Indo-European

Back to the Syllabus

We have seen how English is part of a family that includes Dutch, Italian, and Portuguese but excludes Finnish. The first step in accounting for the history of the English language is to look at the family that English belongs to, the "Indo-European" family of languages.

No records of a language corresponding to primitive Indo-European exist. What we know of it is reconstructed, through a process of inference, from records of the earliest versions of many languages across Europe and Western Asia.

Comparisons of basic vocabulary in Latin, Greek, old Slavic languages, Old English and other early Germanic languages, old Persian languages and old Indian languages like Sanskrit suggest that all of these languages are ultimately dialects of a long-lost prehistoric language we call Proto-Indo-European (for want of any neater way to name it). Languages from Ireland to India are descendants of Indo-European, with notable exceptions like Basque, Magyar, Estonian, Finnish, Turkish, Arabic, Hebrew, Tamil, and others.

Note one very important principle at work here: the fact that people speak languages from the same family does not mean that their biological families are closely related. They could be. All humans are ultimately related, just 120,000 years back or so. But people from widely divergent ethnic backgrounds frequently end up speaking the same language--a look at any modern American community shows that. Work by the population geneticist Luigi Luca Cavalli-Sforza has demonstrated that in some cases, genetic patterns in populations parallel linguistic history. Just as certainly, in many cases language does not parallel genetics. In any case there is no evidence that any group of humans has a different linguistic biology than any other group.

I stress this point because it is easy to connect language to "race" in simplistic and even evil terms. Across the range of the Indo-European languages, non-Indo-European speakers are usually minorities in modern nation-states, often persecuted because their language seems to tag them "racially" as outsiders. In Nazi Germany, the state identified racial traits with the concept of "Aryan," which in turn was identified with original speakers of Indo-European languages. The basic idea was that Germans were racially "original" and part of the pure founding population of the Indo-European world. Many Nazis saw the Anglo-Saxon (hence "Germanic") English as racial cousins.

The Nazi idea is pure nonsense. In most of Europe the local populations speak their languages because of historical political accidents, not because of a racial connection between descent and language. In England, most people today speak English; a few centuries ago, they spoke languages from the very distant Celtic wing of the Indo-European family; a few thousand years ago, they spoke pre-Celtic, non-IE languages. Yet many English people can trace some of their genetic heritage back even further, to early human settlement of Western Europe tens of thousands of years ago. Their languages have shifted as they have been colonized by conquering populations many times during that span. The influx of new people has mixed the gene pool greatly, and the cultural speed with which languages spread and change has mixed the cultural heritage even more greatly.


At some point, several thousand years ago, Indo-European was a single dialect with a fairly limited range, perhaps in the steppes of the Ukraine and what is now Russia north of the Caucasus (that's the best guess). At some unknown point, Indo-Europeans spread their language across a wider area; they, or people who had learned their language, spread westward into Europe, eastward into Asia. This migration provided one basic context for language change--as descendant groups of Indo-European speakers, and colonized people who learned Indo-European lost contact with one another, their languages drifted into the main subfamilies that we observe today.

The history of the drift is mostly lost to us; we see just the results. Some large features of the drift are observable in the features of the descendant languages.

The Satem-Centum distinction

Many words in western Indo-European languages have /k/ sounds (or their derivatives) where Eastern languages have /s/ (or its derivatives).

Avestan satem, for instance, corresponds to Latin centum (both mean "hundred").

Satem languages include Indian languages, Armenian, Slavic languages, and Lithuanian. Centum languages include Greek, Latin, Celtic languages, and Germanic languages.

This was once seen as the great original dialect split between two major IE groups. It isn't anymore (see Lehmann, Historical Linguistics, Routledge 1992: 73), but it suggests that at some time early on, the eastern groups were in close enough contact for a change from an original /k/ sound to /s/ could spread across a wide area, but not to the west, where the sound developed independently.

Grimm's Law

But the English word "hundred" has neither a /k/ nor an /s/. Since English is a Western language, we'd expect /k/. Why isn't it there?

In 1822, Jakob Grimm (more famous as a collector of folktales) noticed that the contrast between Latin centum and English hundred has many corresponding examples. Latin cannabis is English hemp; Latin caput is English head. In fact, English and other "Germanic" languages tend to have an /h/ or /x/ sound where Latin and Greek have /k/.

Grimm noted two other main contrasts--between Latin /p/ and Germanic /f/ (pisces / fish; pes / foot) and between Latin /t/ and Germanic /þ/ (tres / three; tonare / thunder).

What does this mean? That the Germanic languages, in separating from other Western IE languages, underwent a systematic change in consonants that marks them off from all other Western branches (because Latin, Greek, and Celtic languages all contrast to Germanic in this way). And very importantly, it means that Germanic split off from Latin and Greek a very long time ago--several thousand years, at least. Remember that consonants change slowly. For a consonant change to become so widespread, systematic and marked as a dialectal difference, it must go very deep and be very old.

One point I'll keep reinforcing, that's shown here, is that English does not come from Latin or Greek. Some important vocabulary in modern English is influenced by Latin and Greek, but those languages are not the parents of our own; they are very distant cousins.

Top