 Languages can be classified in various ways, syntactically, on the basis of their word order and their relationships between heads and modifiers, phonologically on the basis of several segmental and suprasegmental parameters. The morphological classification, which constitutes the focus of this e-lecture, defines languages on the basis of the inflectional structure of their words. How this is done and what types of languages can be identified using this structural parameter will be discussed in the following. However, first we have to solve one essential question. Why is it inflection and not any other morphological process on the basis of which languages can be classified morphologically? Well, let's discuss this for a moment. The following criteria, which have primarily been based on present-day English, have been put forward to distinguish inflection from word formation, such as derivation. The stability of word class is one criterion. Inflectional processes, as you know from other e-lectures of this channel, do not change the word class of the stem, irrespective of the type of morphological operation involved, prefixing, suffixing, vowel change, and so on and so forth. A second criterion concerns the stability of stress pattern. Inflection, at least in present-day English, we know there are counter examples from other languages, does not involve a change of the stress pattern of the stem. The third criterion is referred to as semantic transparency. Inflection is semantically transparent. That is, the meaning that is created by a morphological operation is stable. And then we have the criterion of productivity. Now a morphological process is fully productive if it applies to all members of a specific class. Hence inflectional processes are highly productive. Word formation processes, by contrast, are by and large not. Let's look at an example. In present-day English, for example, all verbs can have the third person singular or can denote the third person singular. But not all verbs can be nominalized. NMLZ is the abbreviation. So verbs like reiner or seamer, etc., are impossible. And this sort of nominalization process is even different across languages. Thus, it is the last criterion which makes it almost impossible to compare languages on the basis of their word formation principles. All we could do is assess the word formation complexity of a language on the whole. But that's not normally used as a typological parameter. So let us turn our attention to the inflectional classification of the world's languages. Now the inflectional possibilities of most languages can be defined in terms of a continuum which is illustrated by this vertical line here. At one extreme of this continuum, we find the so-called analytic or isolating languages whose words have little or no internal structure. At the other extreme, we find the synthetic languages which do allow the analysis of their words into smaller parts or morphs. Let us look at these two extremes in more detail. Now analytic languages, which are also referred to as isolating, generally do not allow the segmentation of their words. That is, the words of an analytic language cannot or hardly be split into smaller units. Thus, a typical analytic language is a language where there is a one-to-one correspondence between words and morphemes. Still known examples of analytic languages are Vietnamese or Chinese. Well let us look at these two languages in more detail. Here is a sentence in Vietnamese. In fact, it's the sentence most of you know from the VLC language index, the man gives the woman the book. And here is the Vietnamese equivalent. Let's listen first. And as you can see quite clearly, each morph or word, well there is hardly any difference between them, corresponds to one translation or to one grammatical function. So clearly a one-to-one relationship between words and morphemes. Well and here is the same sentence in Chinese. Let's listen again. 男人给女人一本书. And again, we have syllables that stand for words that stand for morphemes. You might want to argue that 男人 and 女人 are compounds, okay. But we said earlier on word formation is not a criterion for the typological classification of the languages of the word. So clearly Vietnamese and Chinese are analytic languages. Let's now look at the synthetic type of languages. Synthetic languages by contrast, as I said earlier on, do allow a segmentation of the words into morphs. Again I would like to illustrate this with two examples. Here is an example from Turkish. Adam ların which means mens in the plural. And clearly this word can be split in to the following morphs. Adam which is the stem and means man, lar which stands for the plural and in which stands for the genitive. So it's genitive plural. And here is a sentence again. The man gives the womb of the book. And as you can see, we can subdivide some words into morphs. For example, Kaduna, where the A stands for the datif, Kitabu, where Kitab is the stem, and U is the accusative case marker, and Verior, which can be subdivided into Veri, the stem, and the present tense singular affix. So Turkish is clearly a synthetic language, and it can be placed down here on our continuum. Let's now look at another example, my own mother tongue German. The word mention can be subdivided into mensch plus its final suffix, normally phonologically represented just by an alveolar nasal consonant, menschen. Well, and here is the sentence, der Mann gibt der Frau das Buch. And as you can see, we can subdivide, well at least one word into two morphs, Gibbs, which where Gibbs is the stem, and the T stands for the third person singular present tense. The remaining words incorporate the various morphological and grammatical functions. But German clearly is synthetic too. But what is the difference between these two languages? Let us now look at the function of the morphs in both cases, German and Turkish, the affixes. In Turkish, we had Adam, Lauren, which can be subdivided into Adam, Lauren, in. And as you can see, we can clearly associate one particular function with the relevant morphs. In German, this is relatively difficult. Now here we have the final suffix, however this time the meaning is not so clear. The final suffix can denote the plural in the nominative case, plural genitive, plural dative, plural accusative, and even the genitive singular. So the difference between Turkish and German is that the bound morphs in Turkish are functionally unambiguous, whereas in German, the bound morphs exhibit several grammatical functions. Thus, depending on the function of the morphs, we can identify in a language, we can define two types of synthetic language. On the one hand, we have the so-called agglutinating synthetic languages, such as Turkish, where the morphs are by and large functionally unambiguous. In such an agglutinating language, some people call it agglutinative, a word may consist of more than one morph, but the boundaries between the morphs are always clear-cut. Furthermore, a given morph has mostly a reasonably invariant shape, so that its identification is in most cases straightforward. Another well-known example of an agglutinating language is Japanese. It's also highly synthetic, so let's place it over here. The agglutinating languages can be contrasted with the fusion of languages such as German. Here, the dependent morphs often have several grammatical functions associated with them. In fusion languages, some people refer to them as inflecting or inflective, there are no clear-cut boundaries between the morphs. Also, several grammatical functions, several grammatical properties are often fused together to give a single, unsegmentable morph. Most Indo-European languages are fusional. For example, Latin, which is highly synthetic, is fusional in many ways, or Russian is also fusional, perhaps less synthetic than Latin, but more synthetic than German. Well, what about present-day English? Well, English has traditionally been defined as a synthetic fusion language too, perhaps over here, but as many of you know from the historical development of English, English is becoming more and more analytic, so perhaps it would be legitimate to place English up there, because English has now many analytic words, words that cannot be subdivided any further, words such as conjunctions, prepositions, pronouns, etc. Note that it is often difficult to assign a language to the agglutinating or fusional type. Many synthetic languages have properties of both types. Further, but very special type of synthetic language is referred to as polysynthetic. Now, a polysynthetic or incorporating language is one in which the verb and the subject or object of a sentence may be included within a single word. In other words, the main functional elements of clause structure are joined, that is incorporated in one word and have no independent existence. There are three types of polysynthetic languages. For example, the concatenative type, which you can see over here. An example would be Eskimo Alliut, where there is a heavy reliance on concatenation, that is chaining of inflectional and derivational affixes in the formation of words, well, or sentences you might argue. Then we have the compositional type, where content words, especially verbal stems, are incorporated into more complex constructions. Examples can be found in the Papuan languages. And finally, we have the slot type, a type of language that determines fixed positions for elements that can be obtained by members of the same paradigm, so slots where certain elements belong. Slot type languages or slot type polysynthetic languages, to be precise, can be found among the North American and Australian indigenous languages. In all polysynthetic languages, the nature of words is extremely difficult to define. So this is another argument for morphemes as linguistic units rather than words. Let's finally illustrate the degree of synthesis of a language. As I said, the distinction between analytic and synthetic is a continuum, ranging from the most radically analytic or isolating type to the most highly synthetic, perhaps polysynthetic type of language. The position of a language on this continuum can be determined by calculating its degree of synthesis, that is the number of morphemes per word in a random text sample of the language. Now the random text sample which we have available in the language index of the virtual linguistics campus is of course the famous story, The North Wind and the Sun, which has been recorded more than a thousand times for you to work with. So let's take the first sentence and only the first clause to illustrate the calculation of the degree of synthesis. Now here is Chinese. This is the first sentence. The North Wind and the Sun were disputing which was the stronger in Chinese. Let's listen first. And here is the same sentence in German. So I can read it myself because I'm a native speaker of German. One dispute between the North Wind and the Sun, which of the two would be the stronger? Well, you might want to stop the e-lecture here and calculate the degree of synthesis by calculating the number of morphemes and divide them by the number of words. So now it's up to you. Okay, have you got the results? Well, here they are. The degree of synthesis in Chinese is clearly 1.0, the lowest figure you can get. We have 19 words and each word can clearly be associated with a particular type of meaning. And so we have a relationship of one to one between morphemes and words. You might want to argue perhaps that Senthai is in fact two morphemes, so about one word, but that's a little bit critical and disputed in Chinese. Well, in German, in German we can clearly count 14 words but 20 morphemes. You see, for example, that we have two morphemes in Stritten, where Stritt is the stem which involves a vocali change, by the way, of Stritten, and N is the aphix which fuses the third person and the singular into its structure. Or look at this one, Enen, them, plus dative plural in the aphix. So clearly German is far more synthetic and the higher this value will be, the more synthetic a language is. Well, here's my suggestion to you. Look at the language index on the Virtual Linguistics Campus, load the language, load the story, the North Wind and the Sun, and calculate the degree of synthesis of that language. Let's summarize. The morphological or structural parameter of language classification is straightforward. In examining the inflectional possibilities of a language, we can classify languages as analytic versus synthetic. And if a language is synthetic, we can further examine the type of synthesis. If the morphemes that make up the words of a language are functionally unambiguous, the language is synthetic agglutinating. If not, it is a synthetic, fusional language. Well, and in specific cases where whole sentences are expressed via inflectional morphemes, we have a special type, the polysynthetic type of language. However, we should never forget that in all cases, the structural parameter is not binary, but continuous.