This article discusses the phonological system of the Bulgarian language.

The phonemic inventory of Contemporary Standard Bulgarian (CSB) has been a contested and controversial matter for decades, with two major currents, or schools of thought, forming at national and international level:^{[1][2][3][4][5]}

The first one considers that there are only 28 phonemes in Contemporary Standard Bulgarian: 21 consonants, 1 semivowel and 6 vowels and that only one of them, the semivowel j, is palatal.^[6][7] This was the general consensus reached by all major Bulgarian linguists prior to the Soviet occupation of Bulgaria in 1944, and still continues to be the view held by multiple modern Bulgarian and probably a majority of Western phonologists.^{[8][9][3][10]}

The second school of thought is based on a sketch of Eastern Bulgarian consonantism made by Russian linguist Nikolai Trubetzkoy in his 1939 book Principles of Phonology, where he introduced palatalization as additional phonemic distinction in Bulgarian, much like in his native language, Russian.^[11] In Bulgaria, the concept was launched in the late 1940s by two younger linguists, Stoyko Stoykov and Lyubomir Andreychin, who proclaimed the existence of 17 new palatalized phonemes, rounding Standard Bulgarian's phonemic inventory to 45 phonemes, 18 of which are palatal.^[12][13]

Both Stoykov and Andreychin had rejected Trubetzkoy's ideas in the early 1940s, before Bulgaria's occupation by the Red Army.^[14][15] The concept got quick approval from the Bulgarian Communist Party and Bulgarian Academy of Sciences. The 17 palatals became a staple of all published standard Bulgarian grammars and phonologies during the totalitarian era.^[16] The consonant model has not fared well abroad, routinely being called into question or outright rejected, including in the 1999 Handbook of the International Phonetic Association, which sided with the traditional interpretation.^[17][18]

Vowels[edit]

According to their place of articulation, Bulgarian vowels can be grouped in three pairs—front vowels: ⟨е⟩ (/ɛ/) and ⟨и⟩ (/i/); central vowels: ⟨а⟩ (/a/) and ⟨ъ⟩ (/ɤ/); and back vowels: ⟨о⟩ (/ɔ/) and ⟨у⟩ (/u/).

Here /ɛ/, /a/ and /ɔ/ are "low", and /i/, /ɤ/ and /u/ are "high".

The dominant theory of Bulgarian vowel reduction posits that Bulgarian vowels have a phonemic value only in stressed position, while when unstressed, they neutralize in an intermediate centralized position, where lower vowels are raised and higher vowels are lowered.^[21][17] This concerns only the central vowels /a/ and /ɤ/, which neutralize into [ɐ], and the back vowels /ɔ/ and /u/, which neutralize into [o].

The merger of /ɛ/ and /i/ is not allowed in formal speech and is regarded as a provincial (East Bulgarian) dialectal feature; instead, unstressed ɛ is both raised and centralized, approaching the schwa (/ə/).^[22] The Bulgarian /ɤ/ vowel does not exist as a phoneme in other Slavic languages, though a similar reduced vowel transcribed as [ə] does occur. The theory further posits that such neutralization may nevertheless not always happen: vowels tend to be distinguished in emphatic or deliberately distinct pronunciation, while reduction is strongest in colloquial speech.

Maximum tongue positions when pronouncing stressed and unstressed Bulgarian vowels

Frequency (F1) distinctions between Bulgarian vowels in stressed and unstressed position

Nevertheless, the hypothesis that high and low vowels neutralize into a common centralized vowel has never been properly studied or proven in a practical setting. Several recent studies by both Bulgarian and foreign researchers, involving volunteers speaking Contemporary Standard Bulgarian, have established—on the contrary—that while unstressed low vowels /ɛ/, /a/ and /ɔ/ are indeed raised as expected, unstressed high /ɤ/ and /u/ are also raised somewhat, rather than lowered, while /i/ remains in the same position.^[23][24][25]

All three studies indicate that a clear distinction is kept between unstressed /ɛ/ and both stressed and unstressed /i/. The situation with unstressed /a/ and /ɔ/ is more complex, but all studies indicate that they both approach unstressed /ɤ/ and /u/ very closely and overlap with them to a great extent, but their average realisations remain slightly more open. One of the studies finds that unstressed /a/ to be practically undistinguishable from stressed /ɤ/,^[26] while another finds a lack of statistically significant difference between /ɔ/ and /u/,^[25] and a third one finds coalescence only in formants for one of the pairs and only in tongue position for the other.

While the difference between all stressed vowels and between unstressed /i/ and /ɛ/ can be heard in almost 100% of cases, the unstressed back and central vowels are perceptually neutralised in minimal pairs, with only 62% identifying unstressed /u/, 59% unstressed /a/ and /ɔ/ and a mere 57% unstressed /ɤ/.^[25]

Semivowels[edit]

The Bulgarian language officially has only one semivowel: /j/. It is traditionally regarded as a semivowel, but in recent years, it has largely been treated as a "glide" or approximant, thus making it part of the consonant system. Orthographically, it is represented by the Cyrillic letter ⟨й⟩ (⟨и⟩ with a breve) as in най- [naj] (prefix 'most') and (тролей [troˈlɛj] ('trolleybus'), except when it precedes /a/ or /u/ (and their reduced counterparts [ɐ] and [o]), in which case both phonemes are represented by a single letter, ⟨я⟩ or ⟨ю⟩, respectively: e.g., ютия [juˈtijɐ] ('flat iron'), but Йордан [jorˈdan] ('Jordan').

As a result of lenition of velarized /l/ ([ɫ]), ongoing since the 1970s, [w] appears to be an emerging allophone of velarized [ɫ] among younger speakers, especially in preconsonantal position: вълк [ˈvɤwk] ('wolf') instead of [ˈvɤɫk]. While certain Western Bulgarian dialects (in particular, those around Pernik), have had a long-standing tradition of pronouncing [ɫ] as [w], the use of the glide in the literary language was first noted by a radio operator in 1974.^[27] A Ukrainian researcher found in 2012 that Bulgarians split into three age-specific groups in terms of [ɫ] pronunciation: 1) people in their 40s or older who have standard pronunciation; 2) people in their 30s, who can articulate [ɫ] but unconsciously say [w]; and 3) younger people who are unable to differentiate between the two sounds and generally say [w].^[28]

A study of 30 graduate students was therefore conducted in 2014 to quantify the trend. The study testified to an extremely wide proliferation of the phenomenon, with 9 out of 30 participants unable to produce [ɫ] in any given word, and only 2 participants able to produce [ɫ] correctly, but in no more than half the words in the study.^[29] Remarkably, not a single participant was able to enunciate [ɫ] between a bilabial consonant and a rounded vowel, e.g., in аплодирани [ɐpwo'dirɐni] ('applauded'), or between a rounded vowel and a velar consonant, e.g. in толкова ['tɔwkovɐ] ('so').^[30] Another discovery of the study was that in particular positions, certain participants enunciated neither [ɫ] nor [w], but the high back unrounded vowel [ɯ] (or its corresponding glide [ɰ]).

The glide [w] can also be found in English loan words such as уиски ['wiski] ('Whiskey') or Уикипедия ['wiki'pɛdiɐ] ('Wikipedia'). The semivowel /j/ forms a number of diphthongs, which are summarized below:^[31][32]

Consonants[edit]

Two schools of thought on Bulgarian consonantism[edit]

The main point of contention between the two schools of thought on Bulgarian consonantism has been whether palatalized consonants should be defined as separate phonemes or simply as allophones of their respective hard counterparts.

The first or the "traditionalist" school of thought developed gradually by consensus over the course of many decades, crystallized in the late 1930s and early 1940s and was reinvigorated after the fall of the totalitarian regime. It posits that Bulgarian has no palatal or palatalized consonants other than /j/, that Trubetzkoy’s 17 palatalized consonants are merely (positional) allophones of hard consonants and that the Bulgarian language therefore has only 28 phonemes.^{[33][34][9][17]} It has proposed alternative notation of palatalized consonants in the form of C-j-V (consonant-glide-vowel) clusters and has made a tentative hypothesis about the decomposition of Bulgarian palatals into consonants + glide using the following arguments:^[4][35][2]

• Unlike the palatal consonants in other Slavic languages, which can usually be found in all positions, palatalized consonants in Bulgarian have very limited distribution: only before central or back vowels, and then only if there is a /j/ before the vowel, noted graphically with the letters ⟨я⟩ [ja]/[jɤ], ⟨ю⟩ [ju] or ⟨ьо⟩ [jɔ].^[10][36]
• The enormous number of palatal or palatalized consonants claimed contradicts the historical development of Bulgarian and the South Slavic languages in general towards depalatalization, especially considering that there were only 9 or 10 palatal consonants in Old and Middle Bulgarian and that four of them have since hardened or are no longer used.^[37][38]
• The reanalysis would make it possible to make sense of sequences like ям [ˈjam] ('(I) eat') ↔ изям [iˈzʲam] ('(I) eat till the last crumb') ↔ доям [doˈjam] ('(I) finish eating something') ↔ отям [oˈtʲam] ('(I) eat to my heart's content'), where a root is considered to be iotated in some cases and to palatalize the previous consonant in others.^[39]
• Stoykov, Andreychin and Tilkov have all provided eyewitness accounts that the pronunciation of so-called palatalized consonants as consonant + /j/ is common among Bulgarian speakers and predominates in Western Bulgaria.^[40][41][42]
• The swift adoption of an approach that looks tailored to Russian but has proven a remarkably poor fit for Bulgarian at a time when Bulgaria was governed by a totalitarian regime completely subservient to the Soviet Union is suspicious and is most likely the result of Soviet duress and/or influence.^[43][44]

The second school of thought came to being rather unexpectedly in the late 1940s, as a refinement of Trubetzkoy's rough draft a decade before. It quickly gained currency in the state apparatus as the only theory, most likely because it used the same approach as in Russian, which was vital for a government so tied to Moscow. It posits that apart from ⟨й⟩ (/j/), there are 17 separate palatal phonemes that are in minimal pairs with their hard counterparts, including дз' (/d͡zʲ/) and х' (/ç/), which are not found in any native Bulgarian words and were excluded from Trubetzkoy's draft.^[45] Thus, only 5 consonants are not in minimal pairs, ⟨ч⟩ (/t͡ʃ/), ⟨дж⟩ (/d͡ʒ/), ⟨ш⟩ (/ʃ/) and ⟨ж⟩ (/ʒ/), which are only hard, and the glide ⟨й⟩ (/j/), which is only soft. They argue that Bulgarian phonemic inventory consists of a total of 45 phonemes, whereof 6 vowels, 1 semivowel and 38 consonants, and present the following arguments:^[46]

• Even though the distribution of palatalized consonants is limited, there is still a large number of Bulgarian words can be distinguished only by the difference in palatalization, e.g. гол [gɔɫ] ('nude/naked') vs. гьол [ˈgʲɔɫ] ('puddle'); лук [ˈɫuk] ('onion') vs. люк [ˈʎuk] ('hatch'); дал [daɫ] ('to have given') vs/ дял [ˈdʲaɫ] ('share');
• Palatal consonants cannot be considered to be formed by their hard counterparts by adding (/j/), as there is a clear auditory difference between pronouncing a soft consonant (as in Russian) and pronouncing a consonant and a glide (as in English). A 2012 comparative study of palatal phonemes in Russian, palatalized phonemes in Bulgarian and CjV clusters in English has deduced that the phonetic and auditory properties of Bulgarian palatals are similar to those in Russian and deviate substantially from the English consonant + glide sequences and that Bulgarian and Russian listeners did not need to wait for formant transitions to identify a consonant as palatal/palatalized unlike English listeners.^[47]

Historical development of Bulgarian consonantism[edit]

Proto-Slavic underwent three separate rounds of palatalization and one of iotation, but the resulting palatal consonants eventually hardened in Western and South Slavic.

By the Old Bulgarian period, there were only four consonants left forming contrastive pairs: р (/r/) and р' (/rʲ/), н (/n/) and н' (/ɳ/), л (/l/) and л' (/ʎ/), с (/s/) and с' (/sʲ/). Three consonants were only hard: к (/k/), г (/g/) and х (/x/), four were only soft: /ʒ/, /t͡ʃ/, /ʃ/ and ꙃ' (/d͡zʲ/), while the remaining eight consonants were generally hard, but could be semi-palatalized: б (/b/), в (/β/), д (/d/), ꙁ (/z/), м (/m/), п (/p/), т (/t/) and ф (/f/).^[48]

Historical phonetician Anna-Maria Totomanova has expressed a slightly divergent opinion: the four hard/palatal contrastive pairs were again /r/ and /rʲ/, /n/ and /ɳ/, /l/ and /ʎ/, /s/ and /sʲ/, 11 consonants, /p/, /b/, /m/, (/f/, /β/), /d/, /t/, /z/, /k/, /g/ and /x/, were only hard, and six consonants, /ʃ/, /ʒ/, /t͡ʃ/, /t͡sʲ/, /d͡zʲ/ and iota (/j/), along with the typically Bulgarian consonant combinations ⟨щ⟩ [ʃt] & ⟨жд⟩ [ʒd], were only soft.^[49] Finally, Huntley mentions 9 palatal consonants: /ʃ/, /ʒ/, /t͡ʃ/, /t͡sʲ/, /d͡zʲ/ and /j/, which were only soft, and /ɳ/, /ʎ/ and /rʲ/, which could also be hard.^[50] Both Haralampiev and Totomanova have noted a marked trend towards consonant hardening.^[48]

Eventually, /ʃ/, /ʒ/ and /t͡ʃ/ hardened permanently, /d͡z/ disappeared from the phonemic inventory, and ⟨дж⟩ (/d͡ʒ/) was borrowed from Ottoman Turkish as only hard. But before that, two phenomena led to the palatalization of more consonants: a second iotation and the dissolution of the yat vowel. As a result of the contraction and closure of the syllable in the Middle Bulgarian period, unstressed /i/ in many cases turned into the semivowel /j/ or attached to a consonant, palatalising it. Thus, Old Bulgarian свиниꙗ [sviˈnija] ('swine') contracted into свиня [sviˈɲa] and братиꙗ [ˈbratija] ('brothers') into братя [ˈbratʲɐ].^[51]

In many dialects, the resulting palatalised т' (/tʲ/) and д' (/dʲ/) turned into palatalised к' (/c/) and г' (/ɟ/).^[52] These were subsequently eliminated from CSB as dialecticisms, e.g., цвет'e [ˈt͡svɛtʲɛ] ('flower')→ цвек'е ['t͡svɛkʲɛ] → Ø. The form accepted in the literary language was instead the unpalatalised цвете [ˈt͡svɛtɛ] based on the Old Bulgarian form.

The dissolution of the yat happened somewhat later, towards the end of the Middle Bulgarian period and had different effects on the various dialects. In most of the East, yat in a stressed syllable softened the preceding consonant and turned into /a/. In the West, however, it led to /ɛ/ in both stressed and unstressed syllables producing no palatalisation anywhere.^[53] This was one of the main factors that led to the markedly different patterns of palatalisation in Western and Eastern Bulgarian dialects, i.e., strong palatalisation of only 5 consonants in the West vs. moderate palatalisation of almost all consonants in the East.

Development of phonological theory before 1945[edit]

The first Bulgarian grammar to mention phonetics is Ivan Bogorov's First Bulgarian Grammar, where he identified 22 consonants, however, including among them ⟨щ⟩ (ʃt), ⟨ъ⟩ and ⟨ь⟩ (no phonemic status at word end).^[54] The first Bulgarian man of letters to correctly identify the 21 consonants in Bulgarian was Ivan Momchilov, in 1868.^[55] According to Momchilov, Bulgarian consonants could sound hard or soft, entirely depending on the vowel accompanying them.^[56]

Phonetics only started developing seriously after World War I, and towards the 1930s, all major Bulgarian linguists had reached consensus that Bulgarian phonemic inventory consisted of 28 phonemes. Out of the six major Bulgarian grammars published in the Interwar period, five explicitly mention the existence of 22 consonants (including the semivowel /j/) and 6 vowels: Petar Kalkandzhiev,^[57] Aleksandar Teodorov-Balan, who suggested 26 certain phonemes + 2 conditional ones (for the non-native and infrequent ⟨дж⟩ (/d͡ʒ/) and ⟨дз⟩ (/d͡z/)),^[8] Dimitar Popov, who posited that the only soft or palatal phoneme in Bulgarian was ⟨й⟩ (/j/),^[58] as well as Lyubomir Andreychin, who considered that even though palatalised consonants had distinctive articulation, they did not deserve phonemic status.^[59] All phoneticians referenced palatalisation extensively, but without ascribing phonemic value to the resulting sounds. Moreover, according to Stefan Mladenov,^[60]

"If we disregard individual cases of old, stronger palatalization, which may be found in Eastern and Western dialects alike, Contemporary Standard Bulgarian has developed a very distinctive "semi-palatalization", which is often neglected."

This was a result of the attempts to unify the extremely divergent patterns of Eastern and Western palatalization into a common standard in the 1800s and early 1900s, which eventually led to its general elimination from the standard language. Examples include the complete elimination of end-word palatals in a number of words ending in ⟨р'⟩ (/rʲ/), ⟨н'⟩ (/ɲ/), ⟨л'⟩ (/ʎ/) and ⟨т'⟩ (/tʲ/), e.g., writing and saying кон [ˈkɔn] ('horse') instead of конь ['kɔɲ], път ['pɤt] ('road') instead of пъть [pɤtʲ]), etc.; the adoption of the hard suffix -не instead of -нье for verbal nouns, i.e., писане [pisɐˈnɛ] instead of писанье [pisɐ'ɲɛ] ('writing'); labelling palatalization before front vowels as dialectal: (поле [poˈlɛ] instead of полье [pо'ʎɛ] ('field'), тиква [tikˈvɐ] instead of тьиква [tʲikˈvɐ] ('pumpkin')), etc.^[61] Thus, the only sanctioned palatalisation in CSB is in syllable-initial position before central and back vowels, i.e., in front of /a/, /ɤ/, /ɔ/ and /u/.

The opinions of Bulgarian linguistics were also shared by a number of foreign Slavicists. French linguist Léon Beaulieux has stated that Bulgarian is characterised by its tendency to eliminate all palatal consonants.^[62] Czech linguist Horalek claimed as early as the 1940s that palatalisation in standard Bulgarian was on its way to disappear through decomposition and the development of a specific /j/ glide and that words such as бял (white) & дядо (grandfather) were pronounced [bjaɫ] & [ˈdjado] (i.e., CjV) or even [biaɫ] & [ˈdiado] just as often as they were pronounced [bʲaɫ] and [ˈdʲado].^[63]

Bulgarian consonantism according to IPA (22-consonant model)[edit]

A graphic representation of the Bulgarian consonant systems according to the International Phonetic Association (22 consonants) follows below:^[17]

As palatalized consonants have very limited distribution in Standard Bulgarian and are only possible in syllable-initial position before central/back vowels, IPA's consonant table above treats them as palatalized allophones of their respective "hard" counterparts + [j] rather than as palatal phonemes and suggests that they can unambiguously be interpreted as CjV (consonant-glide-vowel) clusters.^[7]<^[6] Thus, for example, някой [nʲakoj] ('somebody') can easily be reanalysed as [njakoj]. According to Ternes and Vladimirova-Buhtz:^[71]

"The phonemic analysis underlying the present transcription does not assume the existence of palatalized consonants. An alternative postulates the following palatalized consonants /pʲ, bʲ, tʲ, dʲ, kʲ, gʲ, tsʲ, dzʲ, mʲ, nʲ, rʲ, fʲ, vʲ, sʲ, zʲ, xʲ, lʲ/. The nature of palatalization in Bulgarian is different from that in Russian. Its occurrence is very restricted. Before front vowels and [j], palatalization does not go beyond the degree that is conditioned by the inevitable play of coarticulation. Before back vowels, palatalization may unambiguously be interpreted as C plus [j]. In syllable and word final position, it does not occur."

Among modern Bulgarian phoneticians, strong opinions about the existence of 22 consonants only are held by, e.g., Blagoy Shklifov, Mitko Sabev, Andrey Danchev and especially by Dimitrina Ignatova-Tzoneva, who has consistently argued that palatal consonants, though present in a number of dialects and in earlier stages of the development of the Bulgarian language, have largely been eliminated from Contemporary Standard Bulgarian.^[72][73] All of them have advocated for a CjV reanalysis of palatalization. A large number of other Bulgarian linguists have come out in support of a more traditional view of Bulgarian consonantism, e.g., Kiril Mirchev,^[74] Petar Pashov,^[75] Bozhil Nikolov,^[76] Todor Boyadzhiev,^[77] Борис Симеонов [bg], who has argued that there was no logic that could explain why a consonant affected by yat mutation (e.g., /b/ in бял-бели ['bʲaɫ]-['bɛli]) would be palatal in some of its forms and hard in others, and so on.^[78]

A number of foreign linguists have rejected the 39-consonant model based on an analysis of the distribution and degree of "softening" of Bulgarian "palatals" and the number of speakers pronouncing ⟨bj⟩, ⟨dj⟩ or ⟨fj⟩ instead of ⟨bʲ⟩, ⟨dʲ⟩ or ⟨fʲ⟩. These have included Austrian researcher Merlingen (1957),^[79] Americans Carleton Hodge (1957)^[80] and Joseph van Campen and Jacob Ornstein (1959),^[81] Romanian linguist Alexandru Rosetti, who qualified the degree of palatalization of Bulgarian consonants as "a softening" (1967),^[82] Swiss Max Mangold (1988),^[83] Korean Slavist Gwon-Jin Choi, who has argued about the decomposition of Bulgarian palatalism (into C + j) (1994),^[4][39] as well as phoneticians Ternes and Vladimirova-Buhtz, who have most recently suggested C-j-V notation of palatals, as their limited distribution proved they were allophones rather than phonemes (1999).^[17]

A comparison of the distribution of palatalized consonants in Bulgarian and other Slavic languages and of the number of palatals in each major Slavic languages is of key importance for understanding the issue:

Bulgarian
Position	Distribution	Example
Before back vowels	Y	лют [lʲu̟t] (spicy)
Before front vowels	N	–
Before sonorants	N	–
Before other consonants	N	–
At word end	N	–

Serbo-Croatian
Position	Distribution	Example
Before back vowels	Y	ljut [ʎûːt] (angry)
Before front vowels	Y	polje [pôʎe] (field)
Before sonorants	Y	daljnji [dâːʎɲiː] (far)
Before other consonants	Y	biljka [bîːʎka] (plant)
At word end	Y	kralj [krâːʎ] (king)

Russian
Position	Distribution	Example
Before back vowels	Y	люк [lʲuk] (hatch)
Before front vowels	Y	поле [ˈpolʲe] (field)
Before sonorants	Y	сильно [ˈsʲilʲnə] (strongly)
Before other consonants	Y	полька [ˈpolʲkə] (polka)
At word end	Y	печаль [pʲɪˈt͡ɕælʲ] (grief)

All other palatalized consonants in Bulgarian have the same distribution:

Distribution of consonants in Standard Bulgarian^[85]

Position	Consonant
	p	pʲ	b	bʲ	m	mʲ	f	fʲ	v	vʲ	t	tʲ	d	dʲ	s	sʲ	z	zʲ	t͡s	t͡sʲ	d͡z	d͡zʲ	n	ɲ	r	rʲ	ʃ	ʒ	t͡ʃ	d͡ʒ	j	l/ɫ	ʎ	k	kʲ	g	gʲ	x	xʲ
Before back vowels	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	N	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	N
Before front vowels	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	Y	Y	Y	Y	Y	N	Y	N	Y	N	Y	N
Before sonorants	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	N	N	Y	N	Y	N	Y	Y	Y	Y	Y	Y	N	Y	N	Y	N	Y	N
Before other consonants	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	N	N	Y	N	Y	N	Y	Y	Y	Y	Y	Y	N	Y	N	Y	N	Y	N
At word end	Y	N	N	N	Y	N	Y	N	N	N	Y	N	N	N	Y	N	N	N	Y	N	N	N	Y	N	Y	N	Y	N	Y	N	Y	Y	N	Y	N	N	N	Y	N

Comparison of the distribution of palatal (palatalised) consonants in all major Slavic languages^[86][87][88]

Language	Consonant
	pʲ	bʲ	mʲ	fʲ	vʲ	tʲ	dʲ	sʲ	zʲ	t͡sʲ	d͡zʲ	ɲ	rʲ	ɕ	ʑ	t͡ɕ	d͡ʑ	j	ʎ	c	ɟ	ç	ɣʲ
Russian	Y	Y	Y	Y	Y	Y	Y	Y	Y	N	N	Y	Y	Y	Y	Y	N	Y	Y	Y	Y	Y	N
Belarussian	Y	Y	Y	Y	Y	N	N	Y	Y	Y	Y	Y	N	N	N	N	N	Y	Y	Y	Y	Y	Y
Ukrainian	N	N	N	N	N	Y	Y	Y	Y	Y	Y	Y	Y	N	N	N	N	Y	Y	N	N	N	N
Polish	N	N	N	N	N	N	N	N	N	N	N	Y	N	Y	Y	Y	Y	Y	N	N	N	N	N
Czech	N	N	N	N	N	N	N	N	N	N	N	Y	N	N	N	N	N	Y	N	Y	Y	N	N
Slovak	N	N	N	N	N	N	N	N	N	N	N	Y	N	N	N	N	N	Y	Y	Y	Y	N	N
Slovenian	N	N	N	N	N	N	N	N	N	N	N	N	N	N	N	N	N	Y	N	N	N	N	N
Serbo-Croatian	N	N	N	N	N	N	N	N	N	N	N	Y	N	N	N	Y	Y	Y	Y	N	N	N	N
Bulgarian (22-consonant model)	N	N	N	N	N	N	N	N	N	N	N	N	N	N	N	N	N	Y	N	N	N	N	N
Bulgarian (39-consonant model)	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	?	Y	Y	N	N	N	N	Y	Y	Y	Y	?	N

It is argued that it is highly unlikely for modern Bulgarian to have developed 18 palatalized consonants (incl. /j/) from the 9 or 10 that existed in Old Bulgarian (/ʃ/, /ʒ/, /t͡ʃ/, /t͡sʲ/, /d͡zʲ/, /j/, /ɳ/, /ʎ/, /rʲ/ and /sʲ/), considering that four of those had already hardened or disappeared (/ʃ/, /ʒ/, /t͡ʃ/, /d͡zʲ/).^[89] Townsend and Janda have argued that such a development is at odds with the general development in all South Slavic languages, which had suppressed the development of palatals very early.^[90] If Bulgarian indeed had 18 palatal phonemes, it would be as palatal a language as Russian and Belarussian, which runs counter to auditory experience.

Bulgarian consonantism according to Trubetzkoy (39-consonant model)[edit]

A graphic representation of the Bulgarian consonant system according to the Bulgarian Academy of Sciences and based on Trubetzkoy's ideas follows below (39 consonants):^[46]

The 39-consonant model is inextricably linked to Russian linguist Nikolai Trubetzkoy. A refugee from the Bolshevik Revolution, he settled in Sofia in 1920, where he was granted tenure at Sofia University.^[94] Eventually, he moved to Vienna and became one of the founders of the immensely influential Prague Linguistic Circle.^[95] In his magnum opus, Principles of Phonology, published posthumously in 1939, he referenced extensively Eastern Bulgarian, even offering a model phonemic inventory it.^[11] There he argued in favor of the existence of the distinctive feature of palatalization in Bulgarian, establishing 14 contrastive pairs of hard and palatalized consonants. The consonant inventory suggested by Trubetzkoy consisted of 36 consonants, including й (/j/), but not дз (/d͡z/), дз' (/d͡zʲ/) and х' (/ç/).

Both Stoyko Stoykov and Lyubomir Andreychin, had rejected Trubetzkoy's idea in the early 1940s. It took them nearly 10 years to rediscover it. In the meantime, the country was occupied by the Soviet Union. The new regime reformed the orthography, throwing out all letters not present in Russian "as a manifestation of Great Bulgarian chauvinism"^[96] and introduced many Russian loanwords as part of a propaganda campaign. The atmosphere of pervasive sovietisation and russification of the period 1944–1954 proved far more conducive to the realization of Trubetskoy's ideas. By the turn of the decade, Stoykov had changed heart and consequently published Trubetzkoy's consonant model, adding 15 palatalized consonants to his analysis of the Bulgarian phonemic inventory.^[97] The other major postwar Bulgarian linguist, Lyubomir Andreychin, then quickly suggested another two, /d͡z/ and /ç/, arguing that even though they only existed in foreign proper names like Хюстън /xʲustɤn/ ('Houston') and Ядзя [jad͡zʲa] ('Jadzia') and had no contrastive function, they could have one, if need be.

Stoykov eventually conceded, and after the most distinguished Bulgarian phonetician of the totalitarian period, Димитър Тилков [bg], also agreed to the inclusion ("as they were envisaged by the system"), the 39-consonant system was set in stone.^[98] Tilkov designated /d͡zʲ/ and /ç/ as "potential phonemes", adding ф' (/fʲ/) to them in 1982, as it existed in only a handful of words, all of them borrowings (e.g., фюрер [ˈfʲurɛr] ('Führer')).^[99] The “potential phoneme”approach has not enjoyed much support abroad, where most authors generally omit not only /d͡zʲ/ and /ç/, but also /d͡z/.^{[100][101][102]}

While the consonant model was lauded in the Soviet Union by the likes of, e.g., Yuriy Maslov, acceptance in the West, except for Klagstad, has been lukewarm. Most of those who have opted to go with it rather than with the alternative model routinely call into question parts of it or make caveats. The most prolific Bulgarian phonologist and grammarian in the English-speaking world, Ernest Scatton, notes (1993):^[103]

Alveo-palatal obstruents are weakly palatalized. Palatalized labials are pronounced by many Bulgarians as sequences of [Cj].

In the compilation Common and Comparative Slavic (1996), American Slavist Charles E. Townsend states:^[18]

Palatalization is marked by following vowels as in R[ussian]. Extent of distinctive palatalization is debated; most agree on n/n', l/l', k/k', g/g'. Our inventory lists B[ulgarian] as having some 37 consonants, but this is an idealized number. The real number obviously depends on how many palatalized consonants one recognizes as independent morphophonemes. A great deal of controversy surrounds this question, though, in spite of the large number of phonetic contrasts, phonemic palatalization is more circumscribed than in R. For one thing, phonemic palatalization in B is clearly secondary; we recall that SSL South Slavic Languages in general suppressed the development of palatalization quite early, and not only in SC [Serbo-Croatian], but also Sln [Slovenian] and Mac [Macedonian] (close as the latter is to B) do not show any phonemic contrasts. For another thing, palatalization in B consonants is distinctive only before non-front vowels, and palatalized consonants never occur in final position or before other consonants.

According to Voegelin (1965):^[104]

Hodge and Bidwell treat the palatalized consonants not as separate unit phonemes (as given in the inventory above) but as clusters of consonants + /j/, which occur only before non-front vowels. There is more agreement among the sources in the treatment of /l', n', k', g'/ as palatalized phonemes than in the treatment of the other palatalized phonemes.

Palatalization[edit]

Palatalization refers to a type of consonant articulation, where a secondary palatal movement similar to that for /i/ is superimposed on the primary movement associated with the consonant's plain counterpart.^[105] During the palatalization of most hard consonants (bilabial, labiodental and denti-alveolar consonants), the middle part of the tongue is raised toward the hard palate and the alveolar ridge, which leads to the formation of a second articulatory centre whereby the specific palatal "clang" of the soft consonants is achieved. The articulation of palatalised alveolars /l/, /n/ and /r/ normally does not follow that rule. The palatal clang is instead achieved by moving the place of articulation further back towards the palate so that /ʎ/, /ɲ/ and /rʲ// actually become alveopalatal (postalveolar) consonants. In turn, the articulation of soft /ɡ/ and /k/ (transcribed as /ɡʲ// and /kʲ// or /ɟ/ and /c/) moves from the velum towards the palate, and they are therefore considered palatal consonants.

Articulation of t and tʲ in Contemporary Standard Bulgarian

Articulation of t and tʲ in Russian

However, the only articulatory study of palatalized consonants in Bulgarian, conducted by Stoyko Stoykov via X-ray tracings of vocal tract configurations of hard/palatalised consonant pairs, indicates that the secondary palatal movement is missing (or severely weakened) during the articulation of a number of palatalized consonants.^[106] Only the articulation of bilabial and labiodental consonants (/pʲ/, /bʲ/, /mʲ/, /fʲ/, /vʲ/) is accompanied by a noticeable raising of the body of the tongue towards the palate, but only to a moderate extent.^[107] The articulation of soft /k/, /ɡ/ and /x/ (/c/, /ɟ/ and /ç/) also shows distinctive palatalization, as the place of articulation moves onto the palate.^[108]

However, in denti-alveolars (/tʲ/, /dʲ/, /tsʲ/, /dzʲ/, /sʲ/, /zʲ/), the place of articulation neither shifts towards the palate, nor is the tongue raised. Instead, they are articulated with the blade of the tongue (laminally) rather than the tip (apically), which results in greater surface contact of the tongue front and a modification of the primary articulatory gesture.^[109][110] Stoykov defines them as “weakly palatalized”, while Scatton notes that the position of the mid-tongue in palatalized stops is not much higher than that in their plain counterparts.^[111][112] A comparison with the articulation of the same consonants in a language where palatal consonants indisputably exist, such as Russian, reveals drastically different articulation, with Bulgarian being completely non-conformant with the definition of palatalization.^[113] A comparison of the articulation of bilabials and labiodentals (/pʲ/, /bʲ/, /mʲ/, /fʲ/, /vʲ/) in Bulgarian also reveals much less pronounced secondary palatal gesture than in Russian.

The articulation of /ʎ/, /ɲ/ and /rʲ/ is very similar to that of the denti-alveolars, but with a slight shift of the place of articulation towards the palate and some raising of the mid-tongue towards the palate.^[114] According to Stoykov, /ʎ/ and /ɲ/ are harder than their counterparts in the other Slavic languages, while /rʲ/ is just as palatal.^[114] Based on Stoykov’s study, several foreign and Bulgarian phonologists have noted that distinctive palatalization in Bulgarian can be only claimed in the cases of /c/, /ɟ/, /ʎ/ and /ɲ/,^{[115][116][104]} or /c/, /ɟ/, /ç/ and /ʎ/.^[35]

Moreover, a study of the perception of hard and palatlized consonants conducted by Tilkov in 1983 has indicated that with the exception of palatalized velars (/c/, /ɟ/, /ç/), Bulgarian listeners needed to hear the transition to the vowel to correctly identify a consonant as soft.^[117] All this has raised the question whether Bulgarian palatals have indeed lost their secondary articulatory gesture and have decomposed into CjV sequences, as claimed by Danchev, Ignatova-Tzoneva, Choi, etc.

A 2012 perception study of palatalized consonants in Bulgarian compared with a language where palatalization is indisputed (Russian) and a language where such consonants are undoubtedly articulated as CjV clusters (English) concluded that unlike English listeners, Russian and Bulgarian listeners could identify a palatal(ized) consonant without waiting for the transition to the following vowel.^[118] The study also found similarities in the phonetic shape of palatal(ized) consonants in Bulgarian and Russian and marked differences between those in the two languages and English, disproving the hypothesis for the decomposition of palatalization put forward by Horalek, Ignatova-Tzoneva, Choi, etc.^[118] Nevertheless, based on the phonological distribution of Bulgarian palatals, which was similar to that in English and completely different from that in Russian, the author argued in favour of CjV notation.^[118]

Palatalization of *tj/*gt/*kt and *dj in Bulgarian[edit]

While the results of the three Slavic palatalizations are generally the same across all or most Slavic languages, the palatalization of *tj (and the related *gti and *kti) and *dj in Late Common Slavic led to vastly divergent result in each individual Slavic language.

Bulgarian *tj/*kti/*gti and *dj reflexes ⟨щ⟩ ([ʃt]) and ⟨жд⟩ ([ʒd]), which are exactly the same as in Old Church Slavonic, and the near-open articulation [æ] of the Yat vowel (ě), which is still widely preserved in a number of Bulgarian dialects in the Rhodopes, Pirin Macedonia (Razlog dialect) and northeastern Bulgaria (Shumen dialect), etc., are the strongest evidence that Old Church Slavonic was codified on the basis of a Bulgarian dialect and that Bulgarian is its closest direct descendant.^[120] Though the ⟨ʃt⟩/⟨ʒd⟩ speaking area currently covers only the territory of the Republic of Bulgaria and the eastern half of the wider region of geographical Macedonia, toponomy containing ⟨ʃt⟩ and ⟨ʒd⟩ that goes back to the Early Middle Ages is widely preserved across Northern and Central Greece, Southern Albania, the Republic of Macedonia, Kosovo and the Torlak-speaking regions in Serbia.^[120][121]

Reflexes of Proto-Slavic *tʲ/kt and *dʲ in the wider Macedonian region

For example, in the Struga municipality, the names of 13 out of 43 villages contain either ⟨ʃt⟩ (Kališta, Korošišta, Labuništa, Moroišta, Piskupština, Radolišta, Tašmaruništa, Velešta and Vraništa) or ⟨ʒd⟩ (Delogoždi, Mislodežda, Radožda and Zbaždi).^[120] The same applies to Kosovo, where Russian Slavist Afanasiy Selishchev found а number of place names around the city of Prizren featuring the Bulgarian clusters ⟨ʃt⟩/⟨ʒd⟩ in a Serbian official document from the 1300s (Nebrěgošta, Dobroušta, Sěnožeštani, Graždenikī, Obražda, Ljubižda, etc.).^[122] At present, a total of 8 villages out of 76 villages in the Prizren municipality still feature the Bulgarian consonant clusters ⟨ʃt⟩/⟨ʒd⟩, even though the region has not been ruled by Bulgaria in eight centuries: Lubizhdë, Lubizhdë e Hasit, Poslishtë, Skorobishtë, Grazhdanik, Nebregoshtë, Dobrushtë, Kushtendil. There are also numerous toponyms with the two clusters in the districts of Vranje, Pirot, Knjaževac, etc. in Serbia proper.^[123]

The development of ⟨ʃt⟩ > /c/ and ⟨ʒd⟩ > /ɟ/ in certain dialects in the geographic region of Macedonia is a late and partial phenomenon dating back to the Late Middle Ages, probably caused by the influence of Serbian /t͡ɕ/ and /d͡ʑ/, and possibly aided by the Late Middle Bulgarian's trend to palatalise /t/ and /d/ and then transform them into soft k and g > /c/ & /ɟ/.^{[124][125][126]}

Phonation[edit]

Phonation is a primary distinctive feature for obstruents in Bulgarian, dividing them into voiced and voiceless consonants. Obstruents form 8 minimal pairs: /p/↔/b/, /f/↔/v/, /t/↔/d/, /t͡s/↔/d͡z/, /s/↔/z/, /ʃ/↔/ʒ/, /t͡ʃ/↔/d͡ʒ/, /k/↔/g/.^[127] The only obstruent without a counterpart is the voiceless fricative /x/, whose voiced counterpart /ɣ/ does not exist as a separate phoneme in Bulgarian. The sonorants /m/, /n/, /l/ and /r/ and the approximant /j/ are always voiced.

If the existence of separate palatalised consonant phonemes (39-consonant model) is accepted, 6 more contrastive obstruent pairs are added: /pʲ/↔/bʲ/, /fʲ/↔/vʲ/, /tʲ/↔/dʲ/, /sʲ/↔/zʲ/, /tsʲ/↔/dzʲ/,/ɟ/↔/c/, for a total of 14.

Voicing, devoicing, assimilation, sandhi, ellision[edit]

Like all other Slavic languages apart from Serbo-Croatian and Ukrainian, Bulgarian features word-final devoicing of obstruents, unless the following word begins with a voiced consonant.^[128] Thus, град is pronounced ['grat] ('city'), жив is pronounced ['ʒif] ('alive'). While obstruents devoice before enclitics (град ли ['gratli] ('а city?')), they do not devoice at the end of prepositions followed by a voiced consonant (под липите [podli'pitɛ] ('under the lindens')).

CSB also features regressive assimilation in consonant clusters. Thus, voiced obstruents devoice if they are followed by a voiceless obstruent (e.g., изток is pronounced ['istok]) ('East')), and voiceless obstruents voice if they are followed by a voiced obstruent (e.g., сграда is pronounced ['zgradɐ] ('building')).^[129]

Assimilation also occurs across word boundaries (in the form of sandhi), for example, от гората is pronounced [odgo'ratɐ] ('from the forest'), while над полето becomes [natpo'lɛto] ('above the field').^[130]

The consonants /t/ and /d/ in consonant clusters such as стн [stn] and здн [zdn] are usually not pronounced, unless the articulation is very careful, i.e., вестник tends to pronounced as ['vɛsnik] (‘newspaper’), while бездна tends to pronounced as ['beznɐ]) (‘abyss’).^[131]

Distribution of voiced and voiceless consonants in Bulgarian[edit]

Distribution of Voiced & Voiceless Consonants in Standard Bulgarian^[132]

Position	Consonant
	b	p	v	f	d	t	z	s	d͡za	t͡s	ʒ	ʃ	d͡ʒ	t͡ʃ	g	k	x	m	l	n	r	j
Position I: Before central and back vowels (/a/, /ɤ/, /ɔ/, /u/)	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y
Position II: Before front vowels (/ɛ/, (/i/)	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y
Position III: Before sonorants (/m/, /n/, /r/, /l/)	Y	Y	Y	Y	Y	Y	Y	Y	N	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y
Position IV: Before /v/	Y	Y	Y	Y	Y	Y	Y	Y	N	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y
Position V: At word end	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	Y	Y	Y	Y	Y	Y
Position VI: Before voiceless consonants	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	N	Y	Y	Y	Y	Y	Y	Y
Position VII: Before voiced consonants	Y	N	Y	N	Y	N	Y	N	N	N	Y	N	Y	N	Y	N	N	Y	Y	Y	Y	Y

Consonant classification based on place and manner of articulation[edit]

Place of articulation[edit]

The consonants:

• п (/p/), б (/b/), and м (/m/) are bilabial;
• ф (/f/), в (/v/) and the allophone of /m/ and /n/ before /f/ and /v/, [ɱ], are labiodental;
• т (/t/), д (/d/), ц (/t͡s/), дз (/d͡z/), с (/s/), з (/z/) and the velar allophone of л, [ɫ], are dento-alveolar and sometimes also even described as dental;
• р (/r/) and н (/n/), as well the "light" л (/l/) are alveolar;
• ш (/ʃ/), ж (/ʒ/), ч (/t͡ʃ/) and дж (/d͡ʒ/) are postalveolar;
• й (/j/) is palatal; and
• к (/k/), г (/g/), х (/x/), along with the allophone of /n/ before velars, [ŋ], are velar.^[133][115]

The palatalized allophones of

• п (/p/), б (/b/), м (/m/), ф (/f/) and в (/v/) are pronounced by many Bulgarians as sequences of C+j.;^[134]
• {{Script|Cyrs|т (/t/), д (/d/), ц (/t͡s/), дз (/d͡z/), с (/s/) and з (/z/) remain dento-alveolar but are articulated with the back of the tip of the tongue (laminally) rather than with the tip (apically);
• р (/r/), н (/n/) and л (l) shift towards the palatum and become alveo-palatal; and
• к (/k/), г (/g/) and х (/x/) are articulated at the back of the palatum instead of the velum.^[134]

Manner of articulation[edit]

• Plosives (7): п (/p/), б (/b/), т (/t/), д (/d/), к (/k/), г (/g/);
• Fricatives (7): ф (/f/), в (/v/), с (/s/), з (/z/), ш (/ʃ/), ж (/ʒ/), х (/x/);
• Affricates (4): ц (/t͡s/), дз (/d͡z/), ч (/t͡ʃ/) and дж (/d͡ʒ/);
• Nasals (2): м (/m/) and н (/n/)
• Trills (1): р (/r/)
• Laterals (1): л (/l/)
• Approximants (1): й (/j/)^[133]

Word stress[edit]

Stress is not usually marked in written text. In cases where the stress must be indicated, a grave accent is placed on the vowel of the stressed syllable.¹³

Bulgarian word stress is dynamic. Stressed syllables are louder and longer than unstressed ones. As in Russian and other East Slavic languages, as well as English, Bulgarian stress is also lexical rather than fixed as in French, Latin or the West Slavic languages. It may fall on any syllable of a polysyllabic word, and its position may vary depending on the inflection and derivation, for example:

• nouns – мъ̀ж /mɤʃ/ ('man'), мъжъ̀т /mɐˈʒɤt/ ('the man'), мъжѐ /mɐˈʒɛ/ ('men'), мъжѐте /mɐˈʒɛtɛ/ ('the men')
• verbs – отѝвам /oˈtivɐm/ ('I am going'), отидѝ /otiˈdi/ ('go!')

Bulgarian stress is also distinctive: the following examples are only differentiated by stress (see the different vowels):

• nouns
  • въ̀лна /ˈvɤɫnɐ/ ('wool'), вълна̀ /vɐɫˈna/ ('wave')
  • па̀ра /ˈparɐ/ ('steam'), пара̀ /pɐˈra/ ('coin')
• verbs
  • когато до̀йде /koˈɡato ˈdɔjdɛ/ ('when he comes'), когато дойдѐ /koˈɡato dojˈdɛ/ (when he came')
  • взрѝвен /ˈvzrivɛn/ ('explosive'), взривѐн /vzriˈvɛn/ ('exploded')¹⁴

Stress usually isn't signified in written text, even in the above examples, if the context makes the meaning clear. However, the grave accent may be written if confusion is likely. ¹⁵

The stress is often written in order to signify a dialectal deviation from the standard pronunciation:

• каза̀ ми /kɐˈza mi/ ('he told me'), instead of каза ми /ˈkazɐ mi/
• иска̀ да дойде /iˈska dɐ dɔjdɛ/ ('he wanted to come'), instead of искаше да дойде /ˈiskɐʃɛ dɐ dɔjdɛ/)¹⁶

References[edit]