📚 Grimm's Law: why Latin pater is English father

Latin pater, trēs, centum against English father, three, hundred: the Germanic consonant shift worked out by Rask (1818) and Grimm (1822), and the accent rule Karl Verner found in 1875 to explain its exceptions.

Tags: sound-change, germanic, indo-european, historical-linguistics

Latin pater, trēs and centum line up with English father, three and hundred one consonant at a time: p against f, t against th, and c (pronounced [k]) against h. Latin is not the ancestor of English. Both descend from Proto-Indo-European, and in these consonants Latin, Greek and Sanskrit kept the inherited stops, while the ancestor of the Germanic languages changed them. That change is called Grimm's Law, or the First Germanic Sound Shift.

⚖️ Compare "father" in Latin, Greek, Vedic, Old English and English
LanguageFamilySpeakersFatherWaterOne
LatinIndo-European (Italic)Extinct (liturgical)pater/ˈpater/aqua/ˈakʷa/ūnus/ˈuːnus/
Ancient GreekIndo-European (Hellenic)Extinctπατήρ/patɛ̌ːr/ὕδωρ/hýdɔːr/εἷς/hêːs/
Vedic SanskritIndo-European (Indo-Iranian, Indo-Aryan; Old Indo-Aryan)Extinct (liturgical use, no native speakers)पिता/pitáː/आपः/áːpɐɦ/एकः/éːkɐɦ/
Old EnglishIndo-EuropeanHistorical (extinct as L1 c. 1100; studied liturgically/academically)fæder/ˈfæder/wæter/ˈwæter/ān/ɑːn/
EnglishGermanic~380M L1 + ~1.5B total (with L2)father/ˈfɑːðə/water/ˈwɔːtə/one/wʌn/

Rask in 1818, Grimm in 1822

Friedrich von Schlegel pointed out the match between Latin p and Germanic f in 1806. The Danish linguist Rasmus Rask worked out the full set of correspondences in an essay he wrote in Iceland in 1814 for a prize offered by the Danish Academy of Sciences. The essay won and was printed in Copenhagen in 1818 as Undersøgelse om det gamle Nordiske eller Islandske Sprogs Oprindelse (Investigation of the Origin of the Old Norse or Icelandic Language). Jacob Grimm, who also collected fairy tales with his brother Wilhelm, set them out as a system in the 1822 second edition of the first volume of his Deutsche Grammatik, and added the later shift that sets High German apart from the other Germanic languages. Rask published first, but the name that stuck is Grimm's.

Three steps in a chain

The first two steps cannot have happened the other way round. If *d had become t before the old *t moved on to þ, the two sounds would have merged, and they did not. Either the first step came first, or the two moved together as one chain. The shift also skipped a stop that came right after s. Gothic fisks "fish" keeps the k of the cluster sk, the same k as Latin piscis, even though the first p of the word became f. No inscription dates the shift. It is placed in the first millennium BCE, before the earliest Germanic records.

⚖️ Compare "three" in Latin, Greek, Gothic, Old English and English
LanguageFamilySpeakersThreeWaterOne
LatinIndo-European (Italic)Extinct (liturgical)trēs/treːs/aqua/ˈakʷa/ūnus/ˈuːnus/
Ancient GreekIndo-European (Hellenic)Extinctτρεῖς/trêːs/ὕδωρ/hýdɔːr/εἷς/hêːs/
GothicIndo-European (Germanic, East)Extinct𐌸𐍂𐌴𐌹𐍃/θriːs/𐍅𐌰𐍄𐍉/ˈwatoː/𐌰𐌹𐌽𐍃/ɛːns/
Old EnglishIndo-EuropeanHistorical (extinct as L1 c. 1100; studied liturgically/academically)þrī/θriː/wæter/ˈwæter/ān/ɑːn/
EnglishGermanic~380M L1 + ~1.5B total (with L2)three/θɹiː/water/ˈwɔːtə/one/wʌn/
⚖️ Compare "fish": p becomes f, but sk stays
LanguageFamilySpeakersFishWaterOne
LatinIndo-European (Italic)Extinct (liturgical)piscis/ˈpiskis/aqua/ˈakʷa/ūnus/ˈuːnus/
GothicIndo-European (Germanic, East)Extinct𐍆𐌹𐍃𐌺𐍃/fisks/𐍅𐌰𐍄𐍉/ˈwatoː/𐌰𐌹𐌽𐍃/ɛːns/
Old NorseGermanicExtinctfiskr/fiskr/vatn/vɑtn/einn/ɛinː/
Old EnglishIndo-EuropeanHistorical (extinct as L1 c. 1100; studied liturgically/academically)fisċ/fiʃ/wæter/ˈwæter/ān/ɑːn/
EnglishGermanic~380M L1 + ~1.5B total (with L2)fish/fɪʃ/water/ˈwɔːtə/one/wʌn/

Verner in 1875: the exceptions

The law did not cover every word. If PIE *t became þ, Old English should say *fæþer, but it says fæder, while brōþor "brother" has the expected þ. The Danish linguist Karl Verner found the reason in 1875. His article "Eine Ausnahme der ersten Lautverschiebung" (An exception to the first sound shift) appeared in volume 23 of the Zeitschrift für vergleichende Sprachforschung, pages 97–130, a volume that references date anywhere from 1875 to 1877. The reason was the accent, which Vedic Sanskrit and Greek preserve. Sanskrit pitā́ "father" is stressed on the second syllable and bhrā́tā "brother" on the first. Inside a word, the fricatives f, þ, h and the inherited s became voiced unless the syllable just before them carried the PIE accent. So *ph₂tḗr gave Proto-Germanic *fadēr, while *bʰréh₂tēr gave *brōþēr. Mother followed father: Sanskrit mātā́, Old English mōdor.

The same rule explains the d in hundred. PIE *ḱm̥tóm (Sanskrit śatám) is accented on its last syllable. Its t became þ under Grimm's Law and was then voiced under Verner's, which gives Gothic and Old English hund. Germanic later moved the stress to the first syllable of every word, so the voiced and voiceless variants were left side by side inside verb paradigms. Old English has cēosan "choose" beside the past participle coren, and English has was beside were. In both pairs the voiced s turned into r.

⚖️ Compare "hundred" in Latin, Vedic, Gothic, Old English and English
LanguageFamilySpeakersHundredWaterOne
LatinIndo-European (Italic)Extinct (liturgical)centum/ˈkentum/aqua/ˈakʷa/ūnus/ˈuːnus/
Vedic SanskritIndo-European (Indo-Iranian, Indo-Aryan; Old Indo-Aryan)Extinct (liturgical use, no native speakers)शतम्/ɕɐtɐ́m/आपः/áːpɐɦ/एकः/éːkɐɦ/
GothicIndo-European (Germanic, East)Extinct𐌷𐌿𐌽𐌳/hund/𐍅𐌰𐍄𐍉/ˈwatoː/𐌰𐌹𐌽𐍃/ɛːns/
Old EnglishIndo-EuropeanHistorical (extinct as L1 c. 1100; studied liturgically/academically)hund/hund/wæter/ˈwæter/ān/ɑːn/
EnglishGermanic~380M L1 + ~1.5B total (with L2)hundred/ˈhʌndɹəd/water/ˈwɔːtə/one/wʌn/
⚖️ Compare "mother" from Vedic to English
LanguageFamilySpeakersMotherWaterOne
LatinIndo-European (Italic)Extinct (liturgical)māter/ˈmaːter/aqua/ˈakʷa/ūnus/ˈuːnus/
Vedic SanskritIndo-European (Indo-Iranian, Indo-Aryan; Old Indo-Aryan)Extinct (liturgical use, no native speakers)माता/maːtáː/आपः/áːpɐɦ/एकः/éːkɐɦ/
Old EnglishIndo-EuropeanHistorical (extinct as L1 c. 1100; studied liturgically/academically)mōdor/ˈmoːdor/wæter/ˈwæter/ān/ɑːn/
Middle EnglishIndo-European (Germanic, West)Extinctmoder/ˈmoːdər/water/ˈwatər/oon/oːn/
EnglishGermanic~380M L1 + ~1.5B total (with L2)mother/ˈmʌðə/water/ˈwɔːtə/one/wʌn/

What came later

The th in modern father and mother is not the old þ. Old and Middle English wrote fæder, fader, mōdor and moder, and the d before -er only became [ð] in the 15th and 16th centuries. Gothic is no help with "father" either. Its everyday word is atta, and that is the form the map shows. Fadar occurs only once in the surviving texts, in Galatians 4:6, where it translates the Aramaic abba. In the south, Old High German shifted some consonants a second time: zwei, zand and herza stand where English has two, tooth and heart. This is the second shift that Grimm described in 1822.

⚖️ Compare "tooth": the first and second shifts
LanguageFamilySpeakersToothWaterOne
LatinIndo-European (Italic)Extinct (liturgical)dēns/deːns/aqua/ˈakʷa/ūnus/ˈuːnus/
GothicIndo-European (Germanic, East)Extinct𐍄𐌿𐌽𐌸𐌿𐍃/ˈtunθus/𐍅𐌰𐍄𐍉/ˈwatoː/𐌰𐌹𐌽𐍃/ɛːns/
Old EnglishIndo-EuropeanHistorical (extinct as L1 c. 1100; studied liturgically/academically)tōþ/toːθ/wæter/ˈwæter/ān/ɑːn/
Old High GermanIndo-European (Germanic, West)Extinct (~750-1050 CE)zand/tsand/wazzar/ˈwassar/ein/ein/
GermanGermanic~130MZahn/tsaːn/Wasser/ˈvasɐ/eins/aɪ̯ns/

Sources

More on this

See also: Trivia · Word Map — 168 core words · Han Map — Han character readings

Open the interactive map: Word Map · Han Map · Name Map · Time Map · Word Order · Tree

Part of LangMap — a linguistic visualization project. This is a static, crawlable summary; the interactive maps offer pronunciation audio, filters, and a globe view.

Made by Heuron Inc. · GitHub · cho@heuron.com · LangMap home