So, following on from last Friday’s general introduction, let us start at the beginning.
This is the “family tree” of Indo-European languages. It is slightly simplistic, as it does not take account of languages which have been heavily influenced by other languages (not least English!)
This means that over 400 languages, including all national languages in Europe bar Finnish, Estonian and Hungarian, are derived from a single tongue spoken around 5000 years ago, probably in or near modern Ukraine, which we now call “Proto-Indo-European” (PIE). Half the world’s population speak a daughter language PIE natively. PIE then broke up over the centuries into different dialects as tribes moved geographically and language changed (for a range of reasons from basic language change to coming across new things to describe and, of course, coming into contact with other languages).
So a good start is to have some idea what PIE was like.
Clearly, we do not know precisely what PIE sounded like.
However, we can, through reconstruction, work out that it had a lot of various sounds similar to those typically represented by modern English <h> and <l>. Most of these have been lost, but we can tell they existed from the way words developed subsequently.
We can reliably guess more about consonants than vowels, although we do know the most commonly occurring vowels were /e/ and /o/. Consonants were distinguished not just by “voiced” (e.g. /b/) and “voiceless” (e.g. /p/), but also “aspirated” (as Classical Latin <ph>). There would also have been considerably more of these (i.e. individual consonant sounds) than in most modern languages.
Most noteworthy of all, perhaps, is the clear indication that PIE relied on pitch rather than stress; and that this was applied at the start of words (perhaps with the exception of words with prefixes, which were exempt). This would have given it a markedly more different sound from any Western European language now.
Proto-Indo-European speakers had not, of course, developed the technology of writing. Written forms of the language are, therefore, the reconstructions of academic linguists.
Most of our vocabulary originates from PIE (though in fact this figure is lower for Germanic languages such as English than it is for Romance languages derived from Latin).
- 1 hoi-no-; 2 dwo-; 3 trei; 4 kwetwor-; 5 penkwe; 6 sweks; 7 septm; 8 oktou; 9 newn.; 10 dekm.
Note also k’m.tóm ‘a large number, a hundred’
PIE did have nouns, verbs and adjectives (this is not the case for all languages worldwide). However, other classes were less clear – what are now prepositions in most daughter languages were often postpositions or simply affixes, for example.
Nouns in PIE had eight, perhaps nine, cases – marked by endings to distinguish whether they were being used as subject, direct object, indirect object, possessor, recipient and so on. They had three numbers (singular, dual, plural) and three genders (masculine, feminine, neuter), and fell into a number of classifications. Some were further grouped – those ending -r, for example, often marked family relationship (and generally still do).
Verbs were marked, either by changes to the root vowel or by an ending (or both), primarily for aspect (rather than tense, as such) – whether something is relevant to the present or not. There were also complex moods – essentially marking whether something was certain, optional, counter-factual, and so on. Verbs could also be marked directly for mediopassive – the passive (effectively switching the subject and object around) or reflexive (making the subject also the object). They came in four classes – marked by the stem vowels (i.e. those generally appearing before the ending) /a/, /e/, /i/ or none – and were themselves classified by aspect (as being stative, reflecting a state; imperfective, reflecting something ongoing; or perfective, reflecting something complete – thus, where in English it is correct to say both ‘I boil the water’ and ‘The water boils’, PIE would not have allowed the same form for both).
Common (thematic) verb endings (1st, 2nd and 3rd person):
- Singular: -oh,-esi, -eti
- Plural: -omos, -ete, -onti
Dual also existed, but is not relevant to modern Western European languages.
All adjectives agreed with nouns; it is unclear how much distinction there was between adjectives and adverbs.
Pronouns were markedly different from how we currently understand them. For example, there were first and second person pronouns (‘I’, ‘you’, ‘we’) but not third person (no ‘he’, ‘she’, ‘it’, ‘they’).
Key personal pronouns (in nominative/accusative) were:
- singular h,eg’oH/h,me’, tuH/twe’; plural wei/nsme’, yuH/usme’
Word order was generally SOV, although the range of cases (and other marker particles) would have allowed significant variation for emphasis and there was a shift in some dialects late on to SVO. The key negative particle was ne.
Clearly, it is hard to assess the character of a language spoken thousands of years ago.
We do not know exactly what its own origins were, and whether they were shared with any other language tree (this is keenly debated by linguists, but seems unlikely to me).
We know something about the culture. We can tell from the language that society was clearly patriarchal, for example. Much of this too, however, remains keenly debated.
Let us move forward then to the earliest “Romance” and “Germanic” languages.