Hittite Online

Lesson Text

Translation

Grammar

1 The Sound System

Given the limitations of the cuneiform syllabary and the fact that there are no current speakers of Hittite, the exact pronunciation of Hittite will never be known. However there is consensus on some major points. The following description takes into account some points of controversy but omits many minor details.

1.1 Vowels

Old Hittite had at least three short vowel phonemes and three long vowel phonemes. Subsequent sound change may have altered this system somewhat. At least there are indications that sound change altered the distribution of some of these sounds.

Short Vowels		Front				Back
High		[i]				[u]
Mid		[e]
Low				[a]

Long vowels are indicated by double writing, also called plene writing, for example: da-a-at-ti [dāti] 'you take', pe-e-da-as [pēdas] '(s)he took away', or i-it [īd] 'go!' (see section 2.2). There is no evidence for back rounded mid vowels [o] and [ō]. Indo-European *a and *o had probably merged in a-type vowels by the time the Hittite texts are attested, and there is no evidence that sound change created new o-type vowels. Hittite may also have had two mid front close vowel phonemes that were pronounced like the vowels of English bet and bed, although the existence and distribution of these vowels is not a matter of general agreement.

Long Vowels		Front				Back
High		[ī]				[ū]
Mid		[ē]
Low				[ā]

1.2 Diphthongs

There are several diphthongs, although the exact number is under dispute. [āi] and [āu] were probably long diphthongs, from several sources. It is not clear whether there were also two corresponding short diphthongs [ai] and [au]. There were probably also at least three other diphthongs: [eu], [iu], and [ui], which were created by various sound changes. The length of the vowel element in these diphthongs is not known.

1.3 Stops

There were eight stops, and there is no evidence to indicate that their distribution was affected by sound change during the 500 years during which Hittite was written.

		Labial		Dental		Velar		Labiovelar
		[p]		[t]		[k]		[kw]
		[b]		[d]		[g]		[gw]

Some scholars consider [p], [t], [k] and [k^w] to have been voiceless stops, and [b], [d], [g] and [g^w] to have been voiced stops. Other scholars, however, consider the first series to have been fortis, or long, stop consonants and the second series to have been lenis, or short, stops. Members of the first series are written as double consonants between vowels (for example, a-ap-pa adv. 'back' [āpa]), and members of the second are written as single consonants (e.g., a-pa-a-at 'it' [abāt]). See section 2. There is also some evidence for the existence of true geminate, or doubled, stops, for example perhaps at-ta-as 'father' [attas], though the distribution of such sounds is disputed.

1.4 Glides, Nasals, and Liquids

There were two glides, a palatal [y] and a labial [w]. There were also at least two nasal phonemes [n] and [m] and two liquid phonemes [l] and [r]. Nasals and liquids are written double in a number of contexts, and the evidence suggests that [n], [m], [l], and [r] were opposed to geminate, or doubled, nasals and liquids. In some cases a clear contrast can be established, indicating that the difference was at least in part phonemic, e.g., a-a-ri '(s)he arrives' [āri] as opposed to a-ar-ri '(s)he washes' [ārri].

1.5 Fricatives and Affricates

Hittite had at least one voiceless dental fricative [s], which may have contrasted with a geminate [ss]. The existence of corresponding voiced [z] and [zz] is uncertain. Hittite probably also had at least two h-like phonemes from the Indo-European (IE) laryngeals *h₂ and *h₃. IE *h₃ was retained only initially, and it is unclear whether it was distinct from the descendant of *h₂, which was also retained initially. Between vowels, there are two sounds [h] and [hh], which are both descended from *h₂, and which may have been voiced and voiceless, or fortis and lenis, counterparts of one another. There was at least one affricate, [ts], which is spelled with signs containing Z.

1.6 References on Hittite Phonology

For more detailed current analyses of the Hittite sound system see H.C. Melchert, Anatolian Historical Phonology, Amsterdam-Atlanta, GA, 1994, and S.E. Kimball, Hittite Historical Phonology, Innsbruck, 1999. The two books differ in emphasis and arrangement, and although they are in basic agreement on a number of issues, they disagree on others. This account tries to present some of the differences.

2 The Cuneiform Syllabary

Hittite is written in a form of the cuneiform syllabary, a writing system in use in Sumerian city-states in Mesopotamia by roughly 3100 B.C.E. and used to write a number of languages in the ancient Near East until the first century B.C.E. Most documents in cuneiform were produced by impressing a wedge-shaped stylus cut from a reed into tablets of moist clay, and the name cuneiform, applied to the writing by modern scholars, means "wedge-shaped." In the early third millennium B.C.E. speakers of dialects of Akkadian, a Semitic language, conquered the Sumerian city states, and by approximately 2350 B.C.E. documents were written in cuneiform in Akkadian. Sumerian, a long extinct language, is related to no known language, ancient or modern, and its structure differed from that of Akkadian, which made it necessary to modify the writing system. These modifications are important, because the Hittites borrowed them when they borrowed the writing system, probably from a north Syrian source, in the early second millennium B.C.E. In borrowing this system, the Hittites retained conventions established for writing Sumerian and Akkadian and made few undisputed innovations of their own. For the use of Sumerian and Akkadian in Hittite texts, see section 3.

Unlike an alphabet, in which characters, or letters, represent single sounds, a syllabary is a form of writing in which characters, or signs, represent syllables. Most of the signs that stand for syllabic values are of the shape: V = vowel, CV = consonant + vowel, or VC = vowel plus consonant. There are also a number of signs used to represent sequences of the shape CVC = Consonant-Vowel-Consonant. A standard convention for transcribing Hittite texts is to separate each sign with a hyphen. For example, in the Hittite version of the syllabary the signs AT, TA, and AS (at-ta-as) represent attas 'father' and the signs TAR, RA, AN, ZI (tar-ra-an-zi) represent taranzi 'they speak'.

2.1 Spellings for Consonants

There are some important points to bear in mind about the relationship of the syllabary to the probable pronunciation of Hittite. Because of the spelling practices of the version of Akkadian cuneiform that the Hittites borrowed, the transcription of stops as voiced or voiceless is neither a reliable guide to Hittite pronunciation nor to Indo-European etymology. Signs with T, K, or Q, a sign used for a particular type of consonant in Akkadian, can reflect etymologically voiced or voiceless sounds; for example:

• a-ta-an-zi, a-da-an-zi 'they eat', and e-te-er 'they ate' pronounced: [adantsi] and [ēder] from the IE root *h₁ed-;
• pe-e-da-an, pe-e-di, pe-e-ti 'place' pronounced [pēdan] and [pēdi] from IE *pedom;
• te-e-eh-hi 'I put', da-it-ti and ta-it-ti 'you (sg.) put', da-a-i '(s)he puts', ti-ya-an-zi 'they put' pronounced [dēhhi], [dāiti], [dāi], and [diyantsi] from the IE stem *dheh₁y-;
• ge-e-nu, ke-nu-un 'knee' pronounced [gēnu], [gēnun] from IE *genu-;
• ke-es-sar-az 'from the hand' pronounced [gissrats] from IE *ghesr-;
• du-ug-ga-a-ri, du-uq-qa-a-ri, tu-uq-qa-a-ri 'is important' pronounced [tukāri] from the IE root *tuk-;
• ku-en-zi '(s)he kills' pronounced [g^wēntsi] from IE *gh^wenti.

The reasons for this are complicated. The original version of the syllabary didn't regularly distinguish voiced and voiceless stops in vowel-plus-stop signs. Such signs are transcribed according to the value of a following sign. For example, the sign transcribed as UK or UG in du-ug-ga-a-ri, du-uq-qa-a-ri and tu-uq-qa-a-ri is the same sign, but the signs that follow it are GA and QA. Even where the syllabary theoretically offered ways of indicating a distinction, for example as is the case with most CV signs and many CVC signs, the distinction was not indicated. The original syllabary could distinguish PV from BV, but the Hittites normally used only the former in writing words syllabically. Thus PV signs can indicate [pV] or [bV]; for example: pa-a-ah-hur 'fire' pronounced [pāhur] (IE root *peh₂) but par-ku-us 'high', pronounced [bargus] (IE root *bhergh-).

Between vowels, doubled stops usually indicate etymologically voiceless stops, while single stops indicate etymologically voiced stops; for example: a-ap-pa 'after' pronounced [āpa] (IE *apō) but a-pa-a-as 'that' pronounced [abās] (IE *h₁obho-) or ú-e-ek-ka-an-zi 'they ask' pronounced [wekantsi] (IE *wek-) but te-e-kan 'earth' pronounced [dēgan] (IE *dheghōm).

Liquids, nasals, [s], and [h] may also be written single or double between vowels. Sometimes, they are also written this way before or after consonants, but there is little agreement on the interpretation of these spellings. The glide [y] may be represented with the sign I, for example: i-e-et '(s)he made' [yēt], si-i-e-es-sar 'beer' [syēssar], i-ú-ga-an 'yoke' [yugan], or pi-an-zi 'they give' [pyantsi]. Between a-vowels it is often represented with YA (e.g., sal-la-ya-as 'big' (gen. sg.) [sallayas]), and before [a] it is often spelled Ci-ya (e.g., pi-ya-an-zi 'they give' [pyantsi]). Before [a] it is often represented with I-YA (e.g., i-ya-an-zi 'they make' [yantsi]). The glide [w] tends to be written with WA before [a] (e.g., wa-a-tar 'water' [wādar]) and with the signs U and Ú before other vowels; the latter tends to be favored before I and E, for example: ú-e-el-lu-wa-as 'meadow' [wēllwas]. A spelling ^CU-WA- or Ú-WA-, U-WA- may be used to indicate the sequence [wa], for example: ú-wa-an-du 'let them look' [wantu], u-wa-it-ta-ri '(s)he sees' (3 sg. midd.) [waitari], and ú-wa-a-tar 'water' [wādar]. There is no evidence to indicate that the signs U and Ú ("u-two") indicated different sounds.

Signs transcribed with Z indicate a sound [ts] like the final two sounds of the English word cats, for example zi-ik 'you' (nom. sg.) [tsik], az-zi-ik-kan-zi 'they keep on eating' [atsikantsi], pa-a-iz-zi '(s)he goes' [pāitsi], par-na-az 'from the house' [parnats]. Any evidence for the use of Z to indicate the voiced counterpart of [s] (i.e., with a value like that of English z in zoo) is limited and controversial.

2.2 Spellings for Vowels

Vowels may be doubled by writing CV-V, V-VC or CV-V-VC, a device sometimes known as plene writing. In many cases, this is done to indicate long vowels; for example: da-a-i '(s)he takes' [dāi], ka-a-an-ki 'it hangs' [gānki], i-it (imp.) 'go' [īd], ak-ku-us-ke-e-wa-ni 'we keep drinking' [ak^wskēwani], si-i-e-es-sar 'beer' [syēssar], and as-su-u 'goods' (neut. pl.) [āssū]. The use of plene writing to indicate long vowels is not necessarily consistent. Generally, however, it's used more often in older texts than in more recent ones.

The version of the cuneiform syllabary used by the Hittites did not always allow the front vowels [e], [ē] and [i], [ī] to be distinguished. Signs with front vowel plus stop were indifferent to the distinction. Thus, the sign composed of front vowel plus dental stop could be read as [et], [ed], [it], or [id]. Sometimes plene writing or etymological knowledge allows us to discriminate the vowel. For example, e-et 'eat' (imp.) is from IE *eh₁d and should be read as [ēd], but i-it 'go' (imp.) is from IE *ih₁dhi and is read [īd]. Some CV signs for stop plus vowel did not allow the distinction to be made (e.g., KI can represent [ki], [ke], [ge], or [gi]). Other CV signs for stop plus vowel, however, allowed the distinction to be indicated (e.g., there is one sign for TE and another for TI). Many CV and VC signs with other types of consonants also allowed the distinction (e.g., SE vs. SI, HE vs. HI, ME vs. MI, EL vs. IL, EN vs. IN, and ES vs. IS), but the distinction was not always preserved in spelling. For example, after consonants, the nominative plural ending, which was [-ēs], is found as -CE-E-ES, -CE-ES, -CI-IS, and even -CE-IS and CI-ES. There are a number of reasons for this apparent anarchy. The source from which the Hittites borrowed cuneiform probably did not observe a distinction between [e] and [i] or [ē] and [ī] consistently. Sometimes the Hittite scribes preferred one sign over another regardless of its value in the syllabary. For example, HI, which is a less complex sign than HE, is much more common than HE. ZE, which is more complicated than ZI, is quite rare. Sometimes the use of one sign instead of another seems to have been a matter of habit. DI, for example, is hardly ever found at the beginnings of words, although TI and TE are. Both TI and DI may be found at the ends of words, though TE is relatively rare in this position. It is also possible that at some point the long vowel [ī] and [ē] and the short vowels [i] and [e] had merged, at least in some phonological environments, but such a merger is subject to considerable dispute.

2.3 Spellings for Consonant Clusters

We know that Hittite had initial, internal, and final consonant clusters. But, because the syllabary offers no way of writing a consonant that is not preceded or followed by a vowel, such clusters sometimes had to be written with a so-called "dead" vowel; that is a vowel that is written but not pronounced. In the clearest cases, the designated dead vowel is A, written CA or AC. Some examples are pa-ra-a 'forth, to' [pr:a], par-as-du-us 'foliage' [parstus], e-ez-za-at-te-en 'eat' (2 pl. imp.) [ētsten], ka-a-as-za 'hunger' [kāsts], and ki-is-sa-ri beside ki-is-ri 'in the hand' [gissrī]. Descendents of IE words in initial *s plus stop are normally written with the initial sign IS, (e.g., is-ta-na-a-an-as 'at the altar tables' from the IE root *stah₂-), though forms of the verb meaning "make a libation" (from IE *spend-) are written with SI-PV (e.g., si-pa-an-ti '(s)he makes a libation') except in the early period, where a spelling with IS-P is occasionally found (e.g., is-pa-an-ti '(s)he makes a libation'). It is possible that the initial IS represents a genuine vowel, in which case a spelling is-ta-na-a-na-as would represent [istanānas], indicating a development comparable to the development of an initial vowel before [s] plus stop in Spanish (e.g. espada 'sword' from Latin spatha). However, it is also possible that IS was a sign with the dead vowel used as a fairly conventional spelling for initial clusters of [s] plus stop, in which case the spelling is-ta-na-a-an-as would represent [stanānas].

3 The Use of Sumerian and Akkadian

Since the Cuneiform syllabary was invented by Sumerians and later borrowed and adapted by speakers of the Semitic language Akkadian, the Hittite scribes inherited a number of writing conventions from Sumerian and Akkadian. The most notable of these is the use of Sumerian and Akkadian words as ideograms representing Hittite words and distinguishing grammatical functions. The typical Hittite text is a mixture of syllabic writing of Hittite words, Sumerian, Akkadian, and odd hybrids in which Hittite syllabic writing is combined with Sumerian and/or Akkadian or in which Akkadian and Sumerian are combined.

3.1 Sumerograms

Although Hittite could be spelled with syllabic signs, the Hittite scribes also used Sumerian signs (or Sumerograms) ideographically. In other words, a word might be written in Sumerian, but it was read in Hittite. This is a practice that the Hittites borrowed and adapted from Akkadian cuneiform, which also used Sumerograms. In transcribing Hittite texts, it is conventional to render Sumerograms in capitals. For example, the Hittite word for "king" was usually written as LUGAL, the Sumerian word for "king," but we know that when a text was read, the word was pronounced in Hittite as [hāssus]. Similarly, the Sumerogram EN 'lord, master' was read in Hittite as [ishās], and HUL 'evil, bad' was read as Hittite [idālu]. The Sumerian plural markers MEŠ, HI.A, and DIDLI were also used to indicate plurals. For example, we find ŠEŠ 'brother' (Hitt. nom. sg. [negnas]) beside ŠEŠ^MEŠ 'brothers' (Hitt. nom. pl. [negnēs]), UDU 'sheep' (Hitt. nom. sg. *[pekkus] or [yants]) beside plural UDU^HI.A, or URU 'city' (Hitt. nom.sg. [happeriyas]) beside plural URU^DIDLI. Occasionally, one finds two Sumerian plural markers attached to a word, for example, ERIN^MEŠHI.A 'troops' or URU^DIDLI.HI.A 'cities'.

3.2 Phonetic Complements

Often, probably to distinguish grammatical function, to disambiguate synonyms, or simply to remind the scribe of the Hittite reading, Sumerograms were followed by one or more syllables to indicate the reading. Such syllabic tags are called phonetic complements, and they are another practice borrowed from Akkadian. For example, The sentence LUGAL a-us-zi could mean either "(S)he sees the king" or "The king observes." However, LUGAL-us a-us-zi with the syllabic complement -us indicating that LUGAL is to be read in Hittite as nominative singular [hāssus] clearly means "The king observes" (pronounced [hāssus austsi]). But LUGAL-un a-us-zi with the syllabic complement -un, indicating that the word for "king" is accusative singular ([hāssun]) clearly means that someone sees the king (pronounced [hāssun austsi]). The spelling UDU-us 'sheep' (nom. sg.) indicates that the Hittite word underlying the Sumerogram was a u-stem (probably *[pekkus], which is not attested in syllabic writing). HUL-lu-un (Sumerian HUL 'evil' plus Hittite phonetic complement -lu-un) was read as "evil" (noun or adjective) accusative singular (in Hittite, [idālun]), but HUL-ah-mi (HUL plus phonetic complement -ah-mi) was read as first person singular present [idālawahmi] 'I (will) harm, I (will) do evil'. As with words written in Hittite, Sumerograms are also found with enclitic pronouns and particles, for example: LUGAL-ma-mu UDU pa-is 'But the king gave the sheep to me.' Enclitics could also follow phonetic complements; for example: LUGAL-us-ma-mu a-us-zi 'The king sees me'. (LUGAL = "the king" plus -us indicating nominative singular, plus enclitic conjunction -ma 'but' plus first person singular accusative-dative enclitic personal pronoun -mu 'to me'.)

3.3 Akkadograms and Use of Akkadian

The Hittite scribes also used Akkadian words as ideograms. Such words, called Akkadograms, are conventionally italicized in transcriptions of Hittite texts and they may be capitalized to distinguish them as Akkadograms as opposed to Akkadian words borrowed into Hittite. (The Hittites did borrow some words from Akkadian; for example: Hittite tuppi 'tablet' from Akk. tuppu, ultimately from Sum. dub.) For example, the Hittite word ish:as, 'lord, master' could not only be written in Sumerian as EN but also in Akkadian as BEL (Akk. bēlu). The third person singular preterit of Akkadian ṣābatu 'seize', for example, is sometimes used as an Akkadogram IṢ-BAT to stand for for Hittite ēpta 'seized'. Akkadian was a more highly inflected language than Sumerian, with endings for gender, case and number in the noun and adjective and endings for number, tense, gender and aspect in the verb. These endings were sometimes written in Hittite texts. For example, we find inflected forms of the Akkadogram meaning "lord, master": nom. sg. BE-LU and gen. sg. BE-LAM. Similarly, the word for "father" is found as nom. sg. ABU and gen. sg. ABI. Akkadian also had a special form of the noun, called the construct state (essentially the stem), to which enclitic possessive pronouns were added (e.g., Akk. ab-ī 'my father', abi-ya 'of my father'). The Hittites used these Akkadian possessive pronouns too, but their Akkadian was often faulty, and one finds Akkadograms like ABI-YA, or BĒLI-YA with the pronoun attached to the Akkadian genitive. Sometimes such forms were used syntactically as nominatives, vocatives, or accusatives. Akkadian possessive pronouns could also be used with Sumerograms (e.g., DUMU-ŠU 'his son'). Although Akkadograms were pronounced in Hittite, in writing combinations of nouns and possessive ponouns, the scribes followed Akkadian spelling rules. When possessive suffixes beginning in š were attached to Akkadian words ending in dental stops, the suffixes were spelled with z in the version of Akkadian which the Hittites borrowed, indicating a pronunciation [-ts]. For example, in Akkadian, qāt 'hand' plus the third person singular suffix šu is spelled quazzu. The Hittite scribes followed this practice in writing the suffixes after Sumerograms when the Akkadian equivalent ended in a dental stop, even if neither the Sumerogram nor the underlying Hittite word had a final dental stop. For example, the Hittite word for "land" or "country" was [udnē]. It is often spelled with the Sumerogram KUR, and spellings for "his country" as KUR-ZU, reflecting Akkadian mā-zu for māt 'country' plus possessive suffix -šu are sometimes found.

The scribes used Hittite phonetic complements with Akkadian words less commonly than they did with Sumerian, perhaps because Sumerian was a dead language by the second millennium but Akkadian was a living language used, for example, by the Hittites in diplomatic correspondence. Akkadian could also be used to indicate grammatical relations among Hittite words. For example, the Akkadian prepositions ina, 'in, into' and ana 'to' were used as Akkadograms to indicate the dative and locative as in: IN-A ÉGAL 'into the palace', INA ^URUHattusi, 'in the city of Hattusas', A-NA TUPPI 'on the tablet', or I-NA kisri-ssi 'in his hand'. The Akkadian preposition ša was used as an Akkadogram to indicate the Hittite genitive, e.g., ŠA LUGAL-u-wa-as 'of the king'. Since Hittite could indicate these relations both with case forms (e.g., hāssuwas 'of the king' or tuppiyas 'of the tablet') and with postpositions (e.g., udnē andan 'in(to) the country') such prepositions serve as graphic indicators only; they were not pronounced. One often finds the dative-locative, for example, marked three times (e.g., I-NA par-ni an-da-an 'into the house' = Akk prep. INA 'into' + Hitt dat. sg. parni 'into the house' + Hitt. postposition andan 'into'). Finally, although Hittite had enclitic and free-standing conjunctions, the Akkadogram Ú could be used to indicate 'and'.

Fragments of Akkadian could also be used after Sumerograms as phonetic complements; for example: DINGIR^LUM (from Sumerian DINGIR 'god' plus Akkadian nominative singular ilum 'god'), DINGIR^LIM (from DINGIR plus Akkadian genitive singular ilim), or DUMU-^ru from Sumerian DUMU 'son' plus Akkadian -ru from nom. sg. māru 'son'). Again, the scribes' command of Akkadian was spotty, and a form like DINGIR^LIM was not restricted to the genitive (e.g., with Akkadian phonetic complement from the genitive ilim and Hittite phonetic complement -ni indicating the dative DINGIR-^LIM-ni 'to the god' (pronounced in Hittite as [siuni]).

3.4 Determinatives

A characteristic of Sumerian cuneiform that was borrowed by both the Akkadians and the Hittites is the use of determinatives. These are signs that were not pronounced (though most could be used as independent Sumerograms) that indicated what class of items a particular word belonged to. Although most determinatives precede the words they specify, a few customarily follow. One common way of distinguishing determinatives from words in modern transcriptions of cuneiform texts is to superscript the determinatives. For example, the sign GIŠ, which could be used as an independent Sumerogram meaning "wood" or "tree," was also used as a determinative for objects customarily made of wood, such as: Sumerogram ^GIŠBANŠUR 'table', or Hittite ^GIŠhapūti, a type of chair. Similarly, the sign URU, used independently to write the word for "city," usually preceded the names of cities or settled areas; for example: ^URUHa-at-tu-sa-as '(the city of) Hattusas'. The sign MUŠEN, meaning 'bird', was also used as a determinative following the names of birds, for example, ha-ra-a-as^MUŠEN 'eagle' (or, as a Sumerogram with a determinative, Á^MUŠEN). Divine names were preceded by the determinative DINGIR 'god', which in modern transcription is normally abbreviated D (e.g., ^DTe-le-pe-nu-us, a god of vegetation). Determinatives were also used indicate the sex of persons referred to by proper names or occupational titles. The sign MUNUS, used independently to mean "wife" or "woman," preceded names of women. When it precedes women's names it may be rendered in transcriptions as superscript F (e.g. ^FPu-du-he-pa, Puduhepa, the name of a powerful Hittite queen of the thirteenth century B.C.E. and wife of Hattusilis III). Before the names of men a single vertical stroke was used. This is often rendered in transcriptions as superscript M (e.g. ^MTe-le-pe-nu-us, the name of an early Hittite king). Before male occupational titles, the Sumerian sign LÚ 'man' was used (e.g., ^LÚA.ZU 'physician' or 'ritual healer'), and the sign MUNUS was used before female occupational titles, e.g., ^MUNUSŠU.GI, or, in Hittite, ^MUNUShāsawas 'wise woman, female ritual healer'. The use of determinatives is somewhat like the use of signs like $ or % in English to indicate that what follows is a sum of money or that what precedes is a percentage, although we typically pronounce the non-phonetic signs in, e.g., $100 and 60%, saying "one hundred dollars" (or "one hundred bucks") and "sixty percent."

3.5 Rebus Spellings

Finally, so-called "rebus spellings" involving the use of homonyms were used, especially in writing proper names. This is a strategy comparable to writing the English phrase "I see" with a picture of an eye followed by wavy lines to indicate the waves of the sea, or modern Internet abbreviations such as CU l8tr for "See you later." For example, the word for "man" in Luvian is ziti, and it is an element in a number of names of men in later Hittite. The proper name Uhhaziti is spelled with both Hittite syllables and the Sumerogram LÚ 'man' as ^MU-uh-ha-LÚ in the Annals of Mursilis. One gets the impression that the scribes were having fun with such punning spellings.

4 Active verbs

Compared with the classical languages, Greek and Latin, Hittite has a verb system that is fairly simple. There are two types of endings for active verbs: "mi-endings," which characterize the mi-conjugation and "hi-endings," which characterize the hi-conjugation. Both conjugations include verbs that are transitive, intransitive, and stative. The verb system has two voices, active and middle; the latter is also used as a passive. The verb also has two tenses, present and preterit, or past. The present, often accompanied by appropriate adverbs, is used for the future. The verb system also has two moods: indicative and imperative; the optative and subjunctive of other ancient IE languages are lacking. Modal, aspectual, and other nuances of meaning can, however, be distinguished through the use of adverbial particles, through the addition of derivational suffixes, and through the use of periphrastic constructions. The imperative also has modal functions. Each active conjugation and the middle show some variation in the forms of particular endings. This variation is sometimes the result of the language having inherited variant forms. However, it is often the result of sound changes and analogical innovations, some of which took place in the prehistoric period and some of which can be observed by looking at texts composed at different periods.

4.1 Active Endings

The conjugations take their names from the form of the first person singular present, which is -mi in the mi-conjugation and -hhi in the hi-conjugation. It is only in the singular of the present, preterit, in the third person singular imperative, and partially in the second person singular imperative that the distinction between the two conjugations is maintained; both share a set of endings in the present, preterit, imperative plural, and first person singular imperative. Although we speak of mi- and hi-conjugation forms, each conjugation influenced the other. Some verbs may occur with endings from both conjugations.

Some of the endings provided in the paradigms below have doubled consonants. In the cuneiform syllabary, however, spellings with doubled consonants are only possible after vowels. Such spellings are regular for endings beginning with dental stops or [hh], but the third person singular present mi-conjugation ending may be spelled -zzi or -zi after vowels.

4.2 Inflection of the mi- and hi-conjugations

Present		mi-conjugation				hi-conjugation
1 sg.		-mi				-hhi
2 sg.		-si				-tti
3 sg.		-zzi				-i
1 pl.				-weni
2 pl.				-tteni
3 pl.				-anzi, -nzi
Preterit
1 sg.		-un				-hhun
2 sg.		-s, -ta				-tta
3 sg.		-ta				-s, -ta
1 pl.				-wen
2 pl.				-tten
3 pl.				-ēr, -er
Imperative
1 sg.				-allut
2 sg.		-0, -t				-0
3 sg.		-ttu				-u
2 pl.				-tten
3 pl.				-antu, -ntu

4.3 Paradigms of mi-conjugation verbs

The following are paradigms of two common mi-verbs, ēs- 'be' and daske-, the iterative of dā- 'take'. The former has a stem that alternates between ēs- and as-. The iterative is composed of a root da- to which a suffix that takes the form -ske- or -ska- is added; thus the stem of the iterative is daske-, daska-. In the second person singular imperative of the iterative, the stem final vowel [-e] regularly becomes [-i] in final position. Since not all forms of these verbs are attested, where possible, reconstructions based on attested forms of similar verbs are indicated with an asterisk. Blanks indicate unattested forms for which reconstructions cannot be made with confidence.

Present
1 sg.		ēs-mi		daske-mi
2 sg.		ēs-si		daske-si
3 sg.		ēs-zi		daske-zzi
1 pl.		*as-weni		dasga-weni
2 pl.		*as-teni		daske-tteni
3 pl.		as-anzi		daska-nzi
Preterit
1 sg.		ēs-un		dasga-nun
2 sg.		*ēs-ta		daske-s
3 sg.		ēs-ta		daske-t
1 pl.		ēs-wen		dasga-nun
2 pl.		*ēs-ten
3 pl.		es-ēr		dask-ēr
Imperative
1 sg.		as-allu
2 sg.		ēs		dask-i
3 sg.		ēs-tu		daske-ddu
2 pl.		ēs-ten		daske-tten
3 pl.		as-antu		daska-ndu

4.4 Paradigms of hi-conjugation verbs

The following are paradigms of two common hi-conjugation verbs. The first verb, ār- 'arrive, reach', a consonant stem, shows an alternation between long and short root vowels. The third person singular preterit of ār- is spelled a-ar-sa, ar-sa, and a-ar-as, indicating that it was probably pronounced [ārs]. The verb dā- 'take' has a stem that alternates between [dā-] and [d-]. The ending -umēni of the first person plural preterit indicates that a prehistoric form of the second person plural was *duweni, since a prehistoric sequence of *u followed by *w became um.

Present
1 sg.		ār-hi		dā-hhi
2 sg.		ār-ti		dā-tti
3 sg.		ār-i		dā-i
1 pl.		*ar-weni		t-umēni
2 pl.		ār-teni		dātteni
3 pl.		ar-anzi		d-anzi
Preterit
1 sg.		ār-hun		dā-hhun
2 sg.		*ār-ta		dā-tta
3 sg.		[ār-s]		dā-s
1 pl.		ar-wen		dā-wen
2 pl.				dā-tten
3 pl.		ār-er		dā-er
Imperative
1 sg.
2 sg.		*ār		dā
3 sg.		ār-u		dā-u
2 pl.		ār-tten		dātten
3 pl.		*ar-antu		d-andu

5 Gender and Case

Unlike Greek and Latin, Hittite distinguishes only two genders: animate (or common) and neuter. Some nouns attest both animate and neuter forms. There is evidence that within Indo-European the development of a true feminine gender was later than that of the masculine and neuter. Although it has been argued that the Anatolian languages reflect this early state of affairs, there is evidence that Anatolian did inherit a grammatical feminine but later lost it, and inherited masculine and feminine nouns merged in the animate. In the neuter singular and plural, the forms of the nominative and accusative are identical. Most animate nouns make a nominative singular in -s (from Indo-European *-s) and an accusative singular in -n (from Indo-European *-m). In the animate plural, the regular nominative ending was originally -ēs and the accusative plural ending was -us, though archaic texts attest a rare animate nominative plural ending -as. Both genders share case forms in the genitive, dative, ablative, instrumental and allative. (See subsequent lessons for more thorough discussions of the functions of the cases).

In the singular, there are seven regular case distinctions: nominative, accusative, genitive, dative-locative, ablative, instrumental, and allative. A handful of nouns and adjectives with the suffix -u make a distinct vocative, but the vocative was moribund, even in the early period. For direct address the nominative was normally used instead. A few nouns have a locative formed without a case ending. The allative eventually died out, its function, to indicate direction towards, being taken over by the dative-locative.

The plural shows a tendency to reduce the number of separate case endings. The distinction between the nominative plural in -ēs and the accusative plural in -us is maintained in early texts, but in later texts forms with the original nominative ending -ēs may show up in accusative function and forms with the accusative -us may show up in nominative function. The most common neuter nominative-accusative plural ending is -a. However, a few neuter nouns and adjectives that do not take this ending show lengthening of the vowel of a suffix, at least in early texts. An archaic genitive plural ending -an (from Indo-European *-o:m) is found sporadically in early texts, but it was eventually replaced by the ending -as, identical to the ending of the dative-locative plural. A few nouns attest an instrumental plural ending -it. Sample paradigms of nouns and adjectives made from various stem forms will be provided in subsequent lessons.

5.1 Noun and Adjective Endings

Singular		Anim.				Neut.
nom.		-s				-n
acc.		-n				-n
gen.				-as
dloc.				-i
abl.				-az
inst.				-it
all.				-a
Plural
nom.		-ēs				-a
acc.		-us				-a
gen.				-as
dat.				-as
abl.				-az

The animate nominative singular ending is spelled -s after stems ending in vowels or dipthongs, for example: at-ta-as 'father' = [attas], or ú-e-el-lu-us 'meadow' = [wēllus]. The nominative ending appears as -za, prepresenting a pronunciation [ts] when it occurs with stems ending in dental stops (e.g., ka-a-as-za 'hunger' = [kāsts], or ap-pa-an-za 'captive' = [apānts]). The final consonant of the ending of the ablative singular and ablative-instrumental plural was pronounced [ts]. The ablative endings are normally spelled -az (e.g., ha-as-sa-a-az or É-az and par-na-az, both meaning 'from the house'), although sometimes a spelling -za is found. A few nouns attest an ablative ending -anza, pronounced [-ants] (e.g., lu-ut-ta-an-za 'window'). In some archaic paradigms, the instrumental ending has the form -ta or -da after consonants, (e.g., ú-i-ta-an-ta 'with water', but later ú-e-te-ni-it). After u-stems, the ending may show up as -t. (e.g., ge-nu-ut 'with the knee').

5.2 Sample A-stem Noun Paradigms

A-stems are nouns and adjectives which have stems ending in [-a-] or [-ā-] in the nominative and accusative singular. The first set of paradigms are for the animate nouns lāhha- 'expedition' and anna- 'mother' and for the neuter pēda- 'place'. Since particular forms of some words are not attested, reconstructed forms, marked with an asterisk, are given where reconstructions can be made with confidence. Blanks indicate unattested forms that cannot be reconstructed with confidence.

Singular		Anim.		Anim.		Neut.
nom.		*lāhha-s		anna-s		pēda-n
acc.		*lāhha-n		*anna-n		pēda-n
gen.		*lāhh-as		ann-as		pēd-as
dloc.		lāhh-i		ann-i		pēd-i
abl.		lāhh-az		ann-az		pēd-az
inst.		*lāhh-it				*pēd-it
all.		lāhh-a
Plural
nom.		*lāhh-ēs		ann-ēs		*pēd-a
acc.		lāhh-us		ann-us		*pēd-a
gen.		*lāhh-as		*ann-as		pēd-as
dloc.		*lāhh-as		*ann-as		pēd-as
abl.		*lāhh-az		*ann-az		*pēd-az

5.3 A-stem Adjectives

A-stem adjectives, like a-stem nouns, have a stem ending in -a- in the nominative and accusative singular. As adjectives, they may be declined as animate or neuter, depending on the case of the noun which they modify. The paradigm provided below gives forms of the adjective kūnna- 'right' (direction), 'favorable'. This adjective, though sometimes spelled out as ku-u-un-na-, is often written with the Sumerogram ZAG plus phonetic complement, providing information about the shape of the stem and endings. Forms with an asterisk are not attested but can be reconstructed with confidence on the model of other a-stem nouns and adjectives and by considering the phonetic complements attached to the Sumerogram ZAG.

Singular		Anim.				Neut.
nom.		kūnna-s				kūnna-n
acc.		kūnna-n				kūnna-n
gen.				*kūnn-as
dloc				kūnn-i
abl.				kūnn-az
inst.				kūnn-it
all.				kūnn-a
Plural
nom.		kūnn-ēs				kūnn-a
acc.		kūnn-us				kūnn-a
gen.				*kunn-as
dloc.				*kūnn-as
abl.				*kūnn-az

previous lesson   |   next lesson