A seminal insight into how the Germanic languages diverged from their Indo-European ancestor had been established in the early nineteenth century, and had been formulated as Grimm's law. Amongst other things, Grimm's law described how the Proto-Indo-European voiceless stops *p, *t, *k, and *kʷ regularly changed into Proto-Germanic *f (bilabial fricative [ɸ]), *þ (dental fricative [θ]), *h (velar fricative [x]), and *hʷ (velar fricative [xw]).
However, there appeared to be a large set of words in which the agreement of Latin, Greek, Sanskrit, Baltic, Slavic etc. guaranteed Proto-Indo-European *p, *t or *k, and yet the Germanic reflex was not the expected, unvoiced fricatives *f, *þ, *h, *hʷ but rather their voiced counterparts *β, *ð, *ɣ, *ɣʷ. A similar problem obtained with Proto-Indo-European *s, which sometimes appeared as Proto-Germanic *z.
At first, irregularities did not cause concern for scholars since there were many examples of the regular outcome. Increasingly, however, it became the ambition of linguists like the Neogrammarians to formulate general and exceptionless rules of sound change that would account for all the data (or as close to all the data as possible), not merely for a well-behaved subset of it.
One classic example of Proto-Indo-European *t → Proto-Germanic *d is the word for 'father'. Proto-Indo-European *ph₂tḗr (here, the macron marks vowel length) → Proto-Germanic *faðēr (instead of expected *faþēr). In the structurally similar family term bʰréh₂tēr 'brother', Proto-Indo-European *t did indeed develop as predicted by Grimm's Law (Germanic *brōþēr). Even more curiously, scholars often found both *þ and *d as reflexes of Proto-Indo-European *t in different forms of one and the same root, e.g. *werþaną 'to turn', preterite third-person singular *warþ 'he turned', but preterite third-person plural *wurdun and past participle *wurdanaz.