[5] The languages are known from manuscripts dating from the 5th to the 8th century AD, which were found in oasis cities on the northern edge of the Tarim Basin (now part of Xinjiang in Northwest China) and the Lop Desert.

Scholars studying these manuscripts in the early 20th century identified their authors with the Tokharoi, a name used in ancient sources for people of Bactria (Tokharistan).

A body of loanwords and names found in Prakrit documents from the Lop Nur basin have been dubbed Tocharian C (Kroränian).

The result was a new hypothesis, following the wave model of Johannes Schmidt, suggesting that the satem isogloss represents a linguistic innovation in the central part of the Proto-Indo-European home range, and the centum languages along the eastern and the western peripheries did not undergo that change.

Most scholars reject Walter Bruno Henning's proposed link to Gutian, a language spoken on the Iranian plateau in the 22nd century BC and known only from personal names.

Tocharian B mit), cognate with Old Church Slavonic медъ (transliterated: medŭ) (meaning "honey"), and English mead.

[18] A colophon to a Central Asian Buddhist manuscript from the late 8th century states that it was translated into Old Turkic from Sanskrit, via a twγry language.

In 1907, Emil Sieg and Friedrich W. K. Müller proposed that twγry was a name for the newly-discovered language of the Turpan area.

[23] Sieg and Müller, reading this name as toxrï, connected it with the ethnonym Tócharoi (Ancient Greek: Τόχαροι, Ptolemy VI, 11, 6, 2nd century AD), itself taken from Indo-Iranian (cf.

Ptolemy's Tócharoi are often associated by modern scholars with the Yuezhi of Chinese historical accounts, who founded the Kushan Empire.

[29][30] In 1938, Walter Bruno Henning found the term "four twγry" used in early 9th-century manuscripts in Sogdian, Middle Iranian, and Uighur.

Tocharian is documented in manuscript fragments, mostly from the 8th century (with a few earlier ones) that were written on palm leaves, wooden tablets, and Chinese paper, preserved by the extremely dry climate of the Tarim Basin.

[37][38] It soon became apparent that a large proportion of the manuscripts were translations of known Buddhist works in Sanskrit and some of them were even bilingual, facilitating decipherment of the new language.

In 1998, the Chinese linguist Ji Xianlin published a translation and analysis of fragments of a Tocharian Maitreyasamiti-Nataka discovered in 1974 in Yanqi.

For example, both languages show significant innovations in the present active indicative endings but in radically different ways, so that only the second-person singular ending is directly cognate between the two languages, and in most cases neither variant is directly cognate with the corresponding Proto-Indo-European (PIE) form.

Likewise, some of the verb classes show independent origins, e.g. the class II preterite, which uses reduplication in Tocharian A (possibly from the reduplicated aorist) but long PIE ē in Tocharian B (possibly related to the long-vowel perfect found in Latin lēgī, fēcī, etc.).

The texts were written in Gandhari Prakrit, but contained loanwords of evidently Tocharian origin, such as kilme ("district"), ṣoṣthaṃga ("tax collector"), and ṣilpoga ("document").

[47] In papers published posthumously in 2018, Klaus T. Schmidt, a scholar of Tocharian, presented a decipherment of 10 texts written in the Kharoṣṭhī script.

[49] In 2019 a group of linguists led by Georges Pinault and Michaël Peyrot convened in Leiden to examine Schmidt's translations against the original texts.

They concluded that Schmidt's decipherment was fundamentally flawed, that there was no reason to associate the texts with Krörän, and that the language they recorded was neither Tocharian nor Indic, but Iranian.

In that sense Tocharian (to some extent like the Greek and the Anatolian languages) seems to have been an isolate in the "satem" (i.e. palatovelar to sibilant) phonetic regions of Indo-European-speaking populations.

When referring to humans, the oblique singular of most adjectives and of some nouns is marked in both varieties by an ending -(a)ṃ, which also appears in the secondary cases.

As in Sanskrit, Ancient Greek, and (to a lesser extent) Latin, there are independent sets of classes in the indicative present, subjunctive, perfect, imperative, and to a limited extent optative and imperfect, and there is no general correspondence among the different sets of classes, meaning that each verb must be specified using a number of principal parts.

The imperative likewise shows 6 classes, with a unique set of endings, found only in the second person, and a prefix beginning with p-.

The following shows the thematic endings, with their origin: In traditional Indo-European studies, no hypothesis of a closer genealogical relationship of the Tocharian languages has been widely accepted by linguists.

[66][67][68] As an example, the same Proto-Indo-European root *h₂wrg(h)- (but not a common suffixed formation) can be reconstructed to underlie the words for 'wheel': Tocharian A wärkänt, Tokharian B yerkwanto, and Hittite ḫūrkis.

Michaël Peyrot argues that several of the most striking typological peculiarities of Tocharian are rooted in a prolonged contact of Proto-Tocharian with an early stage of Proto-Samoyedic in South Siberia.