Evolution of ъ & ь in position near j

18 Feb, 2026

1. General Remarks

It includes other topics as i at the start of words, i (ji) after a vowel in the middle of a word, y and i before j; ь or ъ + j within morphems and (for the latter) suffixes.

Originally, in Proto-Slavic (with exception as proto-Russian dialects) yers couldnʼt be before j, where ĭ preserved, and ŭ was changing to y̆, nor after j, where ĭ was preserved. Interpretation of those ĭ and y̆ during yers-declining and alittle later reminds interpretation of strong and weak yers; but more detail analysis shows notable differences. Foremost itʼs about ji at the start of the word; however in the middle of a word i after j had the same principle as for yers: was declined in a weak positions, and turned into a full vowel in a strong position. A vowel y before j had no changes, and i, with some specificity, in general is interpatated as yers.

At this time in some cases — withing boundary morphemes and words — near j yers were introduced. Among them: boundaries pronominalized adjectives in nominal singular masculine (dobrъ + *јь), prefixes (podъ + jimati), sandhi (lʼubimъ + *јь) and so on. Those yers were declined under a general rule too, while there were some specificity.

5. Combination ъ + j withing a morpheme boundaries

Old Ukrainain didnʼt have ъ + j within a morphemes, because in proto-Ukrainian dialects there were no changes ŭ into ъ before j. However, withing a morpheme boundaries the such combination occured during word buildings and combinations. Thereʼre three the most common cases:

1. Pronominalized adjective forms. Most noun forms of adjectives were ending with -й > -ъ in nom. sg. m (and gen. pl.). They were undergoing pronominalizatiion, when a pronoun *jĭ was added. This process happened mostly during late Proto-Slavic period, not attested in written memory, thus itʼs hard to document what was the first: changing -ŭ > ъ created a pronoun type adjectives or reverse. Literature standard of Ukrainian, and dialects point out the latter: *dobrŭ + jĭ > *dobry̆jĭ. Hovewer returning ъ sound from a noun form dobrъ doesnʼt set a problem [too].

   Point pronoun tъ had the similar undergoing, while he pronominalized later, during Old and Middle Urkainian. In this pronoun ъ changed into y before *jĭ (probably, itʼs not about regular changes, but taking an example from adjectives). Occured tyj basically in all dialects. But there were also concurents forms, mostly той and/or тот. In the modern language, a form tyj preserved, obviously, in a word ти́ждень < тый-жє-дєнь (as Polish tyzdień).

2. Within boundary of words -ъ and j-. Could be -ъ and j- counted as non-possible — no possible to find out, but thereʼre some certainty that in most cases ъ sounded as y. Old Ukrainian written history has many examples of the such case, while in Old Church Slavonic itʼs rare, and in Old Russian is almost never. When the combination -ъ and j- written as it in Old Ukrainian written texts, itʼs hard to tell: is it a traditional (morphophonological) writing, or exposure of swaying and facultative in coloquial speeching. From the correct view, choosing -ъ or -ы was facultative, and, obviously, free. Anyway, adequacy of writing with -ы (< -ъ) approved by the fact, that it were still occured after declined of yers. If those letters had passed as -ъ then it would be declined; thus it should be a sound y which also preserved in weak positions. From the other side, letter -ы at most late of ⅩⅣ were starting to lost; probably, it approves of facultative of this sound.

   Declining the such forms with sandhi started, probably, at the end of ⅩⅢ. From this time we see (j)i- with adjectives with -y, forms with -o: ко Ис(ѹс)ѹ (1288), во исходъ 'на вихід' (1307), изо истлѣньѧ 'від зотління' (ⅩⅣ). Forms with -o which had been appeared before weak year without before-j now moved into position j-: there were no reason to differ them then.

3. In positions before a root with j- prefixes with -ъ were undergoing a general rule of ъ → y, but in other positions they often ended with -ъ. With tendention to spreading one form on others, it brings disputes of interpretations. However, a sound inovation, obviously, got the upper hand, because y practically spreaded onto other forms, which were during all Middle Ukrainian and still preserved in other dialects, especially in the nothern, Over-Sian, Lemki, Carpathian (Northen-Western from Mukačevo): if in this or that dialect they distinguish Old Ukrainian reflexes of i and y then for certain is the second (as example, Lemki: одыйти, розыбрати). And more: from prefixes with -ъ (sъ, podъ, nadъ, peredъ), y was spreaded to prefixes, where -ъ was facultative (ot ~ otъ), and to next (vъz, roz, bez, iz), and finally — to words, which even didnʼt started with j (in some dialects, mostly northern, in this sound position, even j was declined).

So, we can say about three stages in prefix history which ends with a consonant (with or no yer in original forms). During Old Ukrainian (at least in texts), in possition before j- thereʼre no ъ → y, even in prefixes with -ъ; other prefixes didnʼt cause changes in roots, and didnʼt changes themselves before a vowel (1). During Middle Ukrainian, y spreaded as a link-element between a prefix and a root with j- or a consonant cluster, while forms with -o were uncommon (2). In the modern language i (and partly o) changed Middle Ukrainian y (3).

So, typical for Middle Ukrainian spreading -y- in prefix form was caused by ъ → и before j, and later y was spreading to others roots with a consonant cluster. But, it omitted prefixes with etymologically -o as po-, pro- and do-. At late ⅩⅦ started new and wider unification: y was changed by o which alternated with i. It became possible, because alternation o : i became stable, and the were strongly linked. In this prefixes thereʼre no cases of y > i, nor ъ > i (while i often appears instead ъ), but about comples process of morphologic and morphonological reasons. Forms with y preserved only in periphery northern and west dialects. In Belarusian language, as in Ukrainian, had the first case (spreading y on all forms), but thereʼre no the second (o : i), because there were no alternations as o : i: Belarusian разыгра́ць, разыхо́дзіцца.

Letʼs show a table (arrows show a scope and direction of spreading a form)

6. Combination ь + j withing boundary morphemes and in sufixes

As with ъ + j, a combination ь + j was unnatural for Old Ukrainian: sound ĭ before j was preserved withouth changing into ь. Hovewer, the such combination occured within morphen and word bounders. Mentioned cases fo ъ + ј are actual for ь + j too, even if in other scope. Also, during analyzing some suffixes some specific problems are occured.

1. For pronoun-type adjective thereʼre the same rules as for ъ + ј, itʼs just less number of lexem. So, a modern adjective of a soft adjective as вечі́рній could be from proto-Ukrainian *večerĭnʼĭ + jĭ or *večerьnʼь + jĭ, later *večerьnʼĭjĭ — the first variation is more credible.

   Later, from Old Ukrainian epoch, the similar process happened with a point pronoun сь, in pronominalized form — *sь + jĭ. The appeared form sij (later syj) widely noted in Middle Ukrainian textes, while thereʼre also concurents as сєй and сєсь.

2. By sandhi -ь before j- regularity of facultative turned into i. In Old Ukrainian textes thereʼre many spelling with и, but a traditional variation ь aslo occured often too. The huge numbers of a form with и (< ь) before j- is a feature of Old Ukrainian. No difference from -ъ → -ы in meaning of time-limits, situation in the same positions.

3. Prefixes and adjetives didnʼt end with -ь with an exception обь which alternated with о ~ об. And existance of exactly this form in Old Ukrainian is under doubt. Probably, ѡбиимашє (Ⅻ) 3p-past-time shows this form with expected ь > і.

4. In gen. pl. noun of m. and f. soft group (with -i base) we could expect -yj < *ĭjĭ, which we can find in many dialects, especially Southern-West (somewhere western to line which going thro Pinsk — west from Rôvne — north to Ternopôlj — north to Xmeljnıcjkıj — South-West from Berdıčı — Eastern to Věnnıcja — thro Hajsın — West Pervomajsjk), also in areas around Xvastova, under Prıpjatj, around Kanôv, Poltava and north of Voroněž Bolast. But in most dialects and in the standard Ukrainain we have -ej: ноче́й, госте́й. This ending could be only from -ьjĭ. We can assume, that even from proto-Ukrainian — or at least before yers-declining — in most dialects ь appeared instead ĭ in this case. This ь exist in non-direct cases in nouns with -i base in plural: dat. pl. — гостьмъ, ночьмъ, instrum. pl. — гостьми, ночьми, loc. pl.--- гостьхъ, ночьхъ. This introduced ь in gen. pl (gostьjĭ, nočьjĭ) was undergone the same evolution as ь in dat., loc. pl.: turned into e after yers-declining. Other dialects saved original -ĭjĭ, which gave -yj. As additional reason, which powered spriding -ej, but occured later, itʼs, probably, reunderstanging gen. pl. nouns with -ĭja as in svinĭja, sudĭja. After yers-declinging they ends with -ij in gen. pl, thus had zero-ending with droping i (later y); usually itʼs e which was drop-out, so it changed i; finally — due reunderstanding — an end-combination -ej started act as an ending. So, during concurentions, some dialects prefered -yj (ночий), others -ej (свиней).

   Old Ukrainian texts have a small number about areal-appearence of gen. pl. -ьі and -ii before yers-declining, or -ej and -ij after. Forms -ii ~ -ij often occures in text from different regions. A choice to writing in Church Slavonic was free, while in Old Church Slavonic Bulgarian redation, -ии was more common.

   Distribution of two forms in texts ⅩⅣ–ⅩⅤ has a diffrent situation. In Southern-West texts prefers -ii (грошии Peremıšj, дѣтии Peremıšlj, людии Smotrıč against грошєй Lucjk, нєприꙗтєлєи Černjaxôv, near Žıtormır).

   From the first glance, this distribuation isnʼt systemic and very different, by with more analyze it show social and areal motivation. In general, -єй is mostly from Northern and Eastern of Ukraine, while -ий is mostly Western lands and Poltava. In Carpathian region itʼs mostly -yj (with an exception Raxôv). This distribution is similar to the current situation occured at late ⅩⅤ (if we donʼt count lands which populated later). Usage in the standard language -ej is a result of Southern-Eastern dialect during its developing at beg. ⅩⅨ, while a high number of -ий during Middle Ukrainian shows an active role of Southern-Western regions at this time culture life.

5. Aside gen. pl, e before j occured aslo in nom. sg. соловей : соловʼя, in dialects also горобе́й ~ веребе́й ~ воробе́й (northen Kıjôv, Černihôv, Sumı, Rôvne, Kobrın). In Ukrainian a suffix -ějĭ introduced in Southern-Western dialects (соловій : соловія), while an expected form *solovyj : *solov(y)ja isnʼt occured during nor Old (with an exception Church Slavonick славий), nor Middle Ukrainian. This forms was non-comfortable, because in nom. sg. it wouldnʼt differ with an pronoun type adj. For omitting this, in many Slavic languages and in some Ukrainian dialects, an other suffix was introduced. It became possible, because a vowel in norm. sg. (Old Urk. *solovij) in non-direct cases alternated with ∅; and in noun-suffixes, a typical alternator was e or o, not i (> y).

   The same problem was for *ručĭjĭ (Russian руче́й, Bulgarian ру́чей), But the problem was solved in proto-Ukrainian dialects by changing *-ĭj- with -ěj- (and after middle-palatal — -aj-), from there modern Ukrainian ручай (also Belarusian руча́й, Polish ruczaj). In солоне́й : солоне́ю with less usage a sound e just didnʼt appeared, as it happened with іне́й and кле́й (Rusian клей, Old Polish klij, comp. Middle Ukr. клий : клию). Different suffixes in ву́лик, dial. ву́лень, вулій and вере́я can show on proto-Ukrainian *ulĭjĭ and *verĭja-, comp. Russian у́лей, Bulgarian у́лей 'ясла'; in Hutsul dialecs still preserved a form улий, which is the same to Middle Ukrainian ѹлий; also Middle Ukrainian вєрѣѫ, вєрѣѧ; Old Church Slavonic вєрѣꙗ 'деревʼяна поперечина', Russian *верея́, Slovak veraj. And word глей doesn't have a suffix (comp. alternation: глей : глина from Indo-European ei : ī). In case of шлея ~ шлейка, itʼs possible a Polish loanword: comp. Polish szla ~ śla, pl. śle, where a plural form took as a main to which -ja was added.

7. Summary

[…] Secondary, created combination yers with j joint to the process at middle Ⅻ with the such results:

• ъ + ј (with ъ → y in strong position): мали́й, Middle Ukrainian одийти
• ь + ј (with ь → і (later i or y) in strong position, but ь → е in gen. pl. in East Ukraine): си́ній, діте́й (West Ukraine діти́й)

[…] In middle ⅩⅦ -y- which linked prefixes with roots was changed into i : o: одійти́, підо́йма.

Any process from above didnʼt bring changes in phonem system. Appeared two alternation — y : ∅, which later lost in prefixes (Middle Ukrainian: одийти : одходити* > Modern Ukrainian відійти́ : відходити), and in sufixes (if they even existed — comp. Middle Ukrainian *solovyj* : solovja > Modern Ukrainian солове́й : словлоʼя́).

8. Spreading area

All changes here which occured withing Ⅻ are actual for all Ukrainian dialects — with exception gen. pl. ending which divided into -ej and -yj by morphology and similar noun sufixes.

+ sameness

− difference

(4) — similarity Polish and Slovak evolution with Western Ukrainian dialects

(7) — lost in the Modern Bulgarian

#linguistic #ukrainian