LING150/University of Oregon

Go to Unit 4 Study Plan

Go to Home Page

4.1.1. The discovery of Indo-European

4.1.2. The Comparative Method

4.1.3. The Indo-European family tree

At one time, scholars believed that there were no consistent similarities between languages. All languages were assumed to be completely unrelated to each other; their differences were thought to be due to

• accident;
• different cultural backgrounds; or,
• divine proclamation, such as the "Tower of Babel" story from the Judeo-Christian tradition.

Then, in the late 18th century, some scholars began noticing that if one listed words from one ancient language side-by-side with the equivalent words in another, there were literally hundreds of systematic similarities. The chart below shows the words for 'foot', 'tooth', and 'heart' in Latin, Greek, and Sanskrit (the ancient literary language of India). Compare these words with each other and with the equivalent words in Basque, a language spoken in Europe that is truly unrelated to any other known European language.

	'foot'	'tooth'	'heart'
Latin	ped	dent	cord (c = [k])
Greek	pod	odont	kard
Sanskrit	pad	dant	syertse
Basque	oin	hagin	bihotz

Notice that the Latin, Greek, and Sanskrit words bear an exceedingly strong resemblance to one another that is particularly striking when we observe the complete lack of resemblance to Basque words. The similarities are especially close for the consonants. Here we see a similar number of consonants and similar consonants in the same position. The vowels are less consistent, so we will be mostly concerned with consonant similarities in our discussion of pre-English history.

The number of similarities between Latin, Greek and Sanskrit seemed so great that it was unlikely to be accidental. In a famous address to the Royal Asiatic Society of Calcutta in 1786, Sir William Jones claimed that the commonalities were so numerous that

"...no philologer could examine all three, without believing them 
to have sprung from some common source, which, perhaps, no 
longer exists." 

Jones' startling suggestion was that the similarities could be best accounted for by assuming that all the languages had descend from a common mother tongue, which was no longer spoken. Later scholars determined that modern European languages such as French, Spanish, German, and English also bore a strong similarity to the ancient tongues and thus were also "descendants" of the same "parent" language.

Darwin's theory of evolution was the "hot theory" of the 19th century and it strongly influenced the way linguists viewed languages. They regarded languages metaphorically as biological organisms which had family trees and ancestors. They postulated that if two or more languages had many similar words with similar meanings, then they had evolved from same parent language. This hypothesis was based on the assumption that languages change in regular, systematic ways. Differences between related languages were due to regular sound changes which could be recognized through careful study.

4.1.2. The Comparative Method

If sound changes are regular and systematic, then one could compare words from modern languages which had similar forms and similar meanings and undo the changes that had occurred over time. By reversing the historical changes, one could determine the original word in the mother language!

Thus, it would be possible to look at modern languages and reconstruct the parent language, which was now extinct. In terms of the "genetic metaphor," this would be something like rebuilding a extinct species by looking at the genetic structure of modern species!

Reconstructing an extinct language depends critically on the assumption that sound change is regular and systematic. That means that, over time, sounds don't move around much the phonetic chart. If they move at all, it will be relatively little and in ways that can be well-understood. For example, it would be reasonable for a [p] to become [b] or [f], since all these sounds involve the lips, but it's unlikely that a [p] would change to an [n] even over thousands of years. There just isn't much articulatory similarity between [p] and [n].

When linguistics attempt to reconstruct an extinct language, they compare cognates. In the example above, the Latin, Greek and Sanskrit words for 'foot'are cognates of each other. The word cognate literally means 'originating together' (co / gn / ate).

It's very important to distinguish cognates and meanings. For example, consider the morpheme men 'think, mind'. The words between the single quotes stand for the meaning component of the morpheme; that is, they represent some abstract, mental concept. Unfortunately, we're not telepathic, so we have to use word forms to convey this abstract concept. Now consider the English morpheme mind 'mind'. ( This may seem redundant, but it emphasizes the distinction between forms and meanings.) The Latin morpheme men 'think, mind' and the English morpheme mind 'mind' are cognates. They have similar forms ( men and mind) and similar meanings ('think, mind' and 'mind'). The English morpheme think 'think' is NOT a cognate of the Latin morpheme men 'think, mind'. Although these two morphemes have similar meanings, they do not have similar forms. The word 'think' is used to help convey the abstract mental concept linked to the form men, but think and men originated from completely different sources.

Linguists put together sets of cognates from as many modern languages as possible. Then they use them to reconstruct the parent language. This techniques is called the Comparative Method.

The parent language that is reconstructed in this way is called a Proto-Language.

The reconstructed proto-language from which most modern European languages descended is called Proto-Indo-European or P.I.E. It is important to remember that P.I.E. is a hypothetical language. Linguists are not suggesting that there was once a language identical to P.I.E. They do believe that at one time a group of people living in a relatively small area, probably between eastern Europe and Northern India, spoke a more or less unified language that is similar in most important respects to P.I.E. They also believe that over the course of several thousand years, most of the modern languages of Europe developed out of this core language.

Here are some reconstructed P.I.E. words. They should look very familiar.

4.1.3. The Indo-European family of languages

P.I.E. is thought to have existed about 6000 years ago. Over time, P.I.E. speakers spread across Europe. As they lost contact with each other, the original language evolved into several distinct branches spoken in different geographical areas. The first division separated the Eastern European languages from the Western ones. The Western branch later evolved into a number of distinct language groups. For our purposes, the most important branches are the Germanic, the Hellenic, the Italic and the Celtic.

The chart below shows some the main features of the Indo-European family. Take a look at it and determine which branches the major languages we've been discussing in this course belong to.

Copyright 1998 by the Department of Linguistics, University of Oregon.