AI Glossary / Graphs and symbolic AI / Intermediate
Atomic sentence
Definition
An atomic sentence is a basic statement in a formal logical language that is not built by combining other statements with logical connectives or quantifiers. In first-order logic, a closed atomic statement commonly applies a predicate to terms or states that two terms are equal.
What it means
A statement such as Approved(Order17) can express that a particular order is approved. The predicate Approved identifies a property, and Order17 identifies the object being described. A relation such as Supplies(VendorA, PartB) connects two objects in a specified order.
Atomic does not mean that the statement has no internal syntax. Its terms can themselves be complex, such as Owner(Order17). The important distinction is that the whole statement is not formed using and, or, not, implication or a quantifier.
Terminology varies slightly across introductory AI materials. Strictly, a sentence has no free variables. Approved(x) is an atomic formula with a free variable unless x is otherwise bound or assigned in a surrounding context. Some teaching sources use atomic sentence more loosely for the basic predicate form. This distinction helps avoid confusing a particular fact with a general claim.
How it works
A logical language specifies constants, function symbols, predicate symbols and formation rules. Terms refer to objects; predicates express properties or relations. An interpretation supplies the domain of objects and the meaning of the symbols.
The truth of an atomic statement depends on that interpretation. Supplies(VendorA, PartB) is true when the ordered pair represented by VendorA and PartB belongs to the relation assigned to Supplies. Reversing the arguments creates a different statement unless the relation is defined to be symmetric. Equality states that two terms refer to the same object under the interpretation.
Logical connectives build more complex statements from these foundations. Approved(Order17) AND Paid(Order17) combines two atoms, while NOT Approved(Order17) is a negated atom and therefore not itself atomic in the usual syntax. A knowledge system must also decide how missing facts are handled; absence from a database is not automatically proof of falsity in classical first-order reasoning.
Three practical examples
- A rule-based purchasing system records Approved(Order17) as a fact about a particular order.
- A knowledge graph uses a relation such as LocatedIn(StoreA, RegionB) to represent a connection between named entities.
- A reasoning exercise distinguishes the atomic equality Owner(Order17) = CompanyA from a rule applying to every order.
Think of it this way
Think of a simple entry on a carefully defined fact card: one predicate applied to specified objects. Several cards can be joined into a rule, but each card has its own meaning.
The analogy is limited because a card's presence does not prove its statement true. Formal syntax specifies how a claim is written; evidence and the interpretation determine whether it accurately describes the intended situation.
Worked example
Building a rule from atomic statements
In an invented order model, A means Approved(Order17), P means Paid(Order17) and R means Ready(Order17). The knowledge base contains A, P and the rule (A AND P) implies R.
A, P and R are each atomic statements in this simplified notation. A AND P is a compound statement, and the implication is a larger compound statement. Given the two facts and the rule, the system can infer R.
If P is not known, the rule alone does not establish R. Nor does failure to derive R prove NOT R in classical logic. Another rule or fact could establish readiness. The example separates the building blocks, the inference rule and the handling of missing knowledge.
Illustrative case study
Fictional case: Willow Inventory builds a small supplier knowledge base. One import writes Supplies(PartB, VendorA), while the documented relation expects the supplier first and the part second. The expression looks syntactically valid but represents the wrong relationship.
The team adds a clear relation dictionary, checks argument roles during import and reviews how source evidence supports each fact. It also distinguishes missing supplier information from an explicit statement that no supplier exists.
The case illustrates why well-formed atomic statements still require consistent semantics and data validation. A reasoning engine cannot repair a reversed meaning simply because the symbols follow the grammar.
Common misconceptions and mistakes
- Atomic means one word with no internal parts. Predicate arguments and function terms can have structure.
- A syntactically valid atom must be true. Truth depends on its interpretation and the represented facts.
- A missing atom is always false. That depends on the system's knowledge assumptions; classical logic does not generally make that inference.
Business use
Atomic statements provide a useful foundation for explicit business facts, knowledge graphs and rule-based reasoning. Their value comes from precise predicate definitions, stable entity references and consistent argument order. A compact formal statement should make an intended claim clearer, not hide ambiguity behind symbols.
When designing a knowledge system, record the source of facts and the treatment of uncertainty, missing information and contradictions. Decide where rules apply universally and where statements concern individual objects. This makes the reasoning trace understandable to people reviewing a consequential workflow.
People also ask
Is a negated atom atomic?
Under the usual syntax, no. An atom or its negation is commonly called a literal.
Can an atomic sentence contain a function?
Yes. A function term can appear as an argument, provided the sentence is otherwise well formed and closed.
Is Approved(x) a sentence?
Strictly, it is an atomic formula with a free variable; a closed sentence requires that no variables remain free.
Check your understanding
Which is atomic: Approved(Order17), NOT Approved(Order17), or Approved(Order17) AND Paid(Order17)?
Answer: Approved(Order17) is atomic. The other expressions use logical connectives and are compound, although the negated expression is a literal.
Related terms
Learn next
Read predicate and literal. Express three facts from a simple business process as predicates with clearly ordered arguments, then write one rule that combines them without confusing missing facts with false facts.
More in Graphs and symbolic AI
- ATMS - ATMS stands for assumption-based truth maintenance system.
- Atom - An atom is a basic logical formula that is not built by combining simpler formulas with logical connectives or quantifiers.
- Atomic representation - An atomic representation describes each state as a single indivisible unit for the reasoning method using it.
- Attribute-based extraction - Attribute-based extraction is the identification of selected properties, such as price, date or color, in source material and their conversion into...
- Augmented grammar - An augmented grammar is a grammar extended with information or conditions beyond simple phrase-building rules.
- Authority - In link analysis, an authority is a page or node that receives strong endorsement through incoming links from useful hub nodes.
Sources
Russell and Norvig, official AIMA Chapter 8 materials: https://aima.cs.berkeley.edu/4th-ed/pdfs/newchap08.pdf and https://aima.cs.berkeley.edu/4th-ed/slides-pdf/chapter08.pdf. Relevant returned text describes predicates, terms, equality and interpretation. Open Logic Project, forall x: Calgary: https://forallx.openlogicproject.org/html/Ch31.html. The closed-sentence distinction is stated explicitly; workflow examples are original illustrations.
Previous: Atomic representation Next: Atomicity (database systems)