A stop word is a frequent function word, such as “the,” “and,” or “are,” removed during text normalization because it contributes little search meaning. Postgres removes stop words while converting tokens into lexemes, alongside case normalization and stemming. section 6.2.1 section 6.1
Removal does not erase positional information. In the sentence “The explorers must save the fragile peace between Earth and the aliens,” “the” and “and” disappear, while “explorers” becomes “explor” at position 2. Preserved positions support phrase and distance operators. section 6.2.1
Observe stop-word removal with to_tsvector
SELECT * FROM to_tsvector('The explorers must save the fragile peace between Earth and the aliens.'); Postgres returns lexemes such as 'explor':2, 'must':3, 'save':4, 'fragil':6, 'peac':7, 'earth':9, and 'alien':12. The articles and prepositions are absent, but their positions still affect phrase matching. section 6.2.1
The same normalization applies to search queries. Functions that produce tsquery values remove stop words before combining remaining lexemes with operators. Phrase queries account for omitted words by using distance operators when needed. section 6.3 section 6.3.2
