As the previous article showed, Grimm himself had already noticed a real case his own table couldn’t fit. In fact, an even more conspicuous exception to the law was known long before that. English father and brother — two words in a similar position, and yet their consonants behave differently for no apparent reason. That small mismatch was resolved, roughly a decade after Grimm’s death, from a direction nobody expected.
Two words that look alike
Grimm’s Law predicts that Proto-Indo-European t should become the voiceless fricative þ (th) in Germanic. And indeed, the th in brother is exactly the sound the rule predicts. But the consonant in father historically descends not from a voiceless th, but from a voiced, d-type sound (written ð in Old English). Both words have a similar shape — a root ending in t, with a suffix attached — so why does only one of them break the rule? Grimm himself never answered that question. Mid-nineteenth-century linguists mostly treated it, to some extent, as an unavoidable “exception to Grimm’s Law.”
The clue was an accent mark, buried in distant Sanskrit
In 1875, the Danish linguist Karl Verner (1846–1896) published a paper that cracked the puzzle. What Verner focused on wasn’t the consonant itself, but a piece of information that seemed, at first glance, unrelated: where the original word’s stress accent had fallen. Unlike Latin or English, Sanskrit’s ancient texts preserve marks showing exactly where the accent sat. Using this, Verner compared Sanskrit pitā́r- (accent on the later syllable), corresponding to father, with bhrā́tar- (accent on the earlier syllable), corresponding to brother.
When the accent did not fall on the syllable immediately before the consonant, the voiceless fricative turned voiced — this was the rule Verner uncovered. In pitā́r-, the accent does not sit right after the root, and the corresponding Germanic consonant became voiced (giving the d-type sound in father). In bhrā́tar-, the accent sits right before the root, and the consonant stayed voiceless (giving the th in brother). Over the centuries, the original Proto-Indo-European accent had already shifted, in Germanic, to a fixed position (the first syllable) — so looking only at the historically attested Germanic languages, this conditioning is completely invisible. It only surfaced once compared against Sanskrit, a language that happened to preserve the accent information that Germanic itself had already erased.
A step toward “laws without exceptions”
Verner’s discovery did more than solve a single puzzle. What looked like an “exception” turned out to be the visible trace of a different, equally regular condition, sitting one layer deeper — and that idea grew, in time, into a founding principle of comparative linguistics itself: sound laws admit no exceptions (any apparent exception must have its own sound law explaining it). The scholars at the end of the nineteenth century who championed this principle called themselves the “Young Grammarians” (Junggrammatiker, the Neogrammarians). In their own slogan, “sound laws know no exceptions” — and Verner’s discovery became the first, and still the most famous, demonstration that this was not just a hope, but an actually testable method. What this school’s work did to Grimm’s legacy is the subject of the next article in this feature.
Summary
| Point | Content |
|---|---|
| The exception | Father’s consonant (voiced) conflicts with Grimm’s Law’s prediction (voiceless) |
| Discovered by | Karl Verner (Denmark, 1875) |
| The clue | Accent-position information preserved in Sanskrit, tracing back to Proto-Indo-European |
| The resolution | A voiceless fricative turns voiced when the accent does not fall on the immediately preceding syllable (Verner’s Law) |
| Significance | The first working example of “every sound law has a findable condition behind it” |
In a nutshell — the “exception” in Grimm’s Law wasn’t proof the law was wrong; it was simply a deeper rule nobody had found yet. And what found it wasn’t the consonant itself, but an accent mark preserved, of all places, in distant Sanskrit.