Prolog Syntax and Terms

Every piece of data in Prolog — from a single number to an entire program — is represented as a term. Prolog recognizes four basic term types: atoms (lowercase-starting constants like paris or 'hello world' in quotes), numbers (integers like 42 or floats like 3.14), variables (uppercase-starting or underscore-starting placeholders like X or _Total), and compound terms (a functor with arguments, like point(3, 4) or date(2026, 7, 10)).

Atoms are Prolog's simplest constant type — symbolic names like red, john, or 'multi word atom' (quoted when they contain spaces or start uppercase) that always refer to themselves, never to something else. Numbers are standard integers and floating-point values usable in arithmetic via the is/2 operator, as in ?- Y is 3 * 7 + 1., while variables are single-binding placeholders within a proof — once a variable is bound during unification, it keeps that binding for the rest of that particular proof, and an unbound variable written as the underscore _ (the anonymous variable) means 'I don't care what matches here.'

A compound term combines a functor (an atom) with one or more arguments, written functor(Arg1, Arg2, ...) — for instance, book(prolog_programming, sterling, 2019) has functor book and arity 3 (three arguments), and Prolog treats terms with the same functor but different arity as entirely distinct, e.g. point/2 and point/3 are unrelated predicates. Lists are a special, heavily-used compound term written with square brackets, such as [milk, eggs, bread] or the recursive [Head|Tail] notation, where Head unifies with the first element and Tail unifies with the remaining list — this pattern is the foundation of nearly all list-processing predicates in Prolog.