Old Tibetan refers to the earliest attested form of Tibetan language, reflected in documents from the adoption of writing by the Tibetan Empire in the mid-7th century to the early 9th century. In , during the reign of Tibetan King Sadnalegs, literary Tibetan underwent comprehensive standardization, resulting in Classical Tibetan.
Old Tibetan is characterised by many features that are lost in Classical Tibetan, including my- rather than m- before the vowels -i- and -e-, the cluster sts- which simplifies to s- in Classical Tibetan, and a reverse form of the vowel letter for i (gi-gu). Aspiration was not phonemic and many words were written indiscriminately with consonants from the aspirated or unaspirated series. Most consonants could be palatalized, and the palatal series from the Tibetan script represents palatalized coronals. The sound conventionally transcribed with the letter à ½ (Wylie: 'a) was a voiced velar fricative, while the voiceless rhotic and lateral are written with digraphs à ½§à ¾² and à ½£à ¾· . Unlike virtually all modern Tibetan languages, the Old Tibetan orthography did not contain silent letters, and the words were pronounced as written.
The following table is based on Hill's analysis of Old Tibetan:
In Old Tibetan, the glide occurred as a medial, but not as an initial. The Written Tibetan letter à ½ w was originally a digraph representing two Old Tibetan consonants .
Vowel Phonemes of Old Tibetan
In Old Tibetan, syllables can be quite complex with up to three consonants in the onset, two glides, and two coda consonants. This structure can be represented as , with all positions except C<sub>3</sub> and V optional. This allows for complicated syllables like à ½Âà ½¦à ¾Âà ¾²à ½²à ½Âà ½¦ "arranged" and à ½ à ½Âà ¾²à ¾ <nowiki/>'drwa "web", for which the pronunciations and can be reconstructed.
A voicing contrast only exists in slot C<sub>3</sub> and spreads to C<sub>1</sub> and C<sub>2</sub> so à ½¦à ¾Âà ½¼ sgo "door" would be realized as while à ½¦à ¾Âà ½´ "body" would be . Final consonants are always voiceless e.g. à ½ à ½Âà ½²à ½Âà ½Âà ¼ <nowiki/>'dzind and à ½Âà ½Âà ½´à ½Âà ½¦à ¼ []. The phoneme in C<sub>1</sub> was likely realized as (or when C<sub>3</sub> is voiced) e.g. à ½Âà ½¦à ¾Âà ¾²à ½º and à ½Âà ½¢à ¾©à ½²à ½¦ . The features of palatalization and labialization can be considered separate phonemes, realized as glides in G<sub>1</sub> and G<sub>2</sub> respectively. Only certain consonants are permitted in some syllable slots, as summarized below:
<sup>ç</sup> In C<sub>2</sub> position, and are in complementary distribution: appears before , , , , , , , and in C<sub>3</sub>, while appears before , , , , , and in C<sub>3</sub>. Additionally, is written before .
Palatalization was phonemically distinct from the onset cluster . This produces a contrast between à ½Âà ½¡ and à ½Âà ¾± , demonstrated by the minimal pair à ½Âà ½¡à ½Âà ¼ g.yaá¹ "sheep" and à ½Âà ¾±à ½Âà ¼ gyaá¹ "also, and". The sounds written with the palatal letters à ½ c, à ½ j, à ½ ny, à ½ zh, and à ½¤ sh were palatalized counterparts of the phonemic sounds à ½ ts, à ½ dz, à ½ n, à ½ z, and à ½¦ s.
Case markers are affixed to entire noun phrases, not to individual words (i.e. Gruppenflexion). Old Tibetan distinguishes the same ten cases as Classical Tibetan:
However, whereas the locative, allative, and terminative gradually fell together in Classical Tibetan (and are referred to the indigenous grammatical tradition as the la don bdun), in Old Tibetan these three cases are clearly distinguished. Traditional Tibetan grammarians do not distinguish case markers in this manner, but rather distribute these case morphemes (excluding -dang and -bas) into the eight cases of Sanskrit.
Old Tibetan transitive verbs were inflected for up to four stems, while intransitive verbs only had one or two stems. In the active voice, there was an imperfective stem and a perfective stem, corresponding to the Classical Tibetan present and past stems respectively. Transitive verbs also may have two passive voice stems, a dynamic stem and stative stem. These two stems in turn correspond to the Classical future and imperative stems.
Old Tibetan has three first person singular pronouns à ¼ á¹ a, à ¼ , and à ¼ , and three first-person plural pronouns à ¼ , à ¼ , and . The second person pronouns include two singulars and -'da' and a plural .