last month
next month

Centre for the Study of Manuscript Cultures (CSMC)

07/2014 manuscript  of the month

How many languages can you count?

A multilingual manuscript from South India

Coming from the very South of India, more precisely from one of the collections held in Pondicherry, this manuscript (RE22704) is testimony to a feature characteristic of the whole subcontinent, in particular to its scholarly environments: multilingualism. Sanskrit was for centuries the language used by the intellectual elite in the area for the composition of texts, but this was soon paralleled by other local and cosmopolitan literary languages with equally sophisticated textual outcomes. How this phenomenon is reflected in the material culture is a topic yet to be fully addressed. This manuscript offers an interesting example of this. Despite its monotone layout, how many languages does it contain?

Fig. 1 Wooden cover and thread of RE22704 > Enlarge

Thousands of manuscripts among the millions hosted in South Asian libraries contain copies, or partial copies, of Amarasiṃha’s Nāmaliṅgānuśāsana (“Teaching on Nouns and [their] Genders”), the most renowned traditional Sanskrit dictionary. Since the time of its composition (or redaction) possibly around the 7th century CE, this has been a fundamental tool for teaching Sanskrit to young students, and a constant reference work for trained scholars.

More precisely, the Nāmaliṅgānuśāsana is a thesaurus, i.e. a synonymic dictionary, in which nouns are grouped by semantic fields. Thus, there is a section for the names of divinities, one for the names of plants, etc. It is for this reason that this work is mostly known as the Amarakośa (“Amara’s Thesaurus”). Furthermore, it is composed in verses in order to facilitate the memorisation of its long sections, as has been done for centuries by traditionally trained scholars of Sanskrit.

Fig. 2 Side image of RE22704. The bundle contains 284
palm leaves    > Enlarge

Owing to its importance in the traditional lore, it comes as no surprise that this work has been at the centre of fervid commentarial activity with textual outputs both in Sanskrit and in the several local literary languages of the subcontinent. South Indian languages are certainly no exception, as many libraries contain copies of the Nāmaliṅgānuśāsana along with Tamil, Telugu, Malayalam, or Kannada annotations.

This manuscript thus makes an interesting case study. It consists of a bundle of almost 300 palm leaves written on both sides, stuck between two wooden covers, and kept together by a string running through its holes. Today part of the collection of the Institut Français de Pondichéry (included in the UNESCO “Memory of the World” Register since 2005), this manuscript was most probably crafted during the 19th century and was written by a man called Kuruṉātayyaṉ from the town of Veḷḷaṅkoḷḷi (present day Kerala?) for a certain Paṭṭaravarkaḷ from Pāḷayaṅkoṭṭai (Tamil Nadu). Unfortunately, we know nothing more about either individuals.

Allegedly, it contains the most complex Nāmaliṅgānuśāsana with Tamil annotations found in a manuscript. According to a well-established commentarial practice, each verse is analysed several times in order to clarify both its grammar and meaning. Remarkably, it introduces the vernacular element by the following procedure: explanations of single words and whole verses are given in Tamil. Thus, annotations become translations too. Furthermore, this manuscript is characterised by a unique feature: the presence of a third linguistic layer. Unpredictably, some Tamil words are preceded by their Telugu translation.

Fig. 3 The three languages of RE22704 (folio 2, recto 2): Sanskrit (yellow), Telugu (red), and Tamil (green). Move the cursor over the image to show details.

Vernacular commentaries to the Nāmaliṅgānuśāsana are usually simple. Not every Sanskrit word is commented on or translated, but single Tamil synonyms are given to certain categories of nouns. This points to an educational environment in which Sanskrit was taught to early stage pupils by means of translating the main semantic categories of the Nāmaliṅgānuśāsana, whereas more sophisticated lexical analyses were reserved for advanced students and carried out by means of commentaries composed in Sanskrit. Deviating in part from this trend, this manuscript offers impressive annotations to verses 2 to 5, which teach how to use the thesaurus, and in particular to verse 1, the invocation to the deity. In fact, the commentary on the latter indulges in a lengthy theological disquisition, trying to persuade the reader that the invocation addresses several religio-/philosophical schools at the same time. Furthermore, it also reflects on its own commentarial activity, looking for a solid justification of what sometimes can look like a wild over-interpretation.

Such an array of idiosyncratic features makes this manuscript a unique exemplar, also suggesting that Kuruṉātayyaṉ might have been the author of the present Tamil (and Telugu) annotations. We can thus make an educated guess and imagine our scribe to have been a Telugu speaking scholar, who worked in an environment in which Sanskrit and Tamil were the main languages of intellectual exchange. This would probably have been a śrīvaiṣṇava environment, where a particular set of Sanskrit devotional and speculative texts connected to the Hindu god Viṣṇu were commented on in a highly Sanskritised register of Tamil (usually referred to as Manipravalam), as is the case in the present manuscript.

Contrary to many other manuscript traditions, the South Indian one does not usually distinguish texts and their commentaries by means of layout, even when these are written in different languages. So it is in this manuscript, where no blank spaces are left between each Sanskrit word and its Tamil, and sometimes Telugu, annotations (only a scant punctuation is provided). Furthermore, all three languages (Sanskrit, Tamil, and Telugu) are written in a mixture of Tamil script and Tamilian Grantha script, two similar writing systems that compel the user of the manuscript to read through each line of the text in order to distinguish its various layers and languages.


NARAYANA RAO, Velcheru (2003): “Multiple Literary Cultures in Telugu: Court, Temple, and Public”. In: Sheldon Pollock (ed.): Literary Cultures in History: Reconstructions from South Asia. Berkeley, Los Angeles, London: Columbia University Press, 383-436.

RAMANATHAN, A.A. (ed.) (1971): Amarakośa [I] with the Unpublished South Indian Commentaries: Amarapadavivṛti of Liṅgayasūrin and the Amarapadapārijāta of Mallinātha. Madras: The Adyar Library and Research Centre.

VENKATACHARI, K.K.A. (1978): The Maṇipravāḷa Literature of the Śrīvaiṣṇava Ācāryas. Bombay: Ananthacharya Research Institute.

VOGEL, Claus (1979): Indian Lexicography. Wiesbaden: Harrassowitz.


Present holder: Institut Français de Pondichéry (Pondicherry, India).
Shelfmark: RE22704 - amarakośaḥ drāviḍaṭīkāsahitaḥ (apūrnaḥ).
Material: Palm leaf bundle; two wooden covers, and one thread.
Folios: Total 284 folios. One title folio (amarapañcakai); one blank guard leaf folio; 276 folios with original numbering from 1 to 276
     (eight lines per folio); two blank guard leaf folios; three guard leaf (?) folios containing part of an unidentified text (possibly a lexicon);
     one blank guard leaf folio.
Dimensions: 25.4 x 4.3 cm.
Provenance: South India (probably Southern Tamil Nadu), ca. 19th century.

Text by Giovanni Ciotti