of
work that had been done in the preceding century on the foundations of mathematics. Since
the middle of the nineteenth century, such thinkers as George Boole (1815–1864), Augustus
De Morgan (1806–1871), Charles Sanders Peirce (1839–1914), Ernst Schröder (1841-1902),
Gottlob Frege (1848–1925), and Giuseppe Peano (1858–1932) had been developing axiomatic
bases for logic and the foundations of mathematics. This research program found its culmi-
nation in Principia, which had a tremendous influence on the development of logic and the
foundations of mathematics in the twentieth century.
Logic is a branch of science that studies correct forms of reasoning. It plays a fundamental
role in such disciplines as philosophy, mathematics, and computer science. Like philosophy
and mathematics, logic has ancient roots. The earliest treatises on the nature of correct
reasoning were written over 2000 years ago. Some of the most prominent philosophers of
ancient Greece wrote of the nature of deduction more than 2300 years ago, and thinkers in
ancient China wrote of logical paradoxes around the same time. However, though its roots
may be in the distant past, logic continues to be a vibrant field of study to this day.
Modern logic originated in the work of the great Greek philosopher Aristotle (384–
322 bce), the most famous student of Plato (c.427–c.347 bce) and one of the most influential
thinkers of all time. Further advances were made by the Greek Stoic philosopher Chrysippus
of Soli (c.278–c.206 bce), who developed the basics of what we now call propositional logic.1
For many centuries the study of logic was mostly concentrated on different interpretations
of the works of Aristotle, and to a much lesser degree of those of Chrysippus, whose work was
largely forgotten. However, the existing logic had no formal basis. All the argument forms
were written in words, and lacked formal machinery that would create a logical calculus of
deduction that is easy to work with.
The great German philosopher and mathematician Gottfried Willhelm Leibniz (1646–
1716) was among the first to realize the need of formalizing logical argument forms. It
was Leibniz’s dream to create a universal formal language of science that would reduce all
∗
Department of Mathematical Sciences; New Mexico State University; Las Cruces, NM 88003, USA;
[email protected]; [email protected].
1
Our webpage http://www.cs.nmsu.edu/historical-projects/ offers more details on the work of
Chrysippus; see the historical project [5].

1
philosophical disputes to a matter of mere calculation by recasting the reasoning in such
disputes in the universal symbolic language of science.
The first real steps in this direction were taken in the middle of the nineteenth century by
the English mathematician George Boole. In 1854 Boole published An Investigation of the
Laws of Thought [3], in which he developed an algebraic system for discussing logic. Boole’s
work ushered in a revolution in logic, which was advanced further by Augustus De Morgan,
Charles Sanders Peirce, Ernst Schröder, and Giuseppe Peano.2
The next key step in this revolution in logic was made by the great German mathe-
matician and philosopher Gottlob Frege. Frege created a powerful and profoundly original
symbolic system of logic, as well as suggested that the whole of mathematics can be developed
on the basis of formal logic, which resulted in the well-known school of logicism.3
By the early twentieth century, the stage was set for Russell and Whitehead to give
a modern account of logic and the foundations of mathematics in their influential treatise
Principia Mathematica.
Alfred North Whitehead (1861-1947), the son of a vicar in the Church of England [6,
p. 23], was born in Ramsgate, Kent, England, and studied mathematics at Trinity College,
Cambridge. In 1884, Whitehead was elected a fellow at Trinity College, and would teach
mathematics there until 1910. After his tenure at Trinity College, Whitehead spent time
at University College London and Imperial College London, engaging in scholarly work in
philosophy. He later emigrated to the United States, and taught philosophy at Harvard
University until his retirement in 1937 [4].
While a fellow at Trinity College, Whitehead met Bertrand Russell (1872-1970), who was
then a student there [6, p. 223]. Russell was born into an aristocratic family. His grandfather,
John Russell, was twice Prime Minister to Queen Victoria [7, p. 5]. Russell graduated from
Trinity College in 1893. He went on to become one of the most influential intellectuals of
the twentieth century, playing a decisive role in the development of analytic philosophy.
Russell was also active in a number of political causes; notably, he was an anti-war activist
and advocated nuclear disarmament. He was a prolific writer, and in 1950 was awarded the
Nobel Prize in Literature “in recognition of his varied and significant writings in which he
champions humanitarian ideals and freedom of thought” [8].
Around 1901, Russell and Whitehead began collaborating on a book on logic and the
foundations of mathematics [6, p. 254–258]. This resulted in an epochal work, Principia
Mathematica, which would later be recognized as a significant contribution to logic and the
foundations of mathematics. Influenced by the work of Frege, Peano, and Schröder, Russell
and Whitehead developed an axiomatic basis for logic and the foundations of mathematics,
and tried to free the foundations of mathematics of the existing contradictions.
In what follows, we will introduce the basic principles of contemporary logic through the
development of Russell and Whitehead’s Principia Mathematica.
2
Our webpage http://www.cs.nmsu.edu/historical-projects/ offers historical projects on Boole, De
Morgan, Peirce, and Peano; see the projects [1, 2].
3
Frege’s formal treatment of propositional logic is discussed on our webpage
http://www.cs.nmsu.edu/historical-projects/; see the historical project [5].

2
2 Propositional Logic
In this section we begin our study of propositional logic from Principia Mathematica. The
chief object of our investigation will be propositions—sentences which are either true or false
but not both. Thus, we are concerned with sentences such as “Benjamin Franklin was the
first president of the United States” and “Two plus two is equal to four.” Clearly the first
of the two sentences is false and the second one is true. Therefore, both of the sentences are
propositions. On the other hand, a sentence such as “Who was the author of Hamlet?” is
not a proposition because it is neither true nor false. Hence, we will not be concerned with
this type of sentence.
To carry out our study of propositions, we introduce the concept of a propositional
variable, which stands for an arbitrary but undetermined proposition. The letters p, q,
r and so forth will be used to denote propositional variables.

2.1 Logical connectives

We now turn to the first major topic in propositional logic, the question of how to form
complicated propositions out of simpler ones. Russell and Whitehead address this question
in the opening pages of Principia Mathematica:

An aggregation of propositions (...) into a single proposition more complex than its
constituents, is a function with propositions as arguments. [9, Vol. 1, p. 6]

Thus, more complex propositions are formed from simpler propositions by means of
propositional functions. What are the propositional functions that yield more complex
propositions? Russell and Whitehead employ four fundamental propositional functions.

...They are (1) the Contradictory Function, (2) the Logical Sum, or Disjunctive Func-
tion, (3) the Logical Product, or Conjunctive Function, (4) the Implicative Function.
These functions in the sense in which they are required in this work are not all inde-
pendent; and if two of them are taken as primitive undefined ideas, the other two can
be defined in terms of them. It is to some extent—though not entirely—arbitrary as
to which functions are taken as primitive. Simplicity of primitive ideas and symmetry
of treatment seem to be gained by taking the first two functions as primitive ideas.
[9, Vol. 1, p. 6]

In modern terminology the Contradictory Function of Russell and Whitehead is known

as negation (“not”), the Logical Sum or Disjunctive Function is known as disjunction (“or”),
the Logical Product or Conjunctive Function as conjunction (“and”), and the Implicative
Function as implication (“if, then”).
Russell and Whitehead mention that the four functions are not independent of each other.
Later on we will see why this is so. For now, let us read how Russell and Whitehead define
these four functions.

The Contradictory Function with argument p, where p is any proposition, is the

proposition which is the contradictory of p, that is, the proposition asserting that p

3
is not true. This is denoted by ∼p. Thus ∼p is the contradictory function with p as
argument and means the negation of the proposition p. It will also be referred to as
the proposition not-p. Thus ∼p means not-p, which also means the negation of p.
The Logical Sum is a propositional function with two arguments p and q, and is the
proposition asserting p or q disjunctively, that is, asserting that at least one of the
two p and q is true. This is denoted by p ∨ q. Thus p ∨ q is the logical sum with p
and q as arguments. It is also called the logical sum of p and q. Accordingly p ∨ q
means that at least p or q is true, not excluding the case in which both are true.
The Logical Product is a propositional function with two arguments p and q, and is
the proposition asserting p and q conjunctively, that is, asserting that both p and q
are true. This is denoted by p.q (...). Thus p.q is the logical product with p and q
as arguments. It is also called the logical product of p and q. Accordingly p.q means
that both p and q are true. It is easily seen that this function can be defined in terms
of the two preceding functions. For when p and q are both true it must be false
that either ∼p or ∼q is true. Hence in this book p.q is merely a shortened form of
symbolism for

∼(∼p ∨ ∼q)

The Implicative Function is a propositional function with two arguments p and q,

and is the proposition that either not-p or q is true, that is, it is the proposition
∼p ∨ q. Thus if p is true, ∼p is false, and accordingly the only alternative left by the
proposition ∼p ∨ q is q is true. In other words if p and ∼p ∨ q are both true, then q
is true. In this sense the proposition ∼p ∨ q will be quoted as stating that p implies
q. The idea contained in this propositional function is so important that it requires
a symbolism which with direct simplicity represents the propositions as connecting p
and q without the intervention of ∼p. But “implies” as used here expresses nothing
else than the connection between p and q also expressed by the disjunction “not-p
or q.” The symbol employed for “p implies q,” i.e. for “∼p ∨ q,” is “p ⊃ q.” This
symbol may also be read “if p, then q.” [9, Vol. 1, pp. 6–7]

The notation used to denote the four fundamental functions of propositions changed
considerably in the decades following the publication of Principia. Today the conjunction
of the propositions p and q is written as p ∧ q rather than p.q, and the symbol → is now
used in place of Russell and Whitehead’s ⊃. Contemporary logicians also refer to the four
fundamental functions of propositions of Principia Mathematica simply as logical connectives.
In what follows, we will adopt modern notation and terminology. We call propositions
p, q, r, . . . elementary propositions, and propositions built from elementary propositions by
means of logical connectives compound propositions. Statements of the form p → q are
referred to as conditional statements or conditionals.

Remark 1. There is an apparent ambiguity in reading propositions like ∼ p ∨ q. The

proposition can be read as either (∼p) ∨ q (i.e., as the logical sum of ∼p and q) or as ∼(p ∨ q)
(i.e., as the result of applying the contradictory function to p ∨ q). This ambiguity is easily
resolved by the agreement that ∼ binds stronger than any of ∨, ∧, →. Thus, ∼p ∨ q is read

4
as (∼p) ∨ q rather than ∼(p ∨ q). Similarly, we agree that ∨ and ∧ bind stronger than →.
For example, the proposition p ∨ q → r should be read as (p ∨ q) → r, and the proposition
∼p ∧ q → r ∨ p should be read as ((∼p) ∧ q) → (r ∨ p).
Our first goal is to obtain a good understanding of propositions and of how the four
logical connectives that Russell and Whitehead introduced yield more complex propositions
out of simpler ones.
Exercise 1. Which of the following sentences are propositions?
(a) The New York Yankees have never won a World Series.
(b) 2 is even.
(c) Please close the door.
(d) The square root of 109.
(e) The sum of two even integers is even.
(f) What is the capital of France?
For the sentences that are propositions, determine whether they are true or false.
Exercise 2. Let p denote the proposition “All mammals have four legs,” q denote the
proposition “All dogs have four legs,” and r denote the proposition “All dogs are mam-
mals.” Represent each of the following propositions using the four fundamental functions of
propositions of Principia Mathematica.
(a) Not all dogs have four legs.
(b) All mammals have four legs and all dogs have four legs.
(c) Not all mammals have four legs or not all dogs have four legs.
(d) If all mammals have four legs and all dogs are mammals, then all dogs have four legs.
(e) If not all dogs have four legs, then not all mammals have four legs or not all dogs have
four legs.
Which of the above are true and which are false?
Exercise 3. Let p denote the proposition “9 is odd,” q denote the proposition “81 is the
square of 9,” and r denote the proposition “81 is odd.” Write each of the following proposi-
tions verbally in words.
(a) p ∧ q → r
(b) q ∧ r → p
(c) ∼r → (∼p ∨ ∼q)
(d) p ∨ ∼(q ∧ r)
Determine which of the above are true and which are false.

5
2.2 Truth-values and truth tables
Having developed a language for discussing the logic of propositions, we turn to the task
of understanding how the notion of truth relates to our symbolic logic of propositions. In
particular, we develop a framework for understanding the truth and falsity of complex propo-
sitions based on the truth and falsity of simpler ones. Principia Mathematica addresses this
issue:

Truth-values. The “truth-value” of a proposition is truth if it is true, and falsehood

if it is false. It will be observed that the truth-values of p ∨ q, p.q, p ⊃ q, ∼p (...)
depend only on those of p and q, namely the truth-value of “p ∨ q” is truth if the
truth-value of either p or q is truth, and is falsehood otherwise; that of “p.q” is truth
if that of both p and q is truth, and is falsehood otherwise; that of “p ⊃ q” is truth
if either that of p is falsehood or that of q is truth; that of ∼p is the opposite of that
of p (...). [9, Vol. 1, p. 8]

Remark 2. In everyday English, propositions of the form “p or q” are usually intended to

mean that either p is true or q is true, but not both. For example, in ordinary English the
“or”-statement “the car was red or blue” means that the car was either red or blue, but not
both red and blue. The logical connective “or” in such statements is called an exclusive “or”.
However, according to Principia Mathematica, the connective “∨” expresses a different sort
of “or”—for “p ∨ q” is true if at least one of p and q is true. The connective ∨ is called
an inclusive “or” because the truth of “p ∨ q” allows for the possibility that both p and q
are true. In logic and mathematics, “or”-statements always employ the inclusive “or” rather
than the exclusive “or”.

Remark 3. The truth-values of the implication “p → q” given in Principia Mathematica are

somewhat different than one might expect. Russell and Whitehead indicate that “p → q”
is true if either p is false or q is true; so, in particular, they regard statements of the form
“if p, then q” as true if p is false. This way of thinking about “if, then” statements may
at first seem unusual, but one may understand the motivation for doing so if one thinks of
“p → q” as being analogous to a promise that if p holds true, then q will also hold true. If
it so happens that p is not true, then the promise is unbroken regardless of the truth-value
of q, and so “p → q” is true. The promise is broken only if p holds true, but q happens to
be false. This is the only situation in which we regard “p → q” as being false. Note that an
implication “if p, then q” which is true because p is false is referred to as vacuously true.

Exercise 4. If the truth-value of p is truth and the truth-values of q and r are falsehood,
compute the truth-values of the following compound propositions.

(a) (p ∨ q) ∨ r

(b) p → (p ∧ q)

(c) (∼(p ∧ r)) → ∼q

We now understand how to determine the truth-value of a compound proposition from

the truth-values of the propositional variables of which it is composed. But we can do better

6
than this. We can design a general method, which will allow us to calculate all possible truth-
values of a compound proposition from all possible truth-values of the propositional variables
of which it is composed. Such a method was devised independently by the great German
philosopher Ludwig Wittgenstein (1889–1951) and the American logician Emil Post (1897–
1954). The method employs truth tables, tables which list all the possible truth-values of the
propositional variables contained in a compound proposition along with the corresponding
truth-values of the compound proposition itself. For example, if p is a propositional variable,
then the truth table of the compound proposition ∼p is as follows:

p ∼p
T F
F T

Here the possible truth-values for the propositional variable p are listed in the column on
the left, and the corresponding truth-values of the proposition ∼p are listed in the column
on the right. Truth is denoted by T and falsehood is denoted by F.
Compound propositions which contain more propositional variables have more compli-
cated truth tables. For example, if p and q are propositional variables, then the truth table
for the proposition p ∨ q is:

p q p∨q
T T T
T F T
F T T
F F F

Exercise 5. Construct truth tables for each of the following propositions.

(a) p ∧ q

(b) p → q

(c) (p ∨ q) → r

(d) p → (q ∧ r)

2.3 Logical equivalence and biconditionals

In some cases, different propositions are, in some sense, logically the same. For example,
the propositions “9 is odd and 81 is the square of 9” and “81 is the square of 9 and 9 is
odd” are somehow alike despite having different symbolic representations. More generally, if
p and q are propositions, the propositions p ∧ q and q ∧ p apparently have the same meaning.
This property of being logically alike, called logical equivalence, is one of the most important
concepts in propositional logic. Principia Mathematica introduces logical equivalence as
follows:

7
The simplest example of the formation of a more complex function of propositions
by the use of these four fundamental forms is furnished by “equivalence.” Two
propositions p and q are said to be “equivalent” when p implies q and q implies
p. This relation between p and q is denoted by “p ≡ q.” Thus “p ≡ q” stands for
“(p ⊃ q).(q ⊃ p).” It is easily seen that two propositions are equivalent when, and
only when, they are both true or are both false. [9, Vol. 1, p. 7]

Here Russell and Whitehead have defined the notion of logical equivalence, but they
have also introduced a new logical connective. In modern terms, this new logical connective
is written using the symbol ↔ and the proposition “p ↔ q” is taken to stand for “(p →
q) ∧ (q → p)”. This new connective is known as the biconditional, and statements of the
form “p ↔ q” are called biconditional statements.
It is very important to distinguish between the biconditional statement “p ↔ q” and
the logical equivalence of p and q. As the basic logical connectives ∨, ∧, ∼, and →, the
biconditional ↔ is also a logical connective. On the other hand, two propositions p and q
are logically equivalent when p and q have the same truth values (both are true or both are
false). When p and q are logically equivalent, we write p ≡ q. We will see later that two
propositions p and q are logically equivalent if and only if the biconditional p ↔ q is always
true, i.e. if the biconditional is true irrespective of the truth of p or q.
How can we determine whether two given propositions are logically equivalent? In order
to do so, we must be able to guarantee that the propositions have the same truth-value
irrespective of the truth-values of the elementary propositions of which they are composed.
At first this may seem like a daunting task. However, the use of truth tables makes it a
simple matter to determine whether two propositions are logically equivalent. In fact, since
the truth tables for the propositions p and q list the truth-values for p and q under all possible
assignments of truth-values for the elementary propositions of which they are built, p and q
are logically equivalent if and only if they have the same truth tables.
An example may better illustrate this point. We shall use truth tables to establish the
logical equivalence of the propositions (p ∨ q) ∧ r and (p ∧ r) ∨ (q ∧ r):

p q r (p ∨ q) ∧ r (p ∧ r) ∨ (q ∧ r)
T T T T T
T T F F F
T F T T T
T F F F F
F T T T T
F T F F F
F F T F F
F F F F F

As the two propositions have identical truth tables, they are logically equivalent.

Exercise 6. Use truth tables to show that:

(a) ∼(p ∧ q) is logically equivalent to ∼p ∨ ∼q.

(b) ∼(p ∨ q) is logically equivalent to ∼p ∧ ∼q

8
These equivalences are called De Morgan’s Laws in honor of Augustus De Morgan.

Exercise 7.

(a) Are p → q and ∼q → ∼p logically equivalent? Justify your answer.

(b) Are p → q and q → p logically equivalent? Justify your answer.

(c) Are p → q and ∼p → ∼q logically equivalent? Justify your answer.

(d) The proposition ∼q → ∼p is often called the contrapositive of the conditional p → q,

the proposition q → p is often called the converse of the conditional p → q, and
the proposition ∼p → ∼q is often called the inverse of the conditional p → q. Are
there any logical equivalences between a conditional statement, its contrapositive, its
converse, and its inverse? Justify your answer.

Now that we have introduced the notion of logical equivalence, we will discuss Russell
and Whiteheads claim (see page 3) that the four logical connectives are inter-definable. For
example, Russell and Whitehead claim that both → and ∧ can be expressed by means of ∼
and ∨; namely, p → q is a shorthand for ∼p ∨ q and p ∧ q is a shorthand for ∼(∼p ∨ ∼q).

Exercise 8.

(a) Show that p ∨ q and p → q are logically equivalent to propositions only involving ∼
and ∧.

(b) Show that p ∨ q and p ∧ q are logically equivalent to propositions only involving ∼ and
→.

(c) What do these results indicate about the logical connectives introduced in Principia?
Are some connectives redundant? If so, which ones?

2.4 Tautologies and contradictions

We have seen that truth tables are a useful tool for determining if two propositions are
logically equivalent, but it turns out that truth tables have a variety of other uses. In what
follows we will explore two other applications of truth tables. The first of these applications
is the identification of certain special propositions. To illustrate what these propositions
are and how truth tables can be used to identify them, we consider the truth table of the
proposition p ∨ ∼p:

p p ∨ ∼p
T T
F T

Notice that the truth-values displayed in the righthand column are always truth; false-
hood does not appear. Thus, p ∨ ∼p is true under any assignment of truth or falsehood to
the proposition p. Propositions with this property—that is, propositions which are true for
any assignment of truth-values to the propositional variables of which they are formed—are

9
called tautologies. From the truth table above, we can easily see that p ∨ ∼p is a tautology.
p ∨ ∼p is referred to as the law of the excluded middle since it asserts that either p is true
or the negation of p is true (so there is no “middle ground” between truth and falsity). In
general, one may test to see if a proposition is a tautology by constructing a truth table: if
the only possible truth-value of the proposition is truth, then that proposition is a tautology.
The tautologies are special propositions that can be identified by the use of truth tables.
Another sort of special propositions that can be identified in the same way are the con-
tradictions. Whereas the tautologies are propositions which are true under any assignment
of truth-values to the propositional variables of which they are formed, contradictions are
propositions that are false under any assignment of truth-values to the propositional vari-
ables of which they are formed. For example, the proposition p ∧ ∼p is a contradiction. This
is illustrated in the following truth table:

p p ∧ ∼p
T F
F F

As the only truth-value displayed in the righthand column is falsehood, the proposition p∧∼p
is a contradiction.

Exercise 9. For each of the following propositions, use a truth table to determine whether
the proposition is a tautology, a contradiction, or neither.

(a) ∼(p ∧ ∼p)

(b) p ↔ ∼(∼p)

(c) p ↔ ∼p

(d) p ∨ (q → p)

(e) (p ∨ q) → (q ∨ p)

(f) (p → q) ↔ (∼q → ∼p)

(g) (p ∧ q) ∧ (∼p ∨ ∼q)

(h) p → (q → p)

The logical laws expressed in (a) and (b) are called the law of non-contradiction and
the law of double negation, respectively. For any proposition p, the law of non-contradiction
asserts that p and ∼p are not both true. We also saw that the law of the excluded middle
asserts that at least one of p and ∼p is true. Hence, when taken together, the law of non-
contradiction and the law of the excluded middle guarantee that exactly one of p and ∼p is
true for each proposition p.
Now, as promised on page 8, we are ready to see that two propositions p and q are
logically equivalent if and only if the biconditional p ↔ q is a tautology.

Exercise 10. Show that p and q are logically equivalent if and only if p ↔ q is a tautology.

10
It is often helpful to be able to recognize commonly occuring logical equivalences at
a glance. For convenience, we summarize some of the most frequently appearing logical
equivalences in the table below. Note that in this table p and q denote arbitrary propositions,
t denotes an arbitrary tautology, and c denotes an arbitrary contradiction.

Commutative laws: p∨q ≡q∨p p∧q ≡q∧p

Associative laws: (p ∨ q) ∨ r ≡ p ∨ (q ∨ r) (p ∧ q) ∧ r ≡ p ∧ (q ∧ r)
Distributive laws: p ∧ (q ∨ r) ≡ (p ∧ q) ∨ (p ∧ r) p ∨ (q ∧ r) ≡ (p ∨ q) ∧ (p ∨ r)
Idempotent laws: p∨p≡p p∧p≡p
Absorption laws: p ∧ (p ∨ q) ≡ p p ∨ (p ∧ q) ≡ p
Identity laws: p∨c≡p p∧t≡p
Universal bound laws: p∨t≡t p∧c≡c
De Morgan laws: ∼(p ∨ q) ≡ ∼p ∧ ∼q ∼(p ∧ q) ≡ ∼p ∨ ∼q
Negation laws: p ∨ ∼p ≡ t p ∧ ∼p ≡ c
Negations of t and c: ∼t ≡ c ∼c ≡ t
Double negation law: ∼∼p ≡ p
Other equivalences: p → q ≡ ∼p ∨ q ∼(p → q) ≡ p ∧ ∼q
p → q ≡ ∼q → ∼p p ↔ q ≡ (p → q) ∧ (q → p)

Exercise 11. Use the equivalences listed in the table above to write each of the following
propositions as a simpler, logically equivalent proposition.

(a) p ∧ (p ∨ (q ∨ p))

(b) p ∨ ∼(p ∨ ∼p)

(c) ∼(∼(p ∨ ∼q) ∧ r)

(d) (p ∧ ∼p) → (p ∨ (q ∨ p))

Exercise 12. Use the equivalences listed in the table above to determine which of the
following logical equivalences hold. Justify your answers, but do not use a truth table.

(a) (p ∨ (q ∨ r)) ∧ ∼(p ∨ (q ∨ r)) ≡ q ∧ ∼q

(b) (p ∧ r) ∨ ((∼p → ∼q) → (q → p)) ≡ p ∧ r

(c) ∼(p → q) ↔ p ∧ ∼q ≡ p → (q → p)

2.5 Inference rules

Up until now we have been concerned exclusively with propositions and their properties.
However, the central concern of logic is not just the study of propositions. Rather the object
of study in logic is inference—the process of drawing correct conclusions from premises. Our
study of inference begins with a simple rule which allows us to deduce the proposition q
from the propositions p and p → q. This rule is referred to as Modus Ponens. Principia
Mathematica describes the rule as follows:

11
Inference. The process of inference is as follows: a proposition “p” is asserted, and
a proposition “p implies q” is asserted, and then as a sequel the proposition “q” is
asserted. The trust in inference is the belief that if the two former assertions are not
in error, the final assertion is not in error. [9, Vol. 1, p. 9]

Russell and Whitehead assert that Modus Ponens is a valid logical rule; that is, if the
premises p and p → q are both true, then the conclusion q is guaranteed to be true. While
this fact is intuitively obvious, a skeptical reader may wonder how we really know this is the
case. Fortunately, the skeptic’s concerns may easily be allayed: we can verify that Modus
Ponens preserves truth. For this we need to be able to guarantee the truth of q based on
the truth of the propositions p and p → q. We thus need only to verify that q is true in
every circumstance in which both p and p → q are true. To do so, we construct a truth table
listing the possible truth values of the propositions p, q, and p → q:

p q p→q
T T T
T F F
F T T
F F T

The only row of the truth table in which both p and p → q are true is the first. Since in
this case q is also true, we know that if p and p → q are both true, then q must be true as
well. Thus, Modus Ponens is a valid logical rule.
So far we have focused exclusively on Modus Ponens, but it is worth noting that we
might just as well have discussed any one of several valid logical rules. There are many valid
logical rules (called rules of inference), and Modus Ponens is distinguished only by being
the simplest of these. In the exercises, we will encounter several other historically important
rules of inference and establish their validity.

Exercise 13. Use truth tables to verify the validity of the following rule of inference: if ∼q
and p → q hold, then infer that ∼p holds. (This rule of inference is referred to as Modus
Tollens.)

Exercise 14. Use truth tables to verify the validity of the following rule of inference: if ∼p
and p ∨ q hold, then infer that q holds. (This rule of inference is referred to as the disjunctive
syllogism.)

Exercise 15. Use truth tables to verify the validity of the following rule of inference: if
p → q and q → r hold, then infer that p → r holds. (This rule of inference is referred to as
the hypothetical syllogism.)

Exercise 16. Consider the following rule of inference: if p → q and q hold, then infer that
p holds. An application of this rule of inference is referred to as affirming the consequent.
Is affirming the consequent a valid rule of inference? If so, use truth tables to establish its
validity. If not, give examples of propositions p and q for which p → q and q are true and p
is not.

12
Exercise 17. Late Saturday night the police are called to Crickwell Manor. That evening
Mr. and Mrs. Crickwell had been holding a dinner party, and at the party Mr. Crickwell had
been murdered. Other than the deceased, there had been four people in attendance: Ms.
Anderson, Mr. Boalt, Mrs. Crickwell, and Mr. Dunham. The police interview each of these
witnesses, and they offer the following statements:

(1) Ms. Anderson says that the murderer must be Mr. Boalt or Mr. Dunham.

(2) Mr. Boalt says that neither Mr. Dunham nor Mrs. Crickwell could have killed Mr.
Crickwell.

(3) Mrs. Crickwell says that she certainly did not kill her husband, and that if Mr. Boalt
were not the murderer then it must have been Ms. Anderson.

(4) Mr. Dunham says that he himself had far more motive to kill Mr. Crickwell than did
Mr. Boalt, so if he did not kill Mr. Crickwell then neither did Mr. Boalt.

Assuming that only one of the people attending the party killed Mr. Crickwell and that
everyone except the murderer was truthful with the police, who killed Mr. Crickwell?

3 Predicate Logic
While the propositional logic developed in the previous section allows us to address a number
of significant issues in logic, it turns out that it is not capable of answering all of the logical
questions which are important to mathematicians. For example, recall that a positive integer
a is a prime number if a 6= 1 and the only divisors of a are 1 and a. A mathematician
may be interested in identifying the conditions under which the proposition “every prime
number greater than 2 is odd” is true. Propositional logic is not very helpful in this regard
because it is not possible to view this proposition as being composed from several elementary
propositions by logical connectives. Thus from the perspective of propositional logic it must
itself be an elementary proposition. However, if “every prime number greater than 2 is
odd” is an elementary proposition, then propositional logic does not provide us with any
information about when this proposition is true—the propositional logic developed in the
previous section says only that elementary propositions are either true or false. In order to
provide a more informative analysis of the proposition “every prime number greater than 2 is
odd,” we must introduce ideas which allow us to discuss the internal structure of propositions
that we previously regarded as elementary.

3.1 Individual variables and predicates

We start by introducing new kind of variables, called individual variables to distinguish
them from the propositional variables of the previous section. Individual variables range
over objects in some domain we are presently concerned with rather than over propositions.
For example, the domain for the proposition “every prime number greater than 2 is odd”
should range over all positive integers. We will use the letters x, y, z and so forth to denote
individual variables.

13
With the notion of an individual variable at hand, we can now give an account of how
“every prime number greater than 2 is odd” may be analyzed in terms of simpler propositions.
Suppose that x is an individual variable, and that we take the expression P (x) to mean that
“x is a prime number greater than 2” and the expression O(x) to mean that “x is odd.” Then
it is clear that the proposition “for every x, P (x) → O(x)” has the same meaning as “every
prime number greater than 2 is odd.” Here the domain of range of the individual variable
x is all positive integers. Thus, if we can formalize the meanings of P (x), O(x), and “for
every x,” then we can give an informative analysis of the proposition “every prime number
greater than 2 is odd.”
We will first formalize the meanings of the expressions P (x) and O(x). These expressions
are examples of what Russell and Whitehead referred to as “propositional functions,” which
Principia Mathematica describes as follows:

Propositional Functions. Let φx be a statement containing a variable x and such

that it becomes a proposition when x is given any fixed determined meaning. Then
φx is called a “propositional function”; it is not a proposition, since owing to the
ambiguity of x it really makes no assertion at all. Thus “x is hurt” really makes
no assertion at all, till we have settled who x is. Yet owing to the individuality
retained by the ambiguous variable x, it is an ambiguous example from the collection
of propositions arrived at by giving all possible determinations to x in “x is hurt”
which yield a proposition, true or false. Also if “x is hurt” and “y is hurt” occur in
the same context, where y is another variable, then according to the determinations
given to x and y, they can be settled to be (possibly) the same propositions or
(possibly) different propositions. But apart from some determination given to x and
y, they retain in that context their ambiguous differentiation. Thus “x is hurt” is an
ambiguous “value” of a propositional function. [9, Vol. 1, p. 15]

Today it is more common to denote the expression φx by φ(x). As the statements “x

is a prime number greater than 2” and “x is odd” become propositions when x is assigned
a particular integer value, P (x) and O(x) as described above are propositional functions.
Today the propositional functions of Russell and Whitehead are referred to as predicates.
Let φ(x) be a predicate and D be some specified domain. Then the collection of objects
a in D for which φ(a) is true is called the truth set of the predicate φ(x) with respect to
the domain D. In other words, the truth set of φ(x) with respect to the domain D is the
collection of those objects in D that have the property that the predicate attributes to them.
It is common practice to denote the collection of nonnegative integers

0, 1, 2, . . .

by N. As usual, we will refer to them as natural numbers. It is also common practice to

denote the collection of integers

..., −2, −1, 0, 1, 2, ...

by Z, the collection of rational numbers (ratios of two integers) by Q, and the collection
of real numbers by R. The sets N, Z, Q, and R are among the most common domains of
the predicates employed in mathematics. For example, it is most natural to consider the

14
predicates P (x) and O(x) discussed above as having the domain N. On the other hand,
if one were considering the predicate Q(x) defined by “x is greater than π,” then one is
most likely thinking of R as the domain of Q(x). However, note that if instead we consider
Z to be the domain of Q(x), then we have specified a much different predicate than if we
consider the domain to be R. This is because if we consider R to be the domain of Q(x),
then propositions such as Q(π + 1) and Q(4π) are true (because π + 1 and 4π are both real
numbers greater than π). However, if the domain of Q(x) is Z, then Q(π + 1) and Q(4π)
are not even propositions because π + 1 and 4π are not integers. From this it is clear that
in order to completely specify a predicate we must first specify its domain. The domains for
the predicates we will discuss will usually be clear from the context.
Exercise 18. Recall that P (x) is the predicate “x is a prime number greater than 2” and
O(x) is the predicate “x is odd.” State each of the following propositions verbally in words.
Also determine whether each proposition is true or false.
(a) P (4) ∨ O(7)
(b) (P (3) ∧ P (13)) ∧ O(12)
(c) P (2) → P (23)
Exercise 19. Identify the truth sets of the following predicates.
(a) A(x) defined by “x is a real number satisfying the equation x2 − 5x + 6 = 0.”
(b) B(x) defined by “x is a rational number satisfying the inequality x2 ≤ 1.”
(c) C(x) defined by “x is a differentiable function whose derivative is x.”
Exercise 20. For each of the following sets, give an example of a predicate with the specified
domain.
(a) Z
(b) The set of all letters in the English alphabet.
(c) The set of all continuous functions defined on the real numbers.
(d) The set of all college students in the United States.

3.2 Universal and existential quantifiers

Now that we have introduced predicates, we are almost able to analyze the logic of the
proposition “every prime number greater than 2 is odd.” Recall that if we let P (x) be the
predicate “x is a prime number greater than 2” and O(x) be the predicate “x is odd,” then
the proposition “every prime number greater than 2 is odd” can be expressed as “for every
x, P (x) → O(x).” In order to fully understand the latter proposition, we only need to clarify
what it means for P (x) → Q(x) to hold “for every x.” To do so, we will introduce the notion
of a quantifier, a logical operator that specifies whether a predicate φ(x) holds for all values
of x or some value of x. In Principia Mathematica, a predicate quantified in the former way
is written as (x).φ(x) and a predicate quantified in the latter way is written as (∃x).φ(x):

15
Thus corresponding to any propositional function...there is a range, or collection, of
values, consisting of all the propositions (true or false) which can be obtained by
giving every possible determination to x in φx...in respect to the truth or falsehood
of propositions of this range three important cases must be noted and symbolised.
These cases are given by three propositions of which one at least must be true. Either
(1) all propositions of the range are true, or (2) some propositions of the range are
true, or (3) no proposition of the range is true. The statement (1) is symbolised
by “(x).φx,” and (2) is symbolised by “(∃x).φx.” No definition is given of these
two symbols, which accordingly embody two new primitive ideas in our system. The
symbol “(x).φx” may be read “φx always,” or “φx is always true,” or “φx is true
for all possible values of x.” The symbol “(∃x).φx” may be read “there exists an x
for which φx is true”...and thus conforms to the natural form of the expression of
thought. [9, Vol. 1, pp. 15–16]

In the rest of the project we will follow the common practice of denoting (x).φx by
(∀x)φ(x). With quantifiers, we may now see that “every prime number greater than 2 is
odd” may be represented symbolically as “(∀x)(P (x) → O(x)),” where P (x) is the predicate
“x is a prime number greater than 2” and O(x) is the predicate “x is odd.”
When we refer to propositions of the form (∀x)φ(x) and (∃x)φ(x), we will always have in
mind some fixed domain of values over which x may vary. This domain is sometimes called a
domain of discourse. Then (∀x)φ(x) means that φ(x) holds for each value of the individual
x in our domain of discourse D. In other words, (∀x)φ(x) asserts that, for each a in D,
the proposition φ(a) holds. Thus, the proposition (∀x)φ(x) is true if, for each a in D, the
proposition φ(a) is true. On the other hand, (∀x)φ(x) is false if this condition fails to hold;
that is, if for some a in D, the proposition φ(a) is false.
Similarly, (∃x)φ(x) means that there exists a value of x in our domain of discourse D for
which φ(x) holds. Thus, (∃x)φ(x) asserts that there is some a in D such that φ(a) holds.
From this it is clear that (∃x)φ(x) is true if there is some a in D such that φ(a) holds, and
that (∃x)φ(x) is false if φ(a) happens to be false for every a in D.
This indicates that there is a close connection between (∀x) and (∃x). We will discuss
this in greater detail later in the project.
The symbol (∀x) is referred to as a universal quantifier since (∀x)φ(x) asserts that φ(x)
holds universally, i.e., for all values of x, and the symbol (∃x) is referred to as an existential
quantifier since (∃x)φ(x) asserts that there exists a value of x for which φ(x) holds.
Exercise 21. Let E(x) be the predicate “x is an even natural number” and O(x) be the
predicate “x is an odd natural number.” Express each of the following statements symboli-
cally using quantifiers and the predicates E(x) and O(x).

(a) There is at least one even natural number.

(b) Every natural number is either even or odd.
(c) Some natural number is both even and odd.
(d) No natural number is both even and odd.
(e) Every natural number that is not even is odd.

16
(f) The square of every even natural number is even.

Which of the above statements are true and which are false? Explain why.

Exercise 22. Let R(x) denote the predicate “x is a real number,” Z(x) denote the predicate
“x is an integer,” O(x) denote the predicate “x is an odd integer,” and P (x) denote the
predicate “x is a prime number.” Translate each of the following statements into English.

(a) (∀x)(Z(x) → R(x))

(b) (∀x)(P (x) → O(x))

(c) (∃x)(P (x) ∧ R(x))

Which of the above statements are true and which are false? Explain why.

Exercise 23. For each of the following statements, define appropriate predicates and write
the statement using predicates and quantifiers.

(a) Every square is a rectangle.

(b) Every real number is either positive, negative, or equal to zero.

(c) Every animal that has a heart has kidneys, and every animal that has kidneys has a
heart.

(d) Some natural number is not a prime number.

3.3 Unary and binary predicates

Our logical analysis of the proposition “every prime number greater than 2 is odd” relied on
our ability to attribute certain properties to an individual variable x, and the introduction
of predicates was what allowed us to accomplish this. Predicates are very versatile: they
can provide symbolic representations for a multitude of interesting properties. For example,
such properties as “being a rectangle,” “being a prime number,” “being greater than 2” are
represented by the predicates “x is a rectangle,” “x is a prime number,” and “x is greater
than 2,” respectively. However, not every property that is of mathematical interest can be
readily expressed using such predicates. For example, one may wish to express the fact that
there exists a natural number n such that n ≤ m holds for all natural numbers m (i.e., the
fact that the natural numbers have a least member).
The most obvious way to formalize this statement is to introduce a symbol P (x, y) that
expresses “x and y are natural numbers such that x ≤ y.” We may then express the existence
of a least natural number by the proposition “(∃x)(∀y)P (x, y).” In this expression, the role
of the symbol P (x, y) is similar to the role that predicates played in our previous discussion.
Indeed, P (x, y) is a sort of predicate, but, unlike the predicates previously discussed, P (x, y)
accepts two variables as arguments rather than one. Such a predicate is referred to as a
binary predicate. Those predicates that attribute a property to only a single variable (such
as those discussed previously) are referred to as unary predicates.

17
Exercise 24. Let N (x) denote the unary predicate “x is a natural number,” Z(x) denote
the unary predicate “x is an integer,” and P (x, y) denote the binary predicate “x ≤ y.”
Represent each of the following propositions symbolically.

(a) There is a least natural number.

(b) There is a greatest natural number.

(c) There is a least integer.

(d) There is no greatest integer.

Which of the above propositions are true? Explain why.

Exercise 25. Let I(x, y) denote the binary predicate “the point x lies on the line y” (if
a point x lies on a line y then x is said to be incident to y). Write each of the following
propositions verbally in words.

(a) (∀x)(∀y)(∃z)(I(x, z) ∧ I(y, z))

(b) (∀x)(∃y)(I(y, x))

What is the mathematical meaning of these two statements? Are they true or false? Explain
why.

Exercise 26. Allowing the variables x and y to range over the domain of all lines in the
Cartesian plane, let P (x, y) denote the binary predicate “x is parallel to y.” Write each of
the following propositions verbally in words.

(a) (∀x)(P (x, x))

(b) (∀x)(∀y)(P (x, y) → P (y, x))

(c) (∀x)(∀y)(∀z)((P (x, y) ∧ P (y, z)) → P (x, z))

When the three propositions above hold for a binary predicate P (x, y), then the relationship
defined between x and y by the predicate P (x, y) is said to be an equivalence relation. Is “x
is parallel to y” an equivalence relation?

Exercise 27. Recall that a natural number x is a multiple of a natural number y if there
exists a natural number n such that x = ny. Allowing the variables x and y to range over
the domain of all natural numbers, let P (x, y) denote the binary predicate “y is a multiple
of x.” Is P (x, y) an equivalence relation? Justify your answer.

Exercise 28. For natural numbers x and y, define x mod y to be the remainder obtained
upon dividing x by y. Allowing the variables x and y to range over the domain of natural
numbers, let P (x, y) denote the binary predicate “x mod 2 = y mod 2” and let Q(x, y) denote
the binary predicate “x mod 3 = y mod 3.”

(a) Is P (x, y) an equivalence relation? Justify your answer.

18
(b) Is Q(x, y) an equivalence relation? Justify your answer.

(b) Fix a natural number n.

(c) 1. If P (n, 2) holds, what must be true of n?

2. What about if P (n, 1) holds?
3. For a fixed natural number n, must at least one of P (n, 2) or P (n, 1) hold?
4. Can both P (n, 2) and P (n, 1) hold for a single natural number n?

(d) 1. List the natural numbers n for which Q(n, 1) holds, list the natural numbers n
for which Q(n, 2) holds, and list the natural numbers n for which Q(n, 3) holds.
2. Do the three lists have any natural numbers in common? Does every natural
number appear in at least one of the three lists?
3. For each of the lists, the collection of numbers appearing in that list is said to be
an equivalence class of the equivalence relation Q(x, y). What are the equivalence
classes of the equivalence relation P (x, y)?

3.4 Logical equivalence of quantified statements

Precisely characterizing when two quantified statements are logically equivalent turns out
to be rather technical, and a thorough treatment of that topic is beyond the scope of this
project. Nevertheless, from the comments above one can see that the proposition (∀x)φ(x)
is true exactly when there is no value of x in our domain of discourse for which φ(x) is false.
That is, (∀x)φ(x) is true exactly when ∼(∃x)∼φ(x) is true. One may likewise note that
(∃x)φ(x) is true exactly when φ(x) fails to be false for all x. In other words, (∃x)φ(x) is true
exactly when ∼(∀x)∼φ(x) is true. As one may expect from this discussion, it so happens
that (∀x)φ(x) is logically equivalent to ∼(∃x)∼φ(x) and that (∃x)φ(x) is logically equivalent
to ∼(∀x)∼φ(x).

Exercise 29. Explain in your own words why

(∀x)φ(x) ≡ ∼(∃x)∼φ(x)

and
(∃x)φ(x) ≡ ∼(∀x)∼φ(x).

One may derive many useful facts about the logical equivalence of quantified statements
from these two laws.

Exercise 30. Use the laws (∀x)φ(x) ≡ ∼(∃x)∼φ(x) and (∃x)φ(x) ≡ ∼(∀x)∼φ(x) to rewrite
each of the following statements. Make sure that in your solution all quantifiers precede any
instances of negation.

(a) ∼(∃x)(∀y)(x < y)

(b) ∼(∀x)(∃y)(x < y)

19
Which of the above two statements is true? Explain why.

Exercise 31. Allowing x and y to range over the domain N, let E(x) denote the unary
predicate “x is even,” O(x) denote the unary predicate “x is odd,” and L(x, y) denote the
binary predicate “x < y.”

(a) Express the proposition “Every odd natural number is less than some even natural
number” using quantifiers and the predicates E(x), O(x), and L(x, y).

(b) Use the laws (∀x)φ(x) ≡ ∼(∃x)∼φ(x) and (∃x)φ(x) ≡ ∼(∀x)∼φ(x) to find the negation
of the proposition obtained in (a). Make sure that in your solution all quantifiers
precede any instances of negation.

(c) Which of the two propositions is true, the one obtained in (a) or its negation? Justify
your answer.

The laws (∀x)φ(x) ≡ ∼(∃x)∼φ(x) and (∃x)φ(x) ≡ ∼(∀x)∼φ(x) are of central importance
in studying the logical equivalence of quantified statements, but a number of other simple
equivalences are also of tremendous importance. For example, it is readily seen that if
P (x, y) is a binary predicate, then (∀x)(∀y)P (x, y) is logically equivalent to (∀y)(∀x)P (x, y)
and that (∃x)(∃y)P (x, y) is logically equivalent to (∃y)(∃x)P (x, y). We describe these simple
equivalences by saying that universal quantifiers (resp. existential quantifiers) commute with
one another.
A far more interesting question is whether universal and existential quantifiers commute
with each other, i.e., whether statements of the forms (∀x)(∃y)P (x, y) and (∃y)(∀x)P (x, y)
are in general logically equivalent. The next three exercises explore this question.

Exercise 32. Allowing the variables x and y to range over the domain of all people, let
L(x, y) denote the binary predicate “x likes y.”

1. Write each of the following two statements symbolically.

(a) Everybody likes somebody.

(b) Somebody likes everybody.

2. Which of the above two statements is true and which is false?

3. Do the above two statements have the same meaning?

4. Are the two statements (∀x)(∃y)L(x, y) and (∃y)(∀x)L(x, y) logically equivalent? Jus-
tify your answer.

5. What can you conclude about the relationship between universal quantifiers and exis-
tential quantifiers from the above example?

Exercise 33. Allowing the variables x and y to range over the domain R, let E(x, y) denote
the binary predicate “x = y.” Write each of the following statements verbally in words.

(a) (∀x)(∃y)E(x, y)

20
(b) (∃x)(∀y)E(x, y)

Which of the above propositions are true? What can you conclude about the logical equiv-
alence of the propositions (∀x)(∃y)E(x, y) and (∃x)(∀y)E(x, y)?

Exercise 34. Give your own example of a binary predicate P (x, y) such that (∀x)(∃y)P (x, y)
and (∃x)(∀y)P (x, y) are not logically equivalent. Justify your example.

We conclude our discussion of the logical equivalence of quantified statements by dis-

cussing several equivalences that hold between quantified conditional statements. We have
seen that in propositional logic a conditional statement p → q is logically equivalent to its
contrapositive ∼q → ∼p, and that the converse q → p of this conditional is equivalent to its
inverse ∼p → ∼q. It turns out that similar equivalence laws hold for quantified statements.
Let φ(x) and ψ(x) be predicates. We call propositions of the form (∀x)(φ(x) → ψ(x))
universal conditional statements. If (∀x)(φ(x) → ψ(x)) is a universal conditional statement,
we define its contrapositive to be the statement (∀x)(∼ψ(x) → ∼φ(x)), its converse to
be the statement (∀x)(ψ(x) → φ(x)), and its inverse to be the statement (∀x)(∼φ(x) →
∼ψ(x)). Just as the conditional studied in propositional logic is logically equivalent to the
contrapositive, so too is the universal conditional (∀x)(φ(x) → ψ(x)) logically equivalent
to its contrapositive (∀x)(∼ψ(x) → ∼φ(x)). Similarly, the converse (∀x)(ψ(x) → φ(x)) is
logically equivalent to the inverse (∀x)(∼φ(x) → ∼ψ(x)).

Exercise 35. Let φ(x) and ψ(x) be unary predicates, and let χ(x, y) be a binary predicate.
Determine which of the following logical equivalences hold. Justify your answers.

(a) (∃x)(φ(x) ∧ ψ(x)) ≡ (∃x)φ(x) ∧ (∃x)ψ(x)

(b) (∀x)(φ(x) → (∃y)χ(x, y)) ≡ ∼(∃x)(φ(x) ∧ (∀y)∼χ(x, y))

(c) (∃x)(∃y)χ(x, y) → (∃y)(∃x)χ(x, y) ≡ (∀y)(∀x)χ(x, y) → (∀x)(∀y)χ(x, y)

3.5 Inference rules in predicate logic

Our development of individual variables, predicates, and quantifiers has provided us with
a language for discussing propositions that is much more expressive than the propositional
language discussed in the previous section. In particular, predicate logic is capable of ad-
dressing questions about complex propositions such as “every prime number greater than 2
is odd” that the propositional logic of the previous section cannot. However, we have not
yet discussed the issue of logical inference in predicate logic. This is the topic to which we
now turn.
Recall that if p and q are propositions, then one can deduce the proposition q from the
propositions p and p → q. We called this rule of inference Modus Ponens, and in the previous
section we proved that Modus Ponens is a valid rule of inference (i.e., if p and p → q are true,
then the deduced proposition q must also be true). As it turns out, there is an analogous
rule of inference for predicate logic.
Suppose that we fix some domain D over which our individual variables may range, that
a is some element in D, and that P (x) and Q(x) are predicates. Then if both (∀x)(P (x) →

21
Q(x)) and P (a) hold, we may deduce that Q(a) holds as well. This gives a valid rule of
inference, which we call Universal Modus Ponens.
A mathematically rigorous verification of the validity of Universal Modus Ponens requires
an analogue of truth tables for predicate logic, and is beyond the scope of this project.
However, we can give an informal argument for the validity of Universal Modus Ponens as
follows.
Suppose that both (∀x)(P (x) → Q(x)) and P (a) hold; we must show that Q(a) holds as
well. Since (∀x)(P (x) → Q(x)) holds, we have from the definition of the universal quantifier
that the proposition P (b) → Q(b) holds for each member b of our specified domain D. In
particular, as a is an element of D, we have that P (a) → Q(a) holds. Now notice that both
P (a) → Q(a) and P (a) are propositions. Since Modus Ponens is a valid rule of inference of
propositional logic, we have that as P (a) → Q(a) and P (a) hold, it follows that Q(a) holds
as well, completing our argument.
Exercise 36. Fix some domain D over which we allow individual variables to range, suppose
a is an element of D, and further suppose that P (x) and Q(x) are predicates. Provide an
informal verification that the following rule of inference is valid: If both (∀x)(P (x) → Q(x))
and ∼Q(a) hold, infer that ∼P (a) holds as well. (This rule is referred to as Universal Modus
Tollens.)
Exercise 37. Fix some domain D over which we allow individual variables to range, suppose
a is an element of D, and further suppose that P (x) and Q(x) are predicates. Consider the
following rule of inference: if both (∀x)(P (x) → Q(x)) and Q(a) hold, then infer that P (a)
holds. Provide an informal argument of whether this is a valid rule of inference.
Exercise 38. A certain city is governed by a council consisting of five voting members: Fay,
Gould, Haber, Ishikawa, and Jacobs. The council frequently votes on issues pertaining to
city government, and several facts about the voting patterns of the council members are
known:

(1) On every proposal, if Fey votes in favor of the proposal then so does Gould.

(2) Ishikawa and Jacobs never both vote in favor of the same proposal.

(3) On every proposal, if Haber does not vote in favor of the proposal then neither does
Gould.

Suppose that a new budget for the city is going to come up for a vote before the council,
and that Fay will vote in favor of the budget. Formalize this situation using predicate logic,
and deduce whether the budget will pass. Justify every step in your deduction. (In order for
the budget to pass, at least three voting members of the council must vote in favor of it.)

4 Looking Forward
The logical system we have studied in this project is today known as classical logic. The
publication of Principia Mathematica led to a considerable amount of research in classical
logic in the first half of the 20th century, and the work of such logicians as David Hilbert

22
(1862-1943), Alfred Tarski (1901-1983), and Kurt Gödel (1906-1978) during this period has
led to a thorough understanding of classical logic. However, original research in logic con-
tinues today. Contemporary research in logic focuses on a variety of so-called non-classical
logics, each of which is in some way a modification or expansion of classical logic. The clas-
sical logic developed in this project hence forms the core of contemporary research, and thus
remains of relevance almost a century after the publication of Principia Mathematica.

Notes to the Instructor

This project has its roots in the authors’ experience teaching discrete mathematics from
primary historical sources. It was designed to serve the needs of college freshmen and sopho-
mores who are meeting mathematical proofs for the first time. However, except for Exercise
19 (which assumes a basic familiarity with first-year calculus), no specific prerequisites are
assumed.
Each section of the project has a large variety of exercises. Instructors can pick and
choose which exercises to assign, depending on what parts of the project they will cover.
Since we cover the whole project, we assign virtually every exercise.
The notation introduced in Principia Mathematica is in some instances antiquated. In
contemporary literature, the logical connectives “and” and “if , then” are usually denoted by
∧ and →. Also, the universal quantifier is more commonly written as (∀x). For pedagogical
reasons, we opted to follow more contemporary usage of these symbols.

References
[1] Barnett, J., Origins of Boolean Algebra in the Logic of Classes: George Boole, John Venn
and C. S. Peirce. http://www.cs.nmsu.edu/historical-projects/

[2] Bezhanishvili, G., Peano Arithmetic. http://www.cs.nmsu.edu/historical-projects/

[3] Boole, G., An Investigation of the Laws of Thought. Macmillan, London, 1854.

[4] Irvine, A.D., Alfred North Whitehead. Stanford Encylopedia of Philosophy.

http://plato.stanford.edu

[5] Lodder, J., Deduction through the Ages: A History of Truth. Available from
http://www.cs.nmsu.edu/historical-projects/

[6] Lowe, V., Alfred North Whitehead: The Man and His Work, Volume I: 1861-1910. Johns
Hopkins University Press, Baltimore, 1985.

[7] Monk, R., Bertrand Russell: The Spirit of Solitude, 1872-1921. The Free Press, New
York, 1996.

[8] “The Nobel Prize in Literature 1950,” available from nobelprize.org,

http://nobelprize.org/nobel prizes/literature/laureates/1950/

23
[9] Whitehead A. N. and Russell, B., Principia Mathematica. Cambridge University Press,
Cambridge, England, 1910 (Vol. 1), 1912 (Vol. 2), 1913 (Vol. 3). Rereleased by Merchant
Books in 2009.

24