Sequence-to-Sequence Models for Grapheme to Phoneme Conversion on Large Myanmar Pronunciation Dictionary

Author(s):  
Aye Mya Hlaing ◽  
Win Pa Pa
Author(s):  
Kathirvel Palanisamy ◽  
Attassry Manoj Kumar ◽  
Murugesan Chinnappa ◽  
M Sabarimalai Manikandan ◽  
K P Soman

2020 ◽  
Vol 24 (3) ◽  
pp. 527-543
Author(s):  
NICOLAS TRAPATEAU

A long /aː/ in pre-fricative and pre-nasal contexts in words such as fast, answer or after is one of the most distinctive phonological features of British RP and, to a certain extent, of Southern Hemisphere varieties of English (Trudgill 2010). The lengthening of /a/ has been particularly gaining ground from the eighteenth century onwards (Beal 1999; Jones 2006). The pronouncing dictionaries published between the eighteenth century and the present day allow us to trace its lexical diffusion (Labov 1994) across the whole lexicon. Drawing on the statistics of the ARCHER corpus, the lexical sets of the ECEP database, the full electronic edition of Walker's dictionary (1791), Wells’ Longman Pronunciation Dictionary (2008) and the Macquarie Dictionary (2015), this article examines the role played by the phonetic environment, word frequency, phonetic analogy and isolated lead words like draught or master in the spread of the lengthening of /a/. The results show that word frequency per se has no clear effect on /a/ lengthening in either pre-fricative or pre-nasal environments in eighteenth-century sources. The article also offers a possible relative chronology of the spread of that phenomenon to each phonetic environment within the bath set.


Sign in / Sign up

Export Citation Format

Share Document