Explanation
Analogical levelling is language change that reduces irregular differences within a paradigm by extending the pattern of frequent, transparent, or systemically aligned forms. Speakers do not process forms only as isolated items but relate them in series. When one member departs from the pattern, a productive formation may be extended to it, making the paradigm more regular while eliminating older alternations produced by historical sound change. Levelling need not affect every cell, and countervailing analogy may create new differences. It is not a conscious decision by individual speakers, although normative influence may aid diffusion. It differs from reanalysis because an existing formal relation is primarily remodelled after another pattern. The movement goes from a divergent member to the perceived paradigm and lets the force of the series act back upon that member's form.
Movement of thought
1. An inflectional or derivational paradigm initially contains related forms, including historically produced alternations or exceptions. 2. Speakers recognise a productive pattern among frequent members and relate a less transparent form to that series. 3. The divergent form is rebuilt after other paradigm cells or similar words and the innovation is repeated in usage. 4. A counter-test examines historical stages and separates levelling from regular sound change, reanalysis, borrowing, and performance error. 5. The paradigm returns in a more regular shape, while its new order records which model forms exerted particular analogical force.
Scientific use case
Historical morphology, dialectology, and corpus linguistics trace variants over time, paradigm frequency, and regional diffusion. Reconstruction compares sound-law outcomes with analogically remodelled forms.
Example
When an older isolated past-tense form is replaced by a new form modelled on the frequent pattern of other verbs, alternation within the morphological system is reduced.
Counterexample
A vowel changes regularly in every word occurring in the same phonological environment, regardless of inflectional paradigm. Such a sound change is not analogical levelling.
Distinction
Distinguish it from analogical innovation generally, proportional analogy, reanalysis, regular sound change, inflection, paradigm, acquisition overgeneralisation, and prescriptive unification; demonstrate model form and historical sequence.
Effect on thought and perception
Analogical levelling increases transparency and predictability within morphological series but may erase older distinctions; competing models create transitional variation and new exceptions.
Resonance
An individual form is drawn into the current of its series; the paradigm answers divergence until resemblance reorganises its internal landscape.
Cite this entry
Reich, Michael (Hrsg./ed.): „Analogical levelling“. Atlas der Sprache, AT-ETY-W0277. ATLAS 260, data model 1.6.0. https://atlasoflanguage.com/concept/AT-ETY-W0277The address remains stable; editorial content may evolve. Editorial reading state: 2026-10-15. Include your access date for precise references.
Historical background
Sapir's 1921 discussion explains analogical levelling through older and oldest replacing earlier vowel-alternating forms. He compares English old/elder with German alt/älter and also describes the expansion of the English s-plural beyond its former class. His account concerns the reshaping of forms according to patterns rather than sound change alone. Its historical reconstruction is documented as Sapir's account, without independent collation of medieval witnesses.
How conceptual history changes the analysis
The documented German–English comparison concerns different trajectories of forms and supports no general ranking of the languages.
References for the historical section
- Edward Sapir, Language: An Introduction to the Study of Speech (1921) · Chapter VIII, foot paradigm, older/elder and generalisation of plural -s
Original English text read in the Gutenberg transcription. No first-coinage claim, full-book verification or automatic endorsement of the author’s 1921 language classifications and generalisations. Examples require modern descriptive checking before reuse as current language data.