Bengali (
/baŋla/) is an Indo-European language (Indic branch) spoken by over 175 million people in Bangladesh and eastern India (Dasgupta Reference Dasgupta, Cardona and Jain2003: 352; Lewis Reference Lewis2009). The speech illustrated below is representative of the standard variety widely spoken in Dhaka and other urban areas of Bangladesh.
Consonants
Plosives and affricates contrast in voicing and aspiration. Although displayed in one column, the articulation of the postalveolars varies by consonant type (see ‘Conventions’ below). Barring rare exceptions, /dɦ h/ do not occur word-finally and /ŋ/ does not occur word-initially (Dasgupta Reference Dasgupta, Cardona and Jain2003: 358–359).
Vowels
Vowels are plotted below based on F1 and F2 frequencies averaged across six speakers. Vowel length is not contrastive, except across morpheme boundaries. The contrast between oral and nasal vowels observed in Kolkata Standard Bengali (Masica Reference Masica1991: 118) is not characteristic of Bangladeshi Standard, presumably due to Eastern dialect influence (Majumdar Reference Majumdar1997: 108).
Diphthongs
Diphthongs contrast with heterosyllabic vowel sequences (Islam Reference Islam2000: 95; Dasgupta Reference Dasgupta, Cardona and Jain2003: 356–357), as in /ɡaj/
‘I sing’ vs. /ɡai/
‘body (emphatic)’ and /ʃo
/
‘lies down’ vs. /ʃoe/
‘having tolerated’. There is no high-mid vs. low-mid distinction in off-glides, so diphthongs such as /a
/ may vary between [a
] and [a
]. Off-glides are produced with a higher tongue position than corresponding nuclear vowels; compare /ʃo/
‘show’ vs. /ʃo
/
‘you lie down’ (cf. /ʃɔ/
‘tolerate!’, /ʃɔ
/
‘you tolerate’). Some speakers monophthongize /ɛ
/ to [ɛ]; compare the two recordings of /nɛ
/
‘justice’. Vowel sequences not included below can only occur across syllable boundaries, e.g. /bie/
‘wedding’, /kua/
‘well’.
Stress
Stress is not marked in broad transcription as it is consistently word-initial. While stress is neither contrastive nor phonetically salient (Chatterji Reference Chatterji1921, Goswami Reference Goswami1944, Ferguson & Chowdhury Reference Ferguson and Chowdhury1960, Anderson Reference Anderson1962, Bykova Reference Bykova1981, Kawasaki & Shattuck-Hufnagel Reference Kawasaki and Shattuck-Hufnagel1988, Hayes & Lahiri Reference Hayes and Lahiri1991, Lahiri & Fitzpatrick-Cole Reference Lahiri, Fitzpatrick-Cole, Kager and Zonneveld1999, Michaels & Nelson Reference Michaels and Nelson2004, Selkirk Reference Selkirk, Lee, Gordon and Büring2006), it is associated with phonological alternations and plays an important role in the intonational system, as it is associated with the postlexical pitch accent (Khan Reference Khan2008: chapter 2).
Geminates
Excluding /dɦ ɡɦ ŋ f s h ɹ/, all consonants can occur geminated within morphemes, and are transcribed with the length diacritic /ː/. Singleton /d dɦ/ do not occur after tautomorphemic vowels except in ideophones and loanwords, e.g. /ɹod/
‘road’.
Conventions
Postalveolars
The postalveolar region includes three distinct tongue configurations: apical for plosives and /ɹ/, laminal for the affricates, and domed for /ʃ/. The plosives /t tʰ d dɦ/ are variously described as ‘cacuminal’ (Islam Reference Islam2000: 90), ‘cerebral’ (Majumdar Reference Majumdar1997: 166), ‘retroflex’ (Ramaswami Reference Ramaswami1999), ‘retroflex alveolar’ (Ray, Hai & Ray Reference Ray, Hai and Ray1966: 6), ‘not true retroflex’ (Haldar Reference Haldar1986: 22), ‘alveolo-retroflex’ (Hai Reference Hai1960), ‘approaching the alveolar region’ (Chatterji Reference Chatterji1970: xxxiii), ‘more apico-alveolar . . . than the true apico-palatals of sister languages to the west’ (Dasgupta Reference Dasgupta, Cardona and Jain2003: 359), and ‘simply alveolar’ (Tunga Reference Tunga1995: 139). The affricates /
/ are variously characterized as ‘palatal affricates’ (Ramaswami Reference Ramaswami1999, Islam Reference Islam2000: 91), ‘apico-dental hissing sibilants . . . [or] affricated plosives’ (Ray et al. Reference Ray, Hai and Ray1966: 81), ‘palatal. . .made with the front of the tongue’ (Chatterji Reference Chatterji1970: xxxii), ‘dental affricates’ (Haldar Reference Haldar1986: 26), ‘alveolo-palatal affricates’ (Dasgupta Reference Dasgupta, Cardona and Jain2003: 359), and ‘dental-palatal plosives’ or ‘dorso-alveolar’ affricates (Tunga Reference Tunga1995: 131). Palatographic evidence indicates dorso-alveolar affrication (Hai Reference Hai1960). The fricative /ʃ/ is described as ‘mediopalatal’ (Alam Reference Alam2000: 43; Islam Reference Islam2000: 91) and ‘palato-alveolar’ (Dasgupta Reference Dasgupta, Cardona and Jain2003: 360).
Rhotic
Bangladeshi Standard Bengali has only one rhotic /ɹ/, as is the case in most Eastern dialects (Dasgupta Reference Dasgupta, Cardona and Jain2003: 359; Masica Reference Masica1991: 97); however, some speakers may maintain a marginal /ɹ/ vs. /ɽ/ distinction in formal registers due to influence from Kolkata Standard. The rhotic /ɹ/ can be realized as a tap [ɾ], especially following dentals (Dasgupta Reference Dasgupta, Cardona and Jain2003: 359), and both realizations can be devoiced [
] in codas. Consonants are often lengthened following /ɹ/.
Fricatives
Unlike Kolkata Standard, Bangladeshi Standard contrasts /ʃ/ and /s/ word-initially, e.g. /siɹka/ 
‘vinegar’ vs. /ʃiɹa/ 
‘syrup’, in medial clusters, e.g. /as
e/ 
‘softly’ vs. /aʃe/ 
‘to come’, and word-finally, e.g. /bas/ 
‘enough’ vs. /baʃ/ 
‘bamboo’ (Masica Reference Masica1991: 98; Islam Reference Islam2000: 91; Dasgupta Reference Dasgupta, Cardona and Jain2003: 360). In word-initial and word-final clusters, /s/ is common and /ʃ/ is rare, e.g. /sɹoa/ 
‘listener’, /desk/ 
‘desk’. As intervocalic singletons and geminates, /ʃ/ is common and /s/ is rare, e.g. /
oʃi/ 
‘guilty’, /oʃːi/ 
‘rogue’. The fricative /f/ is variously realized [f ɸ pʰ] (Dasgupta Reference Dasgupta, Cardona and Jain2003: 360).
Clusters
Although uncommon in native words, clusters are found in initial, medial, and final position, e.g. /sɹi/ 
‘wife’, /ʃɔŋskɹii/ 
‘culture’, /ɡɔnḏʑ/ 
‘marketplace’.
Transcription of recorded passage
Orthographic version
English translation
One day the North Wind and the Sun were disputing who was stronger between them. At that very moment, a traveler wearing a heavy shawl came walking towards them. The Wind and the Sun agreed that the one who could take the shawl off from the traveler's body would be considered the stronger. After this, the North Wind started to blow with all his strength, but the more forcefully he blew, the more the traveler kept his shawl held tightly around him. Defeated, the Wind ended his effort. After this, it was the Sun's turn. The Sun scattered his warmth. The traveler immediately took off the shawl from his body. In the end, the North Wind was obliged to acknowledge that of the two of them, it was the Sun who was stronger.
Acknowledgements
I would like to sincerely thank my primary consultants, Farida Amin Khan and Shafiq ud Dowla Khan, and all my speakers for their time and patience. I would also like to express my gratitude to John Esling, Bruce Hayes, Ewa Jaworska, Sun-Ah Jun, Aditi Lahiri, Aaron J. Lee, and one anonymous reviewer for their valuable comments and insight.
