The paper analyzes some of the 'Tangut' materials as contained in N. Witsen’s Noord en Oost Tartarye (1692, 1705) and represented by a brief Oirat vocabulary given as an illustration to the Tangut Letters. The Oirat words are written in Tibetan script, and the Oirat nature of the text can be clearly seen from the phonetic, morphological, and semantic properties. Therefore, it can be well considered to be the first Oirat-language monument recorded in Tibetan script. The language of the materials referred to by N. Witsen as 'Tangut' is largely Oirat and mirrors both bookish and conversational forms. This is confirmed by evidence as follows: 1) the author meticulously distinguishes between the uvular [χ] and the back [k] using the Tibetan ha and kha; 2) the final -n and short vowels of the last open syllable are retained, e.g., /nüdün/, /šüdün/, /kelen/, etc.; 3) there are such words as malaχai, γodosun, köböün in their Oirat-bookish (Clear Script, or Todo Bichiq) forms; 4) the fact that [e] is expressed through -ie- (emie, kelien, busie, etc.) testifies, in our opinion, of the Oirat nature of the sound - an upper front vowel. Still, the materials contain the words χabar ‘nose’, modun ‘tree’ that rather tend to be typical for Classical Mongolian. But it is widely known that bookish Mongolian words had been borrowed into Oirat through translations from Mongolian, thus making the compositions sound more archaic and scholastic. And many of the words - in the same forms - were introduced into Oirat dictionaries. Especially noteworthy among the peculiar means to convey Mongolic languages via Tibetan graphic elements within the data under consideration are (vowel) 'durations' expressed by the capital ’a. The paper notes that in most cases the 'durations' do not coincide with the long vowels researchers are familiar with, i.e. are not to be found in orthographies of diverse Mongolic languages and written monuments.