As the previous article showed, the Neogrammarians demanded a reason for every exception. In hunting down those reasons, they identified a second major engine of change, standing alongside sound law: “analogy.”

Starting from a child’s mistake

Small children sometimes say things like “I goed to the park” or “I eated my dinner.” This isn’t random nonsense. The child is applying a pattern they already know — the regular past-tense pattern, “walk” becomes “walked” — to a verb whose irregular form they haven’t fully learned yet.

A closely related shift shows up in adult English, too. The older past tense of “sneak” was the regular “sneaked.” Over the past century or so, a newer form, “snuck,” has spread widely, especially in American English — modeled on irregular verbs like “stick/stuck” or “cling/clung,” which already followed that pattern. Bit by bit, an existing pattern from one small group of verbs spread to a verb that didn’t originally belong to it.

Not a sound change, but the copying of a pattern

This is the key distinction. Sound-law change is a process in which a particular sound shifts, mechanically and unconsciously, little by little, over time. Analogical change, by contrast, is a process in which a speaker — usually without realizing it — borrows an existing pattern and applies it to a different word: “this word should work the same way as that group of words.” The first is a change in the sound itself; the second might be called the spread, or contagion, of a formal pattern.

Sound-law change the sound itself shifts mechanically, unconsciously little by little Analogical change an existing pattern is applied to a different word (e.g. "sneaked" → "snuck") Only by pairing sound law with analogy does the full picture of language change come into view
Sound law and analogy — two different forces reshaping a language. Diagram: ToL

Why this looks like an “exception”

If you try to explain the entire history of a language through sound law alone, any form produced by analogy will always look like an “exception” — because that form didn’t arise from a sound-level rule at all, but from copying a different pattern. By treating “sound law” and “analogy” as a paired set of forces, Brugmann’s generation was, for the first time, able to give a systematic reason for a great many exceptions that had previously gone unexplained.

Summary

PointDetail
Sound-law changeThe sound itself shifts mechanically, over time
Analogical changeA speaker applies an existing word-form pattern to a different word
Everyday exampleA child’s mistake (“goed”); the spread of a simpler pattern, as in “sneaked” → “snuck”
ConsequencePairing these two forces makes many “exceptions” explicable at last

In a nutshell — language is reshaped constantly, not only by the mechanics of sound, but by speakers extending patterns they already know onto other words.