Welcome to the third and last presentation on the phonology part of our course on Hittite. In this presentation, we will discuss Supra-segmental phonology. More specifically, we will discuss syllable structure, accent and prosody, and clitics. Let us begin with discussing what are the possible syllable structure in Hittite.

Syllables always have a vocalic nucleus. In Hittite, we find both open and closed syllables. So both types are exemplified in the verb, arta, stands. So you can see here that the first syllable, are, ends with a consonant and is thus closed.

And the second one ends with a vowel, are, and is thus open. We don't know really whether consonants that are double new writing should be considered as the onset and coda of two separate syllables, or only at the onset of the second syllable. So we're not really sure whether the word atas

should have one syllable, at, and another syllable, ta, or one syllable, a, and the following syllable, ta. Now, the second reading is slightly more likely, since as we have already seen in the previous classes, the distinction here is not one of proper gemination,

but rather one of forties and lennies. Complex onsets are also possible, but as we have seen before, they can not be graphically represented as such. So we always find empty vowels in spelling. The following consonant clusters are attested in onset. So we have a stop plus sonorant cluster as in pre,

spelled as para with the empty vowel. We have a fricative plus stop cluster as in the word for pin, which the cluster here is sp, and is spelled as either sepi, sepi, or sapi. And then we have the combination of fricative plus nasal.

Here, the cluster is sm, and is graphically realized as either salmon or salmon. So in the last two cases, again, it is the existence of these spelling variants that suggests that we are dealing with real consonant clusters spelled with graphic vowels. Possible nuclei are vowels or diphthongs,

and possible coda are consonants, with the exception of m, which is never word final. So let us turn to the interpretation of Plina spelling and its relationship to vowel length and stress positioning. As we have already seen, by Plina spelling,

we mean the addition of an extra vocalic sign. Scholars have long debated over the phonetic and phonological interpretation of this spelling practice and have reached the conclusion that Plina spelling has several different functions. So in the first place, Plina spelling can indicate the position of the accent.

Consider the placement of the vowels spelled Plina in different cases of the word for heard. So in the first case, the root vowel te is spelled Plina. This is the nominative accusative of the word. Whereas in the second case, it is the genitive ending, which is spelled Plenas, tecnaas.

So this is consistent with the fact that for etymological reasons, we know that the nominative has accent on the root, whereas the genitive has accent on the ending. In addition, long vowels are always accented. So Plina spelling can also indicate vowel length,

as in the case of the word utar, which is the plural of the word for words and has a long vowel in the second syllable. In some cases, Plina spelling serves to indicate that a vowel is simply phonetically realized as opposed to a graphic vowel. So here you have an example.

The word para, the second vowel is spelled Pline, but this is likely only to indicate that only the second vowel is actually phonetically pronounced, whereas the first one is a graphic vowel. So we should read these as para. In other cases, Plina spelling is used

to disambiguate the reading of ambiguous ee signs. Consider the use of two Pline vowels in the word pieta. So the first Pline ee signals that the first sign should be read as pi and not as pe. And vice versa, the second Pline ee here

indicates that the following sign should be read as et and not as eet. Plina spelling is also often used with monosyllables because one sign syllables are usually avoided with the exception of conjunctions such as nu. So we have the imperative take,

which is spelled da-a with an extra vowel, even though we're not sure whether this is actually long. Finally, there are cases in which Plina spelling on the last word of a sentence indicates that the whole sentence should be read as a question. So there was a common practice in Akkadian texts,

and the Hittite scribes only rarely used it. So we have this sentence nu-me-ma-hi, shall I tell you this is a question, and it is marked as such by Plina spelling of the last vowel of the last word. So being a dead language, it is very difficult to assess where the accent fell in Hittite.

There are two main graphic diagnostics. First, as we have seen, vowels spelled Plina are most often accented. Moreover, there are cases of vowel reduction that we can actually observe that can be taken as evidence that unstressed vowels undergo weakening. So let us see two examples.

So in the first case, we have the alternation in the paradigm of the verb take. We have ep-si, she or he seizes with eh, but in the plural, we have a weakened vowel ah, so we have a-pan-si. And the same goes for the word for mouth. In the nominative accusative, we have eyes, which is mouth with this ah,

and in the dative, the ah is weakened, it disappears, so we have e-si. And note that in this case, we have a Plina spelling of the last vowel. So here again in this case, both Plina spelling and variation in vowel representation can be taken as indication that there is variation in stress positioning.

So the question is, what sort of stress did Hittite have? So recent studies suggested Hittite accent had three main features, comminativity, free stress, and lexical distribution. So with comminativity, we mean that words have only one

prosodically prominent syllable, with free stress, we mean that stress can occur on any syllable of the word without restriction, and with lexical accent, we mean that the position of the stress can be lexically stored for individual morphemes. For example, the derivational affix atar is always accented on the first syllable,

irrespective of the length of the base word it attaches to. So we can see from these examples that the accent can occur on any possible syllable, basically. So in the first column, we have words accented on the first syllable, such as escar here, then we have accent on the second syllable,

as in the case of asu, not that in this case is second and last syllable, but we also have accent on the third syllable, as in these cases, and then even accent on the fourth and the fifth syllable. Note that in case of accent on the fourth and the fifth syllable, this is also because the accent is lexically stored

for the morpheme atar that we had seen before, that attaches to words with different length in the bases. So let us now proceed to our next topic, which is clitics. This is a very important point since hit type features several clitic items, as we will see in a moment.

Most of hit type clitics occur in the so-called Wachernagel position. According to Wachernagel, who based his observation on Vedic and ancient Greek, clitics in ancient European languages tend to occur after the first accented word of the sentence.

Let us see two examples from Vedic and Greek to understand this pattern. So the first sentence comes from Vedic, and here we have this one, they push for you, and for you is realized as a dative plitic te, which is placed right after the first constituent, which is the demonstrative term.

And the same goes for Greek. So we have the sentence two, in the past I have not been accustomed to inquire you, and this you again is realized with a clitic pronoun, se, which occurs in second position right after the sentence initial adverb lien.

So Wachernagel did not consider Hittite in his study, as Hittite was not yet known at the time. However, Hittite conforms to a large extent to Wachernagel though. Among the Indo-European languages, Anatolian languages, including Hittite, are unique because they can feature very long clitic clusters in Wachernagel position.

In Hittite, the various Wachernagel clitics follow a rigid order. So in the first position of the sentence, we find either an accented word, like any lexical word basically, or demodal particle man, or the sentence initial connectives nu, su and ta.

So this is first position. So the first clitic after the first position is the quadratic particle war or wa, which indicates reported speech. Slots two to four are occupied by personal clitic pronouns. The second slot is occupied

by the plural dative accusative clitics. So nas and smas. The third slot by the third person clitic a, this is the stem that we use only in the nominative accusative. Whereas the fourth slot is reserved to dative accusative clitics of the first and second person, so mu and ta or du.

And for the third person singular dative clitic pronoun, se or si. The fifth slot is occupied by the so-called reflexive particle ta, which has several functions that we will see in detail in the lessons on syntax. And finally, the last slot is reserved to the local particles,

which are an Anatolian peculiarity. And we will see them again in another slide. So as already mentioned, several clitics in Wackernagel position can stack onto each other and form rather long clitic chains. So let us see some examples of clitic chain in actual texts. So in example one, the sentence is,

before me or the mata, the son of zita had administered it. And we find in the first position, the at position or preverb piran, followed by three clitics, which form a chain. So we have the connective ma, the third person accusative pronoun at,

and the first person oblique pronoun mu. Note here in this case that we have this curious spelling. So we have the name of armadatta spelled with the numeral for 30. So this is a kind of iconic spelling of personal names that sometimes we find in a text for gods or for other people.

In the second example, we have my brother Muwatalli sat on the throne of his father. So again, here in the first position, we have the name brother, and this is followed by three clitics.

Again, the connective ma, then the reflexive particle ca, and the local particle can. And finally, in the last two examples, we have he was a man, am I not a man too? So on the first part of the sentence, we have the sentence initial connective nu,

followed by two clitics, the quantitative war and the pronoun us in the nominative. And in the second clause, we have not the connective nu, but in this case, a fronted pronoun oog or ook, followed by three clitics. So the connective are, which is the variant of the clitic ma in other cases.

Then again, the quantitative particle wa, which is a variant of war that we have seen before. And then again, the reflexive particle ca, which in this case is only spelled as ca. So as we have seen, the model particle man, has a different behavior than other clitics. In some cases,

man behaves as a vagranial clitic, as in this first example here, in which it regularly stands in second position after the sentence initial verb, i, ami, do. So the sentence is, if I only could do something, it begins with the verb i, ami and man follows in vagranial position.

Note that in this example, we also find the focus particle pat, which is however normally not a vagranial clitic as we will see in a moment. However, in other cases, man can occur in sentence initial position and it can itself as clitics.

As in the second example here, in which we find man in sentence initial position, followed by two clitics, the accusative plural clitic pronoun, us, in this case, and by the local particle can, here. So I have mentioned the focus particle pat as a non-vagranial clitic.

There are other few non-vagranial clitics in Hittite. An example are possessive pronouns, which are typical of old Hittite. These attached to the NP that they modify and are in fact possessive adjectives, as they show full agreement with their head. For example, in this sentence here,

you can see the dative possessive clitic, c, hosted by the dative noun, iski, master. And the whole phrase means their master in the sentence, he or she takes them back to their owner.

Another non-vagranial clitic is the coordinating conjunction, ya, which occur in vagranial position when it is used to connect two entire sentences, but when it is used to conjoin two noun phrases, it follows the second NP. So consider this example here. We report it to the king and the queen.

So the dative nouns, king and queen, are coordinated, and the conjunction appears only after the second noun, that is munus lugul, the queen. So we have it here on the second coordinate. And then we have the particle pat that we have already mentioned before.

So we will return on the function of this particle in another lesson. For now, observe how it cliticizes to the item that it has scope over, which can be either a verb, as in the first sentence here. So we have pat following the verb yermaliyat,

or a noun as in the second example, as in here. So we have pat following the noun damsu, his wife. We also sporadically find local particles that do not occur in vagranial position. So in this example here, the particle can does not occur

at the beginning of the sentence, which is where we would expect it to be, but rather after the noun id, the river. In vagranial position in this very same example, we regularly find the other local particle asta here. So asta relates to noun phrases in the ablative,

such as ista pesnats here from the basin. Whereas the use of can relates to terminative noun such as id to the river here. We return on the function of these particles in another lesson. The study of Hittite prosody is still an open one, and many things still a way to be understood.

Based on careful studies on word order, recent research has suggested that there are a number of other words that show a weak prosodic behavior. These include some adverbs, functioning as preverb or as positions, conjunctions and indefinite and relative pronouns.

So we do not discuss the prosodic behavior of these items in detail today. With this lesson, we end our chapter of our course on Hittite phonology. Today we have focused on suppressing mental phenomena, including syllable structure, stress, and the behavior of clinics. It is time now to move on to the next topic

and venture into the domain of Hittite morphology. Thank you for your attention.