Learning verbs

Hoffner and Melchert and van den Hout present Hittite morphology to their students in different sequences.  Van den Hout presents -mi and -ḫi verbs in the same first chapter, and later add verbs in different stems and featuring ablaut; Hoffner and Melchert's tutorial don't introduce the -ḫi verbs until nearly halfway through the course.  Van den Hout's approach is perhaps more complete--I get a full view of the verbal system.  Hoffner and Melchert choose an approach more in common with the development of the language--if the -ḫi conjugation is a later intrusion into the language, then a learner might be well-served treating it as later, and subsidiary, to the -mi conjugation.  Van den Hout also introduces perfects and mediopassives far earlier. 

Each tutorial presents the various verb classes in different orders, and I don't yet have the perspective on the language to know what order is best.  I do think back to how I learned Greek and Latin, which have much richer inflection systems--in those, authors have to teach verbs beginning with the present tense, and then move into more complex tenses and voices.  Because Hittite verb endings are fairly sparse, I've found it most important to learn them, and jump right into the rest.

To teach in bound transcription or in transliteration?

Hittite is written not with an alphabet but with a syllabary, each sign representing either a single vowel or a cluster: consonant-vowel, vowel-consonant, or consonant-vowel-consonant.  This would obviously create problems for writing a language with sequences of consonants: the third-person plural ending -anzi could be spelled precisely with the two syllables an-zi (see Hoffner & Melchert § 1.7), but at the beginning or end of a word it was more difficult.  Hoffner and Melchert offer the word spikusta, written variously as še-pi-ku-uš-ta, ši-pi-ku-uš-ta, and ša-pi-ku-uš-ta (§1.11).  Even words introduced in the first lesson of in van den Hout's grammar, like paḫšmi, "I protect," must be rendered in cuneiforms with the symbols pa-aḫ-aš-mi.

So the student must be taught with "bound transcription," the attempt to render words like pahs- as they must have been pronounced.  The student must learn which vowels are extraneous.  But almost all Hittitological literature is presented in syllable-by-syllable transliteration and it has been very valuable for me to learn from that sort of transliteration.  So Van den Hout gives practice readings only in transliteration and Hoffner and Melchert offer bound transcription only in the first few lessons.  The awareness both of bound transcription and of transliteration is necessary.  There must be doubt in how certain proper names might look alphabetically; and, very responsibly, the Chicago Hittite Dictionary offers lemmas in both bound transcription and transliteration.

The labiovelar

Hittite apparently has two "labiovelar" consonants, g^w and k^w, transcribed in modern scholarship as ku-.  See Hoffner and Melchert § 1.88.  The labiovelars and other sounds, the laryngeals, are the most exotic aspect of Indo-European phonology to an English speaker.  To reproduce them the student must become comfortable with the pronunciation of a consonant that involves placing the back of the tongue to the velum (as with the English k) while simultaneously rounding the lips (Fortson, Indo-European Language and Culture § 3.5).  (I won't address the laryngeals in this post.)  It's easy for me to imagine k^w before a vowel, but it apparently occurs at the end of words as well (e-ku! = "drink" and -ku ... -ku = "either ... or" are cited in Hoffner & Melchert § 1.88).  I am having difficulty understanding whether it can come before another consonant without a vowel being inserted (epenthetic consonants and vowels are common in Hittite), as I see in verb forms like ekumi ("I drink") and ekutteni ("you [pl.] drink") (Hoffner & Melchert § 12.3 and n.3). 

Van den Hout's Hittite text does not discuss this phoneme.  Yet I wonder whether it might not be better to begin learning Hittite with the most accurate pronunciation.  I've thought, having learned Greek, that I might have been better served if I had learned to aspirate the consonants theta (θ),  phi (φ) and chi (χ)--or rather to de-aspirate the more familiar π = /p/, τ = /t/ and κ = /k/ (e.g., Sidney Allen, Vox Graeca 27).  The Greek theta was apparently pronounced not as the th in "thin" but as a t followed by a puff of air; the same goes for phi and chi.  Meanwhile, the standard π, τ and κ were pronounced with no breath at all.  When ancient Greek is taught in America, most teachers differentiate the letters by opposing th to t, f to p, and ch (as in Scottish loch) to k because the more genuine pronunciations would be so hard to distinguish and to produce.  These sounds almost certainly don't reflect the ancient pronunciation. So now I am forced as I read Greek on my own to check myself, to make sure that I am appropriately aspirating and de-aspirating consonants.  Learning from the beginning to pronounce the labiovelars (not to mention any laryngeals) in Hittite would lead to (lifelong?) good--or at least accurate--habits.

One of the best things about the available textbooks of Hittite is that they preserve for the learner the uncertainty faced by a real scholar.  Our editions of Greek and Latin are "sanitized"--spellings and grammar match what we expect from our textbooks, not the more varied versions of the manuscripts.  Hittite is more fragmentary; there are very many unattested forms, and many variant spellings.  There are four forms given for the nominative plural of antuhsas, "man" (Hoffner and Melchert § 4.2); examples of ablauting verb stems are occasionally given alongside (variant?) forms that do not show the ablaut.  Moreover, there is a wide selection of verb endings for some of the numbers/persons, especially in the mediopassive.  A good challenge.

Hittite flash cards

I've been trying to be as conscientious as possible in memorizing Hittite vocabulary--no one's going to be quizzing me, and so I'm making all the flash cards from van den Hout's chapter vocabularies and numbering them to keep track.  I guess I'll have a thousand or two by the time the course is complete.

It's easier than the last language for which I made flash cards.  In Greek, one's verb flash cards have to have six words each (not to mention the English translation and other notes on the back), one for each "principal part" of each verb, in order to recognize all the forms.  Page 117 here shows the nice and confusing πάσχω and πειθώ.

I try to memorize only three vocabulary cards at once.  It usually takes only a minute or two of repeating the three to help make them stick.  It's useful if the three are very similar phonologically or orthographically.  In Hittite, good examples are parš- (escape/ flee), parḫ- (chase away) and paḫš- (protect/ be careful)--especially good because the meanings, at least for me, are all relatively similar or in the same domain of experience.  I don't find that actually making the cards helps me memorize the words on them; it's really when I sit down and force myself to run through a stack of cards that I learn.

And then there's the cuneiform--I am trying to learn the entire syllabary that van den Hout includes in his textbook.  Hittite cuneiform (borrowed from the Akkadian of the Babylonians) does not represent alphabetic sounds but a series of whole syllables.  Flashcards, with a cuneiform symbol on one side and the syllable (or alternatives) that it represents are the only way I can think of to learn the symbols without years of impressing them in clay myself.  van den Hout is quite correct to present cuneiform symbols of similar shapes together (John Huehnergard's Grammar of Akkadian of course does the same), and, as there's no pattern to how the symbols match the syllables they stand for, there's no way to systematize the learning process or to proceed deductively.  For instance: a horizontal wedge that ends at a vertical wedge, but does not cross the line of the vertical wedge, is pronounced "ka₄" or "qa"¹; if the horizontal wedge does cross the line of the vertical wedge, it is "pár" or "maš."² If we then add another horizontal wedge, piercing the head of the horizontal wedge in pár/ maš, we end up with "an" or "DINGIR" (which indicates the name of a god).  Oy vey, I should say, using monosyllables.


1. The subscript numeral in "ka₄" means the syllabary has at least three other "ka" signs.
2. There's no particular reason, as far as I know, that such disparate syllables as "pár" or "maš" should be represented by the very same sign.

Structure

I'm on the second chapter of van den Hout's book and it's gotten me thinking about the order of teaching elements of a language.¹  Grammars of inflected languages present the grammar in weekly portions, to make them easier to absorb.  (When I post about trying to memorize cuneiform, I'll reflect a little more on the challenges of memorization.)  Now I've learned the declensions for the a-stem (chapter 1) and the i-stem (chapter 2), and the present and perfect indicative active endings.  So far, so good--and because endings are the same throughout the indicative, regardless of stem changes in the verb, I think I'm pretty well prepared to wade into lots of literature.  And because I've studied various languages, I am prepared for the medio-passives, imperfectives and imperatives and have even begun clearing some of the cobwebs from my brain for them.  The Hittites were kind enough to not, apparently, have left us a subjunctive or optative mood (indicating various departures from certainty in statements).

Left to my own devices, I'd probably try to learn or at least become familiar with all the verb tenses at once, not knowing what to expect in my readings, and then just ticking them off as I came to them.  If I did this, I'm sure I'd neglect some rarer ones almost entirely.  It also helps to feel like I'm making some progress--reading the practice sentences at the end of each lesson in a textbook makes me feel like I'm really learning some Hittite, as opposed to having to rush back to the grammars with each sentence.

So I will be consuming these portions as van den Hout and Hoffner & Melchert serve them.  I think this is a very good thing.  Another example: Moving from Latin to Greek (or dabbling in Semitic languages), when it came to verbs I was struck by the decreased didactic importance of the infinitive.  If I had studied Greek without a graded plan, I would certainly have learned infinitives first; then I might have figured out on my own that 1st-person-singulars are a little more helpful for determining forms.  This is especially so for contracted verbs, and for the aorist passives.  1-sgs are, I think, the dictionary lemmas for good reason.


1. Hittite is an inflected language, meaning that Hittite's nouns, pronouns, adjectives and verbs change depending on their syntax. Pronouns, nouns and adjectives are declined, as in he > him > his > they > them; adjectives are conjugated, as in talk > talks > talked.

A summer of Hittite

Subway Hittite is a new blog that will document how I am teaching myself Hittite in the summer of 2013.  You may see me on the New York City subway--with a hittite textbook, or looking through flashcards with cuneiform on them and mumbling to myself.

Hittite has a lot in common with Greek and other ancient languages.  It's attested in cuneiform on clay tablets from the second millennium B.C. (more specifically, according to my textbook, "30,000 tablets and fragments" from about 1650-1180 B.C.).  The books I am using are Theo van den Hout, The Elements of Hittite (2011)¹ and Harry Hoffner and Craig H. Melchert, Grammar of the Hittite Language (2008).²  I hope as I go through the books and write this blog to be able to make suggestions about how best to study the bits and pieces of Hittite, and maybe even what they have in common with the old standard, Johannes Friedrich's sonorously titled Hethitisches Elementarbuch (2d ed. 1960).  I'm on the way to borrow it from the library right now.  Also available, and in my possession, is the 1987 Beginning Hittite by Warren H. Held, Jr., et al., which received negative criticism when it appeared³ and which I don't expect to use much.

Hittite looks like fun, especially because its vocabulary and grammar remind one of other languages that are far better known and attested only much later.  Famously, watar in Hittite means "water" (Greek ὕδωρ, "HOO-dor"); van den Hout early introduces the verb išpantḫi, "I make a libation," which is related to the Latin spondeo.  But more is to come on all this--first impressions can deceive:  My impression that the Hittite antuḫšas, "man," might be related to the Greek ἄνθρωπος (ANT-hropos) was not correct.  The etymological dictionaries of Hittite and of Greek (here and here) show there's no good clue to where ἄνθρωπος comes from; the Hittite etymological dictionary finds that antuḫšas is closely related to the Greek ἔνθεος, and probably means (as the Greek word more or less does) filled with some sort of spirit.

1. Reviewed by Ilya Yakubovich, Classical Rev. 63:1 (2013) and by Mark Weeden, Antiquity issue 336 (2012).

2. Reviewed by Zsolt Simon, J. of Near Eastern Studies 71:338 (2012) and by Philomen Probert, Bryn Mawr Classical Review 2009.05.49.

3. Norbert Oettinger, Historische Sprachforschung / Historical Linguistics 105: 307 (1992); Gary Beckman, J. of the American Oriental Society 111: 658 (1991); Andrew Garrett, Language 67: 402 (1991).