The Germanic tree

Most specialists place the Germanic homeland in southern Scandinavia and the north German coast, between the Elbe and the Oder, in the first millennium BCE. The main argument is geographical economy: the earliest runic inscriptions are found here, and it is the one area from which the attested distribution of the branches can be derived without improbable detours.

Several kinds of evidence are brought to bear. Early loanwords point to neighbours: Finnic languages to the north-east borrowed Germanic words at an early stage, and Germanic itself took words from Celtic to the south, such as *rīks 'ruler' and probably *īsarną 'iron'. Archaeologists have proposed continuity from the Nordic Bronze Age (c. 1700–500 BCE) or a link with the Iron Age Jastorf culture of northern Germany, but neither pottery nor burial customs can show which language people spoke. Ancient-DNA studies show that steppe-related ancestry reached Scandinavia with the Corded Ware and Battle Axe cultures around 2800–2500 BCE. Many scholars connect this with the arrival of Indo-European speech in the region, but that is some two thousand years before Proto-Germanic proper.

The spread is better documented. By Caesar's time Germanic-speaking groups had reached the Rhine. In the Migration Period, from the 4th to the 6th centuries CE, East Germanic peoples crossed the Roman Empire and Angles, Saxons and Jutes settled Britain. Viking Age expansion took Norse to the North Atlantic islands, and from the 17th century colonial settlement carried English, Dutch and German overseas. Hans Kuhn's proposed 'Northwest Block', a region in the Low Countries that he argued was neither Germanic nor Celtic before the Iron Age, remains a minority view but is a reminder that the early map is uncertain.

Beyond the consonant shift and the stress on the first syllable, several features give Germanic languages a recognisable profile.

The verb system is the clearest. Germanic has only two inflected tenses, past and non-past. Future and perfect are built with auxiliaries (I will go, ich habe gesehen). The past is formed in two ways. Strong verbs change their root vowel, inherited from Indo-European ablaut: sing, sang, sung, German singen, sang, gesungen, Swedish sjunga, sjöng, sjungit. Weak verbs, the vast majority and the only productive type, add a suffix containing a dental consonant: English -ed, German -te, Dutch -de/-te, Gothic -da. This dental past has no exact counterpart elsewhere in Indo-European and is one of the family's defining innovations. A small class of 'preterite-present' verbs, including can, may and shall, supplied most of the modal verbs.

Adjectives had two sets of endings, 'strong' and 'weak', chosen by whether a determiner precedes. German still shows it: guter Wein 'good wine' against der gute Wein 'the good wine'. The Scandinavian languages keep a version too (Swedish en stor bil 'a big car', den stora bilen 'the big car').

In word order, every modern Germanic language except English is verb-second in main clauses. The finite verb comes second whatever comes first, so German says Heute gehe ich and Swedish I dag går jag ('today go I'). English has only relics, as in Never have I seen it. German, Dutch and Frisian also put the verb at the end of subordinate clauses.

The earliest writing was runic: the Elder Futhark, adapted from a Mediterranean alphabet. Gothic got its own alphabet, and the Latin script took over elsewhere during Christianisation. Icelandic still writes þ, a rune taken into the Latin alphabet, and ð, a letter first used by Old English scribes.

The broad outline of Germanic is settled; its internal shape and its timing are not.

The first question is how the three branches relate. The majority view is a Northwest Germanic stage shared by the ancestors of North and West Germanic after the East Germanic groups had left. It rests on shared changes such as the lowering of Proto-Germanic *ē to ā (Gothic mēna against Old Norse máni and Old English mōna 'moon'), and on the language of the early runic inscriptions. A minority, following Ernst Schwarz in the 1950s, groups Gothic with Norse instead, citing features such as the Gothic and Norse treatment of *-jj- and *-ww- ('Holtzmann's law'). Others treat the early Germanic area as a dialect continuum that no tree captures well.

West Germanic is a second problem. Don Ringe and Ann Taylor argue that it was a genuine protolanguage with its own shared innovations, such as consonant gemination before *j. Others see a set of neighbouring dialects that converged. The chapters on West Germanic set out both sides.

Dating is a third. Grimm's law is usually placed in the second half of the first millennium BCE, partly on the basis of early Celtic loans. The Celtic source of *rīks already had ī from older ē, yet the Germanic word shows the g → k shift, suggesting Grimm's law was still active when it was borrowed. The argument depends on dates for Celtic that are themselves uncertain.

Finally, Germanic's place within Indo-European is debated. It shares vocabulary with Italic and Celtic, and case endings in -m- with Balto-Slavic (Gothic dagam 'to the days', Lithuanian -mus), where other branches have -bh-. Whether these reflect early subgrouping or later contact is unresolved. The claim that much of Germanic vocabulary comes from a non-Indo-European substrate is discussed under Contested Classifications.