Grimm's law and verner's law

2.3. Grimm’s law
Careful investigation of Grimm’s Law revealed some inconsistencies, which were generally explained as exceptions to the rule. In some cases it is voiced stops rather than voiceless fricatives that correspond in Germanic to IE voiceless stop. For example,

Latin	Greek	Sanskrit	Gothic	Old English
Pa ter	pater	pi tar	fadar	fæ der
	[t ]		[θ] ?? [d]

T he Danish scholar Karl Verner was the first to explain them as the result of further development of Germanic languages. According to Verner, all the early PG voiceless fricatives [f, θ, h] which arose under Grimm’s Law, became voiced between vowels if the preceding vowel was unstressed; otherwise they remained voiceless. The voicing of fricatives occurred in early PG at the time when the stress was not yet fixed on the root-morpheme.
[f – v- b] seofon
[θ – ð – d] O Icel. hundrað – hundert
[h – g] Goth. swaihro –OE sweger
[s – z – r] Lat. auris – Goth. auso – Icel. eyra (ear)
The change of [z] into [r] is called rhotacism.
As a result of voicing, there arose an interchange of consonants in the grammatical forms of the word, termed grammatical interchange. Part of the forms retained a voiceless fricative, while other forms acquired a voiced fricative. For example, heffen (Inf.) - huob Past sg.) heave; ceosan (choose) curon (Past pl.). Some modern English words retained traces of Verner’s Law: death – dead; was- were, raise – rear.
Throughout history, PG vowels displayed a strong tendency to change. The changes were of the following kinds: qualitative and quantitative, dependent and independent. Qualitative changes affect the quality of the sound, for example [o - a] or [p – f]; quantitative changes are those which make long sounds short or short sounds long. For example,[ i – i:]; dependent changes are restricted to certain positions when a sound may change under the influence of the neighbouring sounds or in a certain type of a syllable; independent changes or regular (spontaneous) take place irrespective of phonetic conditions, that is they may affect a certain sound in all positions. In accented syllables the oppositions between vowels were carefully maintained and the number of stressed vowels grew. In unaccented positions the original contrasts between vowels were weakened or lost; the distinction of short and long vowels in unstressed syllables had been shortened. As for originally short vowels, they tended to be reduced to a neutral sound, losing their qualitative distinctions and were often dropped in unstressed final syllables (fiskaz).⁹
Strict differentiation of long and short vowels is regarded as an important characteristic of the Germanic group. Long vowels tended to become closer and to diphthongize, short vowels often changed into more open vowels. IE short [o] changed in Germanic into more open vowel [a] and thus ceased to be distinguished from the original IE [a]; in other words in PG they merged into [o]. IE long [a:] was narrowed to [o:] and merged with [o:]. For example, Lat. nox Goth. nahts; Lat. mater OE modor; Sans. bhra:ta OE bro:ðor .
Problem
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.¹⁰
One classic example of Proto-Indo-European *t → Proto-Germanic *ð 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 *ð 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 *wurðun and past participle *wurðanaz.