Transliteration as seen by Walter Elliott in Madras in 1859

The Madras Literary Society & Auxiliary of the Royal Asiatic Society (MLS-ARAS) appointed a sub-committee comprising Walter Elliot, W.H. Bayley, and M. Norman to explore ways of writing Indian words in Roman characters in early 1858. The Elliot Committee submitted a report (#. 32) to MLS-ARAS, which was published for the public. The cover page of the report indicates that this was printed by H. Smith at the Fort Saint George Gazette Press in 1859. The report was also formally presented at a meeting of the Philological Society (of London?) chaired by F. Pulszky on February 9, 1860.

At the start of the report, a note explains that Bayley and Norman had to leave India because of ill-health before its publication but goes on to state that, before their departure, the Elliot Committee had come to a unanimous conclusion. Nonetheless, the bulk of the text refers to ‘I’ (first person, singular), and, therefore, I felt the major contribution to the report was by Walter Elliot’s.

The objective of the Committee was to lay down a scheme of orthography, which would provide the exact representation of every word occurring in Indian languages. The driving purpose was that MLS-ARAS had recognised that Government reports referred to Indian names and other attributes badly; this report indicates – as an example – that the Murshadabad Survey Register cited Ibrahimpur as either Berahimpoor or Biharipoor.

To achieve the best outcomes, the Elliot Committee self-prescribed the following dictates (p. 3): “(i) That a distinct Roman character as far as practicable be employed to express each established oriental letter; (ii) that the same character should always represent the same letter and should never be employed without some distinguishing sign to designate a second; (iii) that two or more letters should never be employed where a single character could be made to suffice; (iv) that diacritical marks should only be resorted to in the last necessity, and should be of simplest description; (v) that varieties of type, as capitals, italics, black letters, should be considered inadmissible; and (vi) that the scheme should be founded on the system of Sir W. Jones.”

The report starts with extensive remarks on transliterating Dévanagari script, but considering the focus of Madras Musings, I shall restrict my notes and annotations to sections in the report that pertain to Tamil words being transliterated into Roman characters.

The section on The Dravidian Alphabets (pages 10-13) includes the following: “All the Southern races have adapted the Arian (sic Aryan) phonetic system except the Tamil people, and they while retaining their own peculiar system of letters, have formed a second alphabet, founded on the Arian, for Sanscrit literature, which has been extensively cultivated among them. This is called the Granïham character.” From an angle of comparative linguistics, this section includes fascinating remarks on the letters specific to Dravidian languages, some of which have fallen out of use over time. The report refers to the letter zha – the vexata litera of the Dravidian – which has been retained in Tamil and Malayalam only.

Referring to Tamil (pages 20-27), the report particularly refers to –ள–,–ற–, –ன– , and ழ. It says that –ள–, also common in other southern Indian languages, is originally a Sanscrit sound. It is a hard –l–, which the report suggests be marked by a point below, –l–. For – ற–, the report suggests use of –t’t’–. Referring to –ன–, the report says this letter is a genius of the Tamil language and can be represented as –n–. The letter –ழ– represents a sound peculiar to Dravidian languages. It is a combination of j, l, and r. To represent –ழ–, Ziegenbalg (1714) used either rl or rhl, Beschi (1728)–lj; Ellis (1816)–zh; Wilson (1855)–l (Tamizh), r (Malayalam); –ழ– represents a sound altogether sui generis; and according to Wilson “the enunciation is singularly obscure, and cannot be precisely represented by any written characters.”

The sound ‘–ழ–’ creates is by no means unique to Dravidian languages, says Elliot; it occurs in the aboriginal Indo-Chinese tongues of the Himãlaya and Tibet, it has an equivalent in j as used in French (e.g., jeu). Page 25 lists randomly chosen 15 Tamil words that include zh and compares how these words have changed in Malayalam, Kannada, and Telugu. One example is: ézhu (seven) (Tamil), ézha (Malayalam), édu (Telugu), élu (Kannada). In subsequent sections, Elliot goes on to discuss the movement of the letter that sounds zh into la (e.g., uzhundu —> ulundu). For the ‘–ழ–‘, ‘–ள–‘ transpositions Elliot draws an affirmation from Sabda Manjari (the Tamil version of Pãnini’s phonetic grammar).

The following occurs in page 21: “… to provide for differences of sound of the same letter, under different circumstances, as in the cases of ‘–க– ’, ‘–ச–’, ‘–த–’, ‘–ட–’, and ‘–ப–’, which may be read respectively as k and g and ch–t and d–t and d–p and b–r and t. But several of these depend on the position of the particular letter in relation to others, which every Tamil scholar learns by practice.”

The end of this monograph includes a memorandum signed in Bangalore by M. Norman (dated March 21, 1859) that was to accompany the proposed scheme of representation of native words in English characters. Pages 57-64 include an extensive list of alphabets (vowels and consonants) in Sanskrit, Telugu, Kannada, Malayalam, Tamil, and Persian followed by the proposed Roman equivalents, with supplementary remarks.

A fascinating, but a forgotten (ignored–?) monograph, it includes valuable remarks on the comparative linguistics of Indian language, besides the transliteration effort it has attempted to provide.