Explanation
Suppletion occurs when forms within one grammatical paradigm derive from historically different stems. English go–went or good–better–best shows that strong functional relatedness can exist without transparent formal derivation. What speakers synchronically learn and use as one paradigm has grown together diachronically from separate lexical sources. The concept therefore disrupts the expectation that grammatical regularity must always produce visible similarity of sound. Suppletion is not simply any irregular formation: analysts normally require a major stem difference, although boundaries between strong allomorphy and suppletion vary by theory. The movement of thought leads from outward difference to systemic unity and presents paradigms as historically layered functional orders rather than sets generated by one uniform operation.
Movement of thought
1. A grammatical paradigm first creates the expectation that its forms share a recognisable stem and undergo regular changes. 2. For particular, often frequent lexemes, one paradigm cell contains a form that cannot be phonologically derived from the expected stem. 3. Grammatical function, distribution, and speaker knowledge nevertheless bind the outwardly different forms into one lexical unit. 4. A counter-test examines historical relationship, productive sound rules, and degree of allomorphy before classifying the case as suppletion. 5. The paradigm returns as a functional unity whose visible discontinuities preserve the history of its composition.
Scientific use case
Morphology, acquisition research, and historical linguistics compare paradigms, frequency, and etymology. Psycholinguistic experiments test whether suppletive forms share lexical storage or are processed differently from regular inflection.
Example
Speakers learn go and went as forms of the same verb even though the phonological shape of go cannot predict the past-tense form went.
Counterexample
Small–smaller–smallest changes through a productive comparative pattern and is not suppletive. A spelling alternation alone likewise does not establish independent stems.
Distinction
Distinguish suppletion from regular inflection, phonologically explainable stem alternation, and spelling variation. Its boundary with strong allomorphy is theory-dependent; suppletion usually names the extreme of formal discontinuity within a paradigm.
Effect on thought and perception
Suppletion reveals that grammatical unity may be stronger than formal resemblance and that language systems can functionally integrate historical discontinuities.
Resonance
Different word-bodies answer the same grammatical task; their coupling allows one paradigm to arise beyond the boundary of visible kinship.
Cite this entry
Reich, Michael (Hrsg./ed.): „Suppletion“. Atlas der Sprache, AT-103-008. ATLAS 260, data model 1.6.0. https://atlasoflanguage.com/concept/AT-103-008The address remains stable; editorial content may evolve. Editorial reading state: 2026-10-15. Include your access date for precise references.
Historical background
Veselinova’s WALS account distinguishes weak from strong suppletion, illustrated by buy/bought and go/went, without treating them as a strict binary. Its formal criterion is synchronic, not dependent on distinct etymologies. The German IDS entry supplies a complementary terminological account and German paradigms. Neither source establishes the earliest English use of suppletion. The previous requirement to demonstrate separate stem histories was too restrictive for the cited typological framework; explaining a form’s historical recruitment remains a separate task.
How conceptual history changes the analysis
Identify the relevant paradigm and its formal patterns before applying a chosen definition of suppletion. Historical origin can explain the relation but is not a universal prerequisite for diagnosing it.
Additional example
Original teaching comparison: good/better and small/smaller perform comparable grammatical work through different formal relations. The comparison does not itself explain the origin of better.