Skip to main content
Social Sci LibreTexts

4.4: Predicate logic

  • Page ID
    138645
  • \( \newcommand{\vecs}[1]{\overset { \scriptstyle \rightharpoonup} {\mathbf{#1}} } \) \( \newcommand{\vecd}[1]{\overset{-\!-\!\rightharpoonup}{\vphantom{a}\smash {#1}}} \)\(\newcommand{\id}{\mathrm{id}}\) \( \newcommand{\Span}{\mathrm{span}}\) \( \newcommand{\kernel}{\mathrm{null}\,}\) \( \newcommand{\range}{\mathrm{range}\,}\) \( \newcommand{\RealPart}{\mathrm{Re}}\) \( \newcommand{\ImaginaryPart}{\mathrm{Im}}\) \( \newcommand{\Argument}{\mathrm{Arg}}\) \( \newcommand{\norm}[1]{\| #1 \|}\) \( \newcommand{\inner}[2]{\langle #1, #2 \rangle}\) \( \newcommand{\Span}{\mathrm{span}}\) \(\newcommand{\id}{\mathrm{id}}\) \( \newcommand{\Span}{\mathrm{span}}\) \( \newcommand{\kernel}{\mathrm{null}\,}\) \( \newcommand{\range}{\mathrm{range}\,}\) \( \newcommand{\RealPart}{\mathrm{Re}}\) \( \newcommand{\ImaginaryPart}{\mathrm{Im}}\) \( \newcommand{\Argument}{\mathrm{Arg}}\) \( \newcommand{\norm}[1]{\| #1 \|}\) \( \newcommand{\inner}[2]{\langle #1, #2 \rangle}\) \( \newcommand{\Span}{\mathrm{span}}\)\(\newcommand{\AA}{\unicode[.8,0]{x212B}}\)

    Consider the simple sentences in (21):

    (21) a. John is hungry.
    b. Mary snores.
    c. John loves Mary.
    d. Mary slapped John.

    Each of these sentences describes a property, event or relationship. The element of meaning which determines what kind of property, event or relationship is being described is called the predicate. The words hungry, snores, loves, and slapped express the predicates in these examples. The individuals of whom the property or relationship is claimed to be true (John and Mary in these examples) are referred to as arguments. As we can see from example (21), different predicates require different numbers of arguments: hungry and snore require just one, love and slap require two. When a predicate is asserted to be true of the right number of arguments, the result is a well-formed proposition, i.e., a claim about the world which can (in principle) be assigned a truth value, T or F.

    In our logical notation we will write predicates in capital letters (to distinguish them from normal English words) and without inflectional morphology. We follow the common practice of using lower case initials to represent proper names. For predicates which require two arguments, the agent or experiencer is normally listed first. So the simple sentence John is hungry would be translated into the logical metalanguage as HUNGRY(j), while the sentence John loves Mary would be translated LOVE(j,m). Some additional examples are shown in (22).

    (22) a. Henry VIII snores. SNORE(h)
    b. Socrates is a man. MAN(s)
    c. Napoleon is near Paris. NEAR(n,p)
    d. Abraham Lincoln admired Queen Victoria. ADMIRE(a,v)
    e. Jocasta is Oedipus’ mother. MOTHER_OF(j,o)
    f. Abraham Lincoln was tall and homely. TALL(a) ∧ HOMELY(a)
    g. Abraham Lincoln was a tall man. TALL(a) ∧ MAN(a)
    h. Joe is neither honest nor competent. ¬ (HONEST(j) ∨ COMPETENT(j))

    As these examples illustrate, semantic predicates can be expressed grammatically as verbs, adjectives, common nouns, or even prepositions. They can appear as part of the VP, or as modifiers within NP as in (22g).1

    We have seen examples of one-place and two-place predicates; there are also predicates which take three arguments, e.g. give, show, offer, send, etc. Some predicates, including verbs like say, think, believe, want, etc., can take propositions as arguments:

    (23) a. Henry thinks that Anne is beautiful. THINK(h, BEAUTIFUL(a))
    b. Susan wants to marry Ringo. WANT(s, MARRY(s,r))

    4.4.1 Quantifiers (an introduction)

    All the predicates in examples (21–23) have proper names as arguments. Of course we need to be able to represent other kinds of arguments as well. We will discuss this issue in more detail in later chapters, but as a brief introduction let us consider the subject NPs in (24):

    (24) a. All students are weary.
    b. Some men snore.
    c. No crocodile is warm-blooded.

    The italicized phrases in (24) are examples of “quantified” NPs; they contain a special kind of determiner known as a quantifier. Sentence (24a) makes a universal generalization. It says that if you select anything within the universe of discourse that happens to be a student, that thing will also be weary. Notice that the phrase all students does not refer to any specific individual, or set of individuals; that is why we said in Chapter 2 that quantified NPs are generally not referring expressions. Rather, the phrase seems to express a kind of inference: if a given thing is a student, then it will also have the property expressed in the remainder of the sentence.

    Sentence (24b) makes an existential claim. It says that there exists at least one thing within the universe of discourse that is both a man and snores. Actually, this sentence says that there must be at least two such things, but that is not part of the meaning of some; it follows from the fact that the noun men is plural. (We can show this by comparing (25a) with (25b).) Some simply means that there exists something within the universe of discourse that has both of the named properties (e.g., being a man and snoring). Sentence (24c) is a negative existential statement. It says that there does not exist anything within the universe of discourse that is both a crocodile and warm-blooded.

    (25) a. Some guy in the back row was snoring. (at least one)
    b. Some guys in the back row were snoring. (at least two)

    Standard predicate logic makes use of two quantifier symbols: the Universal Quantifier ∀ and the Existential Quantifier ∃. As the mathematical examples in (26) illustrate, these quantifier symbols must introduce a variable, and this variable is said to be bound by the quantifier. The letters x, y or z are normally used as variables that represent individuals. (We can read “∀x” as ‘for all individuals x’, and “∃x” as ‘there exists one or more individuals x’.)

    (26) a. Universal Quantifier:
    ∀x[x+x = 2x]

    b. Existential Quantifier:
    ∃y[y+4 = y/3]

    Quantifier words must be interpreted relative to the current universe of discourse, that is, the set of individuals currently available for discussion. For example, in order to decide whether sentences like All students are female or No student is wealthy are true, we need to know what the currently relevant universe of discourse is. If we are discussing a secondary school for economically disadvantaged girls, both statements would be true. In other contexts, either or both of these statements might be false.

    In the same way, variables bound by one of the logical quantifier symbols are assumed to be members of the currently relevant universal set, i.e., the set of all elements currently available for consideration.2 In mathematical contexts, the universal set is often a particular class of numbers, e.g. the integers or the real numbers. In order to evaluate a proposition involving quantifier symbols, like those in (26), the universal set must be specified or assumed from context.

    Variables bound by a quantifier do not refer to a specific individual or entity, but rather allow for the arbitrary selection of any individual or entity within the universal set. Once a particular value is assigned to a given variable, the same assignment is understood to hold for all occurrences of that variable within the scope of the quantifier (the material inside the square brackets). So for example, if we assume that the universal set in (26) is the set of all real numbers, (26a) can be interpreted as follows: “Choose any real number. If you add that number to itself, the sum will be equal to that number multiplied by two.” The equation in (26b) can be interpreted as follows: “There exists some real number which, when added to four, will be equal to the quotient of that same number divided by three.”

    The value of an unbound (or “free”) variable, that is, one which is not introduced by a quantifier or which occurs outside the scope of its quantifier, is not defined. The variables in (27) are not bound, and as a result the equations in which they occur are neither true nor false; they do not make any claim about the world, until some value is assigned to each variable. (In contrast, both of the equations in (26), where the variables are bound, can be shown to be true.) Of course, it is fairly easy to solve the equations in (27), that is, to find the values that must be assigned to each variable in order to make the equations true. But until some value is assigned, no truth value can be determined for the equations.

    (27) a. x–7 = 4x
    b. y + 2z = 51

    The same applies to variables which occur within logical formulae. A proposition that contains unbound variables is called an open proposition. Such a proposition cannot be assigned a truth value, unless some mechanism is provided for assigning values to the unbound variables.

    The universal and existential quantifier symbols allow us to translate the sentences in (24) into logical notation, as shown in (28). (We will ignore for the moment the difference in interpretation between singular vs. plural nouns with some.)

    (28) a. Universal Quantifier: All students are weary.
    ∀x[STUDENT(x) → WEARY(x)]

    b. Existential Quantifier: Some men snore.
    ∃x[MAN(x) ∧ SNORE(x)]

    c. Negative Existential: No crocodile is warm-blooded.
    ¬∃x[CROCODILE(x) ∧ WARM-BLOODED(x)]

    Notice that all is translated differently from some or no. The universal quantifier is paired with material implication (→), while the existential quantifier introduces an and statement. We will discuss the reason for this difference in more detail in Unit IV, but the fundamental issue is that we want our logical translation to have the same interpretation as the English sentence it is meant to represent. We might interpret the formula in (28a) roughly as follows: “Choose something within the universe of discourse. We will temporarily call that thing ‘x’. Is x a student? If so, then x will also be weary.” This long-winded paraphrase seems to describe the same state of affairs as the original sentence All students are weary. However, if we replace → with ∧, we get the formula in (29), which means something very different.

    (29) ∀x[STUDENT(x) ∧ WEARY(x)]
    ‘Everything in the universe of discourse is a student and is weary.’

    So far we have only considered quantifier phrases which occur as subject NPs, but of course they can occur in other syntactic positions as well. When we translate a sentence containing a quantified NP into logical notation, the quantifier always comes at the beginning of the proposition which it takes scope over, even when the quantified NP is functioning as direct object, oblique argument, etc. Some examples are presented in (30). Note that indefinite NPs are often translated as existential quantifiers, as illustrated in (30b–c).

    (30) a. John loves all girls.
    ∀x[GIRL(x) → LOVE(j,x)]

    b. Susan has married a cowboy.
    ∃x[COWBOY(x) ∧ MARRY(s,x)]

    c. Ringo lives in a yellow submarine.
    ∃x[YELLOW(x) ∧ SUBMARINE(x) ∧ LIVE_IN(r,x)]

    The patterns of inference observed in example (2) above illustrate two basic principles that govern the use of quantifiers. The first principle, which is called universal instantiation, states that anything which is true of all members of a particular class is true of any specific member of that class. This is the principle which licenses the inference shown in (2a), repeated here as (31a). The second principle, which is called existential generalization, licenses the inference shown in (2b), repeated here as (31b).

    (31) a. All men are mortal. ∀x[MAN(x) → MORTAL(x)]
    Socrates is a man. MAN(s)

    Therefore, Socrates is mortal. MORTAL(s)

    b. Arthur is a lawyer. LAWYER(a)
    Arthur is honest. HONEST(a)

    Therefore, some (= at least one) ∃x[LAWYER(x) ∧ HONEST(x)]
    lawyer is honest.

    4.4.2 Scope ambiguities

    When a quantifier combines with another quantifier, with negation, or with various other elements (to be discussed in Chapter 14), it can give rise to ambiguities of scope. In (32a) for example, one of the quantifiers must appear within the scope of the other, so there are two possible readings for the sentence.

    (32) a. Some man loves every woman.
    i. ∃x[MAN(x) ∧ (∀y[WOMAN(y) → LOVE(x,y)])]
    ii. ∀y[WOMAN(y) → (∃x[MAN(x) ∧ LOVE(x,y)])]

    b. All that glitters is not gold.
    i. ∀x[GLITTER(x) → ¬GOLD(x)]
    ii. ¬∀x[GLITTER(x) → GOLD(x)]

    The quantifier that appears farthest to the left in the formula gets a wide scope interpretation, meaning that it takes logical priority; the one which is embedded within the scope of the first quantifier gets a narrow scope interpretation. So the first reading for (32a) says that there exists some specific man who loves every woman. The second reading for (32a) says that for any woman you choose within the universe of discourse, there exists some man who loves her. Try to provide similar paraphrases for the two readings of (32b). Then try to verify that these sentences involve real ambiguities by finding contexts for each sentence where one reading would be true while the other is false.


    This page titled 4.4: Predicate logic is shared under a CC BY 4.0 license and was authored, remixed, and/or curated by Paul Kroeger (Language Library Press) via source content that was edited to the style and standards of the LibreTexts platform; a detailed edit history is available upon request.

    • Was this article helpful?