Turkish MFA dictionary v3.0.0#

Maintainer: Montreal Forced Aligner
Language: Turkish
Dialect: N/A
Phone set: MFA
Number of words: 41,373
Phones: a aː b bː c cː dʒ dʒː d̪ d̪ː e eː f fː h hː i iː j jː k kː m mː n̪ n̪ː o oː p pː s̪ s̪ː tʃ tʃː t̪ t̪ː u uː v vː y yː z̪ z̪ː ç çː ø øː ŋ ɛ ɟ ɟː ɡ ɨ ɪ ɫ ɫː ɯ ɯː ɲ ɾ ɾː ʃ ʃː ʊ ʎ ʎː ʏ ʒ
License: CC BY 4.0
Compatible MFA version: v3.0.0
Citation:

@techreport{mfa_turkish_mfa_dictionary_2024,
	author={McAuliffe, Michael and Sonderegger, Morgan},
	title={Turkish MFA dictionary v3.0.0},
	address={\url{https://mfa-models.readthedocs.io/pronunciation dictionary/Turkish/Turkish MFA dictionary v3_0_0.html}},
	year={2024},
	month={Mar},
}

If you have comments or questions about this dictionary or its phone set, you can check previous MFA model discussion posts or create a new one.

G2P models

Turkish MFA G2P model v3.0.0

Acoustic models

Turkish MFA acoustic model v3.0.0

Installation#

Install from the MFA command line:

mfa model download dictionary turkish_mfa

Or download from the release page.

The dictionary available from the release page and command line installation has pronunciation and silence probabilities estimated as part acoustic model training (see Silence probability format and training pronunciation probabilities for more information. If you would like to use the version of this dictionary without probabilities, please see the [plain dictionary](https://raw.githubusercontent.com/MontrealCorpusTools/mfa-models/main/dictionary/turkish/mfa/Turkish MFA dictionary v3_0_0.dict).

Intended use#

This dictionary is intended for forced alignment of Turkish transcripts.

This dictionary uses the MFA phone set for Turkish, and was used in training the Turkish MFA acoustic model. Pronunciations can be added on top of the dictionary, as long as no additional phones are introduced.

Performance Factors#

When trying to get better alignment accuracy, adding pronunciations is generally helpful, especially for different styles and dialects. The most impactful improvements will generally be seen when adding reduced variants that involve deleting segments/syllables common in spontaneous speech. Alignment must include all phones specified in the pronunciation of a word, and each phone has a minimum duration (by default 10ms). If a speaker pronounces a multisyllabic word with just a single syllable, it can be hard for MFA to fit all the segments in, so it will lead to alignment errors on adjacent words as well.

Ethical considerations#

Deploying any Speech-to-Text model into any production setting has ethical implications. You should consider these implications before use.

Demographic Bias#

You should assume every machine learning model has demographic bias unless proven otherwise. For pronunciation dictionaries, it is often the case that transcription accuracy and lexicon coverage for the prestige variety modeled in this dictionary compared to other variants. If you are using this dictionary in production, you should acknowledge this as a potential issue.

IPA Charts#

Consonants#

Obstruent symbols to the left of are unvoiced and those to the right are voiced.

Manner	Labial	Labiodental	Dental	Alveolar	Alveopalatal	Palatal	Velar	Glottal
Nasal	m Occurrences: 19,604 Examples: * makro: [m a k ɾ o] * amaç: [a m a tʃ] * alman: [a ɫ m a n̪] * nemli: [n̪ e m ʎ i] mː Occurrences: 46 Examples: * emmiş: [e mː i ʃ] * emmim: [e mː i m] * jimmy: [ʒ i mː i] * amman: [a mː a n̪]		n̪ Occurrences: 28,072 Examples: * kondu: [k o n̪ d̪ ʊ] * nesi: [n̪ e s̪ i] * dondu: [d̪ o n̪ d̪ u] * donuk: [d̪ o n̪ u k] n̪ː Occurrences: 99 Examples: * annem: [a n̪ː e m] * anne: [a n̪ː e] * ninni: [n̪ i n̪ː i] * annen: [a n̪ː e n̪]			ɲ Occurrences: 213 Examples: * sanki: [s̪ a ɲ c i] * anki: [a ɲ c ɪ] * angle: [a ɲ ɟ ʎ e] * dink: [d̪ i ɲ c]	ŋ Occurrences: 168 Examples: * banka: [b a ŋ k a] * frank: [f ɾ a ŋ k] * yankı: [j a ŋ k ɯ] * angın: [a ŋ ɡ ɯ n̪]
Stop	p Occurrences: 3,390 Examples: * şüphe: [ʃ y p ç e] * gıpta: [ɡ ɯ p t̪ a] * espri: [e s̪ p ɾ i] * pos: [p o s̪] pː Occurrences: 4 Examples: * züppe: [z̪ y pː e] b Occurrences: 6,496 Examples: * sabun: [s̪ a b u n̪] * başlı: [b a ʃ ɫ ɨ] * meb: [m e b] * elbet: [e ʎ b e t̪] bː Occurrences: 14 Examples: * tıbbi: [t̪ ɯ bː i] * abbas: [a bː a s̪]		t̪ Occurrences: 14,431 Examples: * tabu: [t̪ a b u] * iter: [i t̪ e ɾ] * katip: [k a t̪ i p] * taşın: [t̪ a ʃ ɯ n̪] t̪ː Occurrences: 712 Examples: * yattı: [j a t̪ː ɨ] * hattı: [h a t̪ː ɨ] * tuttu: [t̪ u t̪ː ʊ] * yetti: [j e t̪ː i] d̪ Occurrences: 18,363 Examples: * kadı: [k a d̪ ɯ] * dere: [d̪ e ɾ ɛ] * düşme: [d̪ y ʃ m e] * dirim: [d̪ i ɾ i m] d̪ː Occurrences: 152 Examples: * maddi: [m a d̪ː i] * cadde: [dʒ a d̪ː e] * iddia: [i d̪ː i a] * madde: [m a d̪ː ɛ]			c Occurrences: 7,238 Examples: * keten: [c e t̪ e n̪] * kenti: [c e n̪ t̪ ɪ] * tekel: [t̪ e c e ʎ] * brkiç: [b ɾ c i tʃ] cː Occurrences: 66 Examples: * takke: [t̪ a cː ɛ] * ökkeş: [ø cː e ʃ] * sikke: [s̪ i cː e] * mekke: [m e cː ɛ] ɟ Occurrences: 3,818 Examples: * özgür: [ø z̪ ɟ y ɾ] * genel: [ɟ e n̪ e ʎ] * gelme: [ɟ e ʎ m ɛ] * götün: [ɟ ø t̪ y n̪] ɟː Occurrences: 1 Examples:	k Occurrences: 11,133 Examples: * koku: [k o k u] * şok: [ʃ o k] * uymak: [u j m a k] * parka: [p a ɾ k a] kː Occurrences: 57 Examples: * hakkı: [h a kː ɯ] * hakka: [h a kː a] * dakka: [d̪ a kː a] ɡ Occurrences: 637 Examples: * gıpta: [ɡ ɯ p t̪ a] * gazla: [ɡ a z̪ ɫ a] * algı: [a ɫ ɡ ɯ] * golf: [ɡ o ɫ f]
Affricate					tʃ Occurrences: 3,883 Examples: * açığı: [a tʃ ɯː] * içmem: [i tʃ m e m] * onüç: [o n̪ y tʃ] * çul: [tʃ u ɫ] tʃː Occurrences: 1 Examples: dʒ Occurrences: 4,608 Examples: * gence: [ɟ e n̪ dʒ e] * recep: [ɾ e dʒ e p] * kanca: [k a n̪ dʒ a] * cinsî: [dʒ i n̪ s̪ i] dʒː Occurrences: 9 Examples: * hacca: [h a dʒː a]
Sibilant			s̪ Occurrences: 13,515 Examples: * arsız: [a ɾ s̪ ɯ z̪] * sızdı: [s̪ ɯ z̪ d̪ ɨ] * açsın: [a tʃ s̪ ɯ n̪] * bis: [b i s̪] s̪ː Occurrences: 155 Examples: * bassa: [b a s̪ː a] * hisse: [ç i s̪ː e] * öss: [ø s̪ː] * üssü: [y s̪ː e s̪ e ʏ] z̪ Occurrences: 6,984 Examples: * uzaya: [u z̪ a j a] * sakız: [s̪ a k ɯ z̪] * zevce: [z̪ e v dʒ ɛ] * vezir: [v e z̪ i ɾ] z̪ː Occurrences: 16 Examples: * gazze: [ɡ a z̪ː ɛ]		ʃ Occurrences: 7,626 Examples: * beleş: [b e ʎ e ʃ] * kişi: [c i ʃ i] * shgm: [ʃ ɟ m] * şef: [ʃ e f] ʃː Occurrences: 2 Examples: * eşşek: [e ʃː e c] ʒ Occurrences: 236 Examples: * jakı: [ʒ a k ɨ] * cluj: [dʒ ɫ u ʒ] * juva: [ʒ u v a] * major: [m a ʒ o ɾ]
Fricative		f Occurrences: 1,956 Examples: * falan: [f a ɫ a n̪] * tayfa: [t̪ a j f a] * wells: [tʃ i f t̪ e ʎː s̪] * muaf: [m u a f] fː Occurrences: 27 Examples: * öff: [ø fː] * affet: [a fː e t̪] v Occurrences: 3,157 Examples: * güvey: [ɟ y v e j] * yuva: [j u v a] * varız: [v a ɾ ɯ z̪] * mavi: [m a v i] vː Occurrences: 66 Examples: * kuvve: [k u vː e] * evvel: [e vː e ʎ]				ç Occurrences: 809 Examples: * hile: [ç i ʎ e] * zehir: [z̪ e ç i ɾ] * henri: [ç e n̪ ɾ i] * heey: [ç e e] çː Occurrences: 9 Examples: * sıhhi: [s̪ ɯ çː i]		h Occurrences: 2,982 Examples: * harap: [h a ɾ a p] * hat: [h a t̪] * mhpli: [m h p ʎ i] * sahte: [s̪ a h t̪ e] hː Occurrences: 18 Examples: * ohh: [o hː] * ohhü: [o hː ʏ] * öhhü: [ø hː ʏ] * ahha: [a hː a]
Approximant						j Occurrences: 16,791 Examples: * yazdı: [j a z̪ d̪ ɨ] * reyi: [ɾ e j i] * suyun: [s̪ u j u n̪] * yalar: [j a ɫ a ɾ] jː Occurrences: 24 Examples: * tayf: [t̪ a jː f] * ayyar: [a jː a ɾ]
Tap				ɾ Occurrences: 29,703 Examples: * diri: [d̪ i ɾ ɪ] * asker: [a s̪ c e ɾ] * brady: [b ɾ a d̪ j] * hayrı: [h a j ɾ ɨ] ɾː Occurrences: 62 Examples: * şerri: [ʃ e ɾː i] * zerre: [z̪ e ɾː e] * sırrı: [s̪ ɯ ɾː ɨ] * şerre: [ʃ e ɾː e]
Lateral				ɫ Occurrences: 14,475 Examples: * alım: [a ɫ ɯ m] * lıyı: [ɫ ɯ j ɨ] * fadıl: [f a d̪ ɯ ɫ] * aylık: [a j ɫ ɯ k] ɫː Occurrences: 377 Examples: * allık: [a ɫː ɯ k] * çullu: [tʃ u ɫː u] * yollu: [j o ɫː u]		ʎ Occurrences: 12,655 Examples: * diler: [d̪ i ʎ e ɾ] * talep: [t̪ a ʎ e p] * bil: [b i ʎ] * alemi: [a ʎ e m ɪ] ʎː Occurrences: 496 Examples: * belli: [b e ʎː i] * dille: [d̪ i ʎː ɛ] * kelle: [c e ʎː ɛ] * elli: [e ʎː ɪ]

Vowels#

Vowel symbols to the left of are unrounded and those to the right are rounded.

	Front	Near-Front	Central	Near-Back	Back
Close	i Occurrences: 30,748 Examples: * teki: [t̪ e c i] * bari: [b a ɾ i] * hissi: [ç i s̪ː i] * kadri: [k a d̪ ɾ i] iː Occurrences: 352 Examples: * niğde: [n̪ iː d̪ e] * iyiyi: [iː j i] * değii: [d̪ eː iː] * iğde: [iː d̪ e] y Occurrences: 7,099 Examples: * ürkek: [y ɾ c e c] * üçtü: [y tʃ t̪ ʏ] * hödük: [h ø d̪ y c] * gürdü: [ɟ y ɾ d̪ ʏ] yː Occurrences: 191 Examples: * düğün: [d̪ yː n̪] * büğrü: [b yː ɾ y] * düğüm: [d̪ yː m]		ɨ Occurrences: 2,989 Examples: * zanlı: [z̪ a n̪ ɫ ɨ] * sardı: [s̪ a ɾ d̪ ɨ] * tası: [t̪ a s̪ ɨ] * şahı: [ʃ a h ɨ]		ɯ Occurrences: 22,460 Examples: * yatı: [j a t̪ ɯ] * zıpla: [z̪ ɯ p ɫ a] * ayına: [a j ɯ n̪ a] * tıbbi: [t̪ ɯ bː i] ɯː Occurrences: 968 Examples: * açığı: [a tʃ ɯː] * sığır: [s̪ ɯː ɾ] * işığı: [i ʃ ɯː] * atığ: [a t̪ ɯː] u Occurrences: 12,302 Examples: * kuzu: [k u z̪ u] * uhhuh: [u hː u h] * uzaya: [u z̪ a j a] * havuz: [h a v u z̪] uː Occurrences: 404 Examples: * uğru: [uː ɾ ʊ] * uuuf: [uː u f] * suudi: [s̪ uː d̪ i] * uğur: [uː ɾ]
		ɪ Occurrences: 950 Examples: * deli: [d̪ e ʎ ɪ] * şeydi: [ʃ e j d̪ ɪ] * bilgi: [b i ʎ ɟ ɪ] * nakdi: [n̪ a k d̪ ɪ] ʏ Occurrences: 267 Examples: * usulü: [u s̪ u ʎ ʏ] * müydü: [m y j d̪ ʏ] * yönlü: [j ø n̪ ʎ ʏ] * sözlü: [s̪ ø z̪ ʎ ʏ]		ʊ Occurrences: 897 Examples: * boştu: [b o ʃ t̪ ʊ] * frau: [f ɾ a ʊ] * muydu: [m u j d̪ ʊ] * korku: [k o ɾ k ʊ]
Close-Mid	e Occurrences: 35,016 Examples: * eşine: [e ʃ i n̪ e] * keçe: [c e tʃ ɛ] * ettik: [e t̪ː i c] * yekta: [j e c t̪ a] eː Occurrences: 626 Examples: * eğri: [eː ɾ i] * değse: [d̪ eː s̪ ɛ] * tee: [t̪ eː] * araya: [a ɾ eː] ø Occurrences: 3,206 Examples: * sökün: [s̪ ø c y n̪] * tümör: [t̪ y m ø ɾ] * ömrü: [ø m ɾ y] * öncü: [ø n̪ dʒ y] øː Occurrences: 138 Examples: * öğle: [øː ʎ ɛ] * öğlen: [øː ʎ e n̪] * öğren: [øː ɾ e n̪]				o Occurrences: 8,923 Examples: * koyma: [k o j m a] * oralı: [o ɾ a ɫ ɨ] * port: [p o ɾ t̪] * kol: [k o ɫ] oː Occurrences: 170 Examples: * yoo: [j oː] * oğlun: [oː ɫ u n̪] * oğlan: [oː ɫ a n̪] * boğdu: [b oː d̪ ʊ]

Open-Mid	ɛ Occurrences: 2,477 Examples: * redde: [ɾ e d̪ː ɛ] * bölge: [b ø ʎ ɟ ɛ] * madde: [m a d̪ː ɛ] * lerde: [ʎ e ɾ d̪ ɛ]

Open			a Occurrences: 48,641 Examples: * kanat: [k a n̪ a t̪] * bacak: [b a dʒ a k] * ahh: [a hː] * haddi: [h a d̪ː i] aː Occurrences: 618 Examples: * zağar: [z̪ aː ɾ] * saati: [s̪ aː t̪ i] * bağla: [b aː ɫ a] * bağ: [b aː]