1. Slavic and Old Church Slavonic

1.1. The Slavic languages

1.1.1. Distribution of the Slavic languages

Slavic languages are distributed across Central and Eastern Europe, in the Balkans and in Central and North Asia (Siberia) (see Figure 2). They are spoken by around 315 million people.

Figure 2: Modern distribution of the Slavic languages. Based on Asher & Moseley (2007)

1.1.2. Traditional subgrouping

The Slavic languages (Glottocode: slav1255) are traditionally divided into three main subgroups.

1. South Slavic (sout3147)

2. East Slavic

3. West Slavic

A typical family tree can be seen in Figure 3.

Figure 3: Traditional subgrouping of the Slavic languages.

In practice, the Slavic languages constitute a dialect continuum, with overlapping isoglosses that make it difficult to make a clear phylogenetic model. While the traditional tripartite subgrouping may not reflect the true phylogenetic relationships among the languages, it still constitutes a useful framework for making generalised statements about them. For historical overviews see Blažek (2020); Saenko (2020).

It is difficult to find systematic treatments of the subgrouping of Slavic based on shared innovations (cf. Mareš 1980; Schenker 1993: 60–61; Poljakov 2018: 1588–1589). Criteria that are mentioned in connection with the subgrouping of Slavic are mostly phonological, e.g.:

There are morphological and lexical differences as well, but they are more difficult to get hold of.

I am not aware of any computational studies dealing specifically with the subgrouping of Slavic; those concerned with the entire Indo-European language family are based on lexicon and of dubious value for Slavic (Gray & Atkinson 2003: 437, implying that Polish is more closely related to East Slavic than to Czech and Slovak; Bouckaert et al. 2012: Supplementary Materials p. 24, again with Polish in a weird place; Heggarty et al. 2023: Supplementary Materials p. 57, with OCS and Slovene forming an outlier; for criticism of the two former publications see Pereltsvaig & Lewis 2015: 83–85).

1.1.3. The position of the Old Novgorod dialect

Spoken in the Novgorod–Pskov area in north-western Russia, it seems to be an early outlier in the Slavic family (“North Slavic”) (Nasirpour 2025; Andersen 2026). Some of the features characteristic of the Old Novgorod dialect distinguishing it from the other Slavic dialects are the following:

Old Novgorod dialect does not have the regressive part of second palatalisation (Zaliznjak 2004: 45–47); see 3.4.4.

There is probably no progressive palatalisation of *‍x (and possibly *‍g).

According to a minority view, the Old Novgorod dialect also shows a special development of PS *‍ə̄̆ (from pre-PS *‍ā̆ in final syllables ending in *‍-s) (Zaliznjak, Nikolaev and Dybo apud Zaliznjak 2004: 148–149; Olander 2012): PS *‍ə ə̄ > “classical” CS *‍ŭ y || ONovg. e ě

1.2. Proto-Slavic and Common Slavic

“Proto-Slavic” reconstructions are traditionally written as a kind of lightly modified OCS; for instance, the word for ‘water’ is reconstructed as *‍vodà (cf. OCS voda, Ru. vodá, Po. woda etc.), and the word for ‘whole’ is reconstructed as *‍cě̑lŭ (cf. OCS cělŭ, Ru. célyj, Po. cały etc.).

However, from (1) loanwords to and from Slavic, (2) internal phonological correspondences in the Slavic languages and (3) information from the Old Novgorod dialect we can see that actual Proto-Slavic – in the sense of the language state from which all the known Slavic languages can be derived – looked quite different: the word for ‘water’ probably rather sounded *‍u̯aˈdā in actual Proto-Slavic, and the word for ‘whole’ was *‍ˌkai̯lə (see Olander 2015: 64; Holzer 1997)

1. Loanword relations: Traditionally, “Proto-Slavic” forms are reconstructed with the sequences *‍or ol.

But loanwords from Slavic that contain such sequences seem to always have *‍ar al, suggesting that Proto-Slavic *‍a had not yet become *‍o when the Liquid Metathesis took place. Since the Liquid Metathesis had different results in the Slavic languages, it must have taken place after the dissolution of Proto-Slavic.

Accordingly, forms like *‍golvà, *‍stornà and *‍gȏrdŭ probably never existed; the actual Proto-Slavic forms are *‍gālˈu̯ā, *‍starˈnā and *‍ˌgardə

2. Internal Slavic correspondences: It is easier to understand the Liquid Metathesis in South Slavic (and Czecho-Slovak) as a development *‍ar > *‍rā > *‍ra than *‍or > *‍ra, since if *‍o > *‍a happened before the Liquid Metathesis, *‍or would probably have become *‍rō or *‍ro, but this did not happen (Lindstedt 1991: 114–115).

3. The Old Novgorod dialect (see also 1.1.3): In contrast to all other Slavic languages, the Old Novgorod dialect clearly does not have the regressive part of the Second Palatalisation (see 3.4.4).

Thus, these reconstructions are anachronistic, but in reality give a good picture of the Slavic dialect continuum as it was spoken until the loss of the reduced vowels around 1200 CE (3.4.12). Moreover, it is practical to represent Slavic reconstructions in the traditional notation, partly because it is the way it has always been done, and partly because the words are immediately more easily recognisable. Accordingly, I (and others) refer to the traditional notation as “Common Slavic”, and “Proto-Slavic” denotes reconstructions of the actual common point of departure for the Slavic languages.

Proto-Slavic refers to the stage after the First Palatalisation (e.g. *‍ke > *‍če; 3.4.1) and before changes like the vowel palatalisations (e.g. *‍i̯a > *‍je; 3.4.2), the monophthongisation of oral diphthongs (e.g. *‍ai̯> *‍ě; 3.4.3) and the Second Palatalisation (e.g. *‍kai̯ > *‍cě; 3.4.4) (cf., somewhat differently, Klotz 2023; Olander 2015: 46–67).

1.3. Slavic and the other Indo-European branches

1.3.1. Slavic and Baltic

The relationship between Baltic and Slavic is debated: have they developed in parallel since PIE times, or did an actual Balto-Slavic proto-language exist at some point in prehistory?

At this point it is probably safe to say that Baltic the communis opinio assumes a Balto-Slavic subgroup of Indo-European (for literature on the problem see Olander 2019; Wandl & Olander 2026). Notably, all phylolinguistic studies of the Indo-European language family that I am aware of group Baltic and Slavic together, with no exceptions.

Among the significant common innovations found in Baltic and Slavic are the following:

1.3.2. The position of Balto-Slavic in the Indo-European family

It is disputed what the closest relative of the Balto-Slavic subgroup is. It has been argued that the ruki change (3.2.3) may be a shared innovation of Balto-Slavic and Indo-Iranian (e.g. Ringe & Taylor 2007: 6–7; see also Ringe, Warnow & Taylor 2002; Nakhleh, Ringe & Warnow 2005; but cf. Hock 2021: 594–597, arguing that Ruki is a transitional phenomenon). On the other hand, Balto-Slavic and Germanic share the striking use of *‍m for *‍bʰ in oblique dual and plural case endings, which also has the appearance of a shared innovation.

Some authors give a negative appraisal of the possibility of grouping Balto-Slavic with any other Indo-European branch (e.g. Pronk 2022: 279–285).

A preliminary majority consensus tree resulting from the research project Connecting the Dots: Reconfiguring the Indo-European family tree can be seen in Figure 4.

Figure 4: Indo-European family tree (Connecting the Dots project).

For an attempt at matching population genetic relationships with linguistic branches see Gretzinger (2026: 31). For the genetic background of Slavic-speaking populations see Gretzinger et al. (2025); Gretzinger (2026: 38–39).

1.4. Old Church Slavonic

1.4.1. The OCS corpus

OCS is the first ample direct attestation of the Slavic branch of Indo-European. The OCS manuscripts date to the 10th (the Kiev Leaves) and 11th centuries, which is late compared to many other branches of Indo-European. The Freising Fragments (Sln. Brižinski spomeniki) in Old Slovene are from around 1000 AD.

The language of the OCS corpus is based on a South Slavic dialect closely related to later Bulgarian and Macedonian dialects. OCS is sometimes referred to as Old Bulgarian; cf. e.g. the title of Leskien’s Grammatik der altbulgarischen (alt­kirchen­slavischen) Sprache (Leskien 1919a).

The OCS texts all serve religious purposes. The following shows the OCS manuscript, the script they are written in, and the number of folios they contain (Andersen & Olander forthcoming; cf. Vaillant 1964: 14–16; Schaeken & Birnbaum 1999: 87–91).

1. Gospel

2. Aprakos Gospel (gospel readings for feast days)

3. Apostolos (Acts of the Apostles, Catholic epistles and Pauline epistles from the New Testament)

4. Psalter

5. Liturgical texts (for church services)

6. Edificational literature

1.4.2. The creation of the Glagolitic and Cyrillic alphabets

The following account is a shortened version of the one found in Andersen & Olander (forthcoming).

Around 862 two brothers from Thessaloniki, Methodius (c. 815–885) and Cyril (b. Konstantinos, c. 826–869), were sent on a mission from Constantinople to Moravia (present-day eastern Czechia and Hungary), where they spent three years. The brothers spoke not only Greek, but also a South Slavic dialect; Thessaloniki had a large Slavic-speaking population.

According to tradition, Cyril created the Glagolitic alphabet, and the brothers translated the most important liturgical texts into their variety of South Slavic. The original language of the translations is referred to as “Urkirchenslavisch”, German for “Proto-Church Slavonic” (see Schaeken & Birnbaum 1999: 13–15).

While the mission to Moravia ultimately failed, the followers of Cyril and Methodius relocated to Bulgaria and established a highly influential school there.

While the Cyrillic script is named after Cyril, most scholars agree that it was created later, on the basis of Glagolitic. Most of the Cyrillic letters are derived from Byzantine Greek majuscules (uncial, large book script letters), with Byzantine Greek pronunciation.

Both scripts are based on a phonological analysis of OCS.

The majority of the OCS manuscripts are written in the Glagolitic script; an important exception is the Codex Suprasliensis, written in Cyrillic (see 1.4.1).

Note that in this course we do not use the Glagolitic script; OCS is given in Cyrillic and/or Roman.

Glagolitic does not distinguish between ě and ja, nor e and je.

Table 1: The Glagolitic and Cyrillic alphabets.
Cyr. Glag. translit. name Cyr. Glag. translit. name
а Ⰰ a azŭ ф Ⱇ f frĭtŭ
б Ⰱ b buky х Ⱈ x xěrŭ
в Ⰲ v vědě ц Ⱌ c ci
г Ⰳ g glagoli ч Ⱍ č črĭvĭ
д Ⰴ d dobro ш Ⱎ š ša
є е э Ⰵ e, je jestŭ щ Ⱋ št šta
ж Ⰶ ž živěte ъ Ⱏ ŭ (ъ) jerŭ
ѕ  Ⰷ dz dzělo ы ъи ⰟⰉ ⰟⰋ y jery
з Ⰸ z zemlja ь Ⱐ ĭ (ь) jerĭ
і Ⰹ Ⰺ i i ѣ Ⱑ ě jatĭ
и Ⰻ i iže ю Ⱓ ju –
(ћ) Ⰼ ǵ ǵervĭ я – ja –
к Ⰽ k kako ѥ – je –
л Ⰾ l ljudije ѧ  ⱔ ę –
м Ⰿ m myslite ѩ ⱗ ję –
н Ⱀ n našĭ ѫ ⱘ ǫ –
о Ⱁ o onŭ ѭ ⱙ jǫ –
п Ⱂ p pokojĭ ѯ – ks ksi
р Ⱃ r rĭci ѱ – ps psi
с Ⱄ s slovo ѳ ⱚ θ θita
т Ⱅ t tvrĭdo ѵ ⱛ ü ižica
ѹ у Ⱆ u ukŭ ѡ Ⱉ o otŭ

A transliteration tool for Early Cyrillic to Latin is available at sproghistorie.dk/ocstrans.

1.4.3. Local variants of the Cyrillic alphabet

Variations of the Cyrillic script are still used across eastern Europe and Asia. A reform by Peter the Great (1672–1725) in 1708 changed the shape of the Cyrillic letters to a form resembling the Roman script (see this blog post with illustrations).

In Bulgaria, Macedonia and Serbia local variants of the letter shapes of the Cyrillic alphabet are used, as illustrated in Table 2.

Table 2: ‘Water’ in local variants of the Cyrillic alphabet.
regular italic
Russian вода вода
Bulgarian вода вода
Serbian / Macedonian вода вода

1.5. Handbooks

The Slavic languages in general and OCS in particular have a long and rich tradition of handbooks. The list below is not exhaustive; see Vaillant (1964), 16–18 for references to older handbooks.

1.5.1. Synchronic dictionaries of OCS

1.5.2. Etymological dictionaries of Slavic languages

There are numerous etymological dictionaries of the individual Slavic languages. Here are some of the important ones.

1.5.3. Synchronic grammars of OCS

1.5.4. Historical grammars of Slavic languages