 From two million years before Christ onwards, man gradually became ready for the use of language. However, with no evidence available, the structure of the first language will perhaps never be known. Yet, by a method of language comparison, former systems of human communication, so-called proto-languages, can be reconstructed. The underlying principles of language reconstruction constitute the focus of this e-lecture. In particular, we will discuss the following issues. What exactly is a proto-language? And how can we compare languages? And last but not least, what sort of linguistic data can be used for the comparative method? Let's start with proto-languages. A proto-language is a hypothetical or reconstructed, typically extinct language. It is not known directly, rather it is a linguistic reconstruction that is formulated by applying the comparative method. A well-known example is Indo-European, or to be precise, proto-Indo-European. This is the reconstructed common ancestor of the Indo-European languages. It comprises most European and about one third of all languages and is assumed to have been spoken in Central Europe and in South Asia between 6000 and 4000 years before Christ. The Indo-Europeans probably lived in Central Europe before they began to travel east and west between 3500 and 2500 before Christ. It was reconstructed by August Schleicher, who you can see over here, who published his results in his monumental work, a compendium of the comparative grammar of the Indo-European languages. Other reconstructed languages are, for example, proto-Afroasiatic, proto-Dravidian, proto-Altaic, and proto-Eskimo-Aliot. Some linguists, among them, for example Joseph Greenberg and a number of Russian linguists, discerned similarities among these proto-languages, such as Indo-European, Afroasiatic and the others, and postulated a common ancestor to all these languages, which he called Nostratic. The Nostratic hunter-gatherers would have occupied all of Europe, Northern Africa and large parts of Asia from an origin in the Middle East, perhaps around 15000 years before Christ. Yet, all this is still very speculative. Well, and the story can even be extended. Presupposing a monogenetic view of language origin, that is, a hypothesis that all languages have a common origin, the earliest proto-languages could have a common ancestor. Well, according to married Rulan, a well-known typologist, it could be called proto-Sapiens. You might also want to use the term proto-human, proto-world, proto-terrestric, to name a few, but this is all very, very speculative. So, let us now look at the principles of reconstructing such proto-languages in more detail. In the 18th century, scholars began to compare groups of languages in a systematic way to see whether there were any correspondences between them. Evidence of a common origin for groups of languages was readily available in Europe, and by the beginning of the 19th century the hypothesis emerged that there was once a language from which the many languages of Eurasia derived. As already mentioned, this proto-language came to be called Indo-European, or to be absolutely precise, proto-Indo-European. The standard way of reconstructing such ancestor languages is referred to as the comparative method. Now, the comparative method is a way of systematically comparing a series of languages in order to show a historical relationship between them. Scholars begin by identifying a set of formal similarities and differences between the languages, thus trying to reconstruct an earlier stage from which all forms could have derived. Let us look at the word or the phrase, the book, in the romance languages in order to illustrate the comparative method. Now, you all know that in Spanish we would say el libro. In French it is le livre. In Italian we have il libro, and in Portuguese, o livre is the equivalence of the book. In looking at the word book, we can find amazing similarities. All these forms start with a lateral consonant, le. Then we have an unrounded high front vowel, e, which in some languages may be lengthened in others it is short. The next segment is either a voiced alveolar plosive or some sort of labiodental fricative, so we have this sort of variation, and then we have some sort of r, which can either be alveolar or even uvula or vela. Well, and then in some of these languages we have a vowel, so let us represent it like this. If Latin, the common ancestor language, to these romance languages no longer existed, it would thus be possible to reconstruct a proto-word such as libre or livre. Perhaps with a schwa between lib and r, in order to ease the articulation of the word. And that sort of proto-word would be very close to the real word in Latin, which is of course libre. Exactly this reasoning is used for cases where the parent language no longer exists. For example, when the forms in Greek and Sanskrit form a classical languages are compared in order to reconstruct the Indo-European proto-word, such words by the way are often marked with an initial asterisk. This principle, which we have just applied, is referred to as cognate comparison. This method was introduced in the 19th century by the Neogromarians, a group of German philologists at the University of Leipzig. So let us look at cognate comparison in more detail. Central to this method is the comparison of cognates, that is words from different languages with a similar orthographical and phonological shape denoting the same thing. Numbers are very good in this respect. So let us once more illustrate this principle, this time supported by audio evidence from the language index on the virtual linguistics campus and I will now load the cognate comparison component of the language index. Here we are. So let us simply first of all listen to the number 3 in the following languages. Let us start with Spanish. 3. The next one is English. 3. Let us move on to German. 3. Now here is Polish. 3. Check. 3. And finally let us take Ukrainian. 3. Now what have we found? In Spanish we have 3. In English 3. The German word for 3 was 3. In Czech we have, well somewhere here, somewhere here. In Polish, 3. And finally we heard Ukrainian 3. Now in looking at these words we see amazing parallels. Some languages have an alveolar plosive in the initial position. Others have an african like Czech or Polish. Or a fricative as in English. The next segment is in many languages some sort of r again, which can be an approximate, a trill or a fricative. And then all languages have one vowel which is some sort of high front vowel. So the similarities are amazing. So in such a list of cognates systematic phonological correspondences emerge. They imply that these languages are genetically related. The evaluation of such correspondences allows the reconstruction of earlier phonological stages and eventually the sound system of a now extinct ancestor language, in this case Indo-European. The proto-word for 3 by the way has been reconstructed as something like trace. And in order to signal that this is a proto-word, let's put an asterisk in front of it. Trace. Using this method of cognate comparison, the main philologies of the 19th century, in particular the neo-grammarians established principled relationships between the languages spoken in Europe. And then there was someone called Jacob Grimm, generally known as the author of fairy tales, but of course he's also known as a linguist. And he discovered some principled relationships between the classical languages, Sanskrit, Greek and Latin on the one hand and the Germanic languages on the other. Let's look at his discovery in more detail. In taking the word for father, we will find the following words. In Sanskrit you would say peter, in classical Greek you would say pater, in Latin it would be pater and in English of course father. This of course is British English. Now if we look at the initial elements, we have in the classical languages a bilabial plosive and in English a labial dental fricative. So this is what we have. And if we inspect further data, we can generalize this sort of relationship to the following. If we have an initial voiceless plosive in one of the classical languages, we will get a fricative in with the same place of articulation or a similar place of articulation in the Germanic languages. This is just one observation. Here is another one. Now let's look at the word for brother in these languages. In Sanskrit it would be something like frater, in classical Greek frater, frater, sorry. In Latin it would be frater and in English we get of course brother. And again if we look just at the initial segments you will find that we have some sort of fricative in the classical languages and a bilabial plosive in English and also in the other Germanic languages. So we could generalize this as a sort of rule, some sort of africated bilabial element, plosive or fricative comes out as a bilabial plosive in the Germanic languages and here is the generalization. So this observation with further generalizations has become known as Grimm's law or the first consonantal shift and as I said earlier on it goes back to Jacob Grimm. Jacob Grimm is of course this one here on our picture together with his brother and this first consonantal shift that distinguishes Germanic on the one hand from the classical languages, Indo-European languages on the other has been known as Grimm's law or as I said the first consonantal shift. Now this sort of cognate comparison can even be more complex. Now if we compare some languages like for example from top to bottom English, Dutch, Danish, Norwegian, Icelandic and German and we look at the cognates in it. For example we could look at the representation of the verb give which is marked in red here or we could look at the cognate bread, brood, brood, brood, brausbrot in these languages. Well then we see amazing similarities and in fact all these languages are Germanic languages. However we could even say more. We could now say something about word order principles. Here we have imperative sentences and obviously all these languages have the verb in the initial position. In German the word order is slightly deviant but it is also possible but normally the verb is also in the initial position. Or can we make statements about morphological complexity? Well if we look at these languages and look at the word for daily for example we have an adverb here and you see the different representation of adverbs plus some morphological element to be added illustrate the degrees of morphological complexity of these languages. So in comparing existing with extinct languages or comparing them with other existing languages with each other mind boggling parallels emerge. They can best be explained by assuming principle relationships between them. So that humans could speak a long time ago is undisputed yet the exact structure of their early communication systems will probably never be revealed. However and I hope to have shown this in this e-lecture the principles of language comparison allow us to construct to reconstruct formal languages without having written let alone spoken evidence. Using this comparative method we can reconstruct proto languages such as for example Indo-European or proto-altaic. According to some linguists we can even go further back and postulate a common ancestor language to all of these. We found out that Nostratic could be a suitable name and even further back Nostratic could be grouped with other reconstructed proto languages. Eventually then we could find a first language spoken on our planet well proto-world, proto sapiens, proto-human or whatever name we would like to find for it. However I hope to have shown this that this involves a high degree of speculation.