History of the Polish language

The Polish language is a West Slavic language, and thus descends from Proto-Slavic, and more distantly from Proto-Indo-European; more specifically, it is a member of the Lechitic branch of the West Slavic languages, along with other languages spoken in areas within or close to the area of modern Poland: including Kashubian, Silesian, and the extinct Slovincian and Polabian.^[1]

The separation of Polish as a language is conventionally dated at the second half of the 10th century, linked with the establishment of Polish statehood and the Christianization of Poland.^[2] The history of the language is then be divided into the following periods periods of development: Old Polish (staropolski) with a pre-literate, pre-Polish era up to 1136, the literate era from 1136 with the Bull of Gniezno up to the start of the 16th century; Middle Polish (średniopolski) from the 16th century until the end of the 18th century (1772) with the first parition of the Polish–Lithuanian Commonwealth; New Polish (nowopolski) from 1772-1939; and Modern Polish, since World War II.^[3][4]

Proto-Slavic broke into three dialectal regions, Western Proto-Slavic, Eastern Proto-Slavic, and Southern Proto-Slavic.

Within declension, in North Slavic, the nominative-accusative feminine plural ending of *-ja stems leveled from *-ě (< PIE *-’ās) in the nominative plural and *-ę (<*-’åns < PIE *-’ans) in the accusative to *-ě for both cases: *dȗšę̇ > Old Polish dusze. The leveling of these cases for -ja stems was motivated by the fact that -a stems had the same ending -y for both cases. This also affected the accusative plural of soft-stem masculine nouns (e.g. *koňь, *mǫžь), which was motivated by the homophonous ending -y for feminine and masculine hard-stems.^[5]

Within West-Slavic, *ś shifted to *š: *vьśь gives Old Polish wiesz.^[6] The clusters *tl dl are retained in West Slavic, and elsewhere simplify to *l: *ordlo gives Old Polish radło.^[6] In initial *or-, ol-, as a result of the law of open syllables, metathesis occurs, which usually resulted in the lengthening of *o to *a, giving ra- la-. This occurred in all situations in South Slavic, but in North Slavic only in syllables with an acute, but those with a circumflex, *o remained as *o: *ordlo gives Old Polish radło.^[7] In addition, there was a tendency to raise mid and low long vowels: Old Polish pták, mléko, grzéch, wóz.^[8] There existed a western tendency to fix stress, except in Pomeranian and Polabian.^[9] The pronoun of *tъnъ was created in all of Western Slavic.^[9] Finally, the genitive and dative forms of the complex adjectival declension establishes for the masculine and neuter gender the endings resulting from the contraction of the endings -ego, -emu. This is in contrast with East and South Slavic *-ogo, -omu.^[10]

Within most of Lechitic, the clusters *TorT *TolT *TerT *TelT undergo metathesis: *vȏrgъ gives Old Polish wróg, *mȏldъ gives Old Polish młody, *dȇrvo gives Old Polish drzewo, *melkò gives Old Polish mléko (with a long e).^[11] Notably, both -ro- and -ar- between consonants for TorT a re present in Polabian and Pomeranian, compare reflexes of Kashubian karwa and Polabian korvo.^[12] *ť ď change to c dz: *svěťa > Old Polish świéca, *meďa > Old Poliss miedza, *noťь > Old Polish noc (świeca, miedza, noc).^[8] The nasal vowels are well retained in Lechitic, in contrast to the rest of West Slavic: *pę̑tь > Old Polish pięć, *pętъ > Old Polish piąty.^[13] Combinations of *ъr ъl ьr, ьl volcalize in various ways in Lechitic and Sorbian, but become vocalic liquids in Czech and Slovak.^[12]

Common to Lechitic is the Lechitic ablaut, albeit to varying degrees of intensity regionally: the front vowels *e *ě *ę as well as soft *ьr, ьl into two variants:^[12][14]

1. dispalatalization (dyspalatalizacja) whereby the vowel is fronted and lowered: *a *ǫ *ъr *ъl (later > ar eł) before the dental/alveolar hard consonants t, d, n, s, z, r, ł;

   *bě̃lъ > Old Polish biáły

   *lěto > lato

   *berǫ > biorę

   *nesǫ > niosę

   *květъ > kwiat

   *lěsъ > las

   *zelo > zioło

   *žena > żona

   *měra > miara

   *město > miasto

2. a front variant *e ę ьr, ьl (> ir il) before other consonants and word-finally

   *běliti > Old Polish bielić

   *vъ lětě > Old Polish w lecie.

The alternation equating to Polish eł:il is absent in Polabian due to a merging of *ьl and *ъl into just oł; this process is probably older than the vocalization of liquids (see § Proto-Slavic yers *ъr *ъl *ьr *ьl (*TъrT *TъlT *TьrT *TьlT)), since the merging of soft *ьr, ьl with hard *ъr ъl must have happened before vocalization.^[12] The source of the Lechitic ablaut likely came from Eastern Lechitic dialects, which later spread to the rest of Lechitic as the later e > o abluat is absent from Polabian, which is Western; compare żona, żenić miotła, siodło, and Polabian zéna, métla, sedlǘ.^[12]

The distribution of nasal in Polish were affected due to later sound developments.^[15] See also § Proto-Slavic *e for the articulatory-acoustic motivations for the ablaut.

The Lechitic ablaut likely took place in the 9th and 10th centuries and finished by the 12th century.^[16] As a result of ablaut, soft consonants became phonemic, as previously they only occurred before front vowels, but could now occur before back vowels, with minimal pairs like mara and miara.^[17] Similarly, the opposition of y and i is called into question, as words that were previously minimal pairs, e.g. być and bić could now be understood to differ phonemically in the softness of the consonant.^[18] Finally, ablaut led to many vowel alternations, e.g. e:a: wieźć:wiozę and e:a: (w) lesie:las.^[18]

There are many common morphological and especially lexical developments in addition to these common phonetic developments. This group can be broken into four: the western-most Obotrites and Drevani, the neighboring Veleti, the Pomeranians, including part of Western Pomerania, whereas Eastern Pomeranian (Slovincian-Kashubian) eventually later came under the influence of the rest of Lechitic), and the Polish tribes, which had much closer contact both socio-politically and linguistically as a result of the Polans, and the creation of Literary Polish.^[19]

The earliest attested names from the Middle Ages come mostly from the west, including the Silesian Dadosesani, Bobrans, Silesians, Opolans, Golensizi, and others, the Greater Polish Polans, Kuyavians; from the east only the Vistulans and Masovians (Masurians) are mentioned.^[20] Both the Vistulans and Masovians probably had other smaller associated tribes, records of which have been lost.^[20] Based on similarities between the dialects of these tribes, it is possible to surmise that Standard Polish was formed based on the closely related dialects of the Polans, Silesians, and VIstulans, and the Masovians had a weaker connection, and the weakest was the input from the eastern Pomeranians (Kashubians), as they were drawn into the sphere of Polish influence later.^[20]

These similarities between the Polans, Vistulans, and Silesians include:

1. *ьl > -eł- before front hard consonants (wełna, pełny, zmełty) and -il- before others (wilga, wilk, pilśń), whereas in Masovian and Kashubian -’oł- appears after hard consonants and -’áł- after a soft consonant: wiołna, miołty, Old Kashubian wiáłna, piálny (piáln-i);^[21]
2. Greater Polish and Lesser Polish show voice final consonants appearing before a word beginning with anything but a voiceless consonant (udźwięcznienie międzywyrazowe), e.g kod idzie (kot idzie), bog równy (bok równy), brad niesie (brat niesie), zanióz em (zaniósłem), zmóg em (zmógłem), and Masovian shows devoicing, e.g. kot idzie, bok równy, brat niesie, zaniós em.^[21][22]

Traits that are younger include:

1. Mazuration (mazurzenie), which took place in Masovia, Lesser Poland, and part of Silesia, but not Greater Poland or Pomerania.^[23]
2. The realization of nasal vowels; in Greater Poland they undergo decomposition (bende = będę, kont = kąt, zemby = zęby, zomb = ząb, reŋka = ręka, roŋk = rąk), whereas in Lesser Poland and Masovia nasality was often lost, i.e. in Central Lesser Poland geba, reka, zob.^[24]

Kashubian shows more differences that connect it with its western neighbors, namely:

1. A uniform change of hard *ъl and soft *ьl to -oł- (pôłny, wôłk). Compare the Masovian change of -’áł- < *ьl before front hard consonants like wiáłna, piálny (see above);^[23]
2. A softening of consonants before -ar- from soft *ьr (miartwi), cwiardi (Polish twardy); also compare Masovian siarna.^[24]

These traits, due to their chronology, show that Pomeranian was more divergent than Masovian and show that Pomerania was a transitional area between the rest of Lechitic and Western Lechitic. However, since Pomerania eventually fell under the influence of the rest of Lechitic, albeit later than Masovia, it underwent some common changes.^[25]

1. Development of the slanted (pochylone) vowels in contrast to clear (jasne) ones and their alternations - something not seen in Polabian, seen even moreso in Pomeranian with the present of slanted high vowels;^[26]
2. A change of soft alveolar/dental plosive *t d to palatal affricates ć dź and soft *s z to the palatal sibilants s ź, which in Pomeranian later underwent Kashubation (kaszubienie). However, the change to affricates/sibilants took place in Poland in the 12th century, and this is also what separated Pomeranian from the rest of Western Pomerania (the Veleti).^[26]

Below are sound changes and their motivations with examples from Proto-Slavic throughout the course of history of the Polish language.

Historic Polish lost the Proto-Slavic accent system, modified its stress system, and gradual lost its length distinctions (see § Vowel length).^[27] Slovincian displays the most archaic state of the Lechitic stress system; Polabian in theory does as well, but due to incomplete material it is difficult to use.^[28] See Slovincian grammar for information on the stress system.

The timeline of stress can be divided into three periods:^[29][27]

1. A change of Proto-Slavic free and mobile stress by moving the stress forward or backward due to phonetic and morphological factors, kept in Slovincian and some of Kashubian;
2. A change to initial stress, seen in North-Kashubian, many Goral dialects, and in the Southern Silesian dialects found on the border of Poland and Czechia;
3. A change to penultimate stress, originally secondary, which gained more emphasis, present throughout the rest of the North-East Lechitic region.

During the preliterate (10th-11th centuries) era Polish likely still had free and mobile stress (akcent dynamiczny i ruchomy), then in the 12th-14th century, the accent was still at least mobile, seen in the second and third person singular imperative of verbs formed with the ending -i, e.g. usłysz (Bogurodzica), słysz (Bogurodzica), napełń (Bogurodzica) but zyszczy (Bogurodzica), spuści (Bogurodzica), raczy (Bogurodzica), wstań (Holy Cross Sermons) and pojdz (Holy Cross Sermons), chwali (Sankt Florian Psalter), puści (Sankt Florian Psalter), which could be stressed, and when it was stressed it was kept, but when the ending was not accented it was lost (however notably no imperative forms with -i are kept in Slovincian).^[30][27] Next initial accent was established in the 14th-15th century (compare Northern Kashubian dialects and some Goral dialects like Podhale, where initial accent here is considered a preserved archaic feature), this is also evinced in the loss of medial -i- in words like wielki < wieliki, wszelki < wszeliki.^[27] Finally penultimate stress developed as early as the 15th century^[31] or at the beginning of the 18th century.^[32][31] It is uncertain whether penultimate stress formed earlier in most dialects, or if it spread unevenly; it is more likely that it spread unevenly.^[33]

The development of the initial stress system and the penultimate stress system are likely related in that penultimate stress is likely derived from initial stress, namely, a secondary stress on the penultimate syllable could be found alongside the primary stress on the initial syllable of multi-syllable words, whereas penultimate systems, such as in modern Polish or eastern Slovak often have the secondary stress on the first syllable; this secondary accent become stronger in border regions where the two systems meet.^[34][35] These facts show that initial stress developed first and that the secondary penultimate stress later began regionally to be pronounced stronger.^[34]

Penultimate stress is not entirely regular in Modern Polish due to loanwords with the proparoxytonic stress (on the third to last syllable) from Greek and Latin such as muzyka, logika, polityka, as well as calques such as Rzeczpospolita (< Latin Rēspūblica), however colloquially many of these have penultimate stress; furthermore many French loanwards have word-final stress.^[32] Penultimate stress was also not established in native constructions with clitics such as clitiic pronouns; historically and regionally they in fact shifted stress, e.g. Old Polish dało się but dało, zobaczył go but zobaczył; this continued in the 19th century; and finally compounds originally composed of two accentual units, namely compound past tense forms: chodziliśmy, subjunctive forms: musielibyśmy, compound numerals: osiemset, and compound nouns like graniastosłup; also compare 19th century Kazimierz, Władysław, cudotwór, however in modern Polish there is a strong tendency to fix penultimate stress in these compounds.^[32]

Within the vocalic system, the contrastive length was first devleoped into a system differentiating different qualities, then was generally leveled; the loss of yers plays a part in this as well as in the phonemicization of soft consonants, and closed syllables were introduced into the language as a result as well, ablaut took place resulting in alternations, diphthongs with liquids vocalized in various ways, and nasal vowels underwent large changes.^[36]

Proto-Slavic vowels could be either long or short, including the overshort yers, had pitch accent, and words had free and mobile stress; various changes led to the slanted/clear opposition of vowels.^[37][38] Long vowels are inherited in stressed syllables and post-stress syllables shortened except when there was a neoacute: lato, błoto, męka, krowa, słoma, groch,^[39] and long vowels in pre-stress syllables stayed long only when directly before the stress in disyllabic words, but shortened in longer words: rzéka, świéca, brózda (> bruzda), mąka, łąka, dłóto (> dłuto), ciągnę.^[40] Other instances of slanting are innovative, including compensatory lengthening after the loss of yers and the contraction of two older vowels.^[41]

Compensatory lengthening occurred after the loss of yers due to the reduction of the number of syllables in the word when a vowel occurred in a closed or final syllable, usually before voiced consonants and often liquids, most often seen in vowel alternations caused by the masculine singular -∅ | oblique cases and the feminine genitive plural -∅ | other endings, compare Polish lód||lodu and kłoda||kłód.^[42][43] In Pomeranian, this alternation remains even for high vowels.^[42] Originally this lengthening also occurred before nasals, e.g. dóm, kóń, and during the Middle Polish period these forms were used in both the literary standard as well as in dialects, and Słowacki also shows regionalisms like psóm, dóm, poziómek, korónki - this lengthening (later slanting) was removed due to hypercorrection resulting negative associations with dialects and also through analogy to oN clusters arising from the decomposition of nasal vowels like ząb, dąb, kąty (see also § Proto-Slavic *ę and *ǫ).^[44] Many theories as to why compensatory lengthening did not take place before voiceless consonants exist.

Per Baudouin de Courtenay, it did initially occur before voiceless consonants, and a lack of slanting before tautosyllabic voiceless consonants arose as the result of analogy due to the fact that stems ending in voiced consonants were more distinct, and possibly as the result of word-final devoicing, as similar pairs of words like rok - róg, bok - bóg, płot - płód would phonetically sound more similar to each other: [rɔk] vs [rɔːk], etc., adding a semantic load to the importance of vowel length.^[45][46] This is supported by Old Polish texts, albeit inconsistently. Long vowels were often marked by doubling the vowel, and this can on occasion be seen before voiced consonants and voiceless consonants, for example Jakub Parkoszowic in ‘’Traktat o ortografii polskiej’’ (1440) suggests writing long vowels doubled, and includes examples of long vowels before both voiced consonants and voiceless, e.g. druug, laas.^[45][47] Dialects also sometimes show slanting before voiceless consonants, e.g. lós, stós, kós, potók.^[45] This has further evidence in the fact that compensatory lengthening became widespread in the genitive plural form -∅ of feminine nouns ending in -a, in which slanting became one of the characteristic inflectional features, occurring regardless of the quality of the last consonant of the stem, e.g. noga||nóg, baba||báb (historic or dialectal), cnota||cnót, księga||ksiąg, ręka||rąk, obora||obór.^[45] H. Koneczna in 'Księga referatów II MIędzynarodowego Zjazdu Slawistów, 1934, Warsaw further supports this by claiming that Slavic languages show a significant difference in the length of time between voiceless consonants (the longest), voiced consonants (second longest), and semivowels (the shortest), and these differences then affect the length of the preceding consonants inversely, vowels are the shortest before voiced consonants, second shortest before voiced consonants, and longest before semivowels, resulting in vowels phonemically voicing before voiced consonants and semivowels.^[48][49]

The next source of slanting, contraction, consists of merging two vowels separated by -j- into one. There are two periods of contraction, an older, preliterate period and a younger period which shows dialectal differentiation.^[50][51]

Conditions for older contraction include:^[48][51]

1. Nouns formed with *-ьje (see also § Declension of neuter nouns): picié, zbożé, pisanié (genitive piciá, zbożá, pisaniá)
2. Nouns formed with *-ьja (see also § Declension of feminine vocalic nouns): rolá, głębiá, sędziá, braciá, księżá;
3. The instrumental singular ending for nouns ending in *-a (*-ojǫ > -ǭ) (see also § Declension of feminine vocalic nouns): ręką, nogą, duszą, miedzą;
4. The instrumental plural ending for nouns ending in -ь (*-ьjǫ > -ǭ): kością, nocą; (see also § Declension of feminine consonantal nouns): kości, nocy;
5. The genitive plural of nouns ending in *-ь (*-ь̀jь > -ī (-ȳ after hardened consonants) (see also § Declension of feminine consonantal nouns): kości, nocy;
6. Definite adjectival forms (see also § Proto-Slavic definite forms): dobrý, dobrá, dobré, dobrégo, dobréj, dobrému, dobrą;
7. The word pás (later pas), from earlier pojasъ.

Conditions for younger, dialectically conditioned contraction include:^[52][53]

1. Second and third person singular and second person plural present tense verb forms of verbs with the stem -aje, -eje: działász, działá, działámy, działácie, umiész, umié, umiémy, umiécie, dialectically działajesz, działaje, umiejesz, umieje. Compare the third person plural present forms działają, umieją;
2. Infinitive and past forms of verbs such as chwiać, śmiać się, stać, bać się, chwiáć, chwiáł, śmiáć się, śmiáł się, stáć, stáł, báć się, báł się from older/dialectal chwiejać, chwiejał, śmiejać się, śmiejał się, stojać/stojeć, stojał, bojać się/bojeć się, bojał się;
3. In dialectal forms such as pódę, pódziesz, next to pójdę, pójdziesz, from Proto-Slavic *po-jьdą, * po-jьdešь;
4. Possessive pronoun forms: mégo, mému, má (moja), méj, this occurred the latest contracted and non-contracted forms are kept for different stylistic purposes in the literary standard.

There also exist many cases of slanting outside the above conditions, divided into five groups:^[54][55]

1. In terms such as góra, pióro, skóra, wióry, żóraw (later żuraw), który, wióry, wskórać, wynurzyć (< nórzyć), dialectal zdólić. Seemingly from the r, l, but there are also forms such as nora, pora, zmora, kora, ukorzyć, woleć, pozwolić;
2. rózga and dialectal gróza, potentially from the preceding r and following z, but compare also mrozy, drozd;
3. The ordinal numbers szósty, siódmy, ósmy likely arose via analogy to the old long vowel from the neoacute in the numerals czwárty, piąty, dziewiąty;
4. é and ó in the infinitives niéść, wiéźć, wiéść, piéc, móc, bóść are probably analogy to trzéć, wléc, trząść, róść, where lengthening occurred since the vowel is in a pre-stressed syllable (compare Russian трясти, расти);
5. Some feminine nouns ending in -a gaining slanting via association with nouns formed with *-ьja: rolá, żądzá, władzá, tęczá, paszá, acc. wolą, władzą, żądzą, tęczą, paszą.

Finally some words gained slanting and cannot be placed into any category, namely póki, póty, miesiądz, miesiądza, zając, zająca, pieniąz, pieniądza, mosiądz, mosiądzu, etc., and also verbs formed with -nąć, which was always long whether it was stressed or post-stress.^[56]

The loss of vowel length at the turn of the 15th and 16th centuries had three causes:^[57]

1. An internal factor in that long and short vowels differed from each other also phonetically in quality, shown in Old Polish spellings from the Sankt Florian Psalter like guor (gór), synuov (synów), skutuov (skutków), and this difference in quality weakened the phonemic function of length, and conversely the weakening of length intensified the need for contrastive tensing, which lead to drastic but gradual changes of these vowels unaffected by prosody,
2. An internal factor of the shift from initial stress to penultimate (see § Prosody), three syllable words beginning with zá- (or other long vowels) like zábawa, záloty, kóleczko shortened their vowel after this change as having a stressed vowel after a long vowel caused difficulties;
3. An external factor in the close political-eceonomic contact between the Ruthenians and the Jagiellonians, tied with the weakening of ties with languages with vowel length like Czech or Latin.

The loss of vowel length brought about a new distinction between slanted (samogłoska pochylona) (or tensed (samogłoska ścieśniona)) vowels (á é ó and to an extent ą /ɒ/ /e/ /o/ /ɔ̃/) and clear (samogłoska jasna) vowels (a e o and to an extent ę /a/ /ɛ/ /ɔ/ /ɛ̃/).^[58] This tensing, beginning still when length existed, is a compensatory process, as the energy originally given to hold the length of the vowel for longer is now given to tensing the vowel.^[59] In Polish, old long í ý ú merge with their short variants, but traces of them can be seen in Kashubian and Slovincian in <ë> alternating with i and u.^[60] The further development of slanted vowels is also one of the most important isoglosses between dialects.^[60] The resulting vowel system was rather complicated, and most slanted vowels merged with other vowels in dialects as well as in the standard, likely due to having many vowels near each other, resulting in them not sounding distinct enough from one another.^[61] See § Proto-Slavic *a, § Slanted é, and § Proto-Slavic *o for the further development of slanted vowels in the standard and dialects. In short, slanted á was lost first and in the standard merged with clear a with traces of this already in the 16th century, as in prints the two aren't always distinguished, and most Borderlands poets, e.g. Mikołaj Sęp Szarzyński, Szymon Szymonowic, and Józef Bartłomiej Zimorowic rhyme the two; Jan Kochanowski does not, meaning that Ruthenian nobility influenced the merger of slanted á with clear a, and slanted á disappears ultimately in the 18th century,^[62] slanted é was kept in the general language for longer; the 1891 orthography reform removed it, suggesting its disappearance occurring somewhat earlier; é originally merged with i or y in the 17th century, but clear e was often reintroduced via analogy (e.g. dobrégo becomes dobrego like tego); dialects show much diversity in the development of é, but it is often kept via alternations or as a separate phoneme,^[62] and slanted ó remains in opposition with clear o in the 17th century, after which a tendency to raise it occurs, merging it with u; this process finishes in the 19th century, but traces of it in the standard can be seen in the orthography and morphophonological alternations.^[63]

The loss of slanted á and é per Bajerowa (O zaniku samogłosek pochylonych. Pokłosie dyskusji, Katowice 1978) had both internal, i.e. systematic and social, as well as external causes.^[63] The first factor was the loss of vowel length, as slanting originally accompanied length, and without that support, it lost its ballast; the second factor was the instability, further strengthened by a following liquid l or r, of slanted vowels, being awkward within the morphophonological system and encumbered the standardization of the language.^[64] External factors include a growing social tendency for standardization at the turn of the 18th and 19th centuries, including 19th century orthography reforms, and also influences from Latin as well as Borderlands dialects.^[65]

The chronology of long and slanted vowels can be broken into five stages:^[65]

1. The Old Polish era (10th-15h centuries) where the opposition of long and short vowels took shape and strengthened;
2. The turn of the 15th and 16th centuries, where the opposition of long and short vowels was lost and slanted vowels arise;
3. The first half of the 18th century where slanted á merged with clear a in the general language (see also § Proto-Slavic *a);
4. The 19th century when slanted ó phonetically merges with u (but still morphophonologically alternating with o) (see also § Proto-Slavic *o);
5. The end of the 19th century when slanted é merged with clear a in the general language (see also § Proto-Slavic *e).

The reflex of Proto-Slavic *i is i. Proto-Slavic *i could be long or short, and early Polish likely kept both; compare Pomeranian alternations of i, from old long i, with schwa, from old short i.^[66] It could be possible that a difference in quality accompanied the difference in length, but ultimately we see only i in Polish or Pomeranian i | ë,.^[67]

An important change from Proto-Slavic *i in Polish is to y after sz, ż, rz, cz, dż, c, dz (szyć, żyć, przy, czyn, owcym wrodzy) as a result of the loss of softness in these consonants (see §consonants below).^[66] An exception is the so-called Podhale archaism (archaizm podhalański), whereby i is retained after s z (< s z or sz ż). i can be retained after sz ż in Silesia on the border with Czech, and rzi is kept in Podhale, all of the Beskids, and all of Silesia.^[68] In Podhale, this is sometimes spelled with ⟨ý⟩, see Dialects of Polish.

Proto-Slavic *i also shifted to e or é in the clusters:

1. ir (<*ir): sierota < Old Polish sirota < sirota, szeroki < szyroki, umierać < umirać (in the 15th century)
2. ir (<*ьr: cierń < ciryń, dzierżę < dzirżę, śmierć < śmirć, pierwy < pirwy/pirzwy, wierzba < wirzba, wierzch < wirzch, czerń < czyrń, czerw < czyrzw, czerpać < czyrzpać (in the 12th century).^[66]

This change begins around the 13th century, and the further time from this point the fewer instances of -ir- there are, as the Holy Cross Sermons only has one instance of -ir-, and in a transitionary period from the 13th-16th century much inconsistency and fluctuation occurs.^[69] While the same word could have either -ir- or -er- during the same time frame, generally it had only one form per scribe, i.e. a singular scribe would not use both forms, but instead one.^[69] Some transcriptions from the 16th century suggest both clear and slanted e could exist in er, for example Orzechowski has piérwéj as well as sercem. Some northern polish dialects still show this alternation: śmierć, wiercić || ciérpieć, czérwiec. Therefore perhaps in the Early Middle Polish period ir had two reflexe.^[69] One theory is that the periodization matters: if ir > er|ér took place in forms before the loss of length, then er is the result, because the nearest lower vowel to i was e, but if it took place after the loss of length and quality was the remaining factor, the next nearest vowel was é, provided that that long and short vowels also did not differ in quality phonetically. The preceding and proceeding constants also seem to have affected the reflex of ir:in modern standard Polish, er predominates with a few archaic exceptions (supposedly late borrowings) such as kir, zbir, skir, alongside mir and wir; in Middle Polish forms such as kier and wier exist, also Kazimirz, Kaźmirz.^[70] Similarly the groups il, il < *il can sometimes give e: sieła, mieły, odmienieli, usielstwo. This began in the 14th century and intensified into the Middle Polish period, but never quite to the same level as ir. Standard Polish does not keep this form, but many dialects do. Compare yl, ył below.^[71]

Within inflectional endings there was a tendency to lower high vowels before nasal consonants, resulting in i lowering to e or é, for examples in:

1. The instrumental plural ending -emi or -émi for all genders of pronouns and adjectives, from older -imi. This began in the 14th century and spread in the 17th and 18th centuries: nieścieciemi, wielkiémi. It appears that this change was motivated mostly by phonetic factors, and the predominance of -emi over -imi in the 17th and 18th centuries suggest that this process had more favorable conditions in accented syllables.^[71] See also declenion of pronouns in the plural;
2. -ę, -isz verbs in the present first person plural show - emy|-émy next to etymological -imy: myślemy, mówiémy. This was originally rare in the 15th century and became more common in the Middle Polish era, and adopted into Modern Polish, and is also common in dialects. This process was motivated both by analogy to other paradigms and by the tendency to lower high vowels before nasal consonants, as this only happened in the first person present plural, and in no other forms. The fact this change occurs also suggests it happened in a stressed syllable, i.e. penultimate stress was spreading. Forms with the rarer ending -em such as widziem, musiem, could have arisen based on forms such as widziemy, musiemy;^[71]
3. The masculine/neuter singular ending of pronouns and adjectives -em spread in place ofearlier -im. This was sporadically seen in the 15th century with a small increase in the 16th century, e.g. z czeskiem, jakiem gładkiem słówkiem. But -im not only remained in the nominative singular but even replaced the locative singular for the same genders and number.^[72]

Finally, *i was lost in a few positions, namely:

1. In the infinitive with a change of *-ti > -ci > -ć. Forms with -ci can be seen even in the early Middle Polish period, but exceedingly rarely, generally used for prosodic and rhyming reasons also perhaps under Czech influence, where in the Czech translation of the same text -ti could be seen. It is also possible this i was only orthographically marking the softness of the preceding c in some cases, but was not pronounced. One possible reason why final -i here was lost is the weakness of the morphological-semantic function, i.e. that both the absence of i and the distinctness of the infinitive ending were sufficient in marking the form;^[72]
2. In the imperative. Based on early texts it seems that a series of endings were generalized: -i for the second and third person singular, -imy for the first person plural, and -icie for the second person and third person plural, but even in the oldest texts the imperative without -i is also seen. The loss of -i in the imperative was a slow, gradual process, ending differently depending on the verb in different eras. This process is sometimes explained as being motivated by the general loss of asemantic word-final vowels, i.e. those inert in meaning. Compare also dosyci, maci, tamo, tako, jako, teże, juże, nuże, zasie, daleje, bliżeje, więce; the fact these vowels were unstressed would also support their disappearance, furthermore, verbs that didn’t accent final -i in the imperative lost it. Verbs with that stress, which are compared to other Slavic languages and Pomeranian suggest that it was stressed. Verbs such as chwalić also had a tendency to lose the imperative i to distinguish those forms from the homophonous third person singular chwali. The coexistence of both forms along with the lack of a phonetically motivated reason to have both lead to an immobilization of the stress, which lead to leveling. After the fixing of Polish stress, -i forms were generally lost, motivated also by a lack of a morpho-semantic need. -i was also lost in plural forms by analogy to the singular forms, e.g. niesi : niesimy : niesicie > nieś : nieśmy : nieście, which helped differentiate these forms semantically from the present tense forms.^[73]

Proto-Slavic *y gives y and regionally approaches i, but y can sometimes arise in these areas from old é, or diphthongizes in some western Greater Polish dialects. y occurs after cz dż sz ż r z c dz instead of i, except in some dialects, see above on the Podhale archaism and Silesian, and y changes to i after k and g, except in some dialects, likely in relation to k and g softening before e (< *ъ). The change of ky gy > ḱi ǵi probably began as early as the Psalter of Puławy and ended in the 16th century, except parts of Masovia. Similar to the change of *ir > er is also *yr > er: siekiera < OPl siekira < *sěkyra, occurring in the 15th-16th centuries. In Southern Lesser Poland y sometimes approaches e.^[74]

The clusters yl ył also sporadically change in the Old Polish era, very rarely before that, to el eł: beli, beł, telko. il ił also developed in different ways in dialects, in parts of Silesia and some other places there is no change, some areas have él éł ,and others have il ył, and sometimes with a secondary uł, which also affected forms with y resulting from é. Assimilation to ł (as /w/) can occur in different environments: sometimes in a closed syllable ending in ł (chwiułka but biła) or sometimes even when ł begins the next syllable (chwiułka, biuła). This can often be the result of analogy from other past tense forms. Similar to *im, *imi, *ym and *ymi also change to -em, -emi||-émi in pronominal-adjectival endings. *ym > -em is less common and is not retained. However -emi||-émi becomes more popular in the Middle Polish era.^[75]

Proto-Slavic *e usually results in e with softening of the previous consonant. Initial *e often undergoes prothesis with j- in loanwords, especially in dialects.^[76]

However, *e sometimes ablauts to o, attested even in the Bull of Gniezno, and can be traced to all Lechitic lects, and remained productive even through the 10th century with Latinate borrowings such as Piotr, anjoł. *e > o occurs before synchronically hard alveolar consonants, so *ed et ez es en eł er > od ot oz os on oł or: niosę, biorę, plotę, żona, cios, zioło, meaning e occurs before soft or historically soft consonants: nesies, pleciesz, żeński, cieśla, bierzesz, zielony.^[77][78] The presence of soft e near hard consonants such as bierny, Siedlewit is due to the fact that the consonant at the time was soft due to a following yer, which later disappeared: *berьnъ or as the result of a further neighboring soft consonant (Siedlewit).^[77][79]

This ablaut is motivated by the formants of the vowels, which can have primary and secondary formants, in that e has secondary formants similar to o.^[80][81] Then, if the primary formant is weakened or entirely removed, only the secondary formant remains, which is closer to the back (low) quality; this primary formant is removed by the palatalization of the preceding consonants caused by *e, as palatalization absorbs some of the articulatory movement of *e, resulting in a change to o, meaning that not only did this ablaut occur after the palatalization of consonants before *e, but was also motivated by it.^[82][81] This is also called dispalatalization of e. However if a palatal/soft consonant came after e (i.e. s' z' t' d' r' l' > ś ź ć dź rz l), this difference in position did not occur, thus a need for a vowel change was not present.^[82][81]

Many exceptions exist, often motivated by morphological leveling or other similar processes.^[83][84] The inverse also occurs, etymological czosać, śmiotana, krzosać occur in Old Polish, which were later replaced with -e-.^[85][84] Ablaut is absent however in dialectal forms such as bierę, bierą, but also through analogy to forms such as bierzesz, bierze.^[85] The differentiation of forms in the o direction was still weak in the 16th century, and only became stronger in the 17th and especially in the 18th centuries.^[85] Prepositions with *e did not undergo ablaut due to being clitics.^[86]

Also affecting this was the following consonant, as this change happened in some forms gradually and not simultaneously, and in others not at all, depending on which consonant was next, for example labial consonants, wlokę from the stem < *velk > vlek (compare Old Church Slavonic влѣкѫ (vlěkǫ).^[85][86] Ablaut is frequently absent after labial consonants in dialects, biedro, mietła, piełun, wiesło, wiesna.^[85] It also differs in many examples from the standard, sometimes with an etymological lack of ablaut: siestra, or an unetymological lack: wieska (e from a yer), poziemka (before a labial), or etymological presence: śmiotana.^[87]

The change of e (after softened consonants) > o (after hard) in the dative -ewi > -owi, the derivational suffix -ew- > -ow- (Oleszewo > Oleszowo, as well as adjectival and compound noun endings from this) and the rare genitive -ew > -ow is rather the result of leveling and analogy, as the distribution of o is uneven depending on the dialect and morpheme.^[88] Lesser Poland loses e in these positions the most, less often in Masovia due to influence from the literary standard, and the least often in Greater Poland. Within Greater Poland -e- can even occur after innovated soft consonants; in some Greater Polish dialects in, yn change to iń, yń, so syn becomes syń, and the dative synowi changes to syniewi.^[87] The Łowicz dialect is unique, as sometimes it generalizes e after hard consonants, but also uses o after soft consonants.^[87]

Proto-Slavic *ě had a tendency to raise in children languages, and in most Polish dialects to e. However, this vowel in Lechitic also sometimes sometimes merged with a, resulting in ablaut, bielić : biały, kleć : klatka, kwiecie : kwiat, lecie : lato, mierzyć : miara, mieścić : miasto, strzelić : strzała, świeca : światło, wierzyć : wiara.^[89] This ablaut is attested as early as Bull of Gniezno: Balouanz (Białowąs), Balouezici (Białowieżycy), Balossa (Białosza) but Belina (Bielina), Quatec (Kwiatek), Soledad (Sulidziad), Stralec (Strzałek). The oldest example is Dadosesani as a Silesian name for the strain Dziadoszanie, written by the Bavarian Geographer.^[89] Also here from 1015 is Diadesizi, meaning that this ablaut is prehistoric, affecting also Polabian, and older than *e > ‘o.^[89] This ablaut has the same conditions as *e > ‘o: before *ěd ět ěz ěs ěn ěł ěr > ad at az as an ał ar, and *ě > e before soft consonants or historically soft consonants.^[90] The presence of e < *ě before hard consonants in forms such as wierny, bielmoświetle is explained by the fact that at the time of the ablaut the consonant was soft or by the fact it was under the influence of a yer which was later lost: *věrьnъ, bělьmo, or from influence from a near-by soft consonant, such as świetle, as l is soft.^[90] This ablaut has the same motivations as *e > ‘o, but with /a/ being the back equivalent of /æ/, which was likely the realization of *ě.^[90]

As with the previous ablaut, exceptions exist, and are also the result of morphological leveling, occurring as early as the Old Polish era. These forms are rare and rzezać, lezę, krzesło dominate.^[90] Dialectically oferia and powiedać are used. The term ofiera was originally from Czech, but the form ofiara came to dominate based on native terms, powiedać however is purely a Masovian dialectalism, where ablaut often doesn’t occur after labials: niewiesta, kwiet, wieno, piestun.^[90] The process of leveling e to a completed only in the 17th and 18th centuries, and of course not in all words at the same time.^[91] Alternations such as cza||cze (czas||wczesny, żal||o żeli are not the result of ablaut, but is probably the result of analogy to alternations such as krzyczał||krzyczeli, etc.^[91][92] Such forms with a are original and forms with e are secondary and arose based no widział : widzieli: widzieć, where -e- is actually from *ě and these show etymological ablaut.^[91][86]

At the end of the Proto-Slavic period, there existed originally short *e, lengthened *e, originally long *ě and shortened *ě. Inherited into Polish short *e and *ě could either remain short or be lengthened, and long *e and *ě could either remain long or shorten. Short *e and *ě could lengthen due to Polish innovations mentioned above, resulting in short e and long e, each of which having two origins, that is from short *e *ě and long *e *ě.^[91] Short e < *e: jeleń, leżę, lecę, miedza, mielę, nie, rzekę, ciekę, ciepły, ziele; short e < *ě: krzesło, krzepki, lewy, miech, miesiąc, niemy, wiek, wieża, bielić, w lecie, mierzyć, strzelić; long é < *e: jéż, narzékać, piérze, wiéść, niéść; long é < *ě: biég, césarz, chléb, chléw, kléj, miédź, rzéka, świéży, kléć, śiéca, biélmo, wiém.^[93] e and é also differ in quality: short e generally lowered to its current position, whereas é was higher and slightly more fronted. Some think that é could have been diphthongal, /ei/ and others think it was /je/.^[94]

Aside from é from *e ě, é could also arise from contraction of groups VjV, attested in the Holy Cross Sermons, Sankt Florian Psalter, and Gniezno Sermons:^[95]

1. *eje > é in the nominative, accusative, and vocative singular of neuter adjectival/pronominal soft declensions: pokolenie szukającé, trzeciéć przyrodzenie, napierwszé bogactwo;
2. *eje > é in the present tense of verbs such as umieć, rozumieć: umié, rozumiémy;
3. *eje > é in the phrase nie je (nie jest, nie ma) > nié;
4. *ejě > é in the genitive singular of soft feminine pronouns: w jé świętem żywocie;
5. *ějě > é in the genitive singular of compound feminine soft adjectives: matki bożé obraz, namniejszé pierzynki nie miał;
6. *ějě > é in the nominative, accusative, and vocative feminine singular and accusative masculine plural of compound soft-stem adjectives: wszystki przeciwiającé sie, bojącé się boga, syny bożé, na lędźwie najmocniejszé;
7. *oje > é in nominative, accusative, and vocative neuter singular of compound hard-stem adjectives: w kakié wrzemię, mocné bostwo, mdłé człowieczstwo;
8. *ojě > é in the genitive feminine singular of hard-stem pronouns: krola této ziemie, biskupowie této iste ziemie;
9. *yjě > é in the genitive feminine singular of compound hard-stem adjectives: wieczné śmirci, ludziem dobré wole, z krolewny niebieskié, nijedné gospody;
10. *yjě > é in the nominative, accusative and vocative feminine plural and masculine accusative plural of compound hard-stem adjectives: wielikié przyjaźni, we złé chustki, boleści śmiertné, boleści pkielné, kraje ziemskié, nad syny ludzkié, na święté swoje;
11. *ьje > é in the nominative, accusative and vocative of neuter nouns such as nauczanie and in the numeral trze: włodanié, oświecenié, trzé;
12. *aje > é in the ending *ajego of the genitive singular of compound masculine and neuter adjectives: człowieka grzesznégo, krola cnégo, syna dziewiczégo, nagłégo spadnienia;
13. *uje > é in the dative ending of compound masculine and neuter adjectives: bliźniému, utoka ubogiému, ku siercu wysokiému.

The change of *e > é||e and *ě > || é||e and the raising of é with the loss of contrastive length took place gradually over the 14th-16th century, resulting in slanted e.^[96] Within dialects, é is sometimes kept in all etymological positions, or sometimes leveled in morphological endings. In some dialects, it remains as a separate phoneme, in others it raises to y regardless of the hardness of the previous consonant, in others either raise to i after soft consonants or to y after hard consonants - this was common in the literary standard before the loss of é- or finally it can merge in some dialects with e - this is the rarest.^[97] See various articles on Dialects of Polish for details on individual dialects. In 1518, Zaborowski proposed writing é as ė, which was not widely accepted, but rather é became popular, or sometimes the digraph eé. Murzynowski in 1551 also suggests the use of é. However, only some publishers regularly distinguished the two, such as in Kochanowski’s works, published by Łazarzowa from 1583-1585. Even those that attempt to distinguish the two do it inconsistently, often the fault of typesetters.^[98] Differentiating the two becomes even weaker in the 17th century; Grzegorz Knapski in 1621 complains about those who do not distinguish the two sounds, but he himself marks slanted é with clear e, as slanted å is used for clear a.^[99] As a result, e is used for both phonemes in the 18th century, and Onufry Kopczyński suggests differentiating the two in specific conditions in his grammar from 1778-1783, but many of his suggestions are not etymologically motivated and are the result of his own invention or dialcetal influence.^[100] é sees more use in the 18th century, particularly in certain words or endings, as can be seen in Linde’s dictionary or the Vilnius dictionary, and finally é was removed in the 1891 orthography reform.

Thus, é likely remained in common use in the 16th century and its use was limited over the course of the 17th and 18th centuries.^[100] Phonetics was one factor in this loss, i.e. the neighboring sounds. For example, the ending -éj was often realized as -yj, or -égo in some dialects could be realized as -ego or -égo depending on the previous consonant: dobrego but taniégo.^[100] Analogy also played a major role: zielé > ziele, based on pole, dobrégo > dobrego based on tego.^[100] The Eastern nobility also influenced this, which had either only e or raised é to y, i, seen in Rej’s rhymes.^[100] Another potential cause for this was a feeling that é was unpredictable and difficult to know how to mark in text. Finally the fact it had a low semantic load also contributed to its disapperrance.^[100]

The phonological and phonetic separateness of é began to be lost in the 17th and 18th centuries, and in the 19th century it merged wither with ‘i||y or e, then the awareness of é and where it should be weakened, meaning it became less and less marked in texts, marked usually for grammatical norms; replacing é with ‘i||y weakens in the 19th century, often being replaced more with e, and this is finalized with the 1891 orthography reform.^[101][62] The most likely cause of the spread is likely the result from when Warsaw became the capital in 1800, crossing two literary systems: the Greater-Polish/Lesser Polish system and the Masovian system, as the Masovian realization of é spreads from Masovia in the 19th century to the literary area of Lesser Poland and Greater Poland, replacing i/y.^[102] The orthography reform of 1891 replacing é with e was also likely adopted by schools, where e would be pronounced as written, and the normative influence of schools also likely spread this pronunciation, and realizations with y||i were seen as wrong.^[102] A few words that historically had é, at least in textbooks from the 19th century, sometimes have y||i in colloquial or jocular pronunciation within Standard Polish.^[102] This is the result of maintaining the old pronunciation of the intellectual elite as well as adoption of dialectal substrates and influence in one’s speech, often done for expressive reasons.^[103] The change of -ej > -yj||-ij is nearly common throughout, probably from the result of final -j (tamtyj, zdrowyj, jednyj, jij, nij, drzyj, starzyj, szybcij, only trzej remains without raising).^[104] Slanted é was kept in the general language for longer; é originally merged with i or y in the 17th century, but clear e was often reintroduced via analogy (e.g. dobrégo becomes dobrego like tego); dialects show much diversity in the development of é, but it is often kept via alternations or as a separate phoneme.^[62]

Proto-Slavic *a gives inherited a.^[104] Already in the pre-Polish era, a developmental change of *a took place in whereby the original Proto-Slavic length, long or short, was only sometimes retained, but in other cases long vowels were shortened and short vowels were lengthened, see § Vowel length and clear vs slanted vowels - this means that in pre-Polish and medieval Polish there was long á from late Proto-Slavic long *ā and short a from late Proto-Slavic short *a and long ā which was shortened in Polish, and finally Polish ā from contraction.^[104] Examples of contraction include:^[105]

1. *aja > á in the nominative singular of feminine compound adjectives and the nominative, accusative, and vocative plural of neuter compound adjectives: świętá Katerzyna, słowa znamienitá;
2. *aje > á in the second and third singular and the second plural present tense of -am verbs: pobudz’’’á’’’ ponęcá i powabiá, wyprawiá się, rozpaczász, but okopaje przywitaje is attested as well in the 15th century; to that time -znajesz, -znaje after a prefix: poznajesz, wyznajemy, doznajecie;
3. *oja > á in the stem of the infinitive of the verbs stoję, boję się: Mojżesz wybrany jego stáł, będzie stáć, nie będę się báć;
4. *eja > á in the stem of the infinitive of verbs like wieję : wiać, grzeję : grzać: (Parkoszowic) (ch)wiáł (faal), wiáł (vaal), but in Sankt Florian Psalter chwiejali głową;
5. *ьja > á in the nominative plural in cases where the ending begins with a: sędziá or in collective nouns braciá, and in the declension of nouns of the type przykazanie, zboże (those which originally had the ending *-ьj-): sąmnieniá, przyszciá, głos skurszeniá, od narodzeniá, do widzeniá, pokoleniá.

Long á was rarely marked orthographically with aa in the medieval period, but even circa 1516 you can see aa. Parkoszowic attempted to popularize this, but in the same work (be it either that Parkoszowic made a mistake or his copyist) sometimes writes the same word without double aa, or in words where one would expect aa it is missing.^[106]

In terms of sound quality, short a was likely /a/ and long á was likely /ɒː/, and once length distinction was lost, clear a and slanted á remained.^[106] It’s also likely these two vowels didn’t sound much different, as they often rhyme in Rej’s texts.^[106] However old texts systematically and regularly distinguish the two, imperfectives (frequentatives) often having it: zaráżać and perfectives do not: zarazić, wysłáwiać, wysławić.^[107] The word nie could affect slanting: dáj but nie daj, mász but nie masz.^[108] á also appears in the stem of the non-compound passive participles as a predicative: dán, dána, dáno, but a appears in compound forms: dany, dana, dane.^[108] na(-) and za(-) appear in prepositions and prefixes verbs derived from other verbs and ná(-) zá(-) appear near a nominal.^[108] The dative plural of feminine nouns always have -ám, and the instrumental plural -ami, the locative plural shows -ach ten times more than -ách.^[108]

Generally the letters ⟨á å⟩ more often represented the phoneme clear /a/, and ⟨a⟩ more often represented slanted /ɒ/, but the opposite could occur; this is the opposite of Murzynowski’s suggestions, and dominated from the middle of the 16th century and through the 17th and 18th centuries.^[108] There are also some Borderlands poets which do not differentiate a and á , e.g. Mikołaj Sęp Szarzyński, Szymon Szymonowic, and Józef Bartłomiej Zimorowic rhyme the two; Jan Kochanowski does not, meaning that Ruthenian nobility influenced the merger of slanted á with clear a.^[108][62] There are also cases of inconsistent marking, the result of oversight on behalf of the typesetter or proofreader, as the difference between the two was strongly felt even in the 17th century.^[108]derlands poets, and slanted á disappears ultimately in the 18th century.^[62]

But like slanted é, the difference between the two likely started to blur starting in the 16th century, meaking it the earliest slanted vowel of the three to be lost.^[108][62] Exemplary of this is how the alternation of á||a in the present tense vs infinitive began to disappear (similar to ą:ę, sądzę || sędzić): báczę||baczyć, etc.^[108] This alternations clearly fluctuates in the 16th century, with some verbs preserving it better than others.^[109] The original state could be seen in some infinitives, and in other cases in the imperative.^[110] The pronunciation was not often certain everywhere, and proofreaders did not manage well with these uncertainties.^[109] This process strengthens in the 17th century through the beginning of the 18th century, where the it finishes and á disappears from the literary standard around the middle of the 18th century.^[109][62]

Nitsch explains this as being the result of Ruthenian Poles, who merged á with a.^[109] The north Masovian dialects around Warsaw, which had become the capital, also did not have á.^[109] The loss of á is also associated with the mixing of populations meeting from different places in an area with a low population.^[111]

Dialects preserve á much better than the standard, as the majority of Polish dialects did not merge the two phonemes.^[112] See articles on particular dialects for more.

In early dialects, beginning before the 12th century, the initial cluster ra- changed to re-, found even in the oldest texts a change attested in northern and central Poland, e.g. rano > reno.^[111] The range of ra- > re- shrank with time and now appears only in a few northern dialects.^[111] Similarly initial ja- > changed to je-, but the range of this was less north than ra- > re-, and also affected Silesia, and is attested as early as the 12th century.^[111]

Proto-Slavic *o gives o, with later changes in length similar to those that occurred with slanted é and á. Short o generally remains unchanged.^[113]

In some adverbial pronouns with an accusative neuter singular origin it was lost word-finally already in the Old Polish period: tako, jako, tamo, owako, inako, jednako > tak, jak, owak, inak, jednak, kako was lost entirely.^[113] jako and tako remain in this fossilized expression, and jako as a comparative/copulative word.^[113] Dialectally tamo still exists.^[114]

o often undergoes epenthesis, usually with /w/,^[114] called labialization (labializacja), spelled ⟨ô⟩, which can be seen in most dialects, especially initially; rarely h (/x/) may appear, and rarer still w (/v/).^[114] Fewer dialects have no labialization.^[114] In a large area in the south west of the above mentioned line all the to the mountains in the south labilization becomes stronger and stronger.^[114] Labialization can depend on place, speed of speech, individual carefuless of the speaker, and sometimes on the quality of the preceding consonant;^[114] palatal and front consonants are less likely to be associated with labialization: pôle, kôść are very common, rôście, dôbry are frequent, sôstra, ciôtka are rare.^[114] In places where ł changes to /w/, labialized o and ło can be confused, e.g. pot and płot might sound the same as /pwɔt/.^[114] People with this when trying to speak formally will often hypercorrect and remove etymological /w/, e.g. łopata > opata.^[114] In all of Poland, o often raised to ó before nasals, meaning that the dialectal realizations are etymological with Old Polish texts.^[114]

ó had more changes than o.^[114] Probably already in the preliterate era ó differed in quality from o as well as in length, raising and tensing, especially at the end of articulation, approaching u, giving /oː/ in contrast to /ɔ/.^[115] In the Middle Sges length was lost and the quality difference remained, giving o (/ɔ/) and ó (/o/).^[116]

Orthographically, ó began to be marked differently in the first quarter of the 16th century, and Murzynowski also recommends distinguishing the two, but only a few publishers in the 16th century regularly and consistently distinguish the two.^[116] Kopczyński reintroduced ó into the orthography at the beginning of the 19th century, from which time it has remained.^[116]

Kochanowski has particular rhymes that often suggest ó being closer to o, but sometimes merging with u before r, Rej has a similar tendency, Knapski in 1621 shows there was still a difference in the 17th century, but he likely had labialization for ó, but fluctuation before r also occurs.^[116] Zimorowic (1629) in Roksolanki and Bartłomiej in Roczyzna and Żałoba (1654) have the opposite trend - rhymes where u lowering to ó before r n can be seen in the 16th century in Rej’s and Kochanowski’s texts.^[117] Odymalski (Obleżenie Jasnej Góry Częstochowskiej, pre 1673) shows a tendency to raise ó towards u.^[118] It is difficult to say if these rhymes were accurate of pronunciation, or if they were near rhymes.^[118]

Grammarians’ recommendations from the 17th and 18th centuries do not shed much light, as often phones are not distinguished from letters, sometimes resulting in a perceived phonetic difference where one does not exist; also, sometimes these grammars are written by foreigners or for foreigners, and so sometimes the content of the work is simplified, and sometimes these grammarians generalize observations made about folk speech and attribute them to educated speech.^[118] Mesgnien in Grammatica seu institutio polonice linguae (1649) says that ó approaches u, but is clearly distinct from it, but is closer to u than o.^[118] Cassius in Lehrgebäude der polnischen Sprache (1797) says that the accent on ó is used to distinguish it from “pure o” and to pronounce it like u.^[118] Grammarians of the 19th century also confused letters and sounds, but many from the century comment that ó is barely different from u.^[119] Sochański in Brzmienie głosek polskich i pisownia polska (1861) says that some pronounce ó as u, but the majority pronounce it as a sound between o and u.^[120]

Many efforts were made during the 1936 reform to completely remove ó, but they were generally met with resistance, and ó was removed in a few words such as żuraw, bruzda, pruć, chrust.^[120] The reason ó had more resilience than é and á orthographically is likely because it merged with /u/; once á merged with a then there were no morphophonological alternations; conversely because é regionally alternated with i||y, it remained in the orthography for longer, until it merged with clear e in the standard.^[121] Since ó merged phonetically with /u/ in the standard, there were many morphophonological alternations, giving a phonetic justification for it to remain in the orthography.^[121]

In general the current distribution of ó||o is in agreement with the distribution of old long ó, except before nasals, as o tended to raise to ó before nasals in Old Polish and Middle Polish, e.g. dóm, which is not always seen today.^[121] ó also appears in Old and Middle Polish before ł and r, e.g. aniół, klasztór, sometimes before j, e.g. dójdzie, pójmuje, and before -ż, e.g. kogóż; also in the past tense of verbs with a consonantal stem, e.g. mógłeś, wiódłem, odniósłeś, and the word dróga.^[121]

As a result modern Standard Polish shows much uncertainty, e.g. dwom||dwóm, spódnica||spodnica.^[121] This is the result of phonetic, morphological-analogical, or dialectal factors.^[121]

Sometimes dialects lack slanting where it exists in the standard or vice-versa, forms such as ktory, gora, klotka exist, as well as dróga and Greater Polish szkólny.^[122] Vilnius Polish has much notable deviation, seen in Adam Mickiewicz’s works: zbojca, brzozka, ogrodek, but also spójrzeć, ostróżny, paciórek.^[122]

Thus slanted ó began to weaken in the 16th century in its phonetic uniqueness with, but a clear opposition remained in the 17th century after which a tendency to raise it occurs, merging it with u; this process finishes in the 19th century.^[120][120] Ruthenian nobility likely also influenced this.^[120][63] This process intensifies in the 17th century supported by Masovian and Lesser Polish pronunciations, as ó developed here differently than é.^[120] Then ó realized as u becomes popular enough that it likely becomes part of the literary standard by the 19th century, but remains part of the orthography thanks to Kopczyńśki’s and his successors’ efforts as well as morphophonological alternations.^[120][63]

Proto-Slavic *u gives u. Proto-Slavic *u could be either long or short - as with other vowels in prehistoric Polish this length could change, and both long and short u entered the earliest stages of the language, probably with no difference in quality between the long and short variants.^[122] Only Pomeranian dialects keep traces of this length distinction, and not everywhere.^[122] Traces of short u is best seen after front consonants, where it became ë, like with old short i and y.^[122] After back and labial consonants it usually stayed as short u, sometimes developing as long u: sëchi, szëmiec, łëpic, lëdze, brzëch, cëdzy, and also puscëc, buk, mucha, kupic, gubic, chudy, juńc.^[122] Short u lowered in a similar manner as short i, giving a characteristic reflex: a short low reflex vs a long high one.^[122]

Otherwise u remains unchanged throughout Polish. In a few instances it became i for often uncertain reasons: litować < lutować, Inowrocław < Junowłodzisław.^[123] This is probably motivated by similar articulatory reasons as the e > o ablaut, as i is the front equivalent of u in terms of height, much like the relationship between e and o.^[124] During the Old Polish and Middle Polish period, u often lowers to o before n, ń, m: Dunajec||Donajec, punczocha||ponczocha, słuńce||słońce.^[124] This fluctuation can still occur in more recent Polish: tłumaczyć||tłomaczyć, tłumok||tłomok, tytuń||tytoń, and also dialectal realizations such as grónt, fónt, and sometimes before r: góra, chmóra.^[124]

There were two nasal vowels in Proto-Slavic, front mid *ę and back mid *ǫ, which both could be long and short - the oldest Polish reflexes are found in texts from the 12th century.^[124][125]

During this epoch there were probably nasal vowels of two different qualities, [æ̃] and [ɑ̃] and, based on later reflexes, nasal vowels underwent considerable changes in terms of length, namely, the length of the nasals was shortened in some positions already in the final phase of the development of Proto-Slavic; therefore, in Polish there are long nasals continuing some of the Late Proto-Slavic short nasals or from that part of the Late Proto-Slavic long nasals that were shortened in Polish.^[124][126] Pre-Polish also has long *ǫ that does not continue Proto-Slavic *ǭ, but is the result of contraction of *ojǫ, *ejǫ, *ьjǫ: wodą < *vodojǫ, sobą < sobojǫ, moją < mojejǫ, wsią < vьsьjǫ and long ǫ resulting from compensatory lengthening after the loss of yers (see § Vowel length and prosody).^[124][127] Therefore at the beginning of the literate period in the 12th and 13th centuries the following reflexes can be reconstructed: *ę *ǫ : /æ̃/||/ɑ̃/, /æ̃ː/||/ɑ̃ː/.^[124][128] The quality, which lowered in both cases, was originally determined by the quality in Proto-Slavic, so *ę > /æ̃ː/||/æ̃/, *ǫ > /ɑ̃ː/||/ɑ̃/, except after ablaut, which could result in ǫ occurring after hard consonants (see the Lechitic ablaut, but the results of this were already beginning to blur) of *ęt, ęd, ęs, ęz, ęn, ęr, ęl (all hard) > *ǫt, ǫd, ǫs, ǫz, ǫn, ǫr, ǫl, e.g. Świą́t, Świą́tosz, Trzą́sowo, Trzą́siwuct, Czą́stobor.^[129][130] The quality of these four nasal vowels change in the 13th and 14th centuries, as is reflected in texts from this period, namely, both nasals, originally written from ⟨an⟩, ⟨am⟩, are written identically with the letter ⟨ꟁ⟩, which saw its first use in the 13th century and spread in the 14th century, then in the 15th century, ⟨ą⟩ begins to see similar use for both nasals.^[131] Therefore during the Old Polish era the nasal vowels merged into one vowel in terms of quality, and at the same time, a distinction in terms of length continued: /ã/ < ([/æ̃/ < *ę]||/ɑ̃/ < *ǫ]) and ą̄ < ([/æ̃ː/ < *ę|/ɑ̃ː/ < *ǫ]).^[132][133]

As with other long vowels, there was somewhat of a custom to write the long nasal with a doubled letter, ⟨ąą⟩ or ⟨ꟁꟁ⟩.^[132] Parkoszowic also distinguishes them, but only when it is necessary to avoid confusion, that is contrastively, such as mꟁka and mꟁꟁka.^[132]

The first signs of the modern nasal system can be seen in the 15th-16th century, namely that the long and short nasal vowels begin to differ in quality again, akin to clear and slanted a:á, e:é, o:ó.^[132][134] The result is that ą (ǫ) comes from the old long vowel and ę comes from the old short vowel.^[135] The first document that suggests this change is Psalter of Puławy, which has some of the same psalms as Sankt Florian Psalter, but written orthographically differently.^[132] This development was assuredly gradual and did not yield the same results in all dialects of modern Poland.^[132]

Some texts from the beginning of the 16th century continue the tradition of marking nasals with the same letter: ą or a̩, then in the 1630’s, the front nasal was written e̩ and the back one ą, and in the middle of the century the letters ę ą were used, meaning the front nasal was pronounced as a front, somewhat raised: [æ̃], which, with time and in Greater Poland, moved more forward; in Lesser Poland around the middle of the 16th century however, the realization was probably [ɛ̃] and sometimes even e, which can be seen in rhymes in Mikołaj Rej’s poetry.^[132] Franciscus a Mesgnien Meninski states clearly that ę is as French en as in mien, tien, sien.^[136] The back nasal was probably originally [ã], which later backed and raised to [ɑ̃] or even [ɔ̃].^[137] The realization ą is evidenced by spellings such as ąm, an, au and by the grammarian Jeremiasz Roter (1618), and by rhymes.^[137] The realization [ɑ̃] is evinced by grammarians from the 16th and 17th centuries, Pierre Statorius (1568) says it’s like “dark a”, Ostroróg (c. 1600) claims it to be a diphthong composed of the sound of á and the letter o through the nose.^[137] The pronunciation [ɔ̃] is evinced by spellings such as on, un o, and a remark from Volckmar (1612), and with a similar remark from Mesgnien.^[137] This clearly shows that the realization of the back nasal depended on region, and within the literary standard [ɔ̃] eventually became widespread.^[137] The orthography of ⟨ę⟩ ⟨ą⟩ are established in relation to their original literary realizations (/ɛ̃/ and /ã/), and later when /ã/ raised to /ɔ̃/, the orthography did not change.^[138] The qualitative changes ended in the 16th century, and in the 17th century the modern realization of the decomposition of nasals before non-sibilant consonants to eN, oN, denasalization before l and ł, and denasalization of final -ę occurred.^[139]

Inflection also plays a role in the development:^[140][141]

1. During the Middle Polish era and part of the New Polish era soft-stem feminine nouns ending -a took -ę or -ǫ in the accusative singular. -ę slowly replaced -ǫ during the 19th century and is the current dominant form;
2. Verbs with a nasal stem such as sądzić, wątpić, rządzić, wiązać, żądać, oglądać, pędzić, kręcić showed the alternation of ę:ǫ in the Middle Polish era that was in accordance with etymological changes: ǫ would appear in the present and past tense, and ę in the infinitive and imperative: sądzę, sądził : sędzić sędź etc. Analogical leveling tended to remove this alternation, with one nasal winning in all stems for a given verb;
3. The accusative feminine singular of pronouns til the 19th century had -ę: owę, moję, naszę, now only retained in tę, but this is often colloquially and predominantly tą. In the second half of the century, -’ǫ began to replace it modeled on adjectives. Sporadically the inverse can be seen: starę, zdrowę;
4. The Old Polish divide between mię cię się after pronouns and mie cie sie after verbs was lost in the 16th century;
5. Aside from these categories, there are many words in which leveling towards either ę or ǫ took place, and many orthoepeic dictionaries fluctuate as to which of two forms is more correct: gąsior:gęsior, grzęznąć:grząznąć, okrąg:okręg, zasiąg:zasięg, mięs:miąs, pięter:piątr, piętrowy:piątrowy, siędę:siądę, osiągnąć:osięgnąć, ciążyć:ciężyć, tysięczny:tysiączny, etc. In the past świątość occurred but was leveled with święty, piątno, piądź, w nię, now piętno, piędź, w nią, etc.

Secondary nasalization also occurs as well as loss of nasalization, the causes of which include analogy or a neighboring nasal consonant, otherwise it may happen sporadically, e.g. między < miedzy, tęskny, tesknota < teskny, tesknowa (compare utyskiwanie), częstować < czestować (cześć), paszczęka < paszczeka, jeniec jeńca < jęciec, jęćca (jąć); uczestnik < uczęstnik, Old Pollish szczeście szczesny (część).^[142][143] The dialectal distribution of nasal reflexes or denasalization is highly complex, and is described more in individual articles on Polish dialects.

Proto-Slavic two weak vowels often called yers. The soft yer *ь was front and softened preceding consonants - it gives e in strong position; the hard yer *ъ was back and did not soften the preceding consonant - it gives e in strong position, but distinct from e of other origin, as the consonant before it is hard.^[144][145] There are traces of *ъ causing rounding in early texts, e.g. Lokna (1153), Lukna (13th century) next to Lekna (1136), all modern Łekno.^[146] Similarly ku has u instead of expected e (*kie || k); ku mnie||k tobie.^[146] During the Proto-Slavic period, yers could be either weak or strong, resulting in four yers, the strong hard yer, the weak hard yer, the strong soft yer, and the weak soft yer - these four yers enter the earliest prehistoric Polish vocalic system, but in the 11th century, after the e > o||a ablaut, after which many changes regarding them occur, namely strong hard yer changes e with a preceding hard or functionally hard (ḱ ǵ) and strong soft yer > ‘e with a preceding soft or hardened consonant; weak hard yer disappears and similarly weak soft yer disappears but sometimes leaves behind a soft or softened consonant that preceded it, strengthening the phonemic opposition of hard and soft consonants.^[147][148] The oldest attested signs of the development of yers can be seen in the 12th century in Bull of Gniezno and Bulls of Wrocław, Golec = Gołęk < *golъkъ, Quatec = Kwiatek < *květъkъ,^[149] and there are no traces of them still being vowels in the earliest texts.^[150]

The loss of yers also led to compensatory lengthening, morphophonological alternations of mobile e||∅ the end of the law of open syllables, a strengthening of the contrast between hard and soft consonants, and complex consonant clusters (see § Simplifications of consonant clusters).^[151]

Rules that do not conform to the above include:^[152][153]

1. Final *-ъjь > -y in compound adjectival declensions;
2. Final *-ьjь > -i in compound adjectival declensions, the genitive plural of feminine nouns, now consonantal, e.g. kości;
3. -ьj- > -ij- || -yj- (after hardened consonants) in verbs, e.g. biję, wiję, nouns, e.g. żmija, szyja, and the pronoun czyj;
4. Initial *jь- developed in two ways; if it started a sentence, i.e. was in the absolute initial position (i.e. beginning the utterance) > i, e.g. idę, igra, imać, imieć; if it came after a term ending in a vowel > *j > ∅, e.g. gra, skra, mam, mieć; after a prefix ending in a vowel *jь > j (zajdę, dojdę, najdę). Later either only the i- form was kept, e.g. igła, iglica, ikra, or both are used, e.g. idę, zajdę, sometimes with a changed meaning, e.g. igrać vs grac, sometimes with a stylistic difference, e.g. iskra vs skra;
5. The accusative masculine singular of the pronoun *jь > ji, later replaced in the 16th century by jego, go;
6. If a prefix ended in ъ, and the base started with i-, resulting in ъi, then in Old Polish y (zysk, odydę). Later levellings changed this to odejdę based on other similar verbs.

Resulting forms often undergo analogical leveling, whereby mobile e (|| ∅) often became fixed in its position, e.g. Old Polish sjem || sejmie (locative singular) became fixed sejm || sejmie, aditionally oblique forms of szmer changing to szmer- (originally szemr-, e.g. genitive singular szemra), or psek reshaping to piesek from oblique stems like genitive singular pieska or sometimes a lost vowel was added back in, e.g. deska from original cka.^[154][155] Such processes can be observed as early as Bull of Gniezno, showing mobile e, which often has -ek in the nominative, but sometimes just -k, the result of analogy to oblique forms, and similarly -ec shows a loss of -e- due to similar leveling.^[156] This shows that at that time two tendencies occurs, keeping etymological forms and leveling, but these forms without -e- remain in northern dialects of Poland.^[157][43]

Another deviation is that e resulting from yers sometimes acts as e resulting from *e, sometimes then becoming o: dzionek, dzionka instead of *dnek, *dzienka, wioska from wieś, and also etymological but uncommon osieł alongside more common osioł, kozieł alongside kozioł.^[157][86]

Next to e || ∅ from strong yers is e in a place where there never was a yer, but in a form resulting in a consonant cluster: ogień, węzeł, węgiel, later mydełko instead of mydłko; braterski from bratski, piosenka instead of *piasnka, and in the borrowed suffix -unek, from earlier -unk, as late as the 16th century, from German -ung.^[157][158]

The prepositions/prefixes *vъ(-) sъ(-) > *w(e)(-)/*z(e)(-) also showed uncertainty as to whether the yer accompanying them was weak or strong, as these acted as clitics, being part of the same accentuation unit as another word, meaning that Havlík's law determines the distribution of e or ∅, resulting in the current alternation of w || we and z s || ze.^[157] Alternating forms of prefixes last longer as they are indivisible from the word and were repeated more.^[157] Prepositional forms on the other hand are separate and are used in conjunction with many more words and thus can’t connect accentually to the next word, and as a result show more fluctuation; and so after the loss of awareness of yers, phonotactics then decides more often which form is used, usually in order to avoid uncomfortable consonant clusters, such as doubled ones, e.g. we Wrocławiu, we wtorek, we Francji, ze zwierzyńca, ze strachu.^[159][43] The prepositions przed(e), od(e), nad(e), przez(e), bez(e) underwent a similar process.^[160] Dialects in southern Lesser Poland, Silesia, and to a lesser extent in central Poland and Greater Poland show forms with -e much more frequently.^[43]

Within dialects mobile e is often lost in oblique cases, e.g. bes || besu next to standard bez || bzu, or epenthetic e is inserted, wiater next to standard wiatr, or being in a different position: szwiec, szwieca next to standard szewc || szewca, see also § Declension of masculine nouns.^[160][43] In the north -e- in nominative masculine forms and genitive feminine/neuter plural forms is often missing: podwieczórk, paznokć, krawc.^[160] During the Old Polish era this was likely a regular phenomenon encompassing Masovia, Pomerania, and even north-east Greater Poland.^[160][43]

Proto-Slavic had sonorant diphthongs *ъr *ъl *ьr *ьl, within Lechitic they vocalized in diverse ways from yers, especially *ъl.^[161][162] This vocalization likely took place in the 10th century and not earlier than the 12th century, that is during the same era as the Lechitic ablaut, and after the Proto-Slavic palatalization of *k *g > *č *ž;^[163] *ьl was an early important isogloss between northern and southern Poland,^[164] and further an important isogloss between other Slavic dialects.^[165]

Proto-Slavic *TъrT generally gives TarT with few exceptions, found in the Bull of Gniezno: Carna = Karna, Carz = Kars - this a can be of two lengths, both long TárT and short TarT, which is the result of secondary shortening, see above, and the future development of this long and short a is identical to the development of long and short a otherwise: *bъrkъ > bark, *gъrdlo > gardło. ^[166][167]

Proto-Slavic *TьrT is varied within prehistoric Polish, and the result depended on the following consonant and its length. The clarity of the development is further muddied by later independent developments and leveling, etc.:^[168][169]

1. Before a hard front consonant t d s z n ł *TьrT > TarT, and the cluster hardens, including after a hard consonant and cz ż, e.g. wartki < *vьrtъkъ; the resulting a could be long or short and develops according to later clear and slanted a;
   1. Masovian and Kashubian changes this to 'a, keeping softness, e.g. Masovian siarna (standard sarna) < *sьrna, Kashubian cwiardi (Polish twardy) < *tvьrdъ; this is also present in a few words in Polish like ćwiartka, dziarksi;
2. Before any other consonant group, *TьrT > T’irT or T’irzT, e.g. wiercić < Old Polish wircić < *vьrtiti, which later lowered to e, developing according to clear and slanted e.

As a result of this, there is sometimes an alternation of T’irT||TarT as a result of ablaut, see § Proto-Slavic *ě.^[170] Leveling can occur, ziarna is likely the result of analogy with the locative zirnie, and ćwirdza, ćwirdzić assimilated to twardy, giving twi(e)rdza, twi(e)rdzić.^[170]

The cluster *TьrT always gives either r or rz; rz usually occurring before labials and back consonants, e.g. wierzba, with a few exceptions, such as cierpieć.^[170]

Northern dialects establish ‘ér everywhere except czerwony: śmiérć, piérsi, sérce, ciérpieć, piérszy; in Southern Poland and now including Krajna, er can be seen in some terms: ćwierć, piersi, śmierć, serce, and in others ér: ciérpieć, siérp, piérszy, wiérzba, wiérzch; the amount of terms with ér fluctuates by region.^[170] Forms like myrdać, styrczeć, tyrkować show a regional realization from southtern and central Poland and when these forms entered the general language they were often de-dialectalized to have -er-.^[171]

The word portki instead of *partki is an East-Slavic borrowing; korczak and storczyk have unexplained -or-; czarni instead of expected *czerni likely arose via leveling to other oblique forms of czarny and analogy to words like żarł, darcie, and darł; żerdka likely arose via analogy to żerdź; czerstwy has irregular -er-; ziarno instead of Old Polish zarno arose via leveling to the locative form zirnie; Old Polish ćwirdza, ćwirdzić were morphologically reinnovated based on twardy.^[172]

Exceptionally *TъrT and *TьrT give ur in e.g. kurcz, burczeć, turkotać, gurbić się, kurpie, purchawka; burczeć and turkotać are likely phonologically motivated and partially onomatopoeic, whereas the other cases are likely the result of coarticulation with either a labial or velar consonant.^[171]

Original *TъlT generally gives ł and a vowel, which can vary depending on the preceding consonant:^[173][172]

1. After a dental consonant TъlT > TłuT, e.g. słup < *stъlbъ;
2. After a labial consonant TъlT > TołT mowa < Old Polish mołwa < *mъlva; the resulting o could be long or short and develops according to later clear and slanted o;
3. After a velar consonant TъlT > T’ełT, e.g. kiełbasa < *kъlbasa.

A trace that Proto-Slavic long *TъlT partially retained length and partially shortened can be seen in the Old Polish alternation such as mowić||mówię, which was later leveled, now only see in mowa : mówić; length differentiation could also be seen in puł < poł||pełk.^[173] Kashubian often shows different reflexes, some forms possibly being the influence of the rest of Polish and some being purely Pomeranian.^[173]

The cluster *TьlT generally gives ł, rarely l, and the vowel can depend on the preceding consonant:^[174][175]

1. After a dental consonant TьlT > tłut, e.g. długi < *dьlgъ;
2. After *č *ž, from softened *k g, čьlč > čelč, but after umlaut > čołt||čółt; žьlt > želt, and after ablaut > žołt||žółt, e.g. tłusty < *tъlstъ;
3. After labials, there are two possibilities, depending on the following consonant:
   1. After a labial before a hard dental TьlT > TełT, e.g. pełzać < *pьlzati; but compare northern dialectal wiołna, piołł, see below;
   2. After a labial before other consonants TьlT > T’ilT, e.g. wigla < *jьvьlg.,

Alternations such as zółw but żołna, czółno but dialectal czołen show that length inherited from Proto-Slavic was shortened.^[164] The forms pełnię, pełnić, wełnie are exceptions, and the expected forms would be *pilnię, *pilnić, *wilnie, and are the result of early leveling where in other forms etymological eł is present, e.g. etymology wełna and pełny.^[164][176]

In northern Poland *ьl deviates after labials, as labials kept their original softness, e.g. pielli, wywielga; if e was before a hard dental consonant, then it underwent dispalatalization to o: wiołna, piołny, miołł, umiołty, piołł, and even wywiołka and can be attested as early as the 13th century in placenames like Stolpia (1204) (modern Słupia), Stolp (1240) (modern Słupsk), and Chołm (1136).^[164][177] This can be seen in some places of Masovia and neighboring areas, but even here always wilk, milceć; only in Kashubia you also see wołk, môłczëc; sometimes in Kashubia the softness of the labial is kept.^[164]

Proto-Slavic *or *ol *er *el developed differently depending on whether it was initial or medial; in Polish, initial orT, olT developed in two ways:^[178]

1. If orT olT had a circumflex tone, then metathesis occurred to roT-, loT-: rola < *orlьja, łokieć <*olkъtь;
2. If orT olT had a acute tone, then metathesis and o > a occurred to raT-, laT-: radło < *ordlo, łaknąć < *olknǫti.

Inherited examples of erT-, elT- do not occur.^[178]

Medial *TorT, *TolT, *TerT, *TelT within Lechitic and Sorbian likely saw a yer inserted between the liquid and the consonant with metathesis, so TъroT, TъloT, TьreT, TьleT (see § Proto-Slavic *e); however this insertion of the yer didn’t happen everywhere, and a lack of metathesis can be seen, see below on *TorT, TolT.^[179]

Initial *orT olT gave two results depending on the pitch within the diphthong:^[180]

1. *orT with a acute pitch gave raT: radło< *ordlo, ramię < *ormę;
2. *orT with a circumflex pitch gave roT: robić < *orbit, równy < *orvьnъ, roz- < *orz-, rożen < *oržьnъ;
3. *olT with a acute pitch gave łaT: łabędź < *olbǫdь, łaknąć < *olknǫti, łakomy < *olkomъ, łania < *olni;
4. *olT with a circumflex pitch gave łoT: łódź < *oldi, łokieć < *olkъtь.

Medial *TorT, *TolT likely became TъroT, TъloT. This can be seen in the albeit irregular use of the prepositions and prefixes w(-), z(-), which lost yers in accordance to Havlík's law: we wromotę, ze krolestwa.^[180] The presence of long ó or short o is a continuation of the preliterate and medieval alternation of o and ó, which is the result of the Polish dual result of the original length of the cluster *TorT, TolT, which was either kept or shortened, e.g. brona < *borna, kłócić < *koltiti, płótno < *poltьno.^[181] The reflex *TarT is found only exceptionally, especially only in old texts, occurring much more in Kashubian and even more in Slovincian; whether the reflex was TarT or TroT depends on the intonation of the syllable.^[182] Forms with TarT were later leveled due to analogy with forms with TroT as well as Polish influence.^[182]

Medial *TerT, *TelT is parallel to *TorT, *TolT and gave TreT, TleT, which underwent later changes appropriate for the phones, namely:^[183]

1. r near e softens to rz;
2. e depending on the length changes to e||é, see § Slanted é;
3. e > o in accordance with ablaut;
4. o > o||ó depending on length § Proto-Slavic *o.

This means that the Polish reflexes of *TerT are:^[183]

1. TrzeT||TrzéT, e.g. brzeg < Old Polish brzég < *bergъ, drzewo < *dervo. The prefix prze-, in Old Polish only a preposition (seen in przecz, przebóg), continuing *per-, usually stood before a consonant that began the next word, resulting in *perT, which is why it underwent the same metathesis. The terms trzemcha, trześnia, trzewa, trzewik, and Old Polish trzem with initial trz- come from older czrz-, attested as late as the 15th century;
2. TrzoT||TrzóT, e.g. brzoza < *berza, śród < Old Polish śrzód < *serdъ, including trz- resulting from czrz-, and unexpected trzop. The prefixes/prepositions przed(-) przez(-) did not undergo ablaut perhaps due to the fact they act as clitics, and the word they attached to could start with a hard or soft consonant, meaning that *prezъ *pŕedъ could change to *prezь *pŕedь, and that could stop ablaut, and then analogical leveling from the second type, i.e. without ablaut, spread.

• TelT has the reflexes:^[183]

1. TleT||TléT e.g. mleko < Old Polish mléko < *melko, plewa < *pelva;
2. TłoT||TłóT szłom< *šelmъ, żłób < *želbъ.

Some words, such as brzoza and środa < śrzoda seem to break the chronology with relation to ablaut, as the softness of the consonant suggests an original e, which later changed to o.^[184] In the words człon, szłom, and żłób, o is preceded by hard, suggested original *čolnъ šolmъ žolbъ, but o after these consonants was not present, meaning that the original forms were *čelnъ šelmъ želbъ, with ablaut happened in pre-Polish and before metathesis: *čelnъ šelmъ želbъ > *čolnъ šolmъ žolbъ > *člonъ šlomъ žlobъ.^[184]

In Slovincian, TelT merged with TolT.^[184] Forms such as płóc, młóc are found across all of Kashubia, and even in Greater Poland.^[184]

Polish undergoes major changes in terms of palatalization, changing both inherited palatal consonants and innovating some new ones, especially via the general changes in West Slavic and Lechitic from Proto-Slavic.^[185] Inherited soft consonsants initially remained soft, and later depalatalized; other consonants palatalized near front consonants, and after ablaut and the loss of yers, also phonemicized.^[185]

p remains without change, e.g. płakać < *plakati, spać < *sъpati, chłop < *xolpъ.

b is without change except when word-final or before a voiceless consonant, where it phonetically devoices, e.g. baba < *baba, dąb (pronounced domp) < *dǫbъ, babka (pronounced bapka) < *babъka.^[186] Devoicing, triggered by the loss of yers, began before the 15th century and finished in this century; this is true for all voiced and voiceless consonants.^[187]

If p and b were near a front consonant or j, they palatalized, later phonemicizing to ṕ b́ after ablaut: pić < *piti, piec < *peťi, bić < *biti, biesiada < *besěda.^[188][189] Proto-Slavic had epenthetic *ľ in *pj bj, but this was lost in pre-literate Polish, e.g. kupię < *kupľǫ (first person singular of kupić < *kupiti), gubię < *gubľǫ (first person singular of gubiti), except in a few words: kropla, grobla, przerębla, dialectal grable. Old Polish also has kropia, grobia, rzerębia, and now almost everywhere grabie except in a few dialects.^[187]

ṕ b́ before another consonant or word-finally hardened to p b, e.g. drop < *drop, drób (pronounced drup) < *drobь prać < *pьrati, brać < *bьrati.^[187][190] These soft consonants were written until the middle of the 19th century, but their phonetic hardening began during the Middle Polish era; Roter at the beginning of the 17th century says that Poles should say these soft, but they do not.^[191][192]

ṕ b́ in northern dialects undergo asynchronic pronunciation (asynchroniczna wymowa spółgłosek wargowych miękkich), whereby the palatal element changes into a different spirant, and the labial does not lose its palatalness, e.g. Masurian zrobzić (standard zrobic). ^[193] The degree of asynchronicness may vary, sometimes realized as j, sometimes as [x]/[ɣ], sometimes [ɕ]/[ʑ]; ^[193] see various articles on Polish dialects for details. Hypercorrections of this may also occur when trying to avoid this, for example in Kurpie psiec might be changes to [pʲxʲɛs], but ⟨psie⟩ (standard locative/vocative singular of pies) might be changes to [pʲxʲɛ].^[193] Villages with two or even three types of palatal realizations might be mixed together, making the drawing of borders difficult; it also happens that the villages of the provincial nobility have a pronunciation that is closer to the pronunciation of the educated classes, so they have, for example, the type [pʲj bʲj], and the neighboring peasant has [pʲxʲ] [bʲɣʲ].^[194] Asynchronic pronunciation can also be heard word-finally, except in the imperative.^[195]

t continues unchanged, e.g. trzy < *trьje.^[195]

d remains unchanged, except word-finally or before voiceless consonants, e.g. dobry < *dobrъ, niewód (prononced niewut) < *nevodъ, słodki < *soldъkъ (pronounced słotki).^[195]

If *t *d were near a front consonant, they palatalized to *t́ *d́ (/tʲ/ /dʲ/) and later develop to ć dź, e.g. ciec < *teťi, ciało < *tělo, cichy < *tixъ, dziesięć < *desętь, dziad < *dědъ dzień < *dьnь.^[195][189] The periodization of this change is placed beginning in the 12th century and ending in the 13th century along with ś and ź.^[195][196] dź word-finally and before a voiceless consonant becomes ć, and ć before voiced consonants becomes dź.^[197]

If t́ d́ were before r rz (< *ŕ) n ń after the loss of yers, they hardened, e.g. drę < *dьrǫ (first person singular of drzeć < *derti), drze < *dьretь (third person singular of drzeć < *derti), we dnie < vъ *dьne trę < *tьrǫ (first person singular of trzeć < *terti), trze < *tьretь (third person singular of trzeć < *terti), tnę < *tьnǫ (first person singular of ciąć < *tęti), tnie < *tьnetь (third person singular of ciąć < *tęti), piętnaście < *pętь na desęte.^[197][190] In the imperative before a soft or hard consonant, before which original i occurred or word-finally of these forms after the loss of -i: wróć, gryź; ć and dź assimilate in terms of voicing; final -dź becomes ć.^[190]

If t́ occurred before c after the loss of *ь, then it changed to ć, and then the resulting ćc in the 15th century changed to jc, in other words ćc > jcc > jc, e.g. winowajca, kojca, ojca; d́ in the same position loses its voicing and then changes the same way: rajca, zdrajca.^[197] Old Polish forms such as otca, otczyzna suggest a loss of palatalization in the cluster, unless they were spelled under Czech influence, but this would be a dialectal development.^[197] In later Polish, -jc- is partially kept by tradition, e.g. winowajca, zdrajca, ogrojca, ogrojcu, kojca, kojcem, the latter examples even change their nominative via leveling, from kociec ogrodziec > kojec ogrojec, and morphological innovations sometimes replace inherited forms, e.g. władca radca świętokradca; fluctuations are visible: władzca radzca.^[197]

t́ d́ in the groups st́ zd́ after the loss of *ь before s develop uniquely: *st́s > *śt́s > śćs > śs > jss > js; *zd́z first devoiced to *st́s and then had the same development: miejski < mieśćski. In the nominative plural of the masculine personal passive participle of several -ę, -esz verbs with t d stems have c dz instead of the expected ć dź due to dissimilation: gnieceni, pleceni, uwidzeni, pobodzeni, but gnieciony, gniecione.^[198] Gerunds of the same type of verbs also show this: gniecenie, uwidzenie, etc.^[199]

tj dj originally gave c' dz' (/t͡sʲ/, /d͡zʲ/), and by the middle of the 16th century these had hardened to c, dz (/t͡s/, /d͡z/), but are morphophonologically soft, giving the alternations t : c : ć (światło : świeca : świecic), d : dz : dź (rada : radzę : radzie).^{[199] [192]} See also § Proto-Slavic *ť *ď.

Proto-Slavic *stj *zdj give szcz żdż, e.g goszczę < *gosťǫ (first person singular of gościć < *gostiti), jeżdżę < *jězďǫ (first person singular of jeździć < *ězditi).^[199]

Proto-Slavic *ť *ď are inherited in Old Polish as /t͡sʲ/ /d͡zʲ/, spelled ⟨c⟩ ⟨dz⟩ and dz word finally becomes c: pieniądz.^[200] In the 15th century these were probably still pronounced with softening, and after their desoftening in the 16th century they become morphophonologically hardened in the alternations k : c : cz, e.g. ręka : ręce : rączka and g : dz : ż, e.g. noga : nodze : nożny.^[201] Thus the final development in Polish is c, dz, e.g. cały < *cělъ, cedzę < *cěďǫ (first person singular of cedzić < *cěditi).^[202] The form Polszcze was used until the 19th century - this is probably the result of an archaic exchange of Proto-Slavic *sk before *ě, which change to *šc and later *šč; therefore the form Polsce is from leveling with forms like ręce, matce, where *k + *ě of gives c.^[202] Polszcze can still be heard regionally.^[202] See also § Proto-Slavic *tj *dj.

k and g remain unchanged, and word-finally and before a voiceless consonant changes to k, e.g. kara < *kara, głód < *goldъ, bóg (pronounced buk) < *bogъ.^[203]

Proto-Slavic *ky *gy become [kʲi] [gʲi] in the Middle Ages, e.g. kisnąć < *kysnǫti, gibki < *gybъkъ, and the clusters *kъ *gъ initially gave ke ge, and later in the 15th century softened to [kʲɛ] [gʲɛ], e.g. kieł < *kъlъ, giez < *gъzъ; this process is sometimes called the fourth palatalization.^[204][205] These initially are not phonemes, as their occurrence is conditioned, and Parkoszowic does not differentiate the two; Steiber claims they phonemicize in the 17th century with the denasalization of final -ę, with potential minimal pairs such as drogę ([droɡɛ]) and drogie ([droɡʲɛ]), but also says that the difference is small and calls them "facultative phonemes" (fonemami fakultatywnymi).^[206]

kt in early Polish and in dialects simplifies to cht, e.g. chto < kto, and gd changes to [ɣd], and most often further to d, e.g. kiedy < Old Polish kiegdy < *kogъda.^[204][207]

Some dialects neighboring Kashubian, namely Tuchola Forest dialect, and the western Krajna dialect historically further softened [kʲi] [kʲɛ] [gʲi] [gʲɛ] to ci dzi, e.g. dzipci (/’d͡ʑipt͡ɕi/) (standard gibki).^[204]

v, written in Polish as w, remains unchanged, except word-finally or before a voiceless consonant or notably after a voiceless consonant it becomes f, e.g. woda < *voda, staw (pronounced staf) < *stavъ, wczoraj < Old Polish wczoraj < *vьčera.^[208] The change of v to f after a voiceless consonant began at the latest in the 12th century, which is indirectly shown by the fact that in Lesser Polish and Masovian dialects the cluster chv changed to f as early as the 13th century, which suggests and earlier chf, e.g. chwała > Old Polish fała, and notably some dialects kept voiced w even after voiceless consonants, and others yet have a semivowel realization of /w/.^[208]

In relation to these processes is the phonemization of f, as **f was not part of Proto-Slavic, and early borrowings replace f with p, such as the Old Polish Szczepan from Latin Stephanus.^[208][209] The early dialectal change of chv > f led to minimal pairs like falić vs walić, seen in the beginning of the 13th century.^[208][210] Later borrowings starting from the 14th and 15th century show the presence of f in; folwark||forwark, flak, facelet, farba (next to older barwa), forszt, fryjować, frywołt.^[208][210]

f has a unique origin in the word ufać and related words, as it was originally upwać, containing the prefix o- and the stem -pw-, related to pewny, and this cluster later simplified; compare obfity below.^[211][210]

If v was near a front consonant or j, it palatalized, later phonemicizing to v́ after ablaut:^[212][213]

1. Before a vowel v́ > v́ (no change), e.g. wiara < *věra, wieczór < *večerъ, wilk < *vьlkъ. Epenthetic ľ from vj was lost like like *pľ *bľ: Old Polish dawię < *davľǫ (first person singular of dawić < daviti), mówię < *mъlvľǫ (first person singular of mówić < mъlviti); niemowlę is an exception next to Old Polish niemowię, which is likely the result of Ruthenian influence;
2. v́ after the loss of *ь or *i of the old first and second person plural imperative -imy, -icie before a consonant loses its softness, and if that consonant is voiceless, it assimilates and devoices: owca (pronounced ofca) < *ovьca, mówcie (pronounced mufcie) < Old Polish mówicie;
3. v́ after the loss of *ь or the 2nd person singular ending of the imperative -i in word-finally loses its softness and voicing: krew (pronounced krew), mów (pronounced muf);
4. vj after the loss of *ь word-finally loses its softness and voicing: Jarosław (pronounced Jarosłaf) - this phonetic loss of softness happened during the Middle Polish era, but was sometimes written orthographically;
5. v́ after a voiceless consonant changes to f́: chwila ([xfʲila]);

f́ was not originally a phoneme for the same reasons as f, but became one alongside f before the beginning of the 13th century, and f́ can be seen in loanwords from e.g. the 14th and 15th centuries: fig(a), fiołek, firletka.^[212]

f́ has a unique origin in the word obfity, as it was originally opłwity, which has the prefix o-, the suffix -ity, and the stem -płw-, related to pływać, płynąć, pławić, thus meaning “swimming in something”; compare ufać above.^[214]

v́ f́ have an asynchronic pronunciation like ṕ b́ in Northern Poland, sometimes as [vʲʑ] [fʲɕ], or [vʲj] [fʲj], or [vʲɣʲ] [fʲxʲ]; the geographic spread of these variants is like that of ṕ b́.^[215] For v́ f́, the labial component may be lost entirely, giving just [ɣʲ] [xʲ] or [ʑ] [ɕ].^[215] The spirant resulting from the loss of the labial is never equal to standard Polish /j/, except in the dative ending -owi, as well as the suffixes -owi- -ewi-, thus konoju, Skierniejice, Rojiska.^[215]

The clusters śv́ śf́ in dialects develop differently - rather than decomposing, the labial hardens, but the soft sibilant remains, so śfat, ćwyrć.^[215][210] Pronunciations such as [ɕxʲat] occur, but are less frequent.^[215]

s remains generally unchanged, but before a soft consonant or as a prefix before ć can also undergo softening; skok < *skokъ, śmierć < *sъmьrtь, ściąć > *sъtęti, but spiąć > *sъpęti.^[215] In older forms of Polish and now in dialects it can soften in other positions.^[215]

z remains unchanged, but word-finally and before a voiceless consonants changes to s, e.g. zając < *zajęcь, guz (pronounced gus) < *guzъ, rozpamiętać (pronounced rospamiętać) < pamiętać.^[216] z can also soften before a soft consonant and as a prefix before dź, e.g. gwóźdź < *gvozdь, zdziałać (pronounced ździałać) < działać.^[216] In older forms of Polish and dialects z is ź before other soft consonants.^[216] z changes to dz exceptionally in dzwon and bardzo, still realized as barzo in the 17th century.^[216]

If *s *z were near a front consonant, they palatalized to *ś *ź (initially /sʲ/ /zʲ/),^[216][189] then fully soften to /ɕ/ /ʑ/ beginning in the 12th century and ending in the 13th century along with ć and dź.^[196][216] Softness remains with few exceptions:^[216]

1. Before a front vowel: siec < *sěťi, się < *sę, sidło < *sidlo, ziemia < *zemľa, zima < *zima, ziębić < *zębiti;
2. Before a back vowels that exists as the result of ablaut or the development of [æ̃]: sąsiad < *sǫsědъ, siostra < *sestra, siąknąć < *sęknǫti, zioło < *zelo;
3. Word-finally after the loss of *ь: oś < *osь, gałąź < *galǫzь;
4. Word-medially before a soft consonant after the loss of *ь: w ośle, w koźle, głośni, groźnie, prośbie, groźbie;
5. Word-medially before a hard consonant after the loss of *ь:, prośba < *prosьba, głośny < *golsьnъ, groźny < *grozьnъ, kośca < kosiec. ś ź desoften in a few cases: pismo < *pisьmo, osła (from osioł), kozła (from kozioł). Words like żałosny miłosny have an s resulting from the § Simplification of consonant clusters;
6. In the imperative before a soft or hard consonant, before which original i occurred or word-finally of these forms after the loss of -i: gróź, proś, gróźcie, proście, gróźmy, prośmy.

ś and ź assimilate in terms of voicing; final -ź becomes ś.^[217][190]

Proto-Slavic *ś̌ *ź̌ entered Polish, now spelled ⟨sz⟩ ⟨ż⟩, originally as /ʃʲ/ /ʒʲ/ through the 15th centuries, and by the middle of the 16th century desoften to /ʂ/ /ʐ/, but remained morphophonologically soft seen in the alternations ch : sz as well as s : sz and g : ż as well as z : ż, e.g. mucha : musze, pisać : piszę, rogal : rożek, mażę : mazać.^[217][192] Reflexes include szyć < *šiti, dusza < *duša, kosz < *košь, żal < * žalь.^[217] ż word-finally and before voiceless consonants phonetically changes to sz, e.g. mąż (pronounced mąsz) < *mǫžь.^[217]

Sometimes sz ż correspond to foreign s z in loanwords: msza, nieszpory, koszula, Tadeusz; żegnać, żagiel, Old Polish żemła.^[217]

Older sz changes to ś in the nominative plural ending of nouns, adjectives, numerals, and pronouns of the masculine personal gender: Włosi, głusi, starsi, pierwsi, nasi - this occurred in the 17th century based on morphological and morphophonological leveling, as the typical nominative masculine personal ending became -i with softening of the previous consonant: sąsiedzi, kaci, chłopi, młodzi, łysi, grubi, tępi, and this also encompassed sz ż, but not always consistently, as one typically sees duży || duzi but not boży||bozi, świeży||świezi.^[217]

Proto-Slavic *zž, resulting from older *zǯ́ which continue original *zg+j as well as *zdj gives Polish żdż, which phonetically changes to szcz due to word-final devoicing or assimilation.^[218] A change of zż to żdż also occurs near the prefix z- with original ż, but not secondary, in the onset of the stem if the original two consonants were separated by a yer: żdżegł < zżegł, żdżyć < zżyć, Żdżary < zżary.^[218] Proto-Slavic *sč, which continue original *sk+j as well as *stj gives szcz, e.g. piszczę < *piščǫ < *piskjǫ (first person singular of piskać, piszczeć < *piskati).^[218]

Proto-Slavic *x remains in Polish unchanged, spelled as ⟨ch⟩, e.g. chleb < *xlěbъ.^[218]

The words Chryst(us), chrzest, chrześcijanin, chrzcić, and chrosta originally and dialectically have k, which changed to ch near r rz.^[218]

In Lesser Poland and Masovia chv- and chv́- change to f, f́, fała, fila, fiać, see § Proto-Slavic *v and the rise of Polish f and § Proto-Slavic *v́ *vj and the rise of Polish f́.^[219]

Before a voiced consonant ch phonetically becomes [ɣ]: klechda ([klɛɣda]); this consonant is near [ɦ] which occurs in Ruthenian, Czech, and Slovak Polish, e.g. hańba, hak, herbata, hołota, see articles on Polish dialects, however this is not a phoneme.^[220][221] [ɣ] or [ɦ] occurs sometimes via prothesis or foreign influence (e.g. hańba instead of gańba), and in Borderlands and eastern dialects it may be considered a phoneme of foreign influence.^[221]

ch in Lesser Polish dialects is very weak and intervocalically is barely audible, thus subsequently it was lost in the placename Sudoł (<*Suodoł < *Suchodoł).^[220] Word finally there was a potential for -ch to be lost, but this would lead to morphological complications, such as forms like nogacah becoming noga, robiłach becoming robiła; consequently Lesser Poland final -ch was strengthened to -k, see Lesser Poland dialect group.^[220] This change originally occurred in all of Lesser Poland, but is being lost under influence of standard Polish and now occurs only in parts of Lesser Poland.^[220] This pronunciation fades the most quickly when -ch is the final consonant of the stem and not an inflectional ending, meaning it appears before a vowel before inflectional endings, thus na nogak, na rękak but ruch, dach, groch; stem-final -k is heard now in Goral dialects.^[220] The change of final -ch > -k is also seen in parts of Silesia around Pszczyna and Stalinogród, but this is not a Silesian innovation, but Lesser Polish influence.^[220] In Spisz and around Nowy Targ final -ch becomes not -k but -f, so na nogaf, posłaf, zrobiłef; places with -ch > -k preserve place of articulation, -ch > -f preserve manner of articulation.^[220] The earliest attestations of this change can be seen in the 15th and 16th centuries; it is controversial whether the phenomenon was just emerging at that time and reached its full intensity in the 17th century, or whether it had already flourished then and even earlier, but because this phenomenon encompasses both Silesia and Lesser Poland, it is probable that this began in earlier eras.^[220] Finally Masovian influences on eastern and northern Lesser Poland reintroduced final -ch.^[220]

The Proto-Slavic cluster *chy is generally kept without change, e.g. chyba < *xyba, chylić < *xyliti, and early borrowings in Old Polish with [xi] or [hi] are changed to chy.^[220][206] [xʲ] entered Polish at the latest in the 18th century in loanwords such as Chiny, historia, Hipolit, hipokryzja, as borrowings originally had chy, hy, like chyża, hydra, hymn, and that [xʲ] entered Polish under the influence of German, Italian, and French and then appeared in the 19th century and then only exceptionally in iterative forms such as wysłuchiwać, podkochiwać as well as their derivatives, however soft ch́ remains not a phoneme.^[222][223] che remains in native words without change: uchem, duchem, rozchełstać, chełpić się, chełm, and foreign che, he are also kept: chemia, cherubin, herezja, herbata.^[224] However foreign hie is adopted as [xʲjɛ], e.g. hierarchia, hierarchy, attested since the beginning of the 17th century, Hieronim, hieroglif, hiena.^[224] This is the second Polish innovation of soft [xʲ], meaning that [xʲ] is a Middle Polish innovation which entered Polish via loanwords.^[224]

Word-initially and word-finally j is kept without changes, e.g. jałowy < *jalovъ.^[224]

Sometimes -j was added word finally, namely in dzisiaj, wczoraj, tutaj, indziej, trzej, dwaj, obaj, chociaj, dalej, więcej; the origin of this -j is various and not explained, particularly with regard to dzisiaj, wczoraj, tutaj, indziej.^[224]

In trzej < trzé < *trьje it could have come from a diphthongal pronunciation of é, that is /ei/, and then by analogy to the numeral dwaj, obaj, czterej, see declension of numerals;^[200]

As for the comparative of adverbs, one can assume that in the earliest periods of Polish there were two types:^[200]

1. więcej, krótce, dale, niże, created with the suffix *-je;
2. nowieje, starzeje < *nověje, *starěje, created with the suffix *-ěje.

Within the second type, contraction could have occurred: *-ěje > -é and then >-ej like trzej, or there was a shortening of *-ěje > *-ěj > -ej; then this -ej begins to spread to other comparatives in the 15th century via leveling.^[200] This also could have affected the comparative of adjectives such as ładniejszy, głośniejszy, chytrzejszy via analogy.^[200]

The presence of the alternation of the superlative naj-||naj- is inherited from Proto-Slavic, where it was probably a conflation of two prefixes.^[200] In the Middle Ages na- was used in Lesser Poland and Greater Poland; the appearance of naj- in the translations of psalters is from Czech influence; in the 16th century na- sees more use, only in the 17th century the Masovian prefix naj- comes to dominate, possibly under influence of Ukrainian influence Ruthenian nobility.^[200]

In Old Polish and in dialectal prothetic initial j- can be seen: Jadam, Jarnold, jamroz, Jewa, and dialectically jucho, judo.^[200]

Proto-Slavic *č is inherited into as Old Polish /t͡ʃʲ/, spelled ⟨cz⟩, and remains this way until the middle of the 16th century, the desoftening of sz ż, when it desoftened becoming a morphophonologically hardened consonant in the alternations k : c : cz, e.g. ręka : ręce : ręcznik.^[202][192] Inherited examples include czoło < *čelo, czas < *časъ. Forms with original czrz- change in Middle Polish to trz-, thus czrzemcha > trzemcha; czrześnia > trześnia; czrzewa > trzewa; czrzem > trzem; czrzoda > trzoda; czrzop > trzop; see also § Proto-Slavic diphthongs *or *ol *er *el.^[202] This happened because cz is a fricative containing /ʂ/, and ⟨rz⟩ (see § Proto-Slavic *r *ŕ *rj became /ʐ~ʂ/ in Middle Polish, meaning that /ʂ/ occurred twice, thus /t͡ʂr̝ɛmxa/ > /t͡ʂʂɛmxa/ > /t͡ʂɛmxa/, meaning that the /ʂ/ in cz was lost, leaving only /t/.^[202]

cz becomes dż before voiced consonants: liczba (pronounced lidżba).^[202]

Many dialects merge the series of sibilants and affricates in various ways; mazuration, also sometimes called in Polish sakanie, is the merger of sz ż cz dż with s z c dz (notably /ʐ/ /ʂ/ from rz is unaffected) and is considered an extreme form of depalatalization; jabłonkowanie, also called siakanie, is the merger sz ż cz dż with ś ź ć dź, often realized respectively as /ʃ/ /ʒ/ /t͡ʃ/ /d͡ʒ/; finally kaszubienie is the merger of ś ź ć dź with s z c dz.^[225][192]

These mergers often occur outside of the regions they were named for - mazuration happens in most of Masovia, Lesser Poland part of Silesia, and small islands in Greater Poland, which otherwise does not merge anything; jabłonkowanie occurs in Silesia near Jabłonków and parts of Masovia; except kaszubienie, which occurs in Pomeranian.^[225]

The cause and age of mazuration are unknown - some scholars consider it to be a prehistoric development from the 10th-11th centuries, some consider it a later development, as late as the 15th century; some consider the cause to be the result of foreign substrate, namely Finnish, Prussian, Celtic, and others to be independent, i.e. the result of difficulties distinguishing s š ś, and notably such difficulties can be observed in young children.^[225]

The chronology of mazuration is tied with the rise of the literary standard - if mazuration is old, then the literary standard would have arisen from non-mazurising Greater Poland, but if it is young, from the 14th-15th centuries, there is no specific reason to associate the standard with Greater Poland. ^[225]

Jabłonkowanie is the result of mixing of people groups - in the north the Polish population had contact with mazurising groups as well as non-mazurising groups where remnants of the Old Prussian population could have still been, who had neither mazuration nor ś ź ć dź; in the South, in the Beskids, two Polish colonizational groups collided, the mazurising group from Lesser Poland and the non-mazurising group from Silesia, as well as Slovak peasants without mazuration, but with ś ź ć dź.^[225] It is impossible to tell if foreign influence or the mixing of two native groups speaking differently was more important - perhaps the foreign influence, as mazurising and non-mazurising groups met elsewhere and similar mergers did not happen. ^[225]

r is kept without change, e.g. raz < *razъ.^[226]

ŕ rj and r before a front vowel have the same development in Polish, written ⟨rz⟩, and the prehistoric and early literate realization was probably /r̝/, as it was often written with ⟨r⟩, such as in Bull of Gniezno; in the 13th century the spellings ⟨rs⟩, ⟨rsz⟩, ⟨rz⟩ begin to see use, suggesting a change in pronunciation where the softness of /r̝/ changed to a more fricative realization, probably /r͡ʒ/ (or phonetically /r͡ʂ/ after a voiceless consonant.^[227][228] It is possible this realization began as a dialectal feature of Masovia that spread along with its spelling over the 14th century, when some old documents such as the Holy Cross Sermons still used the spelling ⟨r⟩,^[227] or from Greater Poland in the 15th century.^[196] rz lost its softness probably only during the 16th century, see § Proto-Slavic *ś̌ *ź̌ and § Proto-Slavic *ć̌, and continued as /r͡ʐ/ during the 16th and 17th centuries; this is reflected in orthography, as even in manuscripts, which were less careful in terms of spelling, never used rz for ż like what might happen in modern Polish.^[227] Further evidence is that rz and ż do not rhyme at this time, but on the contrary rhymes such as dzierży - szerzy are attested.^[227] Mesgnien (1649) writes that “rz is pronounced harder and with a certain scraping (durius et cum stridore quodam), which is difficult to explain without the model of a teacher.”^[227] Over the course of the Middle Polish era rz slowly loses its rolled r element and the fricative ż becomes more and more prominent, and by the 18th century ultimately merges with ż.^[227][196] The ż realization is seen exceptionally in the 15th century (1410), in Rej’s works, and with faux mazuration in the work Peregrynacja Maćkowa from 1612, notable as /ʒ/ didn’t change to /z/ later.^[227] Examples of inherited rz include rzeka < *rěka, burza < *buřa.^[227]

The realization of rz differs in some dialects, where the rolled r element is often kept, see articles on various dialects of Polish; it is also kept somewhat in Kashubian and Slovincian, but this realization is giving way to ż everywhere.^[229] This change of /r͡ʐ/ > /ʐ/ results in the reintroduction of /ʐ/ into masurizing dialects. The masuration of rz occurs exceptionally along the line of contact between masurising and non-masurising dialects, among polonized Germans, or in heavily germanized Poles.^[229]

rz word-finally or before a voiceless consonant loses voicing and sounds like sz; rz also loses voicing after a voiceless consonant like w.^[229] This progressive assimilation, as opposed to typical regressive assimilation found in other consonant clusters, is explained by the fact that that old voiced /r͡ʐ/ and voiceless /r͡ʂ/ were allophones and did not create any minimal pairs, but if a voiceless consonant assimilated in voicing to /r͡ʐ/ then it would have to merge with another phoneme, which would cause confusion, e.g. trze would sound like drze, krze like grze, krzywa like grzywa, krzep like grzeb, trzewa like grzewa.^[229]

If rz occurred before ł l ĺ c s it lost its softness and became r: orła, w orle, orli, but orzeł, starca, but starzec, twórca but tworzyć, cesarstwo but cesarz.^[229][192] In standard Polish rz becomes r also before n ń: wierny, wierni, piernik, but in Old Polish rz can be seen: srzebrzne, knąbrzny, knąbrznie; the modern Polish powietrzny, wietrznie, Jaworzno, Jaworznie etc. are the result of mixing and leveling forms with *rьn that gave rn and *rьń which gave rzń.^[229]

Old *sŕ occurred as śrz until the middle of the 15th century, after which śr can be seen in Lesser Poland and Masovia, which dominates in Lesser Poland and Masovia in the first half of the 16th and in Greater Poland old śrz dominates at this time.^[230] Since the middle of the 16th century the spelling śrz is established in prints, so śrzoda, śrzodek, pośrzód, pośrzatnąć as the result of Greater Polish dialectal influence; then around 1820 the spelling śr starts to spread under the influence of Masovian after Warsaw became the capital.^[231][232]

Old *zŕ was kept as źrz until the middle of the 15th century; in the first half of the 16th century in Lesser Poland as źr, in Greater Poland as źrz as jrz, and in Masovia as źr and jrz.^[231] From the middle of the 16th century century the spellings źrz jźrz (jrz) dominate in texts from Greater Poland dialectal influence and at the end of the first quarter of the 19th century the forms źr jrz spread in the standard due to Masovian influence for the same reason as above.^[231]

In parts of Lesser Polish śr źr underwent metathesis, giving rśoda, rśode, rźůdło, rźeb́e, dorźały, sporźał, sometimes with hardening, e.g. wyrze, sporzá; metathesis isn’t equally spread out in all words in this area.^[231][232] This phenomenon is very old, with forms like w posriodku, we rsiodę attested in the 16th century.^[231] In Silesia and Kashubia the development was more or less the same with epenthetic t d giving strz zdrz: strzoda, strzybło, uzdrzeć, zdrzudło; Greater Poland most often has the form śrz źrz, but in areas strz and zdrz are possible; Lesser Poland shows śr źr or sometimes metathesizes to rś rź; Masovia has śr źr.^[231][232] Other variants are possible where środa and źródło did not come to dominate.^[231]

Proto-Slavic *l was originally inherited as a dental lateral liquid /ɫ/, written as ⟨ł⟩, and over the course of the 16th and 17th centuries, completing in the 19th and 20th centuries, it became in most places /w/ through a process called wałczenie; the earliest attestations of this come from 1588 from acts from the Kraków archive with the spellings putora, putrzeci; traces of it from the 15th and 17th century in Lesser Poland, second half of the 17th century in Masovia, and the second half of the 17th century in Greater Poland can be seen.^[233][234] Kochanowski calls ł “foreign” (barbarum), probably in reference to /w/, and that it was seen as strikingly different from /ɫ/, and wałczenie can also be seen at the beginning of the 17th century in Perygrynacja Maćkowa from 1612: okoo||około, psezegnau (przeżegnał), poutory||półtora.^[219] The realization /w/ spread far in the 19th and 20th centuries, and dental /ɫ/ remains in many places around Poland, particularly on the intersection of Poland and Belarus, Ukraine, or Czechia.^[219] Inherited examples include łąd < *ladъ, łeb < *lъbъ.^[233]

Proto-Slavic *ĺ *lj have the same reflexes within Polish as does *l before a front vowel, written as ⟨l⟩.^[219][228] Before i it undergoes palatalization as [lʲ]; it has this realization during the Middle Ages in all other positions as well, somewhat in the 16th century, and then later became [l], while still acting morphophonologically soft.^[219][196] Inherited examples include lipa < *lipa, myśl < *myslь.^[219]

In Greater Poland, Silesia, and large parts of Lesser Poland l before i is [lʲ]; in Masovia however it is pronounced the same as before other vowels, and in dialects differentiating y from i the realization is therefore ly; this can also occur in neighboring dialects experiencing influence from Masovia.^[235]

Proto-Slavic *m continues unchanged, e.g. mak < *makъ.^[236]

If m was near a front consonant or j, it palatalized, later phonemicizing to ḿ after ablaut:, e.g. miedza < *meďa, karmię < *kъrmľǫ (first person singular of karmić < *kъrmiti).^[236][189] Word-final or preconsonantal ḿ hardens to m, e.g. dam < *damь (first person singular of dać < *dati), kłamstwo < *klamьstvo.^[236][190] The hardening of ḿ occurred probably by the end of the 16th century, as Roter in 1616 states that this realization in this position is rare and exceptional,^[236] but some place its end in the New Polish era.^[192] See also § Proto-Slavic *v́ *vj and the rise of Polish f́.

Soft /mʲ/ in northern Poland decomposes either to mj or mń, as other soft labials, see § Proto-Slavic *v́ *vj and the rise of Polish f́, and words with the etymological cluster mń sometimes undergo hypercorrection, e.g. miejszy (standard mniejszy), suggesting a conscious awareness of mń being an proscribed pronunciation.^[236] In dialects with [ɣʲino] [oxʲara] or zino osiara (standard wino ofiara) the labial element completely disappears leaving just ń: niasto, zienia (ziemia), kanień (kamień); elsewhere the shift of ḿ > ń is limited to positions after consonants: śniérć (śmierć), śniga (śmiga), jęcznień (jęczmień), but mniasto, kamnień, ziemnia.^[237] This change of ḿ > mń or ḿ > ń sometimes results in mń in place of ń: mnisko (nisko), mnitka (nitka), mniecka (niecka), but only in the first syllable, so words like kuźnia, tani do not change; ń in place of ḿ may also occur: miecka, mitka, śmiádanie; this change of ń > ḿ is likely the result of hypercorrection.^[238] Pronunciations such as kanień, na kaniéniu were proscribed in Kopczyński’s brochure Poprawa błędów (1808), which was aimed at upper levels of society, not peasants, meaning that even elite sometimes had this pronunciation.^[238] Soft ḿ behaves particularly interestingly in the nominative plural ending -ami as well as the pronominal clitics mi, mię in that within these the soft labial often depalatalizes, giving forms like rękamy, nogamy, (daj) my, (uderzył) me/(uderzył) ma, in other regions pronunciations such as both nogami (with soft m), mi alongside niasto and nogani can be heard.^[238]

Proto-Slavic *n remains unchanged, e.g noga < *noga.

Proto Slavic *ń *nj and *n before a front vowel have common reflexes in Polish as ń, e.g. nieść < *nesti, koń < *koňь.^[239][189] If ń occurred before ń or n due to the loss of *ь, it changed to n: konny< *konьnъ, konnica < *konьnica.^[239][192] In the first half of the 13th century ń in ksiądz, ksieni, księga, and related words changed to ś as the result of dissimilation of two nasal phones ń - ę or ń - ń, that is ń lost its nasality but kept its palatal pronunciation, which after voiceless k became ś; the word knieja however remains unchanged.^[239][240]

Polish inherits many prothetic Proto-Slavic forms, including *j- before front vowels and before the back vowels *a *u in a few words, *v- before the back vowels *ъ *y and sometimes *o and *ǫ.^[241] In the Old and Middle Polish era, the prothetic consonants /j/ before i, e and sometimes a, rarely /v/ before o, a, very frequently /ɣ~ɦ/ before e, i, u, o and unmarked /w/ before o and sometimes u (now written ô û) can be seen in texts from the 14th, 15th, and 16th centuries.^[242]

Due to the loss of weak years many consonants came into contact with one another and within these clusters tendencies of phonetic assimilation occurred based on place of articulation, softness, and sometimes the total loss of a consonant occurred resulting in the simplification of these clusters, namely:^[243][244]

1. *žьs > s: boski, męstwo;
2. *šьs > sL włoski, suski;
3. *sьs > s: ruski, niebieski;
4. *zьs > s: łaski (from Łazy);
5. *dьs > c (orthographipcally dz): ludzki, sąsiedztwo;
6. *tьs > c: kącki, bogactwo;
7. *čьs > c: świadectwo, co (from *ć̌ьso);
8. *zdьn > zn: Gniezno, próżny (Old Polish prózny);
9. stn > sn: miłosny, żałosny, szesnaście;
10. rdn > rn: miłosierny;
11. rdc > rc: serce;
12. *rv > r: topьrvo
13. *stьl > śl: jeśli;
14. *slьs > s: przemyski;
15. *pъv > f in ufać and related words;
16. *xv > x: chory
17. xv > f: fała (see § Proto-Slavic *v and the rise of f as well as v́ and f́);
18. łdn > łn: żołnierz (Old Polish żołdnierz and compare modern żołd);
19. stb > zb: izba;

Two tendencies within the historic development of inflection occur: 1) inflectional endings undergo the above discussed phonetic changes, 2) a reduction in the number of inflectional paradigms via analogical leveling and supported by phonetic changes.^[245] In terms of number, Polish gradually lost the dual seen in nouns, adjectives, pronouns, and verbs.^[246]

Closely connected to the declension of nouns is grammatical gender and the morphologization of it, that is gender being determined by the ending.^[247] Polish inherits morphologized gender from Proto-Slavic replacing thematic declension, for example in the neuter, but also later demorphologizes its endings, that is spreading certain endings regardless of gender, for example -om in the dative plural displacing feminine -am, -ami in the instrumental plural spreading from feminine declension, and -ach in the locative plural also spreading from feminine declension.^[248] Polish declension also underwent semanticization, that is changing under influence of the meaning of the word, seen in the use of genitive for accusative among masculine animate nouns.^[249] Polish further develops this by innovating its modern three-way split of animacy, masculine personal, masculine animal, and masculine inanimate in the New Polish era.^[250] Other examples include full declension of words like to państwo "this nation", te państwa "these nations" versus defective declension like ci państwo "this husband and wife" and the differentiation of proper nouns, e.g. mech : mchu but Mech : Mechu.^[251]

Polish also makes changes to its inherited grammatical number system seen in the loss of the dual number, replacing it with the plural, however traces of the dual can be seen in some irregular declension of body parts naturally occurring in pairs (rękoma w ręku, oczy but oka in other meanings, uszy but ucha in other meanings) and fossilized phrases (mądrej głowie dość dwie słowie).^[246] Polish also has plurale tantum and singulare tantum nouns.^[252] Finally Polish see suppletion, considered a newer development.^[253]

In terms of nominal case, Polish inherits the seven cases seen in Proto-Slavic: the nominative, genitive, dative, accusative, instrumental, locative, and vocative case.^[254]

Masculine nouns come from Proto-Slavic *-o-, *-jo-, *-u-, *-i-, and *-n- stem nouns.^[255]

In Polish *-n stems in the nominative singular were replaced with the accusative: kamień, płomień.^[256][257] In other types the nominative inherits: -∅ after the loss of yers, but the presence of *-ъ, *-ь can be seen in the final consonant of the stem: hard or soft or sometimes functionally soft: róz, wóz, syn, wół, gość koń, mąż, gołąb (< *gołąb́).^[256][257] A Polish innovation among this is within masculine soft-stem nouns, that is the hardening of final soft labials (see § Proto-Slavic *p *b and the rise of *ṕ *b́, § Proto-Slavic *v and the rise of f as well as v́ and f́, and § Proto-Slavic *m, and the rise of ḿ) and the phonological hardening of historically soft sz ż cz dż, which remain morphologically soft.^[257] The new shape of the nominative, which is the result of the loss of weak final yers, is a stem ending in a consonant, and plays a prominent role in the leveling reshapings and through this the original differentiation of *-o/*-u stems ending in *-ъ and *-jo/*-i stems ending in -ь is blurred.^[256][257] The original oposition of *ъ : *ь was morphologically more distinct than the new opposition of hard and soft final consonants in the stem and through this masculine nouns gain another similarity to each other beyond being masculine by having a suffixless stem; this change also influences oblique cases.^[256]

The genitive singular is an example of leveling and mixing of declensional paradigm, a process which began centuries ago and continues now.^[258] The vast majority of Proto-Slavic masculine nouns were *-o and *-jo stems and used the genitive ending *-a; the few *-u stems that existed took the genitive singular ending *-u.^[258][257] It is possible that already in the Proto-Slavic stage the genitive ending *-u began to see use in old *-o stems and this tendency extended to *-o and *-jo stems in preliterate Polish; this can be seen in Middle Ages Polish and still before the 16th century in that many *-o stem nouns took the original old suffix -a and also -u, e.g. wschoda||wschodu and -a can still be seen in the 16th century, e.g. pokoja.^[258] Comparing genitive forms over the course of Old Polish and 16th century Polish shows a gradual replacement of -a with -u, and the use of -a became limited depending on the meaning of the noun in the 16th century and -u becomes the dominant ending in 15-16th centuries.^[259][260] Animate nouns use -a and inanimate nouns use -u, but there is much fluctuation, especially when in a rhyming position, where archaic forms may occur, e.g. człowieka - do wieka but od wieku - człowieku (Kochanowski); this tendency can be seen today, e.g. fotela||fotelu.^[261][262] In modern Standard Polish -a is used for, among others, animate nouns, diminutives, body parts, dishes/vessels, tools, measures, measures of weight, coins, games, dances, magazines, and months whereas -u is predominately used for inanimate nouns, loanwords, mental nouns, collective nouns, material nouns, and the days of the week, as well as abbreviations; there are of course exceptions and words where both endings may be possible, or both occur but are semantics-dependent, e.g. przypadku "accident" versus przypadka "grammatical case", geniuszu "genius as a quality" versus geniusz "genius as a person".^[261][260] Old -u stem nouns often take -a due to leveling, like syna, oćca, ducha.^[261] Original *-i and *-n stem nouns generally take -a, and Masovian and Greater Polish old texts sometimes show -e, which is either the result of a dialectal pronunciation of a as e after j or after original j that softened the preceding consonant or potentially inherited, otherwise -i and -e of *-i- and *-n- stems were lost.^[263][257] Dialects also show fluctuation in the distribution of -u and -a; generally -a sees wider use in dialects and words where -u replaces -a are rare.^[263] Also within dialects mobile e within cases is lost, so mechu, besu, dechu.^[263]

The dative singular shows a similar tendency as the genitive: *-ovi from *-u stems begins to take over the function of -u that of *-o and *-jo stems; the cause for this is that -owi is more distinctive, as it is found only in the dative, and nouns where -u replaced -a also removed -u in the dative for -owi in order to avoid the ambiguity of -u.^[263][264]} -owi also becomes the ending for animate nouns, especially personal, perhaps under influence of etymological synowi.^[265] During the Old Polish and Middle Polish periods many originally *-o stem nouns kept -u, especially one-syllable words and nouns in a prepositional phrase with ku.^[266] Examples of both datives forms are attested in Old Polish: gniewu||gniewowi.^[266] In the 17th century -owi, used more in colloquial speech, spread to the standard and began to be used in the vast majority of words, and nouns whose genitive have -u also begin to take it, except bzu, tchu, snu, whereas nouns that had -a in the genitive except a few words, especially monosyllabic, kept -u: bratu, chłopu, księdzu, pani, ojcu, chłopcu.^[266][267] Place name ending in -ów kept traditional -u after ku: ku Krakowu, ku Tarnowu, ku Charkowu, but otherwise Krakowowi, Tarnowowi, Charkowowi without the preposition.^[266] In dialects -owi can be seen in place of literary -u: bratowi, księdzowi, panowi, psowi; especially in the north; also in the north the new ending -owiu dominates (sometimes pronounced -oju or -oziu from the decomposition of soft labials) as a compromise of the two suffixes.^[266] The ending Greater Polish and Masovian dialectal ending -ewi saw use until the 16th century; originally it only appears after soft and functionally soft consonants, and is an innovation based on an old alternation of *o : *je, e.g. *rabomъ : *konjemъ, *togo : *jego, but in Lesser Polish texts -owi dominates, with Masovia following this first and later Greater Poland, and -ewi generally stops seeing use in the 16th century, except some Greater Polish dialects, which retain this.^[266][262] In the 15th and 16th centuries -ewi sometimes saw use after hard consonants: obrazewi, głosewi.^[266] This ending still sees use in the Łowicz dialect.^[268] Around Sochaczew the realization -eju sees use, the result of combining -ewi and -u, like -owiu/-oju.^[269] *-i and *-n stem nouns since the earliest times have only a leveled ending -owi following original -i: gościowi, gołębiowi, kamieniowi, dniowi.^[269][262]

As a result of sound changes, namely the loss of yers, the accusative singular gained the ending -∅ and this remains the case for inanimate nouns to this day.^[269][267] Animate nouns, on the other hand, especially personal nouns, show a tendency to use the genitive for the accusative; this tendency must have spread in pre-historic Polish, as already in the 14th and 15th centuries the -∅ ending for animate nouns is seen exceptionally, and in the 16th century only for non-personal animate nouns; the -∅ ending remained the longest in Masovia.^[269][267] Within the 16th century, genitive for animate accusative became the exceptionless norm, with the exception of a few fossilizations: iść za mąż, być za pan brat, siąść na koń, na miły Bóg, na święty Mikołaj.^[269][267] This change was semanto-syntactically motivated, because if both the subject and object are animate nouns and had the same form and word order is loose, the meaning of the sentence is blurred, for example ociec widzi syn could mean either “the father sees the son” or “the son sees the father” and a difference in form helps clarify the meaning.^[270] In new Polish the use of genitive for accusative spread in some inanimate nouns, for examples the names of dances, games, money, cigarettes, and in some phrases.^[271][267] *-n stem nouns see the accustive spread to the nominative.^[271][267]

In historic Polish only -em sees use for the instrumental singular, within *-u, *-jo, *-i, and *-n stems this is the etymological form of *-emь, *-ъmь, and *-ьmь, where -e- is the result of the changing of strong yers or inherited.^[271][272] For *-o stems, the use of -em is the result of leveling, replacing *-omъ of *-o-, based on *-u stems; -em coming from -ъmь is seen in the hardness of the final stem consonant, and then this -em with a preceding hard consonant spread because all other stems also had the also had -em.^[271][273]

In the locative singular the inherited Proto-Slavic state remains within nouns - *-o stem nouns have -’e < *-ě, e.g. rodzie.^[274][272] Until the 16th century nouns whose stems ended in -k, -g, -ch took -e (< *-ě) showing softening of *k > c, *g > dz, *x > sz: człowiece, języce, zamętce, pagórce, obłoce, bodze, okrędze, strasze, grzesze, słusze; all other stems from the oldest times show a spreading of -u originally only used for *-u stems, and this ending also takes *-jo stems which only keep the original ending -i exceptionally (gaji, stolcy in the 14th century).^[274][272] Some *-i stem nouns also take it: gościu, gołębiu, and some *-n stem nouns: kamieniu, płomieniu, and only dzień to this day keeps its original ending *-e but only in the fossilized phrase we dnie (i w nocy), as otherwise its w dniu.^[274][272] -u even spread to *-o stem nouns, seen in some rare medieval forms such as sadu, ludu, czasu, baranu, however it remained only in velar nouns: ptaku, Bogu, duchu, replacing the old 16th century ending -’e: ptace, Bodze, dusze.^[274] The word pan is an exception taking -u, likely as a result of leveling in the set phrase panu Bogu.^[274] All *-u stem nouns except syn and dom took -’e from *-o stems.^[274][272]

Polish inherits two main Proto-Slavic endings in the vocative singular, *-e and *-u, also keeping their original range of use with few exceptions.^[275] *-o stems take -e: rodzie, wozie, panie, lesie, chłopie; velar nouns originally took this ending, seen today in some fossilized forms like Boże człowiecze, Old Polish dusze Wojciesze, but even from the earliest times -u can be seen, and becomes the exceptionless ending during the Middle Polish era.^[275][276] Nouns ending in -ec also etymologically take -e, as -ec comes from Proto-Slavic *-ьkъ.^[275] *-u stems except synu and domu also take -e via leveling with *-o stems: wole, miodzie, czynie, and *-jo stems keep -u: mężu, przyjacielu, cesarzu.^[275][276] *-i and *-n stems level to -u instead of *-i: gościu, gołębiu, kamieniu, dniu.^[275][277] Nouns ending in -icz, -ic sometimes took -e in Middle Polish: królewicze, Pryjamicze, ślachcicze, panicze, as the result of assimilation to forms such as ojcze, starcze and later they took -u.^[275]

Many changes occurred in the nominative plural; all endings found in Proto-Slavic are present, and -a (akta, grunta) is introduced:^[278][279]

1. The ending -i (after hardened consonants -y) of old *-o stems remain exclusively or next to -owie through the Middle Ages, the 16th century, and the 17th century among animate nouns: sąsiedzi, chłopi, anieli, biskupi, brytani, lwi, psi, charci, prorocy, krucy, wilce, ptacy. Placenames such as Mydlnicy and Skotnicy also show this ending until the end of the 14th century. Then since the 18th century its use is limited to some personal nouns, in agreement with its modern usage. The process of separating -i (-y) or -owie as a special category of masculine personal nouns finished in the literary dialect only in the 19th century, and there is still much fluctuation in dialects. Personal nouns ending in -ch originally kept inherited -’y (<*-i) with a hardened sz: mnich, Włoch, mniszy, Włoszy, and during the 17th century begin to take -i based on sąsiedzi, chłopi, changing sz to ś: mnisi, Włosi, see § Proto-Slavic *ś̌ *ź̌. Nouns ending in -ec also take -y (<*-i) from the oldest times: mędrcy, kupcy, igrcy, pokarańcy, nowochrzczeńcy, zabilcy next to jeńcowie, łyścowie.
2. The ending *-ove (> -owie) of *-u stems from the earliest times sees extended use beyond its original range, which is one of the most characteristic features of development of the Polish nominative plural. Based on inherited synowie, wołowie, new nominative plural forms with -owie formed, namely old -o- stems of animate nouns: aniołowie, biskupowie, prorokowie, etc. These forms often appear alongside forms with -i||-y: biskupi||biskupowie, lwi||lwowie and are still retained in the 17th century. During the Old Polish era -owie also spreads to -jo stem animate nouns: mężowie, wróblowie. -owie could also be seen in hard and soft-stem inanimate nouns: językowie, ostatkowie, przebytkowie, chodowie, etc. After j and soft consonants -ewie could be seen § Proto-Slavic *e. The use of -owie in this regard reached its peak in the Old Polish era, and sporadically occurs in the 16th century and exceptionally even in the 17th century. Later the use of -owie was limited only to some personal nouns, but this norm is not fully established; personal names mostly use -owie; some words use both: autorzy, profesorzy, fololodzy and autorowie, profesorowie, filologowie.
3. The ending -i for old *-jo stems is not known in Polish outside a group of placenames ending in -icy: Biskupicy, Janowicy, Jarocicy, coming from family names, but already from the end of the 14th century these placenames occur in the form Kiskupice, Janowice, Jarocice, takken from the old accusative. Similarly szlachcicy, dziedzicy, which are retained longer, and rodzicy all the way to the 18th century.
4. *-e was using with the suffixes -ciel, -arz, -anin: przyjaciele, wielbiciele, lekarze, ślusarze, mieszczanie, Zagórzanie, and also old *-n stem nouns: kamienie, płomienie, rzemienie; place names only use -e: Koniarze, Psarze, but already in the 12th century there is a change of -arze > -ary based on placenames like Mydlniki, Skotniki, where the old nominative/accusative form from -o stems were set.
5. The ending -’e < *-ьje comes from *-i stem masculine nouns such as *pǫti and can be seen in czerwie, gołębie, goście, gwoździe, ludzie, łokcie, niedźwiedzie, ognie, paznokcie.
6. -e from *-e as well as *-ьje encompasses -jo stem personal nouns via leveling: krole, męże, bogacze, pieniacze.
7. -e in foreign -ans, -ens, -ons nouns such as romans, kredens, anons occurs as early as the 17th century alongside -y for hard-stem nouns; the motivation of its use here remains to be explained.
8. The vast majority of old *-o, *-jo, and *-u nouns use the accusative for the nominative, including:
   1. Forms ending in -y||-i (after k g) in terms of hard stem inanimate nouns from old *-o and *-u stem nouns: sad, płoty, obrazy. During the 17th century it displaced original -i or -owie in non-personal animate nouns, meaning forms like lwi, psi, krucy or lwowie, orłowie, krukowie disappear and in their place forms such as lwy, psy, orły, kruki appear, which is the current norm, meaning that outside of -y||-i (after k g) only hard-stem masculine personal nouns have -i or -owie. But even here -y can be seen from the 16th century: Longobardy, Szkoty, Bułgary, Serby, filozofy, doktory, biskupy, hetmany. -y (-i) sees further use in the literary language during the Enlightenment also for masculine personal nouns: syny, ministry, Cezary, Wenedy, Tatary, wnuki, naczelniki, Greki. These forms are result of an agelong tendency to use the accusative for the nominative. This tendency won out among inanimate nouns in the Middle Ages, and for non-personal animate nouns (psi > psy) in the 17th century. In the 18th century this also affected personal animate nouns, probably as the result of a tendency of this century to renew the literary language using the colloquial language. But these nominative-accusative forms had taken on a somewhat pejorative, dismissive, or disrespectful tone some time before this, and as Kopczyński states “...gdy czasem myślimy upodlić niecnotliwą osobę ludzką i do zwierząt przyrównać”, and Mrozisńki “Rzeczowniki męskie przybierają niekiedy w 1 przypadku zakończenie rodzaju nijakiego, co im nadaje pomysł upośledzenia: popy chłopy żydy”. By the fourth quarter of the 19th century these forms were neutral in meaning. To this day these forms can be used with nouns of contempt: łobuzy, obdartusy, łotry, snobby, łakomczuchy, wyrodki, nieuki, wyrostki, opryszki, or with a connotation of superiority with regard to immature beings: noworodki, chłopaki, dzieciaki. They can also be used as archaisms for stylistic purposes, seen used by Wyspiański: bohatery, syny, chłopy.
   2. -e < *-ě among soft stem or hardened stem non-personal nouns originally of the *-jo stem: konie, kraje, węże, noże, miecze, płacze, klucze. According to this non-personal nouns with the ending -ec < *-ьkъ take this: końce, kopce, proporce, zwojce, łańce. Contemporarily with the above it also starts to see use among personal -ec nouns with an innovated nominative-accusative ending -e: ojce, Połowce, which is supported by masculine -c(a) nouns (< *-ьka) which in the nominative took -e during the Old and Middle Polish period in agreeance with feminine nouns of the old *-ja stem. Only from the middle of the 18th century do they shed their old ending and take on modern -y (-i) modeled on personal -ec nouns such as starzec, strzelec: wybiercy podatków (from 1744).
9. Since the 15th century the ending -a for hard stems mostly in terms borrowed from Latin, German, and rarely for native terms. In recent decades these forms see less use under influence of normative recommendations, which deny them quality of diligence (które odmawiają im cechy staranności).

The vocative plural of all types merges with the nominative, and the history of both cases is shared.^[280]

The genitive plural endings -ów and -i||-y spread beyond their original use only now *-ъ and *-ь, i.e. -∅, are kept only vestigially.^[281] This was motivated by the loss of yers, which caused the genitive to become less distinct, which in turn caused a need for more distinct forms.^[282][281] -ów gains much use, originally coming *-u stems, and in the earliest times already dominates in *-o and *-jo stems as well as consonantal stems.^[282][281] Soft-stem nouns sometimes had -ew: krolew.^[282] Everywhere already in the pre-literate era the endings -ъ and -ь were displaced, and the vestigial -∅ ending is seen in:^[282][281]

1. In a few words words in the Old and Middle Polish era, e.g. woz (14th century), god (14th century), ząb (1466 and 1570), tysiąc (14th-16th centuries), dziej (15th century), włos (16th century), raz (16th century), sąsiad (until the 18th century), all lost, and in fossilized dotychczas.
2. In names of nations that later became names for areas: do Włoch, Prus, Mieniec, Czech, Węgier, Francuz, Turek which remain largely as archaisms.
3. In names of families or occupations that become placenames: do Krzeszowic, Racławic, Świątnik, Kobiernik, Mydlnik which last to this day.
4. From placenames coming from occupations forms with -arz which took *-ъ in the genitive: do Piekar, Koniar, Owczar, used today.
5. In -anin nouns, which in the plural declined like consonantal stems, so *-ъ in the genitive: dworzan, mieszczan, Rzymian, Zalesian, used today.
6. In the words przyjaciół, nieprzyjaciół created modeled on the genitive of consonantal stems, used today.
7. In reflexes of consonantal stems which resulted from leveling, seen in Old Polish: kamion, jelon, sążon, kmiot, łokiet.

-i||-y (after harded consonants) also sees expanded use, it is inherited in ludzi, gości, gwoździ, gołębki, łokci, paznokci, niedźwiedzi, ogni, czerwi, later it encompasses consonantal and *-jo stems via analogical leveling: koni, dni, kamieni, promieni, miesięcy, pieniędzy, męży, przyjacieli and then competes with -ów and expands among soft-stem nouns and sees more use in e.g. -arz, -aż, -erz, -eż, -orz nouns and -acz, -ocz nouns; in all words -ów is possible but less common.^[283] -ciel nouns except przyjaciel, nieprzyjaciel almost exclusively have -i, but -j nouns predominately have -ów, but pokoi, złodziei occur.^[284]

In the dative plural, the ending *-omъ > -om shows the most durability and expansive capability; its use remains uninterrupted in *-o stems and from the earliest times encompasses *-u stems and consonantal stems, it also displaces -em < *-emъ of *-jo stems and < *-ьmъ of -i stems, which can be seen until the beginning of the 17th century: koniem, gościem, rodzajem, rycerzem, stróżem, winowatcem, ludziem, dzieciem.^[284][285] Next to -om, -óm (-um) appears as the result of a tendency to raise vowels before nasal consonants: klasztorum, ku wrotum (15th century), like Szymun, słuń.^[284] For a short period, -am from feminine declensions appears: kapłanam, pagórkam, koniam, rycerzam, perhaps via nouns semantically masculine but declensionally feminine like wojewoda, sprawca; this begins in the 15th century, sees more use in the first half of the 16th century, after which its use shrinks and eventually falls out of use, seen as late as 17th century when the ending -om becomes the only ending.^[286][287]

The accusative plural changes depending on if the nouns is personal or non-personal, non-personal nouns keep the inherited state, namely:^[288][287]

1. The ending -y || -i (after k g) in old *-o and *-u stems;
2. The ending -e in old *-jo stems;
3. The ending -e in old *-n stems. Only dzień takes the ending -i from *-i stems. Old *-i stems take -e based on the accusative of *-jo stems.

The accusative of personal nouns has two eras - in the first era, ending in the 17th century, the accusative is the same as non-personal nouns: popy, dziewosłęby, anioły, bogi, grzeszniki, niewierniki, sprzeciwniki, Jerozolimczyki, syny, towarzysze, króle, więźnie, ochmistrze, kupce, and -ec and -c(a) nouns, modeled on *-jo stems, take the ending -e: jętce, zbieglce, winobrańce, oćce, powłoćce; -anin nouns keep old -y, and -ciel and -arz nouns keep old -e.^[289] The forms przyjacioły, nieprzyjacioły with -y occur rarely based on hard-stem nouns. The forms ludzi (from etymological *-i stems) and ludzie (based on *-jo stems) fluctuate in this era from the earliest times.^[290] In the second era, from the 17th century onward, the genitive is used as the accusative; in the 16th century there are relatively few examples of this, and only from the 17th century does the usage of genitive as accusative become more common, and the old accusative forms acquire an archaic character and become part of poetic language in the last decades of the 18th century with a particular stylistic function of marking something as upper register or pejorative.^[290][287]

The instrumental plural is divided into two eras, before the 16th century and from the 17th century, with the 16th century, especially its final decades, being a transitional period.^[290] In the Old Polish era the inherited state is generally kept - old *-o stems keep the ending -y||-i (after k g), including -ec nouns; perhaps these are the only trace of -i of *-jo stems, -ic nouns follow them, z dziedzicy, nad rodzicy, ślachcicy, -anin nouns used the old ending -y: miedzy pogany (14th century), z mieszczany, Walentiniany, dworzany (16th century).^[291][292] Old *-u, *-i, and *-n stems take -mi, which could be considered the etymological reflex of *-ъmi and *-ьmi, then -mi spreads in use in the Middle Ages, taking over old *-jo noun, it also appears in -ciel and -arz nouns (compare also nieprzyjacioły).^[293] -mi even spreads to *-o nouns, competing at the time with -y.^[293] The originally dual suffix -oma is rarely seen in the 16th century, and generally nouns with this form are accompanied by a numeral.^[293] Appearing in the 16th century, initially rarely and more commonly towards the end of the century is the ending -ami, originating from the feminine vocalic instrumental plural - within Rej’s works it constitutes 6% of hard-stem nouns and about 2% of soft-stem nouns; within Kochanowski’s works around 11% of poetry and 26% of prose.^[293] The grammarian Strojeński considered it a vulgarism.^[293] From the 16th century it becomes more common and limits -mi to just a soft-stem nouns: końmi, gośćmi, liśćmi, przyjaciółmi and displaces -y, which only remains in some fossilized phrases: dawnymi czasy, perhaps was used for archaicizing purposes.^[293][294] Dialects show a similar situation in that -ami is the predominate ending, -mi occurs less than in the literary standard, but -oma, which occurred only as an rarely of sixteen and seventeen-century texts, sees common use in northern Silesia and Masurian: palcóma, chopóma, bratóna, konióma, płotóma.^[293]

Polish sees three endings in the locative plural, -’ech from *-ěchъ and *-ьchъ in early texts and two innovations, it does not continue *-ich from *-ъxъ; -och, occurring in the preliterate period, and -ach, from the feminine vocalic locative, seen only sporadically in the Middle Ages, then is in clear use from the middle of the 16th century.^[295][294] Old *-o, *-u, and *-i stems originally predominately used -’ech, is used less from the middle of the 16th century, rarely in the second half of the 18th century, and is now is limited to some archaic placenames: Prusiech, Węgrzech, Włoszech as the names of countries.^[295][294] -’ech replaces Proto-Slavic *-ich via leveling.^[295][294] -och, a regional ending of Lesser Poland, Silesia, and Red Ruthenia (whereas in Greater Poland, Kujawy, Łęczyca, and Masovia the ending -ech is seen), can be seen from the oldests texts for various stems, especially frequently in soft and hard-stem nouns ending in -k, -g, -ch displacing old -ech, then in the second half of the 16th century -och disappears.^[296][294] The origin of -och is likely that it arose from leveling with the nominative, genitive, and dative endings -owie, -ow, and -om, because they are used in contrast with the soft-stem endings -ewie, -ew, -em, and -ech causes softening of the final consonant, so as a result there is the series mężewie, mężew, męzem, męzech, and the opposing bogowie, bogow, bogom, bodzech; the form bodzech stands in contrast to the stem bog-, which results in the ending -och, and when soft-stem nouns leveled to hard stem nouns (-ewie > -owie, -ew > -ow, -em > -om), then -och also replaced -ech.^[296][294] Another claim is that -och arose from the leveling itself, so mężewie mężowie, mężew > mężow, mężem > mężom, mężech > mężoch, from which -och also appears in hard-stems.^[296] Nouns ending in k, g, ch keep -och the longest likely in order to avoid changing the stem (k : c, g : dz, ch : sz).^[296] Grappin in the self-published work Historie de la flexion du nom in polonais, Wrocław, 1955 derives -och from Proto-Slavic *-oxъ of *-u stems, considering it an old, native and regional morphological element of Lesser Polish which spread in the 16th century.^[296] The withdrawal of -och is explained by phonetic causes, namely that when in the feminine locative the form -ach (with clear a) became established, it started to replace -och, which speakers assumed to be from -ách, which then was displaced by -ach.^[296] The ending -ach appears rarely already in the Middle Ages, and sees the most use in the 16th century, especially in the second half.^[296] Characteristically of their region of origin, Rej, Orzechowski, and Bielski often use -och, and Kochanowski never, preferring new -ach, and thanks to his high regard greatly influenced the rising popularity of -ach, which from the 17th century sees dominant use, and ultimately exclusiveness.^[297] In 15th century texts this ending is sometimes written with slanted ⟨á⟩. Perhaps this came from Lesser Poland where the ending -och was used transitionally, and -ách was a compromise between -ach and -och, or potentially also taken from the feminine locative under where at the earliest it could have developed under influence of dative -ám.^[297] The spread of -ach in the masculine has various theories, but it is certainly partially due to the tendency to level locative plural forms regardless of gender, as how in pronominal and adjectival declensions there is only one locative form since Proto-Slavic for all three genders; -ach might have had an advantage as it dominated in feminine declension, whereas -ech and -och competed with each other in masculine and neuter nouns and weakened as a result.^[297] Nouns semantically masculine but morphologically feminine such as wojewoda, sprawca could have helped spread -ach to masculine nouns.^[297] Another important factor is the fact that of the two feminine endings, -ach and -ách, -ách approached masculine -och forms, and in the 16th century -ách submits to -ach and -och could have been pulled into this process.^[297]

Feminine vocalic nouns come from Proto-Slavic *-a-, *-ja- stem nouns.^[298]

Old -a sems, which in Polish are hard-stem nouns, take the ending -a in the nominative singular, which in the historic era was short, later clear, and continues now in this state uninterrupted; also in this group are nouns of masculine semantics, e.g. sługa, wojewoda, junoszka, ewanielista, patriarcha.^[299][298] Old -ja stem, which in Polish are soft-stem nouns, also take -a, but here it can be either long or short, later clear a or slanted á, e.g. bania, chwila, but burzá, wolá. This also includes semantically masculine nouns, e.g cieśla, and loanwords, e.g. flasza, sala.^[299] If -a is preceded by -j-, then it is short/clear: kaznodzieja, knieja, nadzieja, szyja, zbroja, żmija; one exception is kopijá based on loanwords, which have slanted á: ambicyjá, Assyryjá, Grecyjá, historyjá, etc., similarly Chananeá, Medeá (Kochanowski).^[299] Nouns with the suffixes -ic(a) < *ik(a) and -c(a) < *-ьka act in accordance with their original hard-stem nature and take clear -a, including masculine nouns.^[299] Over time the loss of -á in literary Polish levels in the nominative of these different types since 18th century and clear -a spreads.^[299] -i inherited from Proto-Slavic is kept in a few nouns: bogini, łani, pani, but sometimes undergo leveling to -á: łaniá, boginiá, gospodyniá, ksieniá (16th-17th century).^[299]

The genitive singular ending -y || -i (after k g) in old *-a stems remains uninterrupted: lichoty, prawoty, prawdy, wody; matki, wargi.^[300][298] -e < *-ě after soft or hardened consonants dominates in the Old Polish era and the 16th century, and in 17th century it slowly yields, and can be seein in the 18th century as an archaism, e.g. nadzieje, ziemie.^[300] In the earliest texts soft-stem nouns whose nominative end in -á take a new ending -éj: woléj, roléj, strożéj, duszéj, żądzéj, Babilonijéj, Idumejéj, and also masculine sędziéj, collective braciéj; and also nominatives ending in -i: paniéj, posełkiniéj (14th century); this ending continues to the end of the 17th century: pochodniéj, studniéj, wieczerzéj, władzéj, bestyjéj, Persyjéj, Grecyjéj (Rej); łaźniéj mszéj, sukniéj, braciéj (Kochanowski); rożéj, kurtuazyjéj, szyjéj, boginiéj, koronacyjéj (17th century), and over the course of the 17th century it falls out of use, attested by the grammarian Woyna (1690) proscribing oracyi, opinii, Maryi.^[300] -éj is a Polish innovation based on the feminine genitivedeclension of compound adjectives leveled to the dative: dobrá pracá therefore dobréj pracéj; however it could also be a phonetic development of -é > -éj, and in practice -éj could have been equivalent to -y||-’i.^[300] Neither of the soft-stem endings -e and -ej were retained.^[300] -i||-y (after hardened consonants) occurs as the result of leveling with the genitive singular of soft-stem feminine nouns, probably also supported somewhat by -y of the genitive of feminine hard-stem nouns; -i||-y can be seen sporadically in the 14th and 15th century prawicy, nędzy, nadziei, rarely in the 16th century: oblubienicy, pieczy (Rej); bogini, łożnicy, stajni (Kochanowski).^[300] This new ending spreads in the 17th century and from the 18th century is used exclusively, but -e is kept in dialects, in these dialects -i is also not uncommon, especially for -á nouns: studniá, rolá - studni, roli, but ziemia świéca - ziemie, świéce.^[301] It is possible that -i arose from old -é, but in ziemie, -e was not slanted and couldn’t phonetically develop into -i.^[302][303]

The dative singular ending *-ě > -e remains unchanged: sierocie, bitwie, chwale, prośbie, dziedzinie, opoce, drodze (14th century); -i (-y after functionally soft consonants) within soft-stem -a and -i nouns is nearly exclusive before the 16th century, dominates in the 17th century, and from the 18th century again exclusive: ziemi, prawicy, Babiloniji (14th century); tajemnicy, stolicy, zbroi, Troi (16th century).^[302][304] Like in the genitive, the innovation -ej occurs in the dative in based on the feminine declension of compound adjectives; it was used for soft-stem nouns ending in -á and -i from the Middle Ages to the end of the 17th century: braciej (14th century.), bestyjej, Lukrecyjej, pracej, wolej, gospodyniej (16th century), żądzej, paniej, ksieniej, Kornelijej (17th century).^[302][304]

The development of the feminine accusative singular is associated with the development of the feminine nominative singular.^[305] Polish hard-stem nouns (in the nominative -a) take -ę: naukę, rękę, (14th century), also -ica, -ca nouns and native -ja nouns: nadzieję, błyskawicę, potwarcę (15th century), studnicę, stolicę, owcę (16th century).^[305] Soft-stem nouns before the 17th century have two endings depending on the nominative ending: -a nouns take -ę, and -á and -i nouns take -ą: duszę, ziemię, jutrznię (15th century), but bracią, sędzią, wolą, pieczą, puszczą, (15th century), pracą, tłuszczą, wolą, wieczerzą, (Rej), leliją, lutnią, rolą, władzą, panią (Kochanowski), but already in the Middle Ages a certain fluctuation can be seen with regard to -ę and -ą as the result of mutual interference of the two accusative endings, e.g. jutrznię, nutrznią, karmię, karmią.^[305][306] The process of leveling reaches its peak in the 17th century, when -ę sees increasing use.^[305] This is explained by the influence of -ę forms supported by the loss of -á, which results in a lack of support for another accusative ending which was supported by a different nominative ending.^[305][307] At the end of the 17th century the grammarian Woyna considered -ą as more elegant and more careful in a nouns ending in -á, all of which borrowed in the form of -ijá, -yjá, and also -i nouns, this is evidence that a certain sense of the old norm was kept since there was a need to support it with grammatical correctness formulations.^[305] -ę spreads in the 18th century, and the most resistant nouns (namely -ija, -yja nouns) also begin to take -ę, but traditional forms using -ą for -á nouns are kept even in the 19th century, and Małecki in his grammar from 1879 recommends -ą for all nouns ending in -ija, -yja, and -i (panią, Zofiją, okazyą), and for a few other soft-stem nouns he claims that any fluctuation is regionally determined: “nie pozostaje przeto, jak na czas bliżej nieokreślony pozostawić to jescze tradycji poszczególnych okolic” (there is no other option but to leave this to the traditions of individual regions for an indefinite period of time).^[305] Kryński in his grammar from 1907 says “ostateczny wynik tej walki, zakończonej za dni naszych, zapewnił stanowczą przewagę formom biernika na -ę nad formami liczebnie mniejszymi na -ą” (the final result of this struggle, which ended in our times, ensured a decisive advantage for the accusative forms in -ę over the numerically smaller forms in -ą).^[305] panią is theonly exception, preserved thanks to its use in fossilized terms of respect.^[305][307] Dialects differentiating -a and -á generally also keep -ą in the accusative where -á is in the nominative.^[305]

In the Polish instrumental singular there is only -ą, which is the reflex of the old long nasal vowel which resulted from contractions (*-ojǫ, *-ejǫ > ǫ): drogą, prawdą, siekirą, błyskawicą, ziemią, bracią, duszą (14th century).^[308][307] In most western and central dialects -ą changed to -ám, pronounced most often as -om. This pronunciation occurs even in dialects where -ǫ is kept, meaning that -om probably did not arise from sound changes alone, but also based on -em from the masculine and neuter instrumental.^[308]

Polish -e continues *-ě in the locative singular hard-stems uninterrupted: ręce, drodze, pysze, robocie, krzywdzie, głowie, chwale, śmierze, glinie.^[308][307] Soft-stem nouns continue primarily *-i, in the form of -i or changes to -y after hardened consonants: nadziei, ziemi, woli, jutrzni, okolicy, dziewicy, świecy (14th century), sbroi, tajemnicy, trojcy, rozkoszy, puszczy, wieczerzy, bóżnicy, wieży (16th century).^[308][307] Alongside these two inherited and dominating endings the innovation -éj appears, also seen in the genitive and dative, based the feminine compound adjectival declension - seen sporadically in the Middle Ages: ziemiej, wolej, and seen somewhat more often in the 16th century: Arabijej, Azyjej, ewangelijej, historyjej, nestyjej, szarańczej, wolej (Rej), Asyryjéj, Frygijéj, gospodyniéj, kancelaryjéj, pracéj, roléj, Trójej (Kochanowcki).^[309][307] Nouns that take this form are soft-stem -á nouns, especially loanwords ending in -ijá, -yjá; this form can be seen as late as in the 17th century: na lutniej, w Hiszpanijej, w Persyjej, armonijej, komedyjej.^[310]

Hard-stem nouns in the vocative singular continue *-o, coro, sławo, chawło (14th century); soft-stem nouns ending in -a keep original -e in the Middle Ages: dusze, dziewice, służebnice, especially gospodze, and very exceptionally are modeled on consonantal feminine nouns: lutni, ziemi (16th century), but as early as the 15th century begin to take -o based on hard-stems: duszo, rożo, nadziejo (15th century) and this leveling tendency finishes in the 16th century.^[310][311] Another tendency is the usage of the nominative for the vocative - it can be seen from the earliest times in hard and soft stems: Bogurodzica, dziewica, Maryja, Zbawicielu nasz nadzieja, chwalcie ji niebo i ziemia, powyszona bądź prawica, wroć sie dusza, wesel się gora, bądź ręka, płyń mołwa, wnidzi prośba (14th century); ustąp melankolijá, piękna Zofijá (16th century).^[310][311] The nominative for the vocative is everywhere for -i nouns: pani, gospodyni, bogini, but sometimes -o appears via analogy: panio (1772).^[310][311] Diminutives with a soft consonantal stem or -l from the 16th century take -u: Kasiu, matulu - this is an innovation based on the vocative of corresponding masculine nouns: Józiu, synusiu, Karolu, but exceptionally Baśku, Zośku also occur.^[310][311]

Polish continues the nominative and accusative plural endings with regard to hard-stem: -y||-i (after k g), soft-stems -e: nominative: dziewki, rzeki, drogi, wargi, wody, mowy, gory, strzały, burze, ziemie, owce, błyskawice, dziewice (14th century); accusative: uliczki, roboty, radyy, ryby, ćmy, obietnice, owce, tłuszcze, role, studnie (14th century). ^[312][311]

The genitive plural in Polish is marked -∅, inherited after the loss of yers: rąk, ksiąg, lichot, prawd, głow, mołw, gor, ran, łez; owiec, tajemnic, dusz, ziem (14th century), -nia nouns (< *-nja) take final -ń o r -n: sukień, sukien, studzień, studzien; forms with hard n appear to be secondary, but see use and by the end of the 17th century they become the only form except in north-east areas, where -ń are still in use in the 19th century, e.g by. Adam Mickiewicz and Juliusz Słowacki.^[312][313] -a nouns (< *-ьja) such as świnia, lędźwia took the ending -i, which is the inherited form of *-ьjь, and from here the ending spreads among nouns ending in -nia preceded by a consonant: luśni, wiśni, kłótni, pochodni, and other soft-stem nouns: duszy, ziemi, żądzy, wieży, chwili (15th-18th centuries), this ending continues even into the New Polish era, and after a period of fluctuation in the 19th century it establishes itself among -nia nouns preceded by a consonant: zbrodnia, kłótnia, kuźnia, dźwignia; among -arnia, -ernia, -alnia, -elnia, and -ownia nouns: kawiarni, cukierni, kopalni, czytelni, warowni, and with others: mszy, rękojmi, wieczerzy; some nouns show alternate forms: kuchni||kuchen, sukni||sukien, studni||studzien.^[314][313] Based on spellings of -ija||-yja nouns such as disputaciy, bestiy (Rej); konstytuciy (Skarga), opiniy (Woyna), it seems they were pronounced as dysputacyj, bestyj, konstytucyj, opinij, that is with -∅, but most likely there was a possible alternation with -i, which is where the fluctuations in pronunciation come from.^[315] Signs of this uncertainty can be found in 19th century grammars: Henryk Suchecki recommends history, lilij, while Adam Kryński oppines: “to nie jest słuszne, gdyż sprzeciwia się powszechnemu wymawianiu” (this is not correct because it contradicts the common pronunciation), and the want to distinguish the genitive singular lilii and genitive plural lilij may not justify such innovations “wobec faktu, że język żywy zgoła rożnicy tej nie wyrobili” (in view of the fact that the living language has not made this difference at all); Antoni Małecki supports the spelling prowincyj, but says that “wymawianie rzeczywiste nasze jednak nie jest ani w liczbie mnogiej herezyj linij, ale jedynie tych herezyi i linii” (our actual pronunciation, however, is neither in the plural herezyj linij, but only of these herezyi i linii).^[315] The modern norm takes -i and is written: herezji, lilii, with only a few exceptions, where the context doesn’t rule out ambiguity of singular vs plural forms.^[315] Native -ja words are notable, as some take -∅: szyj, żmij, and others -i: nadziei, zbroi.^[315][313] In the 17th and especially 18th century the ending -ów taken from the masculine can be see: różów, zbrodniów, boginiów, and in the second half of the 18th century it occurs very often among some of the notable authors, e.g. Krasicki, especially in words adapted with -yja, -ija: komisyjów, sesyjów, and of course also myszów, koniów; this was removed from the literary standard probably under influence of grammarians such as Kopczyński.^[315] -ów in the genitive plural of feminine nouns can also be seen in dialects.

Medieval Polish originally inherits the dative plural ending -am and it replaces *-em < *-ьmъ: kobyłkam, ścieżkam, drogam, nogam, wargam, prawotam, stdzam, duszam, studniam (14th and 15th centuries), and the masculine ending -om can be seen exceptionally: przedawcom, przestępcom (15th century).^[316][313] -om sees an clear increase in use in the 16th century, initially among masculine nouns: drapieżcom, starostom, kaznodziejom, patryjarchom, but also increasingly frequently among feminine nouns, when in the middle of the century a major shift occurs, as Rej uses old -am almost exclusively for masculine nouns; Kochanowski only exceptionally uses -ám, and normally uses -óm: sirotóm, owcóm, poetóm, panióm.^[317][313] The 16th century is a period where -am slowly falls out of use and -om spreads, spreading more from the second half of the century; this process can be explained by the tendency for inter-declension leveling via the spread of formally masculine nouns taking feminine endings.^[317] It is possible that -om as a phonetic development of -am, as in texts two variants occur, -am and -ám, and Rej has consistently -ám, and a raised under influence of m and as such entirely or almost entirely merged with -óm, leaving an orthographical question whether one should write -ám or -óm.^[317] Some dialects prefer -ám, generally those where -ám is clearly distinguished: ludziám, psám, krowám, or even ludziam, kóniam, krowam, e.g. in Masovia; in Greater Poland however this ending is realized in agreement with áN clusters, and not oN: /wɔɛɡoɘn/ (ogon), /koɘŋum/, ludzium, zbiyrum - this would suggest that the feminine dative plural ending -am became preferred, at least in some dialects, as did the feminine ending of the nominative and locative.^[315]

The instrumental plural ending is inherited into Polish as -ami: nogami, wargami, sirotami, wodami, głowami, mołwami, łzami, panoszami, obietnicami (14th and 15th centuries).^[317][318] From the oldest times to modern the sporadic use of -y (-i after k g) via analogy to masculine nouns can be seen: pod wargi ich, z cudzołożcy (14th and 15th centuries), szalonemi wełny, z poety, miedzy zbójcy (16th century) and its use in the 18th and 19th centuries usually has a stylistic function of a non-quotidian nature: trwożysz groźby (Krasicki), pod twojemi nogi fale morza (Krasiński), złotymi farby, gwiazdy niebieskimi, pod ciemnemi jodły, łzy brylantowemi, ze srebrnemi rzęsy (Słowacki), przysięgi straszliwem (Wyspiański).^[319][318] The ending -oma appears in some dialects: nogoma, krowoma, see § Declension of masculine nouns.^[320]

The inherited locative plural ending -ach dominates in Polish, and could have also been -ách, as there are two uncertain fluctuating variants ni the 16th century, Bielski has -ách, Rej -ach or -ách (-ách occurring ten times more frequently), Kochanowski and Skarga -ach, and even in two different publications of the same work alternations occur between -ach||-ách indicating a dialectal nature determined by the typesetter or proofreader, as the marking of slanting depended on them, as 16th century manuscripts usually don’t distinguish it.^[320][321] It is indicative that Kochanowski recommends it only to avoid ambiguity in his orthographical recommendations: żádny (żaden) but żadny (brzydki).^[320] It seems that -ách was of Lesser Polish origin and arose under influence of the dative plural ending -ám; in the 16th century -ách gives way: naukach, drogach, wodach, gorach, ćmach, nizinach, pokusach, tajnicach, wieżach, ziemiach, jutrzniach (14th and 15th centuries), arfach bitwach, ewanielistach, frankach, baniach, bałwochwalcach, zbrojach (16th centuries).^[320] Aside from this ending, which ended up being so expansive it also took over masculine, vocalic feminine, and neuter declension, -ech and -och be rarely be seen via analogy: robociech, szaciech, głowiech (15th and 16th centuries); rękoch, krainoch, sędzioch (16th century), -ech occurs especially after d t s, commented on by Mesgnien.^[320][321]

Feminine consonantal nouns come from Proto-Slavic *-i-, *-u-, and *-r- stem nouns.^[298]

In the nominative and accusative singular Polish has -∅ from the loss of *ьː kość < *kostь: śmierć, żółć, bystrość, ciemność, mądrość, świat, łość, bojaźń, łeż (14th and 15th centuries). ^[322][298] A number of vocalic feminine nouns enter this declension: głębiá, karmiá, łodziá, podróżá, pogoniá, toniá, pieczeniá, tarczá, kolejá which lost -á keeping the original state of the stem: głąb́, karḿ, łódź; a few Proto-Slavic *ū stems: krew, cerkiew, chorągiew, rzodkiew, and originaly -y is kept in some medieval forms: kry, jętry, świekry, cerki, Pełty, Nary, and others, and these in the oldest era show accusative for the nominative: *cь̀rkъvь < cerkiew; and Proto-Slavic *-r stems: mać (simplified from the old nominative *maci < *mati) and macierz (continuing the accusative as nominative *màterь).^[322][298]

Polish inherits genitive singular Proto-Slavic *i stems as -i or -y after hardened consonants: boleści, czystości, głębokości, czci, śmierci, myśli, mocy, rzeczy, rozkoszy, potwarzy (14th and 15th centuries); in the Middle Polish era the ending -ej can be seen from analogy to vocalic feminine nouns in the genitive: przyjaźniej, kradzieżej, or -e: gęśle, czeladzie, słodycze.^[322][298] Old *-u and *-r stems initially keep inherited -e: cerekwie, krwie, macierze, which can be seen sporadically until the 17th century, but already in the oldest texts forms such as cerekwi, krwi, macierzy with -i (-y) based on vocalic feminine nouns occur.^[322][298]

Polish continues -i, -y in the dative singular after hardened consonants from Proto-Slavic *-i from *-i stems: pamięci, postaci, mąrości, sprawiedliwości, czci, mocy, nocy (14th and 15th centuries).^[322][304]

The accusative singular ending -∅ here after the loss of yers.^[304]

Polish continues -ą in the instrumental singular from Proto-Slavic *-ьjǫ from *-i stems: czcią, gorącością, śmiercią, siedmią, mocą, rozkoszą (14th and 15th centuries).^[322][307]

Polish continues -i, -y in the locative singular after hardened consonants from Proto-Slavic *-i from *-i stems: w bladości, we czci, w nienawiści, w szyrokości, w piąci, w bojaźni, w mocy, w pomocy (14th and 15th centuries).^[323][307] The ending -ej can be seen exceptionally in the 16th century as an analogy innovation based vocalic feminine declension: o niewdzięcznościej, w otchłaniej.^[323][307]

Polish continues -i, -y in the vocative singular after hardened consonants from Proto-Slavic *-i from *-i stems: gęśli, żałości, łodzi, mocy, nocy (14th-16th centuries).^[323][311]

In the nominative and vocative plural Polish takes the Proto-Slavic ending *-i, secondarily from the accusative based on feminine *-a, *-ja nouns, where also nomiminative-accusative forms were established: boleści, otchłani, złości, myśli, kaźni, postaci, mocy, rzeczy (14th and 15th centuries); baśni gałęzi, latorośli, pieczęci, przepaści, wsi, twarzy, rozkoszy, nocy, obręczy (16th century); in the 16th century some leveling with vocalic feminine nouns begins to occur and -e sees use, but it only appears in a small number of examples. Only in the second half of the 18th century does the ending -e spread, and a number of nouns take equally -i and -e.^[323][311]} This continues in the 19th century, and the usage of -e intensifies despite resistance from grammarians, such as Małecki, who opine that “zakończenie -i jest jedynym właściwym, -e, przeciwnie, późniejszym wtrętem, który by rad pierwotne -i wyprzeć do reszty” (the ending -i is the only proper one, -e, on the contrary, is a later insertion that would like to supplant the original -i completely).^[323][311] As a result the two endings are still in competition with one another, some have only -i: brwi, maści, napaści, kości, ości, przykrości, własności, or some only -e: łodzie, cerkwie, chorągwie, rozkosze, twarze, and some both, where -e seems more common: garście||garści, moce||mocy, noce||nocy, postacie||postaci, wsie||wsi.^[323]

Polish continues -i, -y after hardened consonants in the genitive plural from Proto-Slavic *-ьjь from *-i stems: boleści, złości, myśli, kaźni, rzeczy (14th and 15th centuries).^[323] From the 15th-18th centuries -∅ occurs based on vocalic feminine nouns: twarz, mysz, rozpacz, podróż.^[324] Another innovation is -ów taken from masculine declension: pieśniów, wsiów, myszów, which appears more often in the second half of the 18th century and first half of the 19th century, also seen in dialects.^[324]

Historic Polish did not inherit the dative plural *-em, which would be the inherited form of *-ьmъ of *-i stems, and -am from vocalic feminine nouns replaces it: czeluściam, kościam, złościam, postaciam, gardzielam, myślam, kaźniam, rzeczam (14th and 15th centuries); głębokościam, kościam, złościam, rzeczam (16th century), and in the 17th century this ending falls out of use; in its place -om spreads taken from masculine and neuter declension, applied here via its leveling found in feminine vocalic declension.^[324] The first examples of this are seen in the Middle Ages: kościom, kaźniom, skroniom (14th and 15th centuries); and in the 16th century, especially the second half (Kochanowski, Skarga), and -om sees more use than old -am and in the next century starts to dominate.^[324]

Polish continues -i, -y after hardened consonants in the accusative plural from Proto-Slavic *-i from *-i stems: boleści, czeluści, kości, lści, próżności, złości, przyjaźni, niemocy, rzeczy (14th century); dźwirzy, baśni (15th century); brzydliwości, pieczęci, goleni, twarzy, roskoszy, niemocy (16th centuries); skroni, łodzi, potwarzy (17th century), and many nouns keep this ending now, but already in the earliest texts the ending -e can be seen: koście, moce, łże, łodzie, dźwirze, karmie, czystoście (14th and 15th centuries); in the Middle and New Polish era these nouns increase, and in modern Polish it appears more and more often as the only ending: kadzie, kiście, łodzie, rozkosze, twarze, or equal: moce, noce, pieczęcie, pięście, postacie, przepaście, wsie. -e came into use under influence of feminine soft-stem -a nouns in the accusative plural.^[324]

The instrumental plural ending -mi < *-ьmi is seen relatively rarely: gęślmi, żyrdźmi, dźwirzmi (14th and 15th centuries); pieczeńmi, pierśmi (17th centuries); it is kept now in kośćmi, nićmi, and -ami, taken from vocalic feminine declension spreads from the earliest times: żerdziami, kościami, ganiebnościami, złościami (14th and 15th centuries).^[325][318]

The ending -ech < *-ьchъ is seen relatively rarely in the Old Polish era: postaciech, światłościech, kaźniech, głębokościech, roskoszech and -ach begins to take its place already from the earliest period, taken from vocalic feminine declension; this process develops in the 14th and 15th centuries, there are 7 examples of -ech in 6 nouns in Psałterz floriański, 3 in 2 nouns in Psałterz puławski, and from the 16th century -ach becomes the only ending; -och sees rare use from the 14th-16th centuries taken from masculine declension: postacioch, głębokościoch, gęśloch, kaźnioch, rzeczoch, drzwioch.^[326][321]

Neuter nouns come from Proto-Slavic *-o-, *-jo-, *-n-, *-t-, and *-s- stem nouns.^[321]

Polish inherits the nominative, accusative, and vocative singular endings -o from Proto-Slavic *-o and *-s stems, -e from *-jo stems, and *-ę from *-n and *-nt stems: siano, mleko, mięso, żelazo, sidło, dziedzictwo, księstwo, świadectwo; oblicze, sierce, słuńce, zboże, morze, czekmanie, pierze, przyjęcie, miłosirdzie; brzemię, ramię, siemię, szczenię; ciało, słowo, ucho.^[327][328] Nouns originally ending in *-ьje sometimes take slanted -é: trzꟁsenyee (Sankt Florian Psalter), pienijé, radowanié, weselé, widzenié, śniadanié, zapalenié, dręczenié, przykrycié, skrzytanié, poruszęnié, zdrowié.^[329][330] In Old Polish there were two types of neuter endings, one with clear -e < *-e and one with slanted é < *-ьje, which can also be seen in the Middle Polish era in Kochanowski’s works, and this difference was later lost due to leveling; sometimes -e leveled to -é for example in Knapski’s and Kopczyński’s works, and sometimes -é levels to -e, seen in the standard; slanted é sometimes raised to -i in nouns, seen at the beginning of the 17th century: ubieżeni, uciekani, miłosiedzi, otpoczywani, rozchwatani, wzdani, ruszeni, pici, wyobrażeni, zboży - this can still be seen in Silesian and northern Greater Polish, and in Silesia, the raised vowel is kept only after soft and hardened consonants; in Southern Silesia it merged with i and elsewhere as -y, and in northern Greater Poland all instances of final -e, even those which never had slanting, raise to -y, which stands in opposition to the rest of Polish, which generalized clear -e, and here slanted -é generalized.^[329][330] The word-forming suffix -e from *-ьje in nouns like chorąże as well as titles formed with pod- can be seen with slanting: podkomorzé, podczaszé, podłowczé, podstolé; the raising of -é as well as association to related adjectival nouns like łowczy, woźny, wojski caused nouns like podkomorze, podstole in the 16th century to take -i||-y in the nominative and decline like masculine adjectives.^[329][330] In some dialects words like cielę, siemię underwent significant weakening; northern dialects made and spread words like cielák, prosiák, zwierzák, and elsewhere words like siemię underwent some modifications, sometimes forms such as (ten) ramień, ramienia or sometimes ramie, ramia, imie, imia, siemie, siemia; but forms like cielę, siemię are kept alongside in many dialects.^[329]

Polish generalized the genitive singular ending -a; the only trace of *-e of *-n, *-nt stems is the softening of the final stem consonant: imienia, cielęcia, as they would have to be softened by a front consonant before -a from *-o and -jo stems spread.^[331][330] The Polish genitive of nouns like ciało, słowo, niebo, koło, oko, ucho does not continue extending *-s-, but adds the ending to the reduced form taken from the nominative: ciał-, słow-, nieb-, ok-, etc., therefore large-scale leveling occurred in the genitive: z błota, od sioła, z żyta, z miasta, od sidła, pątnictwa, obrzęda, sierca, słuńca, do morza; imienia, ramienia, cielęcia, skocięcia, nieba, oka (14th and 15th centuries).^[331][330] Only nouns suffixed with *-ьj- such as bicia, picia, miłosierdzia, pokolenia, weźrzenia, wiesiela had long/slanted -á from contraction of final *-ьj-a; this ending -á can be seen ni all dialects: śniádaniá, szczęśćiá, weselá, kázaniá, however slanting was not kept in the nominative-accusative plural, even though contraction of *-ьj-a > -á occurred in this position.^[331][330]

In the dative singular -u is used everywhere, taken from the original use in *-o and *-jo nouns leveling *-i in *-n and *-nt nouns (*imeni *telęti), but the softening effect can still be seen in the final stem consonants: imieniu, cielęciu, as -u was applied secondarily; a small number of exceptions occur, dziecięci, książęci; individual exceptions starting from the Middle Ages to now with -owi can be seen: imieniowi, południowi.^[332][333]

The instrumental ending -em is inherited *-jo and consonantal stem nouns continuing *-emь and *-ьmь: siercem, słuńcem, plemieniem, imieniem, dziecięciem, panięciem (14th-16th centuries).^[332][333] Polish has an additional innovated ending -im occurring in nouns suffixed with *-ьj- resulting from contraction of *-ьjemь: przyścim, wiesielim, badanim, rządzenim, pienim, wyobrażenim, miłosiedzim, cirnim (14th and 15th centuries), naczynim, malowanim, spustoszenim, pożegnanim, staranim, zdrowim, drżenim, pierzym, imienim (from imienie, “property”), rozumienim, obliczym (16th century); -im is used often until the end of the 15th century with very few exceptions, but over the course of the 16th century under influence of other neuter nouns that take -em it starts to be confused, initially rarely, and around the midpoint of the century more and more often, and towards the end it gives way as an archaism.^[334][333] The Proto-Slavic *-o stem ending *-omь was not kept in Polish, but based on masculine *-u stems (*-ъмь) as well as perhaps the rest of neuter nouns already in the prehistoric era they took -em: łęczyskiem, miastem, dziedzictwem, pogaństwem, źrzebrem, and secondarily (wtórny) złotem, gardłem, kadzidłem, winem (14th and 15th centuries), -em was added relatively late based on the fact that final consonants are not softened by it.^[335][333]

Proto-Slavic *-o stem nouns keep their inherited locative singular ending, that is -e < *-ě: mieście, prawie, mnożstwie, pogaństwie, gardle, jezierze, łonie, mięsie, żelezie; the only change that occurs is within nouns whose stems end in -k, -g, -ch, as until the end of the 15th century they take -e: w mlece, w jebłce, na rusze (15th century), this change was possibly spurred by a want to avoid stem alternations of k : c, g : dz, ch : sz, and sporadically already in the 15th century and commonly in the 16th century -u from the locative of neuter -e and -ę nouns is applied.^[335][336] Old *-jo stems lose original *-i and gain innovated -u, taken from masculine declension: siercu, słońcu, morzu, polu (14th and 15th centuries); etymological forms are seen exceptionally: morzy, siercy, słuńcy (14th and 15th centuries).^[335][333] -i can be seen somewhat more often in nouns suffixed with *-ьj-, therefore where -u comes from *-ьji: kazani, odpuszczeni, narodzeni, pisani, czynieni, umęczeni, wabieni, wiesieli, zdrowi (14th and 15th centuries), but innovated -u is used much more often: biciu, skryciu, miłosierdziu, przewodziu, wiesielu, myśleniu, obeźrzeniu, pokoleniu (14th and 15th centuries), and *-n and *-nt stems also take exclusively -u via analogy: imieniu, ramieniu, dziecięciu, cielęciu (15th centuries), and traces of original *-e can be seen in the softening of the final consonant.^[337][338]

Within the nominative, accusative, and vocative plural, nouns inherit -a Proto-Slavic: dzieła, jeziora, biodra, lata, świadectwa, sierca, miłosierdzia, założenia, ramiona, zwierzęta (14th and 15th centuries); titles of officials such as podstole, chorąże also have -a until the 16th century: podkomorza byli (15th century), later -owie or -e: chorążowie, podczasze, and in the accusative -e: ma pod sobą podchorąże.^[337][336]

The most frequent ending in the genitive plural is -∅, inherited after the loss of *-ъ for *-o, *-n, and *-nt stems as well as -ь of *-jo stems: lat, ust, stad, bogactw, skrzydł, zioł, sierc, miłosierdź, kazań, brzemion, imion, cieląt, książąt; some soft-stem nouns take -i||-y: podziemi, bezkrólewi, bezprawi, wezgłowi, poddaszy, obliczy, wybrzeży, podnóży, podwórzy, rozdroży, zaciszy - these are prepositional constructions built modeled on *-ьj- nouns such as danie, wiesiele, miłosierdzie; some deviations occur, some without a preposition: pustkowi, nozdrzy; some compounds: półroczy, and prefixed deverbals: narzeczy; on the other hand prefixed nouns with -∅ occur: zbóż, pokoleń, nieszczęść, or both endings: przedmieść||przedmieści, przysłów||przysłowi.^[337][336] The origin of these forms is uncertain and they are not seen in the Old Polish era; they could possibly derive -i from *-ьjь which occurred in the genitive of nouns such as bicie, przykazanie, miłosierdzie; but these nouns due to their semantic nature were rarely used in the plural, and when they were, they took -∅: kazań, miłosierdź, pokoleń, stajań, zbóż (14th and 15th centuries); this would mean that -i||-y is not an archaic reflex of *-ьjь but rather a newer ending taken by modeling on genitive forms like gości, kiści, koni, żołnierzy, koszy, kości, nocy.^[339][336] From the 16th-18th centuries -ów from masculine declension can be seen: piekłów, dziełów, cłów, stadów, gusłów, ogniwów, zwalisków, bagnów, igrzysków, niebiosów, which can also occur in dialects.^[340][336]

Before the 15th century the ending -om for the dative plural of various origins spreads: custom, dziełom, słowom, świadectwom, zwierzętom (14th century).^[340][341] Within old *-o stems it is inherited from Proto-Slavic and from here the tendency to level took over all other dative neuter forms removing the original prehistoric *-emъ in *-jo stems and *-ьmь in consonantal stems.^[340][341] In the 15th century the feminine ending -am sees some use: dziatkam (15th century), latam, słowam, krolewstwam, miastam, ciałam, pismam, przykazaniam, polam, obawieniam, pokoleniam, zwierzętam, jagniętam (16th century); it appears that this tendency was at its strongest from the transitional periods between the 15th century and 16th century and from the middle of the 16th century it starts to give way as Kochanowski doesn’t use it at all, and it falls out of use with the end of this century, and the only used suffix becomes -om, sometimes with slanting -óm, like Kochanowski: latóm, drzewóm, książętóm, oczóm, uszóm.^[340][341]

Neuter hard-stem nouns in the instrumental plural, i.e. old *-o stem and consonantal stem nouns inherited -y in the Middle Ages and to the middle of the 16th century almost without exception: laty, pióry, żyty, ciały, zwierzęty, książęty (14th century); słowy, prawy, gasidły (15th century), drzewy, jarzmy, państwy, skrzydły, stady, zioły, bliźnięty, imiony, znamiony (16th century); -y is also used for nouns suffixed with -c(e): jajcy, siercy, mieśćcy.^[342][341] Old *-jo stems from the earliest times have -mi instead of *-i taken from masculine declension (synmi, gośćmi) and feminine (kiśćmi): polmi, miestcmi, wołańmi, zbożmi, pokoleńmi (15th and 16th centuries); -mi can sometimes be seen in neuter hard-stem nouns ending in a liquid consonant: piórmi (14th century); jeziormi, ciałmi, działmi; kołmi, prześcieradłmi, winmi, ramionmi (15th and 16th centuries).^[343][341] Before the 16th century -ami, from feminine declension, can be rarely seen: przedmieściami, pokoleniami, odzieniami (15th century); działami, piorami, kolanami, pętami, prawami, ziarnkami (first half of the 16th century); over the course of the 16th century this ending spreads to neuter nouns of various origin, and in the 17th century becomes the dominant form, -y and -mi forms can be seen later, but as archaisms, such as modern przed laty, innymi słowy.^[343][341]

Among three originally inherited locative plural endings, only -ech < *-ěchъ of old *-o stems sees wide use, however the ending *-ich < *-ichъ of old *-jo stems is not used, and the ending -ech < *-ьchъ of consonantal stems very rarely: niebiesiech, ptaszęciech (14th and 15th centuries); -ech sees dominant usage until the middle of the 16th century: błogosławieńśtwiech, piśmiech, skrzydlech, święciech, jaślech, okniech, drzewiech, stadziech, leciech, żeleziech, mieściech, czelech, cielech, słowiech (14th-16th centuries), but by the end of the 16th century it falls out of use.^{[343][321][344]} From the 14th to the middle of the 16th century all types sometimes take -och of masculine declension: prawoch, kolanoch, latoch, siercoch, mieścoch, poloch, znamienioch, imienioch, dobytczętoch, książętoch.^[343] Already in the Middle Ages -ach||-ách sees uses: swawidłach, siedliskach, państwach, bractwach, polach, przykazaniach, siercach, imionach, ramionach (14th and 15th centuries), and in the 16th century becomes the dominant ending in soft-stem nouns; in hard-stem nouns it sees co-usage with the ending -ech and by the end of the century entirely displaces it: Rej’s Apokalipsa has a ratio of -ech to -ach of 33:9, and -ech occurs in Kochanowski’s works exceptionally.^[343][344]

Mixed noun declension are the result of semanticization (see § Nouns.^[345] They are considered an irregular declension.^[346]

There are two main groups of masculine nouns ending in -o: the first are diminutive nouns attested as early as the Middle Ages such as Mieszko, Jaśko, Walko, Dobko - these are most likely Polish innovations based on neuter nouns it is doubtful that these are the result of artificial latinization as this type of noun is also found in southern and eastern Slavic dialects.^[347] Later innovations of this type are given names such as Jasio, Kazio, Franio, etc - all of them decline as masculine nouns.^[347] The second type are family names such as Jagiełło, Sanguszko, Matejko, Fredro - in the singular they are kept in masculine forms until the 17th century: Jagiełła, Sanguszka, Tarła; Jagiełłowi, Sanguszkowi; Jagiełłem, Fredrem, etc. and in the 17th century they begin to take feminine endings: Fredry, Fredrze, Fredrą, but archaic masculine forms can be seen in the 19th century.^[347]

Masculine nouns ending in -a include hard-stem nouns: wojewoda, sługa, starosta, mężczyzna, niecnota, monarcha, poeta, Kmita, Kostka; soft-stem nouns: sędzia, hrabia, rękojmia, kaznodzieja; names: Zawisza; and nouns ending in -c(a): rajca, zastawca, przyczyńca, złoczyńca, jedynowłajca - in the singular they generally decline according to feminine paradigms: wojewodzie, patriarsze (dative singular), rarely do they take masculine endings: patriotowi, stworcowi (dative singular).^[347][345] The nominative plural in the Middle Ages is usually from feminine declension: sługi, wojewody, patriarchy, starosty, mężczyzny, kaznodzieje, przyczyńce, but already in the 15th century, more often in the 16th century, and predominantly from the 17th century the masculine endnigs -owie, -y occur: starostowie, rybitwowie, poetowie, Kostkowie, sędziowie, margrabiowie, zdrajcy, naśmiewcy.^[347] The genitive plural originally had a normal, feminine -∅ ending: wojewod, starost, poet, panosz, sędź, kaznodziej, mężobojec, radziec, and modern mężczyzn, but hard stems over the course of the 15th century and soft-stems from the beginning of the 16th century take the masculine ending -ów, also in dialects: lutnistów, wróżbitów, Kmitów, sędziów, złoczyńców.^[348] Other plural forms in the Middle Ages take feminine endings, and from the 16th century share their fates with masculine endings, with dative -om, instrumental -ami, and locative -ach.^[348]

Collective forms such as bracia, księża are now understood to be the nominative plural of brat, ksiądz, but originally were nominative singular with a collective meaning: the original ending was *-ьja and contracted to -á: braciá, księżá; however already in the earliest epoch these forms can syntactically connect with adjectives, pronouns, and verbs in the plural due to their collective meaning: uźrzą bracia twoi i moi, naszy mili bracia, and based on this forms like bracia became the nominative plural, braci as the genitive plural (like gości), and also the accusative plural.^[348][349] These new associations were further supported by soft-stem nouns like gość lead to innovations in other cases: braciom, braćmi, braciach - this process probably took placein the Middle Polish era.^[348][349] Dialects not only keep old braciá, księżá (as the plural), but also created new muzykanciá, adwokaciá, swaciá.^[348]

Neuter nouns ending in foreign -um from the earliest times don’t decline in the singular, and in the plural they decline like neuter nouns, except in the genitive plural, where -ów is used.^[348][346]

The Proto-Slavic dual forms in historic Polish show a slow but complete disappearance, and in the oldest texsts they aren’t necessarily used to mark dualness or parity, but rather were used instead of normal plural forms, and only objects naturally in pairs, that is certain body parts were expressed in the dual: oczyma twym uznamionasz, uciekła pod jego skrzydle (14th and 15th centuries), but conventional parity, i.e. frequent but not necessary, and incidental parity are accomponied by the numeral dwa or oba, and the loss of a dual meaning is demonstrated by examples such as trzema palcoma, trzema dnioma.^[350][351] The end of relative productivity of dual forms lands the second half of the 16th century, with more forms occurring the further back in time, and individual instances can be seen in 17th century.^[350][351] Dual forms in Polish are generally inherited frmo Proto-Slavic forms with only some leveling and simplification.^[350][351]

In the nominative and accusative dual:^[352][351]

1. In the masculine:
   1. -a < *-a: dwa szczyta, męża dwa, dwa kmiecia, dwa miecza, dwa konia, dwa młodzieńca, oba końca, oba przyjaciela;
   2. -y < *-y; *-u stems: dwa syny, dwa woły, sometimes replacing -a via association: rozdzieliłasta się dwa braty;
2. In the vocalic feminine:
   1. -e < *-ě in hard stems: dwie rybie, dwie żyle, dwie kiełbasie, dwie siekirze, dwie strzale, dwie fidze, dwie dziewce, obie naturze; ręce is kept in a plural meaning in the nominative and accusative;
   2. -i||-y < *-i: dwie duszy, dwie świecy;
3. In the consonantal feminine:
   1. -i||-y < *-i: dwie gałęzi, dwie piędzi; oczy and uszy belong here despite being neuter nouns, as they were originally *-i stems (compare Lithuanian akis ausis) and create their dual according to the feminine consonantal dual;
4. In the neuter:
   1. -e < *-ě in hard stems: dwie lecie, dwie świetle, dwie ciele, dwie słowie, obie wojsce, obie kolanie, also the Old Polish innovation oce usze;
   2. -i||-y < *-i in soft stems : dwie poli, dwie słońcy.

In all declensions paradigms in the genitive and locative inherited -u occurs: dwu rowu, dwu rodu, dwu aniołu, dwu koniu, dwu miesięcu, na obu boku, na dwu łanu, o dwu strożu; ręku, dwu dziewku, obu dziedzinu, na obu stronu, we dwu ziemiu, obu wsiu, we dwu sieniu; dwu pokoleniu, na dwu drzewu; the locative ręku is kept in a singular meaning; genitive oczu uszu in the plural.^[353][351]

Dative and instrumental dual:^[353][351]

1. In the masculine:
   1. -oma < *-oma of *-o stems generalized to all types; *-ema from *-jo stems; *-ъma from *-u stems; *-ьma from *-i stems: dwiema ławnikoma, dwiema biskupoma, dwiema kroloma, dwiema strumienioma, obiema końcoma;
2. In the vocalic feminine:
   1. -ama < *-ama rękama, dwiema dziurkama, dwiema dziedzinama, obiema niewiastama;
   2. -oma from the masculine: rękoma, dwiema drogoma, dwiema winnicoma, dwiema wieżoma; rękoma is kept in a plural meaning, see also Declension of masculine nouns in the instrumental plural;
3. In the consonantal feminine there are no instances of inherited *-ma < *-ьma; the preserved and inherited from Proto-Slavic instrumental forms oczyma uszyma are exceptional and built by adding the etymological ending -ma to the full nominative forms oczy uszy:
   1. -ama based on vocalic feminine nouns: dwiema częśćiama;
   2. -oma based on masculine nouns: dwiema rzeczoma, obiema garścioma;
4. In the neuter:
   1. -oma < *-oma from *-o stems, generalized instead of -ema < *-ema from *-jo stems: kolanoma, dwiema latoma, dwiema wojskoma, czołoma, dwiema morzoma, nozdrzoma.

Polish makes many changes to the pronouns of Proto-Slavic: *onъ *ona *ono become personal pronouns, *jь sees a considerable restriction in use, *sь is lost except in derivations like dzisiaj, latoś, the indeterminate ending -ś, or fossilized phrases like do siego roku, ni to, ni sio.^[354]

Changes in declension in the masculine and neuter singular comes from a Proto-Slavic phonetic process of the influence of *j on the following vowel, resulting in *-jo > -je, *jě > ji, *jъ > jь, *jy > ji; the Polish tendency was to level both types according to the soft-stem endings.^[355][356]

In masculine nominative, after the loss of yers Polish gained -∅: ów, on, mój, nasz. The pronouns *tъ, *jь were extended with *nъ: *tъnъ, *jьnъ. ten remains in use, jen falls out of use over the course of the 16th century.^[355][357] There are no changes to the neuter nominative and -o is inherited.^[355][357]

The genitive form -ego spreads via leveling to the soft-stem: mojego, also contracted mego, twego, swego, naszego, jego, owego, onego, tego; the late age of the leveling is evidenced by the hardness of the consonant near the front vowel -e.^[358][359] The archaic exception togo is attested three times in Kazania świętokrzyskie, once as a masculine genitive, once as a masculine genitive-accusative, once as a neuter genitive, onogo (genitive-accusative).^[360] Masovian dialects have slanted tégo based on adjectival declension.^[360]

In the dative the ending -emu spreads via leveling to the soft-stem: mojemu, also memu, twemu, swemu, naszemu, jemu, owemu, onemu, temu; the late age of leveling is evidenced by the hardness of the consonant near the front vowel -e; there is a single instance of archaic tomu (neuter dative) in the Holy Cross Sermons.^[360][357]

The accusative is equal to the nominative.^[360] The archaic accusative masculine form ji is the reflex of Proto-Slavic accusative *jь and was rare already in the 15th century, in the 16th century it is more often replaced by the form go, probably the result of reduction of the genitive form jego, as the stress was originally on the last syllable; another trace of this old accusative form can be seen in prepositional constructions + -ń such as zań, weń, przedeń, as this -ń < *njь is the old accusative of *-jь with epenthetic n, which arose in Proto-Slavic from combinations such as *sъn jimь, *vsъn jimь, *kъn jemu as the result of perinteɡration, i.e. reinterpretting the -n as belonging to the next word.^[360] Already in the earliest texts the masculine genitive is used as the masculine accusative if the pronoun refers to a man or is in a syntactical phrase with a masculine-personal noun: tego poniża a onego powysza; tego prześladował jeśm; oddalił jeś bliźniego mojego, gdy pwam w boga mego (14th and 15th century).^[360][359]

The instrumental from the oldest era to the middle of the 15th century the ending -im or -ym, generalized to the soft-stem, dominates: moim, aso mym, twym, swym, naszym, tym, onym, owym etc. and in the 15th century the instrumental ending begins to mix with the locative ending -em.^[360][359] The pronoun jen in the instrumental (j)im is kept until the end of the 15th century and is kept now in the fossilized construction im - tym; in the 16th century nim occurs next to im and nim becomes the only form used in the 17th century.^[360][359]

In Old Polish the locative ending -em, generalized based on soft-stem declension, is used: w tem, w samem, w żadnem, w jem, na naszem, then in the 15th century the instrumental ending -im, -ym sees use in the locative, therefore the instrumental and locative had distinct endings in the earliest era and begin to merge starting from the second half of the 15th century into one form with the fluctuating form -im, -ym or -em.^{[360] [357]}Orthographic works attempted to deal with this situation, for example Kopczyński arbitrarily introduces differentiating genders without regard to case instead of the old differentiating of case regardless of gender and sets -im, -ym in the instrumental/locative masculine and -em in the instrumental/locative neuter., and the orthography reform of 1936 recommends the forms -im, -ym for both cases and both genders.^[361][357]

No changes occurred in the nominative and the ending is -a: ta, owa, moja, also ma, twa, swa, nasza, etc.^[362][357]

The Proto-Slavic genitive endings *-ojě, *-ejě contract in prehistorsic Polish into -é: oné mowy, z této ziemie, z oné straszné postawy, z té powieści, żádné sprawy, słáwy twé, swé głowy (14th-16th centuries); młodzieży mojé, macierze mojé, w jé świętem żywocie, od nié, twojé miłości (14th-16th centuries).^[362][357] Next to these forms already in the oldest texts the genitive form -ej occurs often, arosing via assimilation to the dative and locative etymological -ej, and was further phonetically motivated as -é sounds similar and probably had a phonetic final -j, [ej]: pełność jej, nieprawdy mojej, dziewki twojej, z ręki twej, męki swojej, dusze swej, dziedziny waszej, wszej ziemie (14th and 15th centuries).^[362] As a result of raising slanted é towards y or i as well as the loss of final j a third ending occurs sporadically: -y after hard consonants or -i after soft consonants: z takowy powieści (16th century), połowa większa flotty ony zginęła (17th century), ty ji pieniądze (15th century), fortuny swoi (17th century).^[363][357]

The dative and locative share the ending -ej, which is the reflex of Proto-Slavic soft-stem ending *-eji, which simplified by removing the final vowel; tendencies to level also meant that this ending replaced hard-stem *-oj(i) in prehistoric times: krasie mojej, prawicy twojej, prawdzie swojej, naszej duszy, wszej postaci, bydlić będę w niej, we śmierze mojej, w drodze twojej, w sławie swojej, w ziemi naszej, na drodze tej, w tej trojcy (14th century).^[364][357] The ending -e for the dative-locative also occurs via assimilation of the dative-locative to the genitive, e.g. in dative: teto kaźni posłuszen, jebyło to imię, anjoł jest ci się je był ukazał, ucieczmy się k nie (14th century), ktoré jął dziękować, ojtczyznie swé (16th c.); and in the locative: w oné stajni, w té nędzy, w niéktoré epistole, w ktore jest wypisan, w swe przyjaźni, po śmierci twoje (16th century).^[364][357] These forms were not kept as the opposite leveling won out, that is leveling the genitive to the dative-locative.^[364][357] The difference between genitive do te głupie baby and dative-locative ty głupi babie was only kept in southern Silesian, Lesser Poland, and in various places in the western periphery of Greater Poland.^[364][357]

The Polish accusative has -ę < *-ǫ: jedzinaczkę moję, rozciągnąłeś rękę twoję, drogę swoję, prze chudobę naszę, wszytkę radę, nawiedzi winnicę tę (15th century) with one exception, ją odjęli ją, na drodzęe jąż wybrał (15th century), after a preposition nię wpadli w nię, przez nię mężnie poczynali, na nię ciężko p rzyszło (15th and 16th ceneturies).^[364][357] Only in the second half of the 19th century is the accusative -ą taken from the feminine accusative adjectival form; Małecki considers it a mistake, saying in 1879 “Formy moją siostrę, naszą, waszą, czyją matkę, których niektórzy pozwalają sobie w mowie i piśmie, są to błędy stanowcze. Odwoływanie się do brzmienia ją, wyjątkowo urobionego, nie byłoby tu na miejscu” (The forms moją siostrę, naszą, waszą, czyją matkę, which some people allow themselves in speech and writing, are definite errors. Referring to the sound of ją, exceptionally developed, would not be appropriate here).^[364][357] The archaic accusative form tę remains, but more often in writing than in speech, and is often replaced now with tą.^[364][357]

Polish has the instrumental ending -ą < *-ojǫ, *-ejǫ and remains from the oldest times without changes: między bracią moją, z dziedziną twoją, podpira swoją ręką, nade wszą ziemią, wszystką myślą.^[364][357] The fom ją sees use in the 16th century: czarty ją (wodą), odganiając; abyśmy się ją zapalali, and after prepositions since the oldest times nią, which in the 16th century begins to displace ją, and since the 17th century is the only form.^[364][357]

The masculine nominative is -i from the oldest times or -y after functionally soft nouns, but only among masculine personal nouns: ci co znają, oni zginą, nie będą powyszeni sami w sobie, wszytcy rzeką, jiż wykupieni są nieprzyjaciele moi, kapłani twoi, oćcowie naszy (14th century); the range of this form was wider in the past, and until the 17th century it occurred exclusively in masculine animate nouns: ci ćwicy, niektorzy ptacy, wszytcy ptacy (16th century), however during the Old Polish era it could occur also with inanimate nouns if the syntactically dependant noun had an ending originally nominative in origin, and not accusative as nominative: wszytcy krajowie, wszytcy końcowie, śladowie moi, sądowie twoi, wozowie twoi, ołtarze twoi, dniowie naszy.^[365] Outside of these conditions the accusative acts as the nominative, in soft stems -e: palace twoje, grzechy twoje, dni nasze (14th century); in hard stems -y||-i (after k g): wszytki kraje, wszystki sądy, ty pieniądze (14th century), ony ogni, samy zwony, ty sądy wszytki narody, ony losty (16th century); from the 17th century this ending gives way to -e taken from the soft declension, giving the modern forms te domy, owe sady, one dni, etc.^[366] The inherited masculine personal forms naszy, waszy (< *naši *vaši), e.g. oćcowie naszy, ojcowie waszy, naszy wymowce polscy, bracia naszy mili (14th-17th century) fall out of use over the 17th century and the innovations nasi wasi, occurring from the beginning of this century are from the nominative masculine personal plural adjectival declension, where thematic *š alternates with soft ś.^[366] The feminine soft-stem nominative continues uninterrupted as -e < *-ě: moje nogi, kości moje, ścieżki twoje, wargi nasze; the hard-stem ending is kept until the 17th century as -y: wszytki drogi (14th century), ty rzeczy, rzeczy, ktory ku ubiorowi należą (15th century), wszytki rozkoszy moje, jeśliby się ony (rozmowy) podobały, ty rzeczy, samy Amazony (16th century); sporadically from the 15th century, then more often from the middle of the 16th century, and commonly from the 17th century and used today is the ending -e: te, owe, które etc., introduced based on the nominative soft-stem declension, also influenced by the adjectival nominative feminine plural ending.^[366][357] The neuter rarely takes -a in medieval texts: a są ta ista słowa zmowiona (Holy Cross Sermons), wszytka działa jego, lata moja, świadectwa twoja, lata nasza, książęta wasza (Sankt Florian Psalter), and more often forms that assimilated to the nominative inanimate masculine and feminine nouns occur with -y in hard stems: tyto książęta, only drzewa, wszytki źwierzęta (15th and 16th centuries); soft-stem nouns take -e, which is used today: nasze dusze, moje starania.^[366] Over the course of the16th century hard stems also take -e based on soft-stem declension: te, one, owe pola, and this state continues from the 17th century without change.^[366][357]

In the genitive and locative -ych, -ich after k g, is kept without change in hard-stem pronouns and -ich, -ych after hardened consonants, in soft-stem pronouns: od tych co mnie gonią, jarzmo jich, jedno z nich, drog mojich, snow twojich, dział swojich, podług grzecohw naszych (14th century); -ich is the reflex of Proto-Slavic *-ichъ, but -ych in hard stems is an innovation instead of Proto-Slavic *-ěchъ, which arose via tendencies to assimilate hard-stem forms to soft-stem ones, which, as can be seen, was influenced also by singular forms.^[367][357]

Historic Polish has the dative plural -ym, -im after k g, in hard stems and -im, -ym after hardened consonants, in soft stems: przeciwo tym, daj sie wiesielić wszytkim, posłał jim, mołwić k nim, wrogom moim, obietam twoim, łajaniam twoim, duszam swoim (14th and 15th centuries; -im is inherited from Proto-Slavic *-imъ, and -ym is an analogous innovation like the genitive; the only kept archaism with etymological -em < *-ěmъ is the form ciem in the Sankt Florian Psalter: prawda jego nad syny synowymi ciem, co chowają ustawienie jego (page 102).^[367] As in the instrumental singular and plural, the dative sporadically shows fluctuation of the vowels i, y, or e before m: wszytkiem żywem utnę szyję (15th century); niektórem zdrowie dawał, ku tem paniam, anioł rzekł k niem, powiedział jęm, soli tám tém krájóm dosyć dawa (16th century); mieśce pannie z panem młodym dajcie, jemci z sobą być (17th century).^[367]

Within the accusative, masculine and feminine soft-stems have inherited -e < *-ě without change: zbawiony uczynisz je, oświeci czyny moje, poznaj drogi swoje, prze sądy twoje, kaźni twoje miłował jeśm, pod nogi nasze, nad syny wasze (14th century); in hard-stems the etymological reflex -y remains until the 16th century: w ty jiż sie nawracają, rozproszy ony, zgubisz wszystki, widział jeś wszytki syny, tyto krole powalił, wszytki ludzie potracisz, prze ty dwa bogi o ony gwałty, zowią ty czas (14th and 15th centuries); over the course of the 16th century -y gives way to soft-stem -e through leveling, which is in common use from the 17th century.^[368]

The neuter ending -a is partially kept: odpuści wszytka dopuszczenia, słuszał jeś wszytka słowa, położył jeś rammiona moja, miłował jeśm świadectwa twoja, ostawią cudzym bogactwa swoja, położcie sierca wasza, ta słowa pisze (14th century), but more often, already in the oldest texts, forms modeled on the masculine and feminine accusative -y in hard stems and -e in soft stems: słyszali wszytki słowa, na wszytki zwierzęta, słyszeć ty słodkie słowa, ty drzewa ujrzał, strzec serca i ciała nasze, we świadectwa twoje (15th century), and in the 16th century, -y forms slowly give way to -e from the contemporary accusative soft-stem of pronouns and all adjectives; from the 17th century to now it is the only form.^[369]

Before the 16th century -imi, -ymi after hardened consonants, occurs for all genders, and -ymi, -imi after k g, in hard stems: z tymi, tymi słowy, miedzy przyjacioły mojimi, z panoszami swoimi, nad grzechy naszymi, ze słzami mojimi, z ciały swojimi (14th and 15th centuries); -imi is the direct reflex of Proto-Slavic *-imi and -ymi is an innovation based on prehistoric leveling of expected forms such as *ciemi, *owiemi (<-ěmi) based on moimi, naszymi, with keeping the hardness of the stem in most cases, but in the Sankt Florian Psalter, page 119, 6 the exceptional form cimi, z cimi jiż są nie naźrzeli p okoja, był jeśm pokojen occurs.^[369] In the 16th century -emi sees more and more use resulting from phonetic changes of -ymi, -imi and strong lowering of y, i before m: -ymi, -imi, > -émi > -emi: témi káráktermi, obyczajmi swemi, miedzy niektoremi uczonymi, z inszemi się uda, z szledziami… ktoremi chłopy zarzucają, miedzy inszémi rzeczami, wszystkiemi siłami, miedzy wszytkiemi inemi niebezpieczeństwy, z swemi państwy (16th-18th centuries).^[369] Old -ymi and -imi remain in equal use next to -emi then in the 19th century orthographic works artificially regulate this by limiting -ymi, -imi originally to the masculine gender, later only masculine personal, and -emi to feminine and neuter or masculine non-personal: tymi chłopcami, psami, temi robotnicami, książkami, dziećmi, and now only -ymi, -imi are written, but -emi can still be heard, but there is no functional difference between the two, as one can hear both temi kobietami and temi chłopcami.^[369]

Relatively few dual forms are seen til the end of the 15th century; they fall completely out of use in the 16th century.^[370]

Masculine pronouns in the nominative and accusative take -a: pręt twoj i dębiec twoj ta jesta mnie ucieszyła, Jozef z Maryją jesta ona była przyszła, dwa syny twa moja będzieta (14th and 15th century); sama dwa krola biłasta się (16th century).^[370] Feminine hard stem pronouns ake -e: wypuści światłość twoję i prawdę twoję cie jesta mie przewiedle; o nie poczęlesta płakać (14th and 15th century); soft-stem feminine pronouns take -i: ręce twoi gospodnie uczynilesta mnie, ręce swoi umyje, rozciągnąłem ręce moi (14th and 15th century).^[370] Neuter hard-stem pronouns take -e: za cie dwieście grzywien; soft-stem pronouns take -i: widziele oczy moi, oczy moi mdlesta byle, podźwigł jeśm oczy moi (14th and 15th century).^[370]

For all genders the ending -u is used in the genitive and locative: świeca oczu moju, cień skrzydłu twoju, w działu ręku twoju, działu ręku naszu, nie posawił tu dwu wołu, w oczu naszu, w ręku twoju, w tu dwu niedzielu (14th and 15th century).

For all genders the ending -ima||-yma for the dative and instrumental is used: dam sen oczyma moima, zawoław farao Mojżesza i Arona rzecze jima, zawoła Mojżesza a Arona a rzecze k nima, rzekł swyma żonama, uszyma naszyma słyszeli jesmy, swyma rękama dotknąć, patrzyć moima oczyma (14th-16th century).^[370]

All mordern forms of kto and nikt(o) continue from the oldest times with the exception of the locative, which originally was kiem: przetoż je nam chwalić słusza, w kiem jeść koli dobra dusza (15th century); the only change from Proto-Slavic was in the locative, as locative and instrumental later merged, with instrumental form kim being established.^[370][371] The instrumental form kim instead of expected cem < *cěmь arose based on soft-stem instrumental pronouns and adjectives; the locative kiem is also assimilated to them instead of *komь; all other cases are inherited: kto < *kъto, kogo < *kogo, komu < *komu.^[370][371]

The only changes that occurred in historic Polish for co, nic(o) are in the instrumental and locative; the instrumental before the 16th century was czym, and the locative czem; later there was a clear tendency to use one form for both, but their forms are mixed, czym or czem; orthographically in the 19th century the forms czem, czém, czym are used, and in the end czym won out.^[372][371] In comparison to Proto-Slavic there is one difference, which is that co < czso is used in the nominative and accusative, and derives from the Proto-Slavic genitive form *čьso||*česo; the original nominative-accusative form čь was phonetically weak by the loss of yers, meaning that Polish could use the form czso, which was rendered unnecessary by the new genitive form czego, created from prehistoric times.^[372][371] The old accusative form cz is kept in Polish only supported by a preposition, przecz, zacz, (w ni)wecz.^[372][371]

Most forms are inherited:^[373][374]

1. Nominative: ty < *ty, my < *my, wy < wy;
2. Genitive: mnie < *mene||*mьne, ciebie < *tebe, siebie < *sebe, nas < *nasъ, was < *wasъ;
3. Dative: mnie or mi < *mьně, tobie or ci < * tebě||tobě or *ti, sobie or si or se < *sebě||*sobě or *si; the forms mnie, tobie, sobie are used when stressed, but eastern Poles near the border may use them unaccented. The forms mi, ci, and si (rare in Old Polish) and dialectal or colloquial se (<sobie) from the 18th century are enclitics.

Other forms and developments include:^[373][374]

1. The nominative ja < já < *azъ underwent prejotation and shortening; in prehistoric times probably both jaz and ja existed next to each other, and the use of one or the other depended on the initial sound of the next word, then ja spread in use and jaz is noted only once in the Sankt Florian Psalter, page 108, 3: ale jaz modlił jeśm się, most certainly a Bohemism;
2. The enclitic use of accusative forms as genitive: mię cię, się, and the opposite anaology of genitive as the accusative in accented forms;
3. In the accusative, in an accented position genitive forms mnie, ciebie, siebie appears; some texts keep old inherited mię < *mę, cię < *tę, się < *sę, but in enclitic position only mię, cię, się or nasalless mie, cie, sie were used. This differentiation occurred in dialects: Greater Polish and Masovian had only nasal forms mię, cię, się, but both nasal and nasalless forms occurred in Lesser Poland depending on the neighboring term, near a verb mie, cie, sie, not near a verb, especially after a preposition, mię, cię, się. In the 17th century full forms appear after prepositions, mnie, ciebie, siebie, but only in the 19th century the current situation establishes itself, and only these forms in this position are possible. Eastern Poles near the border may use mnie, ciebie, siebie, even unstressed;
4. The accusative forms nas, was are genitive in origin, and the original forms ny, wy are lost, but poświęci ny attested once in Gniezno Sermons;
5. Instances of dual forms are attested: nominative first person dual wa: wszystko co wa przysięgła (15th century), osieł z wołem: wa nie umiewa oracyj (16th century). wa is an innovation instead of expected wie < *vě probably under influence of the first person dual ending for verbs;
6. The genitive-locative naju < *naju, waju < *vaj and the dative-locative nama < *nama, wama < *vama. The forms naju, waju, nama, and wama were kept in some dialects, but these forms now have a plural meaning: u naju, u waju meaning u nas, u was.that subtly is quickly fading.
7. się/siebie is defective and lacks a nominative; it is also often used to create a mediopassive voice or impersonal passive voice;
8. Dative forms of pronouns are used to create the dativus ethicus;
9. The Proto-Slavic pronouns *onъ *ona *ono were originally demonstrative pronouns and become personal pronouns in Polish; their oblique forms are suppletive from *jь;
10. Like nouns, pronouns also differentiate animacy;
11. Genitive forms of third person pronouns become possessive pronouns.

Proto-Slavic adjectives, except possessional adjectives, had definite forms, used when a noun was definite, and indefinite forms when a noun was indfinite;^[375] the definite form was built from the indefinite form extended with the pronoun *jь.^[376] Polish did not keep any marking of definiteness, but instead changes their semantic function; the indefinite form is possible only in the predicate, and only the definite form is used attributively.^[375] Only a few vestigal exceptions exist.^[375]

Proto-Slavic adjectives stems could also be soft or hard; continued into Polish.^[377]

The declension of indefinite adjectives consisted of adding the ending appropriate to the noun declension to the adjective, i.e. masculine and neuter hard-stem adjectives take masculine and neuter endings of *-o stems, and masculine and neuter soft-stem adjectives atke feminine *-a stems, feminine soft-stem adjectives take feminine *-ja endings: star stara staro, stara stary, staru starze, star starę staro, starem starą, starze; this declension is sometimes called a nominal declension because of this.^[378] Mostly nominative forms are kept as a result of the fact that it was possible to use indefinite non-possessive adjective forms only in the predicate.^[379]

Indefinite forms differed from definite forms initially only in length and later in clearness and slanting: the feminine nominative singular and neuter nominative plural took short, clear -a in the indefinite, and long, slanted -á in the definite, similarly, the neuter nominative singular of soft stems and feminine nominative-accusative-vocative plural of soft-stems took short, clear -e in the indefinite and long, slanted -é in the definite.^[379]

The difference of length or slantedness were not marked in manuscripts, and prints marked it only sometimes and often incorrectly; Łazarz’s print marks entirely correctly distinguishes indefinite forms: Twym rozumem ták miernie ziemiá usadzoná, że na wiki nie będzie nigdy poruszoná… Pełná jest wszytká ziemiá twej szczodrobliwości (where á marks clear a); in Wietor’s print from 1542 “Sąd Parysa” (á = clear a, a = slanted a): pożywaj wesela, pókiś młodá; ja bábá stára, ale miedzy młodzieńcy jára - młodá” - in the first example is a predicate with a indefinite form, and in the second an attributive in a definite form; but a loss of the sense of distinguishment can be attested via sentences such as: przymiż odemnie jabłko, Wenus młodá, or at a different point in time: miła pani Wenus - both examples have an attributive in a definite form; similarly not clearly marked is a in the sentence: jestem ci bárzo godna, w wieńcu bárzo cudná, where in both cases the adjective is in the predicate and should have indefinite clear -á.^[379] In the masculine nominative plural in both forms the ending is -i||-y.^[379]

Already in the oldest epoch indefinite adjectival forms in the attributive are rather rare and exceptional, from the 16th century they are not seen at all except adjectives with the formant -ow and -in. Only possessive adjectives with the formant -ow from personal names or common nouns dominate in the short form in the attributive, and in the predicate adjectives from the oldest epoch take either the definite or indefinite form; from the 16th century indefinite forms are rare.^[380] The adjectival participle in the predicate has a indefinite form to the end of the 15th century; fluctuation begins in the 16th century, and from the 18th century definite forms dominate almost exclusively.^[381][382] Both in the attributive and in the predicate adjectives of various word-formation build appear in indefinite forms, most often with the formant -ow- in the attributive, with the formant -en||-n in the predicate.^[376] Some adjectives from the oldest epoch show a particular tendency for non-comound forms in the predicate, now they mainly have this form exclusively: rad wart, or optionally: pełen, wesół, zdrów, winien, gotów, łaskaw.^[376] Traces of indefinite declension is kept in a few categories of terms, which have since long stoped being adjectives in Polish:^[376]

1. Above all else adverbs endings in -o: młodo, staro, boso, mało, cicho, blisko, wysoko, daleko, which are the old nominatives or accusatives neuter singular (some innovated soft-stems based on hard-stems occur, e.g. głupio, tanio, dostatnio; adverbs with -e: dobrze, pięknie, źle, dokładnie, which some consider the old singular locatives and others consider to be formed under influence of the nominative neuter singular of comparative adjectives; forms with w such as wkrótce, wcale are undoubtedly from the old locative;
2. Prepositional phrases in the genitive singular: z cicha, z chłopska, z głupia (frant); in the dative: po cichu, po polsku, po prostu, in the genitive plural: z dawien (dawna); these are all sued adverbally;
3. Placenames formed with -ów, -owa, owo, -in, -yn, -ina, -ino, -sk, -*jь: Kraków, Częstochowa, Wejherowo, Teresin, Szczuczyn, Lusino, Radomsko, Przemyśl, Poznań, Oświęcim, which were originally possessive adjectives with an attributive function naming an ellided noun such as gród, osada, wieś, pole, etc.;
4. Nouns referring to women from masculine surnames or masculine professions with the ending -ina||-yna: Sroczyna, staro