As we have seen, Central Asia in general, and the Tarim Basin in particular, have provided the backdrop for the rise and fall of numerous empires and for the ebb and flow of a varied system of trade. In such a fluid and vibrant region, one naturally expects the confluence of numerous peoples of varied backgrounds, and this is exactly what appears in the historical record. In particular the historical documents of the Chinese Han period provide descriptions of a wealth of cultures inhabiting the general region, many of whom scholars are still trying to correlate with known peoples from archaeological or other historical records. We take a moment here to discuss them in broad outline.
Chinese records of the Han Dynasty mention a people called the Xiongnu. If they may indeed be linked with the northern 'barbarians' mentioned in slightly earlier documents, then this nomadic people proved a constant threat to the Chinese state from the 8th century BC onwards. They only enter the written record by name in the 3rd century BC. This culture predominantly raised horses, cattle and sheep, and to a lesser extent other animals such as camels and mules. They were a fearsome group that believed it was every male's duty to fight in war and to repose in relative leisure otherwise. They were skilled horsemen fond of the bow and arrow, and employed iron swords in close combat. They pressed those captives who escaped the sure release of death into indentured servitude. The Xiongnu early inhabited regions to the north of the Chinese, and followed a lifestyle still reflected in modern Mongolian culture, with the felt-walled yurts and tending of livestock. From these northerly regions they descended upon the Chinese with such fury that the latter built barricades to repel their assault. These were later joined in the 3rd century to form the 3,000-mile-long structure we today recognize as the Great Wall. This channeled their expansion toward what the Chinese termed the Western Regions, where they took to the eastern edge of the Tarim Basin. Here among other things they found the great iron deposits needed to produce their weapons. The Xiongnu presence in the Western Regions prompted the initial westward expansion of the Chinese empire in the 2nd century BC. The Chinese overpowered the Xiongnu for a time, but the latter rose again to assert their dominance. Power in the region continued to shift between the two, until finally in the 1st century AD the Han dynasty was able to effectively drive the Xiongnu from the region.
The Xiongnu do not seem to have entered the Tarim Basin in peace, but rather subjugated the Wusun and Yuezhi upon arrival. The Wusun apparently inhabited a region to the northeast of the Tarim Basin, likely in the Ferghana region, perhaps near Lake Balkhash. They formed a remarkably large population: 120,000 households, comprising 630,000 individuals, of which 188,800 could bear arms. In this early period the Chinese pronunciation of their tribal name was likely *osǝn or *uo-suǝn, leading some scholars to associate them with Asiani mentioned by Pompeius Trogus in the 1st century BC, or with the Issedones mentioned even earlier by Herodotus (Mallory & Mair, 2000). The Wusun evidently made their presence felt over a large expanse, since one Wusun band destroyed a Chinese outpost in the 1st century BC which some scholars suggest might have been located near Lopnur. Early in the second century BC the Wusun remained under Xiongnu dominion, but by the end of the first century they had shaken off Xiongnu rule. To further their own interests, the Chinese entreated the Wusun to settle ever more toward the east, thereby occupying the land claimed by the Xiongnu and thereby forcing the latter out of the region. The Xiongnu quickly understood the motives of the Chinese and commenced entreaties of their own with the Wusun.
One particularly interesting mention of the Wusun occurs in the work of Yan Shigu (AD 579--645): 'Of all the Rong of the Western Regions the Wusun looked the most peculiar. Those of the present Hu [Barbarians] who have cerulean eyes and red beards and look like Mi monkeys are their descendents.' (Mallory & Mair, 2000) This description is often interpreted as identifying European physical characteristics as viewed through Chinese eyes.
The Yuezhi fell victim to the aspirations of the Xiongnu and their alliance with the Wusun. As the Xiongnu, assisted by the Wusun, descended into the Gansu, the region east of the Tarim Basin, the Yuezhi fled west. In their retreat they split into two groups, Great and Lesser. The latter pushed south and was assimilated into the Qiang culture of the Tibetan region. The Great Yuezhi however were driven by the Wusun through Ferghana, until the Yuezhi finally impinged on Sogdiana and in turn drove the Sakas into Bactria. The Yuezhi themselves finally pushed into Bactria near the close of the second century BC. They divided the region into five provinces. One of these subsequently expanded to dominate the remaining states: this province was known as Guishuang, the Chinese rendering of Kushan (Kuśāṇa).
At the time the great annals of the Han Dynasty, the Hanshu, were composed, the Yuezhi inhabited the region north of the Oxus river and west of Ferghana. The capital Jianshi contained some 100,000 households, comprising 400,000 individuals, of which 100,000 could bear arms. They were thus a force to be reckoned with, and they exercised considerable influence in the Tarim Basin. The Chinese accounts say that, though originally nomadic, their cultural practices were similar to those of the Parthians. The people were evidently characterized by a 'reddish-white color' of the skin, possessed numerous horses, and were skilled in horse-mounted archery. (Mallory & Mair, 2000) The Chinese had already successfully requested their influence in the first and second centuries AD in preventing Sogdian assistance in Kašghar and in attacking Turfan. As the Yuezhi came to inhabit Bactria, their empire saw its greatest extent under the rule of Kaniṣka: from the Aral Sea and Indus river in the west to the Ganges, Kashmir, and the frontier of Xinjiang in the east. The culture was profusely cosmopolitan: coins depicted Greek, Iranian and Indian gods; the king enjoyed such titles as Indic mahārājasa 'great king', Iranian rājatirājasa 'king of kings', a translation of a Chinese epithet devaputrasa 'son of heaven', and even kaïsarasa 'Caesar'. (Mallory & Mair, 2000) Though Zorastrianism, Jainism and Hinduism all coexisted within the empire, the Kushan empire in addition provided the vehicle for the Great Vehicle, Mahayana Buddhism, to spread into the Tarim Basin and beyond into China and ultimately Japan. Through this religious influx Buddhist monks and the Prākrit language spread throughout the region, carrying along with them a distinctive art that united a range of cultural influences. As Yuezhi influence declined over the course the 4th century AD, a Turkic people known as the Hephthalites invaded in the fifth century and pushed their conquests as far south as India. They held the remnants of the Kushan empire for the better part of a century, until they too were dislodged.
The Hephthalites were eventually displaced by Turks invading in the mid-sixth century AD. This brought them face to face with the Sogdians, whose homeland Sogdiana was situated between the Amu Darya and Syr Darya, with capital in Samarkand. We have already seen that the Sogdians gave none other than Alexander the Great a run for his money during his campaigns in the region. In the intervening time the Sogdians had fractured into a number of smaller states, but they maintained an entrepreneurial spirit which led them to dominate commerce along the Silk Road. Their Iranian-derived language provided the lingua franca along the Silk Road for much of the Middle Ages. (Mallory & Mair, 2000)
By the middle of the seventh century AD the Chinese had largely diminished the strength of the Turks from Samarkand eastwards. But in Mongolia a series of revolts culminated in the seventh century with the Turkic Uyghurs presenting the Chinese emperor with the head of the last king of the Turks. The Uyghurs themselves were of Turkic linguistic heritage. They adopted Manicheism as the state religion in the mid-eighth century, but a century later Kirghiz incursions forced them from their seat of power in Mongolia. Some fled east to China, some west to the Gansu, establishing a small state that lasted until 1028; and still others fled to the Tarim and Turfan. There they set aside Manicheism for Buddhism, and they flourished in a multicultural empire until the thirteenth century.
It is against this rich tapestry of varied ethnicities and languages that we find the Tocharian documents.
The following excerpt continues the Tocharian B text B107 (THT 107) from the previous lesson.
We see in line 18 of this selection the use of the subjunctive in a subordinate clause denoting a general characteristic: B se ññissa śpālmeṃ tākaṃ cwi aiścer 'Will you give (them) to him who would be better than me?', or in more colloquial English, 'Will you give them to whoever is better than me?' or '... to someone better than me?' The subjunctive in this context signals that there need not be any specific person in mind: if there is any sage better than Indra, whoever it be, give the alms to him.
Note also in line 18, as well as previously in line 13, the form B weñā-me-ś. We see here the secondary case ending applied to enclitic pronouns. Moreover this provides an illustration of the use of the allative with verbs of speaking to denote the person addressed.
This passage also employs several times the term B ylaiñäkte. This term literally means 'king-god' and is used to refer to the god Indra. The second member of the compound is clearly -ñäkte 'god, divine being', also found in this passage in B Bra(h)m-näkte 'brahma-god'. The first member ultimately derives from the same root as A wäl B walo 'king'. See Section 47 of this lesson for further details.
17 - Indre kārpa rṣākäññe weṣ myāskate stām ñor cau lmoṣ lyakāre. śilāre-ne oṅkarñai wñār-ne purwar wesanmeṃ pinwāt rṣāka.
18 - snai epiṅkte bramñikte kārpa totka maṃtstsaś aśrāmne peñiyacce yaknesa lyama ylaiñikte rṣākäññe weṣ memisku weñā-me-ś ṣerśkana ñi aiścer ce pintwāt epe se ññissa śpālmeṃ tākaṃ cwi aiścer.
19 - Nānda Nandābala weñāre se cisa śpālmeṃ tākaṃ cwi aiskem. ylaiñikte bramñikteś mant ṣerpsa-me weñā-me-ś ṣerśkana sam rṣāke ñissa śpālmeṃ ste cwim nai kalas.
20 - toy kakkāccuwa bramñikteś maitare wināṣṣar-ne oṅkarñai ṣarnene eṅkuwa weskeṃ-ne-ś.
17 Indre kārpa rṣākäññe weṣ myāskate stām ñor cau lmoṣ lyakāre. śilāre-ne oṅkarñai wñār-ne purwar wesanmeṃ pinwāt rṣāka.
18 snai epiṅkte bramñikte kārpa totka maṃtstsaś aśrāmne peñiyacce yaknesa lyama ylaiñikte rṣākäññe weṣ memisku weñā-me-ś ṣerśkana ñi aiścer ce pintwāt epe se ññissa śpālmeṃ tākaṃ cwi aiścer. 19 Nānda Nandābala weñāre se cisa śpālmeṃ tākaṃ cwi aiskem. ylaiñikte bramñikteś mant ṣerpsa-me weñā-me-ś ṣerśkana sam rṣāke ñissa śpālmeṃ ste cwim nai kalas. 20 toy kakkāccuwa bramñikteś maitare wināṣṣar-ne oṅkarñai ṣarnene eṅkuwa weskeṃ-ne-ś.
17 Indra descended (and) took the guise of a sage (and) those seated under the tree saw him. They brought him the porridge (and) said to him, "Partake of our alms, sage." 18 Suddenly the brahma-god descended (and) a little below he sat in the ashram in splendid manner. Indra, having taken the guise of a sage, said to them: "Sisters, will you give me these alms or will you give (them) to him who would be better than me?" 19 Nanda (and) Nandabala said, "We will give (them) to whoever would be better than you." So Indra motioned to the brahma-god (and) said to them, "Sisters, this sage (here) is better than me; give (them) to him." 20 Rejoicing, they went to the brahma-god (and) worshipped him. Having taken the porridge in their hands they speak to him.
Tocharian possesses an indefinite pronoun A saṃ B ksa 'who-, which-, whatever' or 'who-, which-, whatsoever'. As with its cousins the interrogative and relative pronouns, the indefinite pronoun is mercifully indifferent to distinctions of gender and number. For example, A saṃ B ksa refers to a noun in the nominative, be the referent masculine or feminine (or neither), or be it singular or plural. The following chart lists the full, but brief, set of forms.
The two Tocharian languages differ slightly in their usage of the pronoun. In Tocharian A, the indefinite typically follows the negative particle (A mā saṃ), a demonstrative (compare Gk. hós-tis), or the pronominal adjective A ālak 'other' (A ālak saṃ 'someone, anyone', similar to Lat. ali-quis 'someone, anyone'). In Tocharian B, the indefinite typically follows a relative or interrogative pronoun (B kuse ksa (nom.), kuce kca (obl.), etc.) or a demonstrative (B su ksa).
The forms of the indefinite in Tocharian pose some difficulties in terms of their historical evolution. In Tocharian A, the situation appears straightforward: the nominative and oblique forms reprise those of the demonstrative, while the genitive is identical to that of the interrogative. The Tocharian B genitive form B ket(a)ra likewise stems from the interrogative, with the addition of a final particle B -ra 'also'. The form B ketra generally appears in poetic language, while B ketara, derived from the accent falling on the anaptyctic vowel PToch *ketärā, generally appears in prose.
The nominative and oblique forms in Tocharian B bear an obvious similarity to the corresponding forms of the interrogative (B kuse, kuce, cf. A kus, kuc), but with a conspicuous difference in final vocalism and a lack of the labial element u of the initial consonants. As a consequence, the origin of these forms remains somewhat obscure. One proposed derivation looks to the PIE antecedent of generalizing phrases such as Lat. quis-quis and Hitt. kuiš-kuiš 'anyone' (Pinault 2008): PIE *kʷi-s kʷi-s. If we recall (Section 41, Lesson 9) that the form *kʷi-s likely reflects a collocation of an interrogative stem *kʷi- and the deictic *so-, then we may suppose an archaic phrase something like *kʷiso kʷiso. If stressed as a unit, we might expect *kʷisó kʷisò. The result in Proto-Tocharian would be PToch *kʷyäsæ kʷyäsæ > *kwäsæ ksæ, with stress on the vowel in boldface. This would further result in *kwäsæ ksæ > *kwäsæ ksā if we assume that the final unaccented vowel evolves along the same lines as the negative particle PIE *meH₁ > *mē > PToch *mæ > AB mā; this would be a rule peculiar to monosyllables (contrast Section 11.5.4, Lesson 3). Such a sequence of changes would finally yield PToch *kwäsæ ksā > B kusé ksà, from which we could imagine that speakers isolated the form ksa and extended it to other contexts.
Tocharian declension Class VII contains but a single noun: A wäl B walo 'king'. This noun ultimately derives from the Proto-Indo-European root *welH- 'be strong'. The paradigm shows two different stems: PToch *wälõ in the nominative singular, *lānt- elsewhere. Even the oblique stem, however, derives from prior *wlānt-, since PToch *w- regularly falls away immediately preceding a liquid. The following chart gives the paradigm, together with the historical evolution of the particular forms.
Note in the above paradigm that Tocharian A has generalized the palatalized stem *lāñcä proper to the nominative plural to the remaining plural forms.
The nominative forms have provided a major source of scholarly controversy. This arises in part because we final in nominal compounds the forms A wlāṃ- B *yälai-: e.g. A ylai-ñäkte B wlāṃ-ñkät, literally 'king-god', where both elements are generally taken to be nominative singular. To account for the alternation A w- B y-, we suppose the Proto-Tocharian form contained a palatalized *wy-, and so the form likely derives from a full grade of the root: PIE *welH-. Consider the following.
|2. *welḤ-ōn(ts)||*wyäl-ōn||*wlõ||wäl||wlo, wálo|
In (1a) we see the development based on the full grade of the root and zero grade of the suffix. Alternatively, in (1b) we may see the suffix as imported from the accusative. As first member of a compound we expect the accent to fall on the second syllable; either development yields the forms we find in composition. Note the typical evolution *-ān(ä) > B -ai in an accented syllable. In (2) we see that, if we admit the full grade of the root, then we must nevertheless have in the Proto-Tocharian period depalatalization of the initial *wy-. This would follow upon deletion of the reduced vowel *ä, leaving in Tocharian B a form wlo. This latter we do in fact find as a common by-form in poetry, alongside the form B wálo with accent on the PToch reduced vowel.
We find derivatives of this word based on the stem *lānt-. In particular, the word for 'queen' follows a straightforward feminine formation: PIE *wlḤ-nt-iH₂ > PToch *wlā-nt-yā > *wlāntsā > A lānts B lāntsa. We also find A lāñci B lantuññe 'royal', among others.
Tocharian possesses one infinitive ending, identical in the two languages: AB -tsi. The formation, however, differs slightly between Tocharian A and Tocharian B: Tocharian A appends the infinitive suffix -tsi to the present stem, while Tocharian B applies it to the subjunctive stem. The form remains invariant regardless of voice, active or mediopassive.
Inasmuch as the infinitive is a verbal noun, it may take secondary case endings. In Tocharian A we only find the infinitive with the allative suffix, and rarely at that. Tocharian B is more flexible in this regard, employing the infinitive with case endings for allative, perlative, locative, and even genitive. Consider the following examples:
As the invariable infinitive suffix -tsi does not distinguish voice, the resulting verbal noun may be construed with either an active or mediopassive sense. For example:
The Tocharian languages also employ the infinitive as a predicate. Consider
The infinitive, even though built to the base stem of a verb, may nevertheless occasionally be construed as causative. Take the following examples:
We generally find phonetic changes where the infinitive suffix encounters a root-final sibilant, *-s-tsi > -ssi. When the stem-final sibilant results from the palatalization of one of the present or subjunctive class endings, we find depalatalization: *-ṣṣ(ä)-tsi or *-ṣ(ä)-tsi both yield *-s-tsi > -ssi. In Tocharian B, however, the affricate or palatalization may reappear. We find a similar loss of palatalization, and gemination, with present and subjunctive suffixes -ññä- and -ñä-. Consider the following examples.
|PToch||AB||Example||PToch||Toch A||Toch B||Meaning|
|*-s-tsi||-ssi||*wäs-tsi||wassi||'to put on', also 'clothing'|
As we see in the last example, where Tocharian B shows depalatalization *täṅkwäññ-tsi > B tänkwantsi, Tocharian A occasionally restores the original palatalized nasal: *tuṅkiññ-tsi > A tuṅkiñtsi.
In the written documents, we often find that later texts show a simplification of the infinitive suffix -tsi to -si. For example, for the root śu-/śwā- 'eat' we often find infinitive AB śwāsi instead of the expected AB śwātsi.
The specific origin of the infinitive ending remains somewhat unclear, though the source most likely derives from a PIE nominal formation in an oblique case such as the dative, a sense reflected in English by the to of infinitives such as 'to do' or 'to speak', and analogous to what we find in others of the early IE languages. One common argument takes as starting point a nominal formation in PIE *-ty- with dative ending *-ey; another argument supposes an original *-dʰyōy, where the *-ō- evolves to *-u as found before some finals (cf. Section 11.4.4, Lesson 3), and then the genitive-dative ending *-i is imported from the nominal paradigms during the Proto-Tocharian period. The following chart summarizes the evolution.
|*-dʰyōy||*-tsu + -i||*-tsäi||-tsi||-tsi||Ved. -dʰyai|
Certain verbs in Tocharian do not superficially conform to the patterns outlined in previous lessons for the formation of the various tenses and moods. For this reason, we term these verbs 'anomalous', or sometimes 'irregular'. But as we have seen over the past several lessons, the entire process of historical linguistics can be viewed as one whereby the scholar employs seeming irregularities in a language at a given point in time to reconstruct actual regularities which obtained at an earlier point of the same language. This in a nutshell is the process of linguistic reconstruction. Viewed in this light, 'irregular' verbs a no longer truly 'irregular'; they are simply no longer regular. To the linguist, they are precious jewels whose glimmer represents earlier hoards of treasure. Below we unearth some of the more precious treasures concealed within the Tocharian languages.
When a particular verb occurs in a language with great frequency, we quickly learn its forms by rote: the continual exposure to the verb allows us to correct our mistakes. For this reason in English we are able to assimilate the forms of the verb 'to be': am, is, are, was, were, be, been. When a verb occurs less frequently, its paradigm often shifts as speakers forget or avoid previous forms and must invent them anew: is the past tense of hang really hung, or is it hanged? Sometimes the process of invention occurs by adopting a pattern current elsewhere in the language. For example, compare OE byrnan, with preterite beorn, to its modern reflex burn, with preterite burned; as the inherited strong preterite (beorn) was eventually replaced by a form (burned) built according to the more prevalent and productive weak preterite paradigm. In other instances, however, where speakers find forms of a certain verb deficient for some reason, often they co-opt forms of another verb with a related meaning. Look for example at the some forms for the present and preterite in Modern (New) English, Old English, German, Latin and Sanskrit as listed in the following chart.
|1 Sg.||am||eom, bēo||bin||sum||asmi|
|1 Pl.||are||sind(on), bēoþ||sind||sumus||smaḥ|
In both Latin and Sanskrit we find full paradigms for the PIE root *H₁es- (*Ø-grade *H₁s-): this yields Lat. es- (Ø-grade s-) and Skt. as- (s-). The lengthened ā- in Sanskrit derives from an additional augment, PIE *(H₁)e- > Skt. a-, coalescing with the initial vowel of the root. But as we see in German and Old English, forms with initial b- have begun to make their presence felt. In fact, at one point in the history of English, evidently there were many available forms for this root. Ultimately this derives from a separate PIE root *bʰuH 'be, become', which invades the Latin paradigm elsewhere (e.g. perfect fuī 'I was'), and which retains a completely independent paradigm in all tenses in Sanskrit (e.g. present bʰavāmi 'I become'). Moreover German and English show importation of yet another root, with initial w-, in the preterite. This ultimately derives from a completely different PIE root *wes- 'dwell, live', preserved in certain Latin nominal forms (e.g. Ves-ta 'goddess of the hearth', astus 'craft (practiced in a town)'), and as a completely independent verbal paradigm in Sanskrit (e.g. vásati 'dwells').
Scholars employ the term suppletion to denote this process of adopting forms from other roots to fill gaps left in the paradigm of a given root. The resulting hodge-podge of forms culled from various roots to correspond to a given meaning is sometimes termed the verbal complex corresponding to that meaning. Thus, in English, 'to be' is a suppletive verb whose verbal complex consists of the forms as, is, are, was, were, be, been, originally spliced together from the three different PIE roots *H₁es-, *wes-, *bʰuH-. The process of suppletion occurs in languages the world over and is not unique to English. What remains unique to a particular language, however, is the particular root or roots affected, and the manner in which the verbal complex is filled out.
Tocharian naturally exhibits the results of suppletion in numerous verbal complexes. The most common situation encountered is one in which two roots combine to fill out the verbal complex corresponding to a certain basic sense or meaning. As we see above with English, where forms divide along the lines of tense, we also find a certain regularity in Tocharian concerning where a given root picks up after another leaves off. Rather than finding a division along the lines of tense, however, we typically find a split along the lines of verbal aspect (see Section 45 of Lesson 9 for a fuller discussion of verbal aspect). In particular, in a Tocharian suppletive verbal complex, one root typically provides forms with imperfective aspect (hence present), while the other provide perfective forms (hence preterite and subjunctive). Thus the most prevalent situation is one in which two roots combine to fill out a given verbal complex. The following list provides the most salient examples.
|Meaning||Imperfective Stem||Perfective Stem|
|lead||AB āk-||AB wā(y)-|
|stand||AB käly-||A ṣtäm- B stäm-|
|sit||AB ṣäm-||AB läm-|
|eat||AB śu-/śwā-||AB tāp(p)-|
|drink||AB yok-||AB tsuk-|
|see||AB läk-||AB pälk-|
We also encounter in Tocharian suppletion of three roots into a single verbal complex. The aspectual split between roots in such instances is not necessarily so clear-cut. The following chart provides examples.
|give||A e-||A e-||A wäyā-|
|B ai-||B ai-||B wäyā-||B *t(s)-|
|carry||A pär-||A kām-||A kām-||A kām-|
|B pär-||B ās-||B kām-||B ās-|
As the above chart shows, the particular interplay between suppletive stems at times differs between Tocharian A and B. Of course Tocharian possesses a number of other important suppletive verbs, corresponding to the senses 'make, do', 'go' (cf. English go, went), and 'be' (cf. English am, is, are, was, were, be, been). We turn to a discussion of each of these below.
One of the single most common verbal stems in Tocharian is AB yām- 'make, do'. This stem appears only in the preterite and subjunctive in Tocharian A, the remainder of the verbal paradigm being filled by the stems A ya- and A ypa-. In Tocharian B, the stem B yām- occurs throughout the paradigm. The following chart provides the present, imperfect and imperative forms of this verbal complex for both Tocharian A and B.
|AB yām-||A Pres.||B Pres. IX||A Impf.||B Impf.||A Impt. iii||B Impt. iii|
The chart below provides the remainder of the paradigm, listing the forms for the subjunctive, optative and preterite.
|AB yām-||A Subj. ii||B Subj. i||A Opt.||B Opt.||A Pret. III||B Pret. IV|
In part the ubiquity of this verb results from its use in periphrastic denominative formations. That is, Tocharian very often employs the formula noun + yām- 'do (noun)' to derive a verb from a given noun. This parallels the use of the verbal root kṛ- 'do, make' in Sanskrit or kar-nā in Hindi to derive verbs from nouns: e.g. Hindi fon karnā 'to make a (tele)phone call, to call, to phone'. In Tocharian the resulting verb may be transitive or, less frequently, intransitive. The following chart lists some denominative verbs in Tocharian B employing B yām=.
|Part of Speech||Meaning||Denominative Verb||Meaning|
|ankaiṃ||adjective||false; reverse||ankaiṃ yām-||to vomit|
|apākärtse||adjective||manifest, apparent||apākärtse yām-||to be visible, be manifest|
|ārwer||adjective||ready||ārwer yām-||to prepare oneself|
|onmiṃ||noun||remorse, repentance||onmiṃ yām-||to repent|
|ārwer||adjective||ready||ārwer yām-||to prepare|
|kṣānti||noun||forgiveness||kṣānti yām-||to forgive|
|pākri||adjective||clear, obvious||pākri yām-||to make public|
|saim||noun||support, refuge||saim yām-||to take refuge in|
The compound verb noun + yām-, when transitive, typically takes a direct object in the oblique. With some such verbs, however, the object may be in a different case, most typically the genitive.
At present the etymology of AB yām- and the remaining stems in the verbal complex remains obscure. Scholars have proposed various etymologies for the pair ya-, ypa- (Pinault, 2008). We note that the stem ypa- occurs only in those positions in the paradigm where we would expect the PIE thematic vowel *-o- > PToch *-æ-. Given this, the evolution could perhaps follow from either PIE *yew- or PIE *yeH₁- as given below.
The root PIE *yew- has the meaning 'hold, draw to oneself, attach', while PIE *(H)yeH₁- means 'throw, put (by throwing)'. Either solution presents its own problems, and there is no shortage of alternate views: for example ypa- may derive from *pi-yeH₁- by metathesis (Adams, 1999). In any of these situations, however, scholars tend to view yām- as deriving from one of the above roots via a process of extension with *-m-, perhaps by analogy with other roots such as *H₂emH₃- 'take hold, seize' (cf. Lat. amā-re) which would also show initial *y- in Tocharian.
Another verb ubiquitous in the Tocharian languages is AB i- 'go'. Though typically listed in the dictionary under i-, the form *yä- in fact forms the basis for the majority of the paradigm. The following chart provides the present, imperfect and imperative forms.
|AB i-||A Pres. I||B Pres. I||A Impf.||B Impf.||A Impt. vi||B Impt. vi|
|1 Pl.||ymäs||ynem, ynemo||*yemäs||yeyem|
In Tocharian A the verb i- takes a suppletive stem kälk- in the subjunctive, optative and preterite. Tocharian B, by contrast, shows the suppletive stems mäs- and mit- in the preterite only. The verb AB mit- 'to get started' actually shows a full paradigm -- present, subjunctive and preterite -- in both Tocharian A and Tocharian B. The stem A kälk- may be related to the verb B kalāk- 'follow'. The following chart list the remainder of the paradigm for AB i-: subjunctive, optative and preterite.
|AB i-||A Subj. v||B Subj. = Pres.||A Opt.||B Opt. = Impf.||A Pret. I||B Pret. I/III|
Note in the above that the Tocharian B preterite form mas(s)a derives from preterite Class III, whereas the remainder of the preterite forms derive from Class I. Moreover, in Tocharian B the present forms double as subjunctive forms; there is no formal distinction between the two. This extends to the infinitive stem as well.
In Section 24.1, Lesson 5 we already met the present forms of AB i- as an illustration of straightforward athematic present conjugation in PIE and the result in Tocharian. We reprise here the Tocharian forms, together with the corresponding forms in Greek, Latin, and Sanskrit. In the two central columns labeled 'PIE', the leftmost provides the likely PIE progenitors of the Tocharian forms, while the rightmost the progenitors of (a majority or indicative sample of) the forms in other sister languages.
|AB i-||Toch A||Toch B||PIE||PIE||Greek||Latin||Sanskrit|
We have already seen that the second person singular forms pose difficulties in Tocharian (a possible source would be PIE *H₁i-Ø-tH₂e, with an ending imported from the PIE perfect). Likewise the third person singular forms, which possibly show in their endings the remnants of enclitic particles. We also encounter in Tocharian B a nasal formation: third person plural PIE *H₁i-no-nt > PToch *yänæn > B yaneṃ; similarly in the first person plural. This could perhaps even provide the source of the third person singular of Tocharian B, removing the need to invoke an enclitic particle in this instance. This nasal element also seems to appear in the second person dual imperative in Tocharian A: A pines < PToch *pä-yän-ais. We find this nasal suffixed form of this root preserved in both Latin and Hittite. The imperative likewise presents difficulties: *H₁i-dʰi should yield PToch *yäcä, rather than the *yäṣä required for A piṣ B paṣ. The plural form does seem to correspond, however, with the result in Tocharian A: PIE *H₁ité > PToch *p(ä)-yäcä > A pic.
Tocharian possesses a rather idiosyncratic manner of expressing the verb 'to be' if we consider A nas- B nes- 'be' in conjunction with the copula B ste, etc., discussed below in Section 49.4. Even within the paradigm of the verb listed as A nas- B nes- we find a suppletive paradigm. In particular, A nas- B nes- provides the stem solely in the present finite forms. The imperfect forms derive from a stem A ṣe- B ṣai-, while the complete remainder of the paradigm derives from the stem AB tākā-. The following chart lists the present, imperfect and imperative forms.
|nas-/nes-||A Pres. II||B Pres. I||A Impf.||B Impf.||A Impt. i||B Impt. i|
The chart below provides the remainder of the paradigm, listing the forms for the subjunctive, optative and preterite.
|nas-/nes-||A Subj. v||B Subj. v||A Opt.||B Opt.||A Pret. I||B Pret. I|
We note in the Tocharian A imperative forms the appearance of ṣ- preceding the initial t- of the stem. This provides our first clue as to the PIE origin of the stem tāka-. The most likely candidate is PIE *(s)teH₂ 'stand', with *s-mobile, in zero-grade and with the addition of a stem-final *-k- reminiscent of that found in the Greek perfect active: PIE *(s)tH₂-k- > PToch *tāk-. Note we find in the Tocharian A subjunctive an abbreviated stem tā-.
The imperfect stem A ṣe- B ṣai- most probably derives from the optative of PIE *H₁es- 'be'. With zero grade of the root and full grade of the optative suffix, we seem to find a secondary recharacterization with another optative suffix: PIE *H₁s-yeH₁- + -iH₁- > *H₁s-yē-iH₁- > PToch *syæ-yä-, yielding A ṣe- and B ṣai-.
Perhaps most problematic is the present stem: A nas- B nes-. Scholarly opinion remains divided over the ultimate origin of this root (Pinault, 2008; Adams, 1999). Some see the PIE root *nes- 'return home', as seen in Gk. néomai. The Tocharian forms would therefore reflect an original PIE *o-grade *nos-, elsewhere unattested as an athematic present. Others see in this stem the remnants of a locative 'be' verb, where the ubiquitous verb *H₁es- 'be' employs the locatival prefix *H₁eno- 'in', parallel to Gk. én-esti 'is in, is among, is present': we would have PIE *H₁eno-(H₁)s- > PToch *næs- > A nas- B nes-. This requires, however, that the initial laryngeal of the zero-grade root *H₁s- fall away, preventing an evolution *-o-H₁- > *-ō- > PToch *-ā-.
When Tocharianists refer to 'the copula', this refers less frequently to a class of verbs equating subject and predicate, and more often to a specific verb that appears solely in Tocharian B: B ste, star- 'be'. In both Tocharian A and B, one encounters equation of subject and predicated effected by simple juxtaposition, akin to modern English stereotypes of the contact language that arose between Native Americans and pioneers of European descent in the midwest United States: 'This man -- good.' Modern English always requires the use of a form of 'to be' or equivalent verb in such situations, e.g. 'This man is good'; but the Tocharian languages had no such requirement per se. Rather the employment of a verb of existence in such situations derives from other considerations, such as the desire to express
In each of these three instances, Tocharian A uses the appropriate form of A nas- to provide the copula. Tocharian B, however, differs in that it uses special forms for the copula solely in the present tense; outside of the present, Tocharian B likewise employs the appropriate forms of B nes-. The forms of the so-called copular present are listed in the following chart.
|B ste||B Pres.||B Pres. Suffixed|
|3||stare, skente||stare-me, skentar-ne|
As the above chart makes clear, these verb forms rarely appear outside of the third person. Moreover, different forms are employed when the copula is suffixed with an enclitic pronoun.
It will by now come as no surprise that the historical origins of these forms continue to invite scholarly debate (Pinault, 2008). Speakers of Spanish will be happy to know that one look-alike, Sp. estar 'to be', might also provide an etymological cousin: one hypothesis for the PIE origin of the Tocharian B copular present is PIE *steH₂- 'stand', which yields Lat. stare and later Sp. estar. According to this line of argument, the Tocharian forms B ste and stare would derive from the mediopassive forms of the root aorist, as below.
Note, however, that the shift *-ā- > a of the root vowel suggests the verb form itself would have been enclitic, an atypical situation. From here the forms skente, skentär- are viewed as dialectal innovations.
A second view suggests the copular present forms derive from PIE *H₁es- 'be' (here Spanish speakers may content themselves to know Sp. ser 'to be' would now be the cognate, though that is perhaps less useful as a mnemonic device). In particular, scholars suggest an origin in the zero-grade of this root, augmented by the ubiquitous *-sḱ- suffix in PIE. This parallels Gk. éskon and Lat. escit, escunt. The evolution would be as follows.
This view has the appeal of deriving all the forms at once, the first two from the PIE injunctive, the second two from the indicative.
Studies have shown (cf. Pinault, 2008, for discussion) that the form stare appears limited to those text found in the easternmost region (Turfan), specifically to texts which are later and popular. The form skente appears to be in more widespread use for the plural throughout the entire region of Tocharian B texts. This prompts a third view (Pinault, 2008), whereby skente remains a form inherited from PIE, but *skæntär > B skentär would arise as a secondary extension of this form modeled on verbs such as B mäsk- 'find oneself, be': *mäskæntär > B mäskentär.
As we have discussed elsewhere (Section 45, Lesson 9), tenses place an action in time relative to the moment of utterance. In that regard, they refer to factual assertions, since those actions actually exist in time. Such statements or assertions are said to be (in the) indicative. Tocharian contains three indicative verbal formations: the present, preterite, and imperfect. We discuss their uses below.
The present tense in Tocharian, naturally enough, represents an action contemporaneous with the moment of utterance. We also find a usage termed the historical present which is familiar to colloquial English. Specifically, a speaker of English when relating past events may opt to employ the present tense: as a stereotypical example (often imagined with a New York accent), "I was working yesterday and these two guys walk up to me and they say... ." We would expect, respectively, the past tense forms walked and said. This particular habit of speech typifies many Indo-European languages, particularly early languages such as Latin and Greek; Tocharian is no exception. Often grammatical accounts motivate this shift to the present tense by suggesting this is a rhetorical technique designed to portray the action as more vivid in the mind of the listener. Alternatively, it may harken back to a possible PIE habit to employ the least-marked verb form whenever a preceding form has made all the relevant information (tense, aspect, etc.) clear (Sihler, 1995). Consider the following examples in Tocharian:
Tocharian moreover employs the present tense in prohibitive clauses negated in Tocharian A by mar and in Tocharian B by mā. For example: A mar yat mar yat mar slākkär naṣt yaṃtrācāre 'Don't do (it)! Don't do (it)! Don't be downcast, mechanic!'; B mā traṅko yamas-ne 'Don't commit sin(s)!'
The imperfect tense in Tocharian fulfills two basic functions. The first role played by the imperfect is the marking of a past action with durative aspect. That is, the imperfect denotes an action occurring before the moment of utterance and which is viewed as ongoing, e.g. with no fixed beginning, no fixed end, or some definite span of time between beginning and end. (See Section 45, Lesson 9.) This parallels the English past progressive, as in He was walking home from the store. If we compare the English simple past -- He walked home from the store -- the latter views the action as a complete whole: begun, executed, and terminated. The person mentioned reached home. The former, expressed by was walking, does not have the same interpretation: the person mentioned may not have reached home, i.e. the verbal action as expressed by was walking does not necessarily include the endpoints of the action. The Tocharian imperfect parallels this construction. Consider the following examples:
The second role played by the imperfect is the marking of past repeated action. That is, the imperfect serves to denote a sequence of individual actions which occurred prior to the moment of utterance. These individual actions may themselves be perfective (instantaneous or complete) or imperfective (ongoing or unfinished); the imperfect has no particular connotation in this regard. An English parallel would be the verb phrase was shooting. The verb shoot itself is lexically perfective or punctual (there is no extended notion of 'shoot'; it is by the nature of the verb an action instantaneously over and done with). The simple past He shot denotes a past occurrence of this action as a complete whole. But when we employ in English the construction he was shooting, the combination of lexical punctuality with morphological durativity forces the listener to understand a sequence of past occurrences of the verb shoot. The Tocharian imperfect fills the analogous role in Tocharian. Consider the following examples:
The Tocharian preterite in many respects parallels the German perfect. The German morphological perfect has developed such that the same phrase Ic+h habe das Mädc+hen gesehen can have either of two possible interpretations: analogous to the English simple past I saw the girl, or analogous to the English perfect I have seen the girl. Tocharian, like German, employs the preterite in both of these senses. The former use embodies a past action with perfective aspect, i.e. a situation viewed as an indivisible whole and completed before the moment of utterance. The latter notion, that of the perfect, generally stands apart from the perfective-imperfective distinction: it denotes an action which occurred in the past but which has 'present relevance'. Often general linguistic references term this state of affairs an anterior. The fact that some part or sense of the past action persists until the moment of utterance leads nicely to an intersection with the notion of a stative formation: simply enough, a stative verb form denotes a state of being, such as be standing or be alive. General scholarly opinion regarding developments within PIE views perfects encountered in daughter languages such as Latin, Greek and Sanskrit as deriving from original statives in PIE: the connection lies in the fact that certain past formations, such as stood (up), imply a logical, naturally associated resulting state, e.g. be standing.
The Tocharian preterite may generally denote any one of the above situations: past perfective, perfect or anterior, stative. Interestingly, the Tocharisches Elementarbuch (Krause & Thomas, 1960-1964) enumerates categories based in part on terms related to the role in discourse. The categories maintain close ties to the past perfective, perfect, and stative. In particular, the past perfective lends itself to the relation of a sequence of past events, hence a narrative use. We also find the Tocharian preterite employed in stage directions: this parallels the English use of the morphological present. That is, English they leave does not denote an event happening contemporaneous with the moment of utterance; rather it denotes a general action: they leave (whenever the going gets tough), they leave (on the first of every month). English naturally employs this "tense" in stage directions inasmuch as these pertain to no specific time, but generally hold true for the life of the play. Tocharian employs the preterite, likely as a result of origins in the PIE stative, in the analogous role. The scholarly term constative denotes those actions, or more generally assertions, which relay information subject to interpretation as true or false. Consider the following examples.
Not all utterances contain facts or statements purported to be factual. Some utterances deal with hypothetical assertions, said to be non-indicative. Since these are hypothetical and therefore do not necessarily exist in the real world, it makes little sense to ask whether they definitely precede, follow, or are contemporaneous with the moment of utterance. Tense therefore does not apply in the strict sense. A similar situation applies with direct commands, which to not fall so easily onto the timeline of past, present and future. To handle such utterances Tocharian contains three so-called non-indicative moods: subjunctive, optative, and imperative. We discuss their uses below.
The subjunctive plays an important role in the Tocharian verbal system. Part of the ubiquity of the subjunctive derives from its use as a simple future tense. Tocharian contains no separate morphological formation for indicative statements concerning the future. We have seen that the present tense may denote future time, but Tocharian also employs the subjunctive without any modal connotations to denote future actions. This likely arises from the historical antecedents of subjunctive forms exhibiting perfective aspect. This use as a pure future occurs in both main clauses and subordinate clauses. Consider the following examples:
In addition to the above, Tocharian employs the subjunctive as a true non-indicative mood. That is, the subjunctive often represents verbal actions that do not treat of assertions of fact. The subjunctive may denote actions associated with suppositions, exhortions to action, prohibitions -- in general statements of general action (or inaction), not tied to any particular time relative to the moment of utterance. We find the following uses in main clauses:
The subjunctive also finds employment in subordinate clauses. The clauses often serve to delineate characteristics of a constituent of the main clause, or they might list the presumptive purpose of an action in the main clause. In particular they frequently denote a possible situation which must obtain for a certain real action or outcome to occur, and in this role they serve to denote actions in the subordinate ('if'-) clause of conditional ('if... then') statements. As such conditions outline general circumstances and not factual situations, they typically do not pertain to a specific moment in time. But when the subjunctive is involved, it typically signifies that the condition is to be checked for validity at or after the moment of utterance. Consider the following examples.
The Tocharian optative provides another non-indicative mood. As with the subjunctive, the optative does not represent actions as factual, but rather potential. In independent clauses, if the subjunctive carries a connotation of generality, expectation, and exhortation, the optative by contrast carries the connotation of possibility, wish, and prescription. The speaker typically employs an independent optative (i.e. one in a main clause) to mark a non-factual assertion that she wishes to be true, or not true. Consider the following examples.
Unlike the subjunctive, the optative has no generally agreed upon indicative use. At times however the distinction between an optative and an imperfect can become somewhat blurred, as in English (cf. Section 50.1.2 above).
Tocharian also employs the optative in subordinate clauses. Succinctly, these uses parallel those of the subjunctive, but with regard to an imagined time frame of applicability preceding the moment of utterance. That is, they refer to possible or general assertions which might have been true prior to utterance. In this regard, the decision to employ subjunctive or optative in a subordinate clause often rests on the time frame delineated by the verb of the main clause: what would have been a subjunctive subordinate clause with a present time frame in the main clause will often be rendered by an optative subordinate clause when the main clause shifts to a past time frame. Consider the following examples.
Tocharian employs the imperative in direct commands, as with English Do it!, Get out!, Break a leg!. (Note however the usage with the present tense outlined in Section 50.1.1 above.) Both languages employ the imperative in the second person, singular and plural (Section 44, Lesson 9). Moreover Tocharian B employs a dual form in the second person mediopassive. Though Tocharian usage of the imperative departs little from expectations one would have based on similarities with English or other Indo-European languages, we do nevertheless find a couple idiosyncrasies with a typical Tocharian flair: