A uniform orthography and early linguistic research in Australia

David Moore
University of Western Australia

Introduction

A Uniform orthography can be defined as one which is segmental and phonographic. Each graphic segment is pronounced and has a distinct value. Internal consistency in transcription is achieved by defining each segment and the sound that it represents in the orthography. Each sound of a language is assigned a segment: a letter or a combination of letters which is outlined in statement of the orthography or ‘phonetic key’. By contrast, a non-uniform writing system involves writing languages where the value of each segment is unspecified. If the language of transcription is English, there is a poor correspondence between the letter and sound. The problem is particularly acute with English vowels. The five vowel letters of English are polyvalent; that is they each represent a number of English phonemes. Ten English phonemes are represented by <a> in English (Coulmas 2003: 186). Also, each English vowel phoneme can be represented by different graphemes. The spelling may be at the word level and based on what Dench (2000:59) says is ‘subjective impression of similarity to particular English words’. Individual segments in this ‘logographic’ spelling have little or no phonetic interpretation.

Uniform orthographies were the forerunners of the International Phonetic Alphabet (IPA). The first Australianist linguist to use the IPA appears to have been John McConnell Black (1855-1951), for a language of the Western Desert (Black 1915). I claim that some early investigators of Australian languages used Uniform orthographies in their writing of Australian Aboriginal languages and avoided the problems of English-based spelling.

Antecedents: Uniform orthographies in the late 1700s

Amongst English-speaking researchers of the late eighteenth century there were opposed views of the way in which words from Arabic, Persian and Sanskrit should be transliterated from their original scripts to the characters of the roman alphabet. Halhed and others advocated the use of English sound values (Master 1946). The year 1788 is significant as the date of the first British settlement in Australia, and is also the date of publication of the Dissertation on the Orthography of Asiatick Words in Roman Letters (Jones 1788). Its author, Sir William Jones (1746-1794), realised that the languages of the Indian subcontinent couldn’t be written consistently using the sound values of the English alphabet. He wanted to represent vowels with their ‘continental’ (particularly Italian) values. There was little difference between Jones and his opponents in the representation of consonant sounds. According to Jones, ‘each original sound may be rendered invariably by one appropriated symbol, conformably to the natural order of articulation, a perfect system of letters ought to contain one specific symbol for every sound used in pronouncing the language to which they belonged’ (Cannon 1990:249). I use ‘roman’ in lower case for the twenty six letters of the roman writing system, as distinct from that of a particular language.

The widespread adoption of the Jonesian system

The Dissertation excited great interest in Britain and Europe (Master 1946:7). There were many attempts to represent the sounds of hitherto unwritten languages according to Jones’ conventions. John Pickering (1777-1846) recommended the use of the Jonesian system for recording American Indian languages (Pickering 1820). In 1807 the Reverend John Davies sent the Congregationalist London Missionary Society (LMS) a copy of his manuscript for a “Tahitian Spelling Book” for publication, using the occasion to argue for the ‘continental’ system (Schütz 1994:107). The LMS adopted the system for Tahitian. Lancelot Threlkeld (1788-1859) worked for the LMS in the 1820s and used the ‘South Sea Islands’ orthography in his Australian Grammar (Threlkeld 1834).

Royal Geographical Society conventions 1836

Travellers needed a guide to the correct pronunciation of foreign names. Number three of the Journal’s aims was the development of a ‘more uniform and systematic orthography than has hitherto been observed, in regard to the names of cities and other objects; and a more precise and copious vocabulary than we at present possess, of such objects’ (Prospectus of the Royal Geographical Society: xi). The first explicit statement of the RGS orthographic conventions appears to have been in a footnote to Observations on the Coast of Arabia between Rás Mohammed and Jiddah (Wellsted 1836). See also Aurousseau (1942:245). Wellsted’s use of the RGS conventions was to write place names of Arabic, an Afro-Asiatic (Semitic) language with a Classical literature.

The vowel letters would be pronounced as for Italian and as read in selected English words: the <a> in ‘far’, <e> in ‘there’, <i> in ‘ravine’, <o> in ‘cold’ and the <u> in ‘rude’.

The consonants were to be pronounced as for English. The footnote reveals the writer’s familiarity with the sounds of British languages and English dialects and show an attempt to explain the unfamiliar by the familiar. The ‘Northumbrian r’, probably [ʁ], the use of the voiced uvular fricative allophone/variant of post-alveolar approximant /ɹ/ (Hughes, Trudgill and Watt 2005:124) is compared with a ‘guttural’ sound of Arabic, probably غ gayn, the voiced velar fricative [ɣ], transliterated <gh>. The voiceless velar fricative ch [x] of Welsh and Scots is compared with Arabic, خ xā’, and transliterated <kh>. Digraphs and the apostrophe were used to represent sounds for which there was no conventional spelling in European languages. Ain ع, a voiced pharyngeal fricative [ʕ] was transliterated with an apostrophe.

The use of the RGS conventions in Australia

The RGS conventions were used for the transcription of hitherto unwritten languages and this was the case in Australia where all the Aboriginal languages of the continent were unwritten until the arrival of Europeans. In Australia, as elsewhere, some early researchers used uniform orthographies and others wrote sounds according to their knowledge of English.

Dr John Lhotsky (1795-1865) collected language data in the Australian Alps from January to March 1834 and from Tasmania in 1836. His contribution to the Journal (Lhotsky 1839) appears to be the first example of the application of RGS conventions to an Australian language.

Recognition of consonant sounds

Early researchers had a pioneering role in recording sounds on paper for the first time. Awareness of foreign sounds increased with time spent in the field. Threlkeld (1834:6) had already encountered the velar nasal [ŋ] word-initially, based upon his hearing of Polynesian languages in which the sound commonly occurs at the beginning of a word. His recognition of the velar nasal probably led later other Australian researchers to recognise the sound in Aboriginal languages.

Lyon (1833) heard a sound which he compared with Classical Hebrew ע ayin, a voiced pharyngeal fricative [ʕ]: ‘the Ain of the Hebrew, the pronunciation of which has been so long a desideratum to the philologists of Europe, these people (Nyungar speakers) seem to possess in perfection’ (Lyon, Perth Gazette 30th March 1833). Lyon doesn’t identify the sound in Nyungar which the ayin resembles. Although [ʕ] doesn’t occur in Nyungar, this instance shows Lyon attempting to describe the unfamiliar by the more familiar. However, ayin had become a ‘silent letter’ in Hebrew and its sound value would have been difficult for Lyon to recover. Lhotsky (1839) noted that would be ‘more accurately rendered by Polish z’ which has no equivalent in the English tongue’.He was referring to /ʑ/, the voiced palatal fricative.

The consonant letters of the roman alphabet which were discarded are as significant as those which were included. Early researchers edited their work and made orthographic choices, using a selection of letters. Redundant letters such as <c> and <q> were discarded. Lyon (1833), comparing sounds in the Nyungar language with the Classical Hebrew alphabet noticed that the Nyungar language had ‘neither the Zain [z], the Samedi [s], nor the Schin [ʃ] of the Hebrews. The letter s, they are incapable of pronouncing’. He felt that he was ‘obliged to throw out every letter which was in the least allied in sound to the letter s’. George Fletcher Moore (1798-1886) read the RGS Journal and utilised the RGS conventions (Moore 1842:vii). In 1833 he obtained the name ‘carrar’ for the black goanna (Varanus tristis) from Weeip, one of his Nyungar-speaking informants (Cameron 2009). The entry was changed to ‘kardar’ for the Descriptive Vocabulary (Moore 1842:56) after <c> had been eliminated from the orthography. Salvado (1851) wrote <c> in his transcription of Nyungar words for the palatal plosive /c/ which Moore represented with <dj>:

Table 1: Transcription of Nyungar word for ‘Wedge- tailed Eagle’.

Moore 1842 waldja
Salvado 1851 ualce, ualge

Although the IPA eventually adopted [c] for the palatal plosive, the use of non-English consonantal values would not have helped the English-speaking reader. The users of the RGS were English-speaking Britons and so the use of English consonantal values, where this aim was realistic, was practical. Even though researchers were encouraged to use the values of English consonants they found ways to be creative in representing unfamiliar sounds.

Scott Nind (1831:47) noted that the Nyungar language ‘abounds in vowels’ but the values of the individual segments are not spelled out and no phonetic key to his transcription is provided, as can be seen in comparison with that of researchers who used the RGS:

Table 2: Early Transcriptions of Nyungar word for ‘Black Duck’

Nind 1831 wainern
Lyon 1833 goona-na
Grey 1839 ngoon-un
Moore 1842 ngwonana
Salvado 1851 n-unan
Travel and levels of education

The use of uniform orthographies presupposes a level of language awareness, often lacking among settlers in the frontier Australian communities of the nineteenth century. Then, as now, there were those who sought to write Aboriginal words according to English spelling conventions. Educated and travelled researchers were familiar with European languages and the languages and English dialects of Britain and Ireland. They had often engaged with what Clarke (1959:176) describes as ‘that form of foreign travel which we know as the study of the classics’, including Hebrew, Greek and Latin. They tended to be more aware of the limitations of English spelling because they had experience of foreign language scripts in which each letter had a distinct pronunciation which had to be learned.

The late nineteenth century

The RGS conventions were published as a ‘System of orthography for native names of places’ in the Proceedings (RGS 1885). The RGS aimed for a pronunciation of the ‘true sound pronounced locally’. Characters were selected from the twenty six letters of the roman alphabet. Of the thirty-four characters of the RGS, four digraphs represented consonant sounds and four other digraphs represented dipthongs. Hyphens were not allowed, and accent marks were only allowed for indicating stress.

As more sounds were discovered, the resources of the roman alphabet were found to be inadequate. Augmentation of the roman alphabet was necessary to write new sounds. The consonant digraphs were increased to nine in the 1892 revision. The RGS authors remained committed to the letters of the roman alphabet with limited additions: ‘Those who desire a more accurate pronunciation of the written name must learn it on the spot by a study of local accent and peculiarities’ (RGS 1885). The letters of the 1885 orthography:

1885 Orthography vowels

1885 Orthography

The RGS in the history of Australian linguistics

In Australian linguistic writing, there has been little appreciation that uniform orthographies represented a genuine advance on English-based spelling. Uniform orthographies and their value to linguistic research are unacknowledged in the literature on the history of Australian language research. Dixon, Ramson and Thomas (1990:7) dismiss the work of R.H.Mathews (1841-1918) and others: ‘Unfortunately, these early recorders were not trained linguists and wrote down Aboriginal words in terms of English sounds rather than in a phonetic alphabet’. However they (1990:15) also claim that Aboriginal languages are now written according to a Roman (sic) alphabet, implying that uniform orthographies utilising roman characters were only created in the twentieth century by professional linguists. Surely R.H. Mathews used the resources of the roman alphabet (Koch 2008). As argued in this paper, the aim of the RGS orthography was to circumvent the need for English spelling conventions.

Conclusions

It is fortunate for language documentation that many Aboriginal languages were written according to the RGS orthography. There are differences in quality in early wordlists and it is more likely that words were transcribed accurately with a uniform orthography. The RGS orthography was in use from the 1830s and gradually refined over the following century. The RGS system enabled English-speaking researchers to transcend the inconsistencies of English spelling. The extent to which the researchers recorded sounds accurately according to the RGS conventions must be answered in individual cases.

Bibliography

Aurousseau, M (1942). Suggested Principles for the Use and Spelling of Geographical Names. Part II. The Geographical Journal 100, 245-256.

Black, J.M (1915). Language of the Everard Range Tribe In S.A. White (ed.), Scientific Notes on an Expedition into the North-western regions of South Australia in Transactions of the Royal Society of South Australia 732-735. Adelaide.

Cameron, J.M.R. (2009). The Millendon memoirs: George Fletcher Moore’s Western Australian diaries and letters, 1830-1841. Carlisle, Western Australia: Hesperian Press.

Cannon, Garland (1990). The life and mind of Oriental Jones: Sir William Jones, the father of modern linguistics. Cambridge [England]; New York: Cambridge University Press.

Clarke, M.L (1959). Classical Education in Britain 1500-1900. Cambridge: Cambridge University Press.

Coulmas, Florian (2003). Writing Systems: An introduction to their linguistic analysis. Cambridge: Cambridge University Press.

Dench, A.C (2000). Comparative Reconstitution. In John Charles Smith and Delia Bentley (eds.), Historical Linguistics. Amsterdam/Philadelphia: John Benjamins.

Dixon, R.M.W, W.S Ramson and Mandy Thomas (1990). Australian Aboriginal Words in English, their origin and meaning. Melbourne: Oxford University Press.

Grey, George (1839). Vocabulary of the Aboriginal Language of Western Australia. Perth.
Hughes, Arthur, Peter Trudgill and Dominic Watt (2005). English Accents and Dialects: An introduction to Social and Regional Varieties of English in the British Isles, (fourth edition). London: Hodder Arnold.

Jones, Sir William (I788). A dissertation on the orthography of Asiatick words in Roman letters. By the President. Asiatick Researches , Transactions of the Asiatick Society I 1-56.
Koch, Harold (2008). R.H. Mathews’ schema for the description of Aboriginal languages. In William McGregor (ed.), Encountering Aboriginal languages: studies in the history of Australian linguistics 179-218. Canberra: Pacific Linguistics.

Lhotsky, John (1839). Some Remarks on a Short Vocabulary of the Natives of Van Diemen Land; And Also of the Menero Downs in Australia. Journal of the Royal Geographical Society of London 9, 157-162.

Lyon, Robert Menli (1833 30th March-20th April.). A glance at the manners and language of the aboriginal inhabitants of Western Australia with a short vocabulary Perth Gazette and Western Australian Journal 51-64. Perth.

Master, Alfred (1946). The Influence of Sir William Jones upon Sanskrit Studies. Bulletin of the School of Oriental and African Studies. Cambridge University Press Vol. 11 798-806.

Moore, G.F. (1842). A descriptive vocabulary of the language in common use amongst the natives of Western Australia. London: W.S. Orr & Co.

Nind, S. (1831). Description of the natives of King George Sound (Swan River Colony) and adjoining country. Journal of the Royal Geographical Society of London 1, 21-51.

Pickering, John (1820). An Essay on A uniform orthography for the Indian Languages of North America. Cambridge: Hilliard and Metcalf.

Royal Geographical Society (1831). Prospectus of the Royal Geographical Society, Journal of the Royal Geographical Society of London, Vol. 1, pp. vii-xii

Royal Geographical Society (Aug., 1885). System of Orthography for Native Names of Places. Proceedings of the Royal Geographical Society and Monthly Record of Geography, New Monthly Series 7, 535-536.

Salvado, R. (1851). Two native dialects of the New Norcia district. Rome: De Propaganda Fides.

Sampson, Geoffrey (1985). Writing Systems: a linguistic introduction. London: Hutchinson.

Schütz, Albert J. (1994). The Voices of Eden: A history of Hawaiian Language Studies. Honolulu: University of Hawai’i Press.

Threlkeld, E.L (1834). An Australian Grammar, comprehending the principles and natural rules of the language, as Spoken by the Aborigines in the Vicinity of Hunter’s River, Lake Macquarie &c, New South Wales. Sydney: Stephens and Stokes.

Wellsted, R. (1836). Observations on the Coast of Arabia between Rás Mohammed and Jiddah. Journal of the Royal Geographical Society of London 6, 51-96.

How to cite this post

Moore, David. 2013. ‘A uniform orthography and early linguistic research in Australia’. History and Philosophy of the Language Sciences. https://hiphilangsci.net/2013/10/16/a-uniform-orthography-and-early-linguistic-research-in-australia

Tagged with: , ,
Posted in 19th century, 20th century, Australia, Field linguistics, History, Linguistics, Phonology
13 comments on “A uniform orthography and early linguistic research in Australia
  1. David Nash says:

    William Dawes departed England the year before Jones 1788 was published. Dawes’ record of the Sydney Language, made up until he left Sydney in 1791, used a uniform orthography, which he set out at the beginning of one notebook http://www.williamdawes.org/ms/msview.php?image-id=book-b-page-1 Dawes did in practice add to his table, e.g. using td for the interdental stop, but his orthography is based on Latin (or Italian), augmented by ŋ. I wonder where Dawes had picked up ŋ from?

  2. moored03 says:

    The ŋ eng or engma symbol must have been known for some time? I wonder whether later Australian works acknowledged Dawes and were influenced by Dawes 1791?

  3. moored03 says:

    The vowels in Dawes’ manuscripts look as though they are based on English, compared with e.g, the RGS 1885 orthography. Did Dawes claim that used Latin or Italian?

  4. David Nash says:

    Yes sorry you’re right about Dawes’ vowel symbols David; they have a specified English value, and for vowels I should not have said ” his orthography is based on Latin (or Italian)”. I don’t think we have any comment by Dawes about his methods. Jeremy Steele in his 2005 MA thesis http://hdl.handle.net/1959.14/738 carefully analysed Dawes’ system of transcription (he detects three successive variants) and I think it counts as a uniform orthography.
    As to the influence of Dawes’ work, Steele (2005:39) summarised it thus: “When Dawes left the colony in 1791, with him departed many of those who knew him best and knew his work, and others such as Governor Phillip and David Collins were to follow. If memory of Dawes and his contribution to learning the indigenous language lingered, it did not last a generation.” The transcription that was in a position to be influential was Collins’ through his vocabulary published in his widely read ‘Account of the English colony in NSW’ (1798, 1804), though Collins’ spelling falls short of being uniform. He should’ve paid (even?) more attention to Dawes!

  5. moored03 says:

    Thanks for the link. Steele gives us a good insight into Dawes’ methods. It appears that Dawes’ model of grammatical description was influenced by his study of Latin and French. I agree that Dawes’ orthography is uniform and he might have refined it over time if he had stayed in New South Wales.
    The Dawes manuscripts show that the ŋ was used earlier than the nineteenth century IPA and Ellis-Pitman systems. According to Pullum and Laduslaw in The Phonetic Symbol Guide, ‘It occurred in this form with this value in Isaac Pitman’s 1845 Phonotypic alphabet’. Albright in the IJAL Vol. 24 (January 1958), page 11: ‘There seems little doubt this excellent symbol, usually credited to Pitman and Ellis in the nineteenth century, was invented by Holder two centuries earlier’.
    William Holder, Elements of Speech: An Essay of Inquiry into the Natural Production of Letters with an Appendix concerning Persons Deaf and Dumb (London: John Martin, 1669).

  6. David Nash says:

    More on ŋ: as well as its promotion by Pitman, Benjamin Franklin earlier used the symbol in his ‘Reformed Mode of Spelling’ http://archive.org/stream/politicalmiscell00franrich#page/470/mode/2up (http://en.wikipedia.org/wiki/Benjamin_Franklin%27s_phonetic_alphabet), which was published in London in 1779 (Franklin lived in France from 1776 to 1785 as US Ambassador).
    It is notable that the RGS 1885 orthography explicitly used the digraph ng for both [ŋ] and [ŋg] while recognising the “two separate sounds”. A small irony is that the coastal dialect of the Sydney Language, the one Dawes was learning, is unusual among Australian languages in lacking [ŋg] — inland ŋg corresponds to coastal ŋ by Dawes’ Law (as I dubbed it).

  7. moored03 says:

    At least unlike wouldn’t be confused with a consonant cluster. The RGS orthography may have influenced Baldwin Spencer who also confused [ŋ] with [ŋg]. The Permanent Committee on Geographical Names (Gleichen 1921) is even more confusing: ‘ng has three separate sounds, as in engrave, finger and singer’ The Permanent Committee and the revised RGS (called Geographic II) was used by Australianists such as Tindale,
    Spencer in the ‘Glossary of Native Terms Used’ in The Arunta (1927: 610): ‘there are two separate sounds. If the letters occur in the middle of a word, they are actually sounded separately as in mangrove. If at the beginning of a word, they are sounded as in English. Both sounds occur in such a word as Ngurangura, ‘evening’.’
    Ng occurring at the beginning of a word is exactly not like English. Only ŋ occurs in Ngurangura. The glossary is full of inconsistencies. The example words in the phonetic key are all English which is difficult when a sound of Arrernte is not like any of the sounds in English.

  8. moored03 says:

    In the first line I meant to say: “At least unlike couldn’t be read as a consonant cluster.”

  9. David Nash says:

    When Spencer (1927: 610) wrote “at the beginning of a word, they are sounded as in English” he presumably meant “at the beginning of a word, they are sounded as [the ng is] in [the word] ‘English'”.

  10. moored03 says:

    It appears that is what he meant, the ng in ‘ring’ and ‘singer’. He misses the ŋ in alkna ‘eye’ and oknia ‘father’ and the cluster in atninga ‘avenging party’. Sometimes [ŋ] is represented with as in erilknabota, ‘a very wise man’, but often not. The correct pronunciation can’t be derived from Spencer’s transcription, because as the preceding examples show, its not predictably inaccurate. And this is Spencer 1927 not Spencer at the beginning of his Central Australian fieldwork! Its the difference between who only spent a few months in contact with speakers and Carl Strehlow who spent 26 years in contact with speakers of the language and whose orthography is consistent. Without a detailed knowledge of the language, Spencer would have been reliant on accurate transcription of words which were used in books like The Arunta- its in phonetic transcription that he fails most noticeably.

  11. moored03 says:

    Correction: Sometimes [ŋ] is represented with , as in erilknabota, ‘a very wise man’

  12. David Nash says:

    I’ve just chanced on this commentary on the RGS conventions, in the entry for the WA town ‘Dwellingup’ in http://www.landgate.wa.gov.au/corporate.nsf/web/History+of+country+town+names When the name was proposed as ‘Dwellingupp’ in 1909,

    The double p spelling in the original gazettal of this name was used because the Lands and Surveys Department had adopted a system for spelling Aboriginal names developed by the Royal Geographical Society. A number of Aboriginal names ending in “up” were for a time spelt with the “upp” ending (including Kirupp, Kulikupp, Manjimupp and Mungalupp). The RGS system had a rule that vowels are pronounced as in Italian and consonants as in English. This would have meant that names ending in “up” should have been pronounced as “oop”, because the Italian “u” was a long “u”, as in flute. These Aboriginal names were meant to be pronounced as “up”, and the Department asked the RGS for a rule to assist in correct pronunciation. The RGS solution was that doubling the following consonant shortened the preceding vowel, and this meant the “upp” ending ensured the “up” pronunciation. However, this particular rule was rescinded in 1915 for SW towns with the suffix “up”, as the Australian way of pronouncing the letter “u” was almost always short, and rarely the Italian “oo”.

  13. moored03 says:

    Thanks for that link. There are numerous references to the official Western Australian orthography. Although the Department had some success in implementing its guidelines with ‘Kunjin’ and ‘Kunanalling’ among many others, it was frustrated by ‘cacographic’ English spelling. I’ve written about the WA conventions in a longer version of the above post:

    The RGS conventions were adopted officially in Western Australia but place names continued to be spelled inconsistently. In 1901 Western Australia issued The Spelling of Native Geographical Names following correspondence between the Western Australian Minister for Lands and the Royal Geographical Society in London. Robert Cecil Clifton (1854-1931) became Western Australian Under Secretary for lands in 1891, a position he held for twenty-seven years. He attempted to implement a Uniform orthography for recording place names in the Western Australian goldfields, but was frustrated by ‘the objection of officials and people generally to adopting the system’. Settlers would move to a new centre in the goldfields and adopt its Aboriginal name. Official spelling guidelines were ignored in favour of spellings based upon English. The Society referred Clifton to the conventions of 1885 and reassured him that the RGS orthography would be able to record all of the place names of Western Australia. Clifton said that the Lands Department officials would attempt to implement the RGS system ‘but require the assistance of the public, and more particularly the Press to back them up’.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: