XXI

__________________________________

An Outline of the History of the Polish Language

_______________________________________

CHAPTER II

_________________________________

AN OUTLINE OF THE HISTORY OF THE

ARTICULATORY-AUDITORY SYSTEM OF POLISH

<. . .> §13. This is a history of the phonemes (or sound-representations) and their smallest components, i.e., their psychologically indivisible articulatory-auditory elements on the one hand, and the various combinations of phonemes on the other.

But are we justified in comparing in a historical framework the differences and similarities between phonemes and their components apart from the morphemes, or morphological elements, of which they are a part? Can we speak of an identical history of the Polish phonemes associated with the letters or graphemes c, e, s, d, etc., just because all the phonemes associated with each letter are identical from an art.-aud. (articulatory-auditory) viewpoint, and because they can be reduced to the same art.-aud. representation?

We cannot.

In the present linguistic system, the phoneme c has a different historical origin and different variants in the morphemes świec-, płac-, piec-, tłuc-, rzec, noc-, moc, móc, strzec . . . , ręc-, rzec-(rzek-a, rzec-e), owc-, kupe-, lic-, plac-, tac-a, cegła, cesarz, car.

Despite its art.-aud. identity, the Polish phoneme e has a different history in the morphemes wiezie, wiedzie, bierze, plecie, miedza . . . , wiedza, świeci . . . , wieś, len, cześć . . . , sen, mech, łeb, krew . . ., pełny, bełtać . . ., ser, sierota, zabierać . . . , cera, telegraf, etc.

The phoneme s (or ś) of sad, syn, siedem ... is historically connected with the proto-IE phoneme s; the very same phoneme s of słoma, pros- (prosić), pros- (prosię), sto, osiem, dziesięć, wieś ... is historically related to the IE phoneme k₁ (fronted k), whereas the phoneme ś of pleść, kłaść . . . alternates with (i.e., is a variant of) t, d and developed historically from t, d. The phoneme s, furthermore, is a part of many loan-words, or rather, of assimilated morphemes.

The phoneme d of dać, dom, wiadomo . . . has its origin in a nonaspirated IE d; whereas the d of dym, wdowa, broda, rudy . . . goes back to an aspirated dh; in addition, the phoneme d occurs in many assimilated words.

Thus we can see than an art.-aud. history of a language must consider the phonemes and their components not as art.-aud. elements independently of meaning, but as art.-aud. elements endowed with semasiologized and morphologized functions.

§14. One could shorten and simplify the exposition by presenting the history, not of the whole, psychologically complex phonemes, but of the psychologically indivisible components, i.e., of the simplest articulatory and auditory elements making up the phonemes. I shall explain this with examples.

In the consonantal phoneme which is associated in Polish with the grapheme s, we can single out the following ultimate, indivisible elements: (1) the art.-aud. element which is connected with the action of the tip or front part of the tongue (the art. element) and the acoustic effect corresponding to this position (the acoustic element); (2) the art.-aud. element of constriction (the art. element) which produces the effect of hissing or friction (the aud. element); (3) the art.-aud. element of the separation of the vocal cords (the art. element) which produces the expulsion of air without acoustic vibration (the aud. element); and (4) the art.-aud. element resulting from the activities of raising and pressing the soft palate toward the lower and rear opening of the nasal cavities, activities which prevent the production of an acoustic or auditory resonance in the nasal cavities.

The phoneme corresponding to the grapheme ą includes the following constituent elements: (1) the art. element of contracting and rounding the lips, combined with the activity of the back section of the tongue. Together they modify the passive position of the oral cavity to produce : (2) an aud. element of resonance due to this very modification of the oral cavity; (3) an art.-aud. element consisting of the contraction of the vocal cords (the art. element), producing a corresponding acoustic vibration (the aud. element); and (4) the non-participation of the soft palate, i.e., its lax position (the art. element) which allows the expelled air to pass through the nasal cavity and produce a nasal resonance (the aud. element).

Each psychologically indivisible articulatory element we shall call a kineme, and each psychologically indivisible acoustic element, an aconsmeme, whereas the acoustic result corresponding to the articulatory activity we shall call a kinakeme. This is the indivisible and twofold art.-acoustic element, which, Janus-like, has one face turned toward the physiological and motor activities of the speech organs, and the other toward the sphere of acoustics, i.e., the sphere of vibrating air and auditory impressions produced by these vibrations.

The above terms may at first glance seem strange and will undoubtedly startle the reader, and I only use them with reluctance. But I cannot avoid using them, since any newly formulated seientific concept requires its own verbal symbol.

§15. In order to give a historical outline of the art.-aud. aspect of the Polish linguistic system, whose direct predecessors include Slavic and proto-Slavic, and the more remote predecessors, IE and proto-IE, we must first of all establish the art.-aud. make-up of proto-IE and then show how its elements developed in the course of time, by either remaining more or less the same as in their initial period, or by changing their form, i.e., by changing the place or manner of articulation, by splitting into two or more variants, or by disappearing altogether.

§16. The juxtaposition and comparison of the various IE languages which constitute the branches of a once common language enables us to infer that original IE contained three categories of phonemes which differed in syllabicity, the configuration of the oral cavity and the articulatory-acoustic activities of the speech organs. These categories were:

1) pure consonants, i.e., phonemes which are incapable of forming syllables and are, from an articulatory viewpoint, occlusive, with the oral cavity being completely or partially closed. The corresponding auditory element results from the specific manner of closure of the lips or the tongue (the graphic symbols of these phonemes are p, b, t, d, k, g. . . , s. . .);

2) pure vowels, i.e., syllabic and open sounds which are produced when the movements of the lips and some part of the tongue modify the oral cavity, making it into a resonator for the vibrating air expelled from the lungs: a, o, e, u, i. . . ;

3) semiopen sounds, or phonemes, which are intermediate between consonants and vowels with regard to both syllabicity, the behavior of the lips and tongue, and the shape of the oral cavity. These include the phonemes which are represented by the graphic symbols r, l (“liquids”), m, n (nasals), u, i (the semi-vowels , or w, j). All these phonemes may function both as non-syllabics (consonants) or as syllabics (sonants).

§17. The proto-IE consonants may be assumed to have had the following distinctions:

1) five kinakemes differing as to their position in the oral cavity, namely: (1) labial (p, b, ph, bh); (2) dental-alveolar (t, d, th, dh, s); and three kinds of retracted (“back-lingual”) consonants (k, g, kh, gh): (3) fronted (palatal), (4) medial, and (5) back (“deep-back” velar);

a sixth medial (mide-tongue) position can be assumed for the nonsyllabic i;

2) the distinction between a kinakeme of friction (s) and that of occlusion followed by an explosive release (t, d, and the other consonants);

3) two distinctions in the articulation of the occlusive consonants: (1) one between the kinakeme of voicing (b, d, g₁, g₂, g₃) and voicelessness (p, t, k₁, k₂, k₃), and (2) another between aspiration (ph, th, kh . . . , bh, dh, gh . . .) and lack of aspiration (p, t, k . . . , b, d, g...);

4) In their consonantal, or nonsyllabic function, the half-open phonemes m, n were distinguished through the kinakeme of a lowered soft palate with the corresponding acousmeme of nasal resonance, as opposed to the kinakeme of a raised soft palate with the corresponding acousmeme of the lack of nasal resonance characteristic of the other consonants;

5) Proto-IE contained only one autonomous, independent, separate phoneme of friction, the asyllabic (voiceless) s. Its voiced counterpart z could have occurred only before a voiced consonant (e.g., zb, zd, zg). The other spirants, the labial v (w) and the palatal j were only a special case of the nonsyllabic function of the vowels u and i.

6) Proto-IE lacked consonantal diphthongs (affricates) of the type which are represented in Polish spelling by the graphemes or groups of graphemes c, dz, cz, dż, ć, dź; in Polish these arose historically from back or front phonemes which became centralized or “softened” under the influence of the adjacent palatal phonemes i, e . . ., j. . . .

7) Proto-IE also lacked the palatal consonants ć, dź, ś, ź, ń . . . , p’, m’, w’ ..., k’, g’, ch’, which are found in modern Polish. Palatal j was the vowel i in its nonsyllabic function, and not an independent consonant.

The proto-IE vowels, or open syllabic phonemes are assumed to have carried:

1) the distinction between simple vowels (a, o, e, u, i) and vocalic diphthongs (au, ai, ou, oi, eu, ei ...);

2) the distinction between short (a, o, e, u, i ...) and long vowels (ā, ō, ē, ū, ī);

3) three different positions of narrowing of the oral cavity: labial, combined with a back articulation (o, u), back (a, ə), and palatal (e, i);

4) two degrees of narrowing: medium and extreme, pertaining to the labial vowels (o, u), and the palatal (e, i), and back vowels (a - ə [narrow a]);

5) supplementary back narrowing of the labial vowels (o, u) and the absence of labial narrowing of the palatal vowels (in other words, the absence of the kind of phonemes known to modern German and French, where they are rendered by the graphic symbols ö, ü, eu and eu, u respectively).

6) the absence of nasal vowels.

The sonants or syllabic “liquids” (r, l) and nasals (n, m) may be assumed to have been distinguished by:

1) palatal narrowing (the “soft” syllabic r’, l’, n’, m’) and its absence (“hard” syllabic r, l, n, m);

2) by shortness (short syllabic r, l, n, m of both the “soft” and “hard” varieties) and length (also of both types).

§18. According to established custom, the simplest art.-aud. elements which receive graphic representation are not the kinemes, acousmemes, or kinakemes, but the phonemes. In adhering to this custom, we posit for proto-IE the following system of phonemes, i.e., of semasiologized and morphologized units which are independent of the influence of their environment, which constitute the point of departure for the history of the Polish phonetic system:

The short and long sonants l, r, n, m can be divided into those without (l°, r°, n°, m°) and those with a palatal kinakeme (lⁱ, rⁱ, nⁱ, mⁱ).

From the standpoint of the proto-IE system, as well as from a historical viewpoint, we must posit, along with the individual phonemes, the following dipthongs and combinations:

These were pronounced with a different intonation, that is, showing different nuances and yielding historically different results. Thus it can be assumed that the syllabic and nonsyllabic phonemes had, in addition to quantity, a changeable accent, i.e., different degrees of stress and intonation.

The history of an art.-aud. system includes, finally, the history of the structure of syllables and whole words, as well as the history of the relation between the art.-aud. structure of morphemes and the division of words into syllables, i.e., the difference between tautosyllabism (when the final phonemes of a given morpheme are a part of the same syllable) and heterosyllabism (when the final syllabic phoneme of a given morpheme marks the end of the syllable and the consonant or consonantal cluster following the vowel marks the beginning of the next syllable).

§19. Let us consider now from a historical viewpoint what happened to the art.-aud. elements (kinemes, acousmemes, kinakemes) of proto-IE when they passed into proto-Slavic and the subsequent Polish linguistic system.

First it should be noted that some semasiologized and morphologized art.-aud. elements (kinemes, acousmemes, kinakemes) making up the phonemes may disappear from the system of subsequent generations. This may occur when the environment and position in the word group cause the disappearance of the phonemes containing these elements, or when entire words or morphemes containing the phonemes and their elements disappear.

If, however, the minimal art.-aud. elements do not disappear, but are, on the contrary, reproduced together with the phonemes that contain them, we are faced either with straightforward linguistic conservatism, or with transformation and a new arrangement of the elements.

§20. The following art.-aud. distinctions have been preserved without change from proto-IE to the present:

1) There is not a single case of change of nonsyllabics or consonants into syllabics or vowels, or conversely, of vowels into consonants. This is true not only of the first two clearly differentiated categories of phonemes, namely the pure consonants and vowels (to, wozu, wóz, piek-, widzi, ducha. . .), but also of the third category, the so-called sonants (which may function both as nonsyllabics and syllabics). If any of these transitional phonemes was nonsyllabic (consonantal) in proto-IE, it remained nonsyllabic also at a later period (ma-t-ka, nos, dymu, syna, raz, łoże. . .). If they were syllabic, however, they could have maintained their syllabic function or lose it, as is the case of the Polish phonemes l, ł, r, rz, n, m.... In Polish the syllabic function is performed either by a full vowel that has replaced the entire sonorant phoneme (cf. the nasal vowels in piąć, na-pię-cie, czą-ć, żąć...), or by the syllabic element of the original sonant which is otherwise preserved in the adjacent consonant (kark, targ, smark, mrug-a, pierści-eń, na-parst-ek, wilk, pełn-y, długi).

§21. 2) If we turn to the consonantal and sonant phonemes in prevocalic position, we observe that they preserved the distinction between the kineme of a lowered soft palate (with a corresponding acousmeme of nasal resonance) and the kinakeme of a raised soft palate (with the corresponding acousmeme of the lack of nasal resonance). The kineme of a lowered soft palate has been preserved in morphemes which in Indo-European had a consonantal, nonsyllabic m or n before a vowel of the same morpheme or in morpheme: e.g., ma—t—k—a, morz-e, nos, nowy, mięs-o, nie. . . ; dom-, dyrn-, syn-, jun-.. . . On the other hand, there is not a single morpheme in which the kineme of a raised soft palate was historically replaced by a lowered soft palate, or where nasal resonance replaced its absence. The IE phonemes of the type b, p. . ., d, t. . ., g, k. . ., s. . . , r, l. . . , a, o, e, i, u. . . have remained nonnasal until this day.

The nasality of the nonsyllabic m, n has been absorbed into the art.-aud. composition of the preceding vowel of the same syllable, giving rise to proto-Sl. nasal vowels. In such a manner, the IE syllabic nasal sonants have in most cases yielded a proto-Sl. nasal vowel ç. The nasal resonance disappeared only in those cases where the syllabic m was replaced in pre-Sl. by a short u, and in proto-Sl. by a short y. The reflexes of proto-Sl. nasal vowels are in Polish almost always nasal vowels or their substitutes, sequences of vowel plus nasal consonant: wąs-, wącha, kąsa, gęś, męski, węch, mięso, grzężnie, więzy ..., wiązać, pląsa..., ząb-, dąb-, zęb-, dęb..., sąd, wątek. . . , męty, tędy. . . , sądzić, mącić. . . , sędzia, pięć. . . , piąty. . . ; ręka rąk. . . , cięgi, ciągnąć. . . , spiąć, spięcie. . . , -cząć -częcie. . . .

§22. 3) The so-called liquid phonemes which, like the nasals, could function in proto-IE as either nonsyllabics or syllabics, were even more resistant to historical change. As we have seen the nasality of the nasal phonemes could have been lost historically in some environments, whereas the phonemes r and l (which were either nonsyllabic or syllabic in the IE period), though they could mix with and replace each other, have been preserved to this day as a separate pair. We can thus formulate the stability of a “liquid” / / “nonliquid” opposition: on the one hand there is r and ł, 1, as in ra, rok, ruch, ryba, rudy, rydz, rdza. . . , łamać, łuna, łoże, łysy, łeb. . . , leży, lice. . . , prosi, grodzi. . . , słoma, głowa. . . , lie. . . , Ignie, łiy; and on the other hand, all the other morphologized and semasiologized phonemes.

It should, however, be noted that in Polish the liquid phonemes have been subject to some constraints: in certain positions they depended on the following or preceding consonant, becoming either voiced (rdza, łgać, łże, lży, lgnie, łba. . .) or voiceless (rtęć, łkać, lśni. . . , wiatr, zmysł, myśl. . ., jabłko. . .).

But a similar dependence and weakening also affected the nasal consonants (mgła, mgli, mknie, mchu, mści. . . , pasm, drachm, widm, piosnka, pieśń, waśń. . .).

The phoneme r underwent palatalization (“softening”) both in the proto-Sl. period when it became r before j (burza, morze, stwarza. . .), and later, in the period of a separate Polish language, when it “softened” before the palatal vowels i, e, ç (parzy, dworzec, rzeka, rzecze, rzçsaג rzqdג grzebie, grzçda, brzeg, drzewo. . . , przy, trzy, krzywy, prze, trzebaג krzemien. . .). This ŕ has, in turn, become rz, i.e., a consonant proper which was at first a diphthong (an “affricate”), and then a spirant.

The phoneme l which was originally fronted or “hard” (ł), has in the pronunciation of most Poles become a nonsyllabic, bilabial vowel (w or nonsyllabic u).

§23. 4) Among the consonantal phonemes, the kinakeme of labiality was transmitted without change from IE to contemporary Polish. The labiality of the phonemes p, ph, b, bh has been preserved in proto-Sl. and Polish: po, piana, ból, baśń, być. This psychologically autonomous noise-producing action of the lips has also been preserved in the phonemes v (w) and m in their nonsyllabic, consonantal function: wóz, widzi. . . , morze, masło, mięso. . . .

Thus the consonants have preserved without change the difference, or we should rather say, opposition between the kinakeme of labiality and any lingual kinakeme. The history of Polish offers no instance of interchange between labial and lingual consonants. Such an interchange takes place, however, in the case of fricatives or spirants where ƒ may be replaced by ch, chw or vice versa; but these are special cases which do not affect the general regularity of the historical continuity of linguistic material.

§24. 5) The kinakeme of fronting (przedniojęzykowość) which characterized some IE phonemes as components of morphemes, was also transmitted via proto-Sl. to contemporary Polish: to, stoi, ty, dać, dom, widzi, dym, broda, wdowa. . . . The number of fronted elements was, furthermore, increased in proto-Sl. when some back phonemes acquired a fronted point of articulation, either spontaneously, i.e., owing to the peculiar make-up of these phonemes, or combinatorily, i.e., owing to their phonetic environment (s, z instead of IE k₁, g₁, g₁h; cz, c, ż, dz instead of IE k₂, k₃, g₂, g₃; cf. §§28, 29). The series of fronted consonants (przedniojęzykowość) suffered only a minor loss when IE s changed under certain historical conditions to proto-Sl. x (ch).

§25. The following IE elements and distinctions of IE were preserved without change in proto-Sl. and in pre-Polish: 1) the nonpalatal (“hard”) and palatal (“soft”) varieties of syllabic r and l. The transition from proto-Sl. and pre-Polish to a separate Polish period was marked by the breakdown of the simultaneous combination of the consonantal r or l elements with the reduced (nonpalatal or palatal) vowel elements into two successive, nonsimultaneous elements, i.e., into a vowel and consonant or consonant and vowel; e.g., garb, kark, smark. . . , mrug-a, mruk. . . , czatn-, żarn-, martw-,—parst—, wart-, śmier-ć, pierści-eń, wierc-i. . . , pierw-, wierzb-, wierzg-, pierzch-, wierzch-. . . ; wilk-, milk-, wilg-. . . , pełn-, bełt-, pełz-. . . , czołg-, czółn-, żółt-. . . , dłub-, dług-, tłust-. . . .

§26. 2) The quantitative difference between short and long vowels and syllables. The IE short vowels (a, ə, o, e, u, i) are continued in proto-Sl. as short (o, e, y, i), and the IE (or pre-Sl.) long vowels (a, o, e, u, i) and diphthongs (au, ou, eu, ai, oi, ei. . . an, on, en.. . ar, or, er, al, ol, el. . .) are continued in proto-Sl. and Common Slavic as long vowels, or as half-open long sonants (a, e, y, i; u, e, i; ą, ę, and syllabic long r and l with a palatal or nonpalatal vocalic element).

The transition from proto-Sl. and Cm. Sl. to later stages was characterized by the merger of the original long and short vowels. Differences in accent and intonation brought about the shortening of original long vowels, whereas the loss of the short y and i caused the lengthening of certain short vowels.

Pre-Polish may be assumed to have inherited from IE the distinction between originally short or shortened vowels and originally long or lengthened vowels. At a later period, and especially in contemporary Polish, the psychological ability to distinguish the temporal quantity of vowels is lost. Polish vowels are neither short nor long; they are quantitatively neutral.

The original distinction between long and short vowels is partially reflected in Polish as a qualitative distinction in the vocalic configuration of the oral cavity.

§27. 3) IE had a morphologically mobile accent, i.e., an accent which accompanied different syllables of the word and differentiated certain morphemes. This morphologized and semasiologized art.-aud. property was transmitted with some modifications to proto-Sl. and Cm. SL, and, in one way or another, to the later stages of the separate Slavic languages. However, in Polish this property was suppressed and lost, leaving only some traces of its former existence. The Polish accent became fixed and relinquished its morphological role; now its function is purely syntactic, in that it serves to distinguish different words.

Such a transformation in the art.-aud. structure of the language could hardly have occurred without foreign influence. In this respect we must regard Polish as an ethnically mixed language. In Kashubian dialects and elsewhere in the Lekhitic linguistic community, we find phenomena of morphologized accentual mobility.

§28. In passing from IE and the oldest reconstructible stage of Polish to its subsequent stages, certain art.-aud. features underwent qualitative or quantitative change. Some elements of the linguistic system disappeared without a trace, and others were replaced. These shifts in the mutual relation of art.-aud. elements also changed their numerical ratio. While the number of elements of one category increased, the number of elements of another category decreased or disappeared completely. On the one hand, some elements disappeared in the neighborhood of other elements as, for example, certain consonants before other consonants (e.g., the simplification of IE vr into Slavic r), and at the end of words or syllables; on the other hand, certain distinctions between phonemes were lost.

In the transition from IE to proto-Sl. there disappeared:

1) the distinction between aspirated and unaspirated stops. On the basis of the Slavic languages alone, no one could have thought that the historical source of the phonemes b (in ból), d (in dar, dom, siada), z (in ząb, zna), going back to the IE unaspirated voiced stops b, d, g,, differs from that of the phonemes b, d, z (in by-ć, ba-śń, dym, wdowa, brod-a. . . , woz-u, zim-a, zorz-a), going back to the IE aspirated stops bh, dh, gh.

2) the distinction between three series of IE velar stops k₁, g₁, g₁h; k₂, g₂, g₂h; k₃, g₃, g₃h. The first series moved into the class of fronted fricatives s, z (słoma, słowo, sto, sroka, pros-; for examples with z, cf. 1 above).

The ratio of the original semasiologized and morphologized phoneme s was, at the same time increased by s from k₁, whereas the change of g₁, g₁h to z gave rise to a new independent phoneme, the fronted voiced fricative z. The remaining two series k₂, g₂, g₂h and k₃, g₃, g₃h remained in the class of velars, but without any difference in position [localization], so that proto-Sl. has only one psychologically defined velar (tylnojęzykowe) position k, g (and only one g, because of the loss of distinction between aspirated and unaspirated consonants); e.g., the suffix -k-; krzyw-, krew, ko-go, wilk-..., góra, śniegu, gawiedź. . . , gorzeć, goni. . . .

3) the distinction between an IE narrow a (i.e., ə) and a short open a: ə merged with a which, in turn, merged with o; proto-Sl. short o thus continues all three IE phonemes (stoi. . . , ostry, oś, sol-..., ośm-, owca...).

4) The kinakeme of labiality of long u and short u was lost (delabialization); these vowels were replaced by proto-Sl. long y and short y (dym, być, syn, wymię; krew, mech, sen, łeb. . .).

The labiality of long o was likewise lost, merging with a long a, whose ratio increased: (dać, dar, dwa, nagi. . .).

The ratio of short o increased, however, at the expense of short a, which had merged with it. (Examples under 3).

5) The IE vocalic diphthongs became long vowels (§26).

Thanks to all these changes, which marked the transition from IE to proto-Sl., the latter acquired new phonemes and series of phonemes; among others:

1) more fronted consonants; more positions and combinations of fronted consonants;

2) more spirants (z, ż. . ., ch. . .);

3) a series of palatal [średniejęzykowe] consonants which appeared as a result of the palatalization [ześredniojęzykowienie] of back consonants in a palatal environment (cz, z, sz; c, dz) or as a result of the merger of fronted and labial consonants with j (tj, dj, sj, zj; pj, bj, vj, mj, rj, lj, nj);

5) consonantal diphthongs cz, c, dz, dż. . . ;

6) the vowel y (long and short);

7) nasal vowels (§26).

§29. Some of these art.-aud. innovations were the result of combinatory, i.e., environmental influence which split an originally single phoneme into two or more phonemes. These include:

1) the appearance in some environments of proto-Sl. x (ch) from IE s, which in other environments remained unchanged (e.g., such, duch-, mech, wierzch, pych-a, chory. . . , the locative plural ending -ch. .., side by side with s in such-, syn, bos-, nos-, sad-...).

2) the split of velar consonants into velars and palatal spirants or consonantal diphthongs: k vs. cz (piek-piecze, rok roczy, wilk - wilczyca. . ., mog– - moż, śnieg– - snież, bog– - boż—. . ., such– - susz, strach– - strasz, much– - musz. . .).

3) the split of the combinations of dental and labial consonants with the vowel i (ti, di, si, zi, ni, ri, li, pi, bi, mi, vi) into the same combinations in syllabic position (i.e., before consonants and in word final position) and into palatal consonants before t’, d’ (mod. Polish c, dz), sz, ż, n’, r’, l’, pl’, bl’, ml’, vl’, or p’, b’, m’, v’ before a vowel (e.g., mod. Polish płaći - płac, chodzi - chodz–, nośi - nosz–, wozi–, woż, gani - gań–, tworzy - tworz–, chwali - chwal–, topi - top’–, lub’i - lub’–, tłum’i - tłum’–, łow’i łow’–. . .).

The proto-Sl. phoneme t’ (from tj) also developed from the group kt (kt, gt) before i (mod. Polish: piec, noc, tłuc, wlec . . . lec, moc, móc, strzec. . .).

4) in the pre-Sl. combinations jo, ju, later jo, jy, the vowels became fronted under the influence of the preceding palatal j, yielding the combinations je, ji . . . (morze, pole. . . , kraj, koń, mąż, płacz. . .). The ratio of the vowels e, i was thus increased at the expense of o, y.

§30. Passing from IE to proto-Sl., certain phonemes and series of phonemes merged, whereas others, previously autonomous, split into two or more phonemes. But here we must distinguish merger and splitting of phonemes in immediate contact with other phonemes from the same kind of processes which take place in the psychologically existing phonetic system of a language.

The following mergers took place in immediate contact:

—combinations of consonants with j (i.e., kj, gj, chj, . . . tj, dj, nj, lj, rj, . . . sj, zj. . .) (§28);

—one-syllable combinations of vowels with nasal consonants which changed into nasal vowels (§26).

The following splits took place under the same conditions:

—palatal velar consonants changed into consonantal diphthongs (k’ - cz, c; g’ - dz) (§28);

—labial consonants, palatalized through merger with the consonant j, changed into the groups pl’, bl’, ml’, vl’ (§29).

§31. The following phonemes merged in the phonetic system:

—aspirated with nonaspirated (§28);

—velar stops of the series k₁ g₁ with the fronted spirants s,z (§28);

—the vowel a with the vowel o;

—the combinations jy (ju), jo with ji, je;

—vocalic diphthongs with long vowels.

The following phonemes and phonemic series split in the phone tic system:

—the consonant s into s and ch (§29);

—the velar consonants (resulting from the merger of k₂, g₂, g₂h with k₃, g₃, g₃h) into velar k, g and fronted consonants cz, ż, which later became c, dz (§29);

—the diphthong oi into long e and long i.

§32. Considering all the above-mentioned changes which marked the transition from IE to proto-Sl., we may represent the art.-aud. system of proto-Sl. by the following table of phonemes and historically indivisible groups or combinations of phonemes:

This presumed phonetic system of proto-Sl. and Common Slavic represents, from the viewpoint of Slavic comparative phonetics and Polish historical phonetics, a hypothetical cross section of the art.-aud. representations which had formed over a certain period of time.

§33. We may now examine the subsequent fate of the individual components of this system and of the system as a whole as it evolved into an autonomous Polish linguistic system.

Certain of the art.-aud. components and distinctions of proto-Sl. have remained the same up to the present, whereas others have undergone either qualitative or quantitative changes. In examining the invariance and changes, we shall first consider those that were shared by all branches of the Slavic linguistic world, by all “Slavic languages,” second, those which Polish shared with some other Slavic languages, and third, those which are characteristic only of Polish in contradistinction to the other Slavic languages.

Before analyzing the common and individual Slavic changes, let us first single out those elements, groups, and series of elements which underwent neither common nor separate change. To some extent we will have to repeat what was said above (§§20—24) about the preservation of the IE elements and distinctions in the present state of the (Polish) language.

Thus, in the evolution of proto-Sl. and Common Slavic to later stages, including Polish, the following elements remained unaltered:

—labiality of the consonants p, b,v, m;

—frontedness of the consonants t, d, s, z, c, sz, ż, cz, r, l, n;

—backness of the consonants k, g, x (ch) (insofar as these did not become palatalized (“soft”);

—palatality of the consonant j;

—occlusive articulation of the consonants p, t, k, b, d, g, m, n;

—the specific oral articulation of s, z, c, r, l, j;

—closure of the nasal cavity and the nonnasal acoustic character of all proto-Sl. nonnasal consonants and vowels, and openness of the nasal cavity and the nasal acoustic character of the nasal consonants m, n.

By contrast, there is not a single vowel that has not changed in some way or other.

§34. Among the common changes and invariance of the proto-Sl. state in its later phases, including Polish, we may list:

—the gradual change of the consonants going back to the original groups sj, zj, which finally became sz ż (noszę, wożę);

—the gradual change of palatalized (“softened”) consonants going back to k, g, ch, which first became sz, ż, cz and then c, dz;

—the disappearance of the autonomous, psychologically distinct vowel y, which may be regarded as the last step of a pre-Sl. process. Originally the long u had become delabialized into a back vowel y, which was psychologically distinct from other vowels; this stage of development is still reflected in OCS records. Later, however, this vowel lost its independent status and either merged completely with the reflex of the vowel i, or became only a variant of that vowel, which varied according to the preceding consonants.

§35. Another change shared by the entire Slavic world was the general phonetic shortening and weakening of words, which was realized in two ways:

1) the original proto-Sl. long vowels a, i, e, y, u, ą, ę became shortened (this stage is reflected in the oldest OCS records);

2) the short vowels y (from IE u) and i disappeared or became phonetic zeroes under certain conditions, particularly in word final position.

The latter historical-phonetic process had the following consequences:

1) the shortening of a large number of words by one or more syllables;

2) the appearance of new closed syllables, i.e., syllables ending with a consonant;

3) the appearance of new consonantal groups;

4) the creation of new conditions of phonetic accommodation (phonetic adaptation) between adjacent words, i.e., at the juncture of the end of a word and the beginning of a following word;

5) the appearance of new syllabic “liquid” sonants r, l and nasal sonants n, m (which are alien to Polish, but which are found in Czech and in some other Slavic languages), or of voiceless, nonsyllabic “liquid” and nasal consonants, such as are found in the Polish wiatr, jabłko, myśl, piosnka, pieśń, pasm. . . ;

6) lengthening and “compensatory” strengthening of vowels in syllables preceding those syllables in which the short vowels y and i became phonetic zeroes, e.g., bóg, wóz. .., śniég, chléb.. ., sen, pies, wieś... from older *bogy, *vozy... (with short y), *snegy, *chleby... (with long e and short y), *syny (with the first y long and the second short), *visi, *pisy (with short i and y).

§36. Among the changes, as well as manifestations of conservatism which Polish shares with some, but not all Slavic languages are:

—the reflexes of the consonant clusters kv, gv in combination with front vowels among the Northwestern Slavs (including the Poles) (kwiat, kwilić . . . gwiazda. .). The Southern and Eastern Slavs have replaced these groups by cv, zv;

—a similar territorial division corresponds to the treatment of the proto-Sl. consonant groups dl (dł, dl), tl (tł, tl). The Northwestern Slavs (including the Poles) have retained these groups without change (mydło. . ., padła. . ., plotła....), whereas the other Slavs simplified them to l (ł, l);

—the proto-Sl. nasal vowels which have been preserved, albeit with some modifications, only in Polish, were also used by those Slavs whose speech provided the basis for the OSC graphic-visual language or literature, i.e., by the early Bulgarian and Macedonian Slavs who occupied the Eastern part of the Balkan peninsula. Only the OCS literary documents reflect the proto-Sl. stage in which the ą (continuing the pre-Sl. group on) and the ę (continuing the pre-Sl. group en or the syllabic sonant n) were differentiated. Among the other Slavs, including the linguistic descendants of the early Bulgarians and Macedonians, we find only traces of nasality in the reflexes of the proto-Sl. nasal vowels. In Polish the two nasal vowels have merged; ę and ą continue both proto-Sl. ę and ą. But if these vowels are preceded within the same morpheme by a consonant, the consonant itself provides a clue to the historical origin of the nasal vowel. If the preceding consonant is non palatal in modern Polish, and was also hard in the past, the nasal vowel in question stems historically from a proto-Sl. ą (e.g., tępy, będzie, bądź, tędy, stąd, sąd, sędzia; ząb, zęby, ręka, rąk, męka, mąka, wązki, zwęzić, gęba, gąbka...). If, on the other hand, the consonant is palatal or can be traced back historically to a palatalized or “softened” consonant, the nasal vowel in question goes back to a proto-Sl. ę (pięć, piąty, więzić, wiązać, mięso, miąższ, cięży, ciąża, dzięeioł, dziąsło, sięgać, siąkać, ziębi, ziąb, lęka się, zląkl się . . . , część, cząstka, żęty, żąć, rzędy, rząd, przędzie, prząść...). Of course, after the consonant j it is impossible to determine the origin of the nasal vowel, since j as a nonsyllabic i has ever since IE been an autonomous palatal consonant not subject to palatalization; in proto-Sl. it could be followed by either ę or ą (jęczeć, jąkać się. . .). (The word język has its own particular history);

—the proto-Sl. palatal phonemes t’ (from tj and kt) and d’ had the same development in Polish as in Slovak (traeę, świeca. . . , chodzę, miedza. . .), whereas Czech has c from tj, like Polish and Slovak, but z, not dz from dj;

—Polish shares only with the Lusatian languages its treatment of the pre-Sl. tautosyllabic groups or, er, ol, el (broda, grodu, proch, krowa, prosię, droga, sroka, mrowie. . . , przodu, drzewo, brzeg, źrebię. . . , głowa, kłos, płochy, błoto, tłok, dłóto, słoma, złoto, młot. . . , mleko, plon. . . , człon, żłób. . .). The other Slavic languages have different reflexes of these diphthongs.

§37. For Cm. Sl., as opposed to IE, we have established two types of palatalization (“softening”) of consonants: (1) two successive palatalizations of velar consonants caused by adjacent palatal phonemes; (2) palatalization of dentals and labials which have merged with a following j (nonsyllabic i ) (§§28-30). Later, North Slavic, including Polish, acquired a new series of palatal consonants as a result of the palatalization of dental and labial consonants before front vowels and front syllabic sonants: t’, d’, s’, z’, r’, l’, n’, p’, b’, v’, m’. In some of these consonants, the palatal element was subseąuently strengthened; conseąuently the palatals t’, d’, s’, z’ became ć, dź, ś, ź with a prominent element of friction; r’ became rz; and the phonemes p’, b’, v’, m’ yielded the diphthongs p’j, b’j, v’j, m’j with a characteristic fricative j (e.g., ci, ciało, ciężar, ciemny, ćma, cierń. . . , dziw, dzieci, dzierżeć. . . , siła, siano, siodło. . . , zima, zięć, ziarno. . . , darzy, brzytwa, rzec, rzeka, rzadki ..., lizać, lewy, lęk, len. . ., ni, niesie, dzień. . ., pisze, piechota, piasek, pięć, pies, pierwszy. . ., bić, bierze, biały. . ., widzi, wieś, wierci, wilk. . ., miły, mięso, milknąć....

Of all Slavic languages, Polish and the Northwestern Slavic languages in general, have felt most strongly the influence of consonants upon the quality of vowels and sonants. Of paramount importance here is not only the distinction between palatal (“soft”) and non palatal (“hard”) consonants, but also that between fronted (dental) and nonfronted (nondental) consonants. It is sufficient to point out the difference in such words as grzebać, trzepać, piekę, niech, ulegać, niemy, mleko, pierwszy, wierzch, cierpnąć. . . , w lecie, w lesie, w mieście, na czele, wiedzie, niesie, plenić, twierdza, smierć pierścień. . ., but lato, las, miasto, czoło, wiodę, niosę, plon, twardy, martwy, naparstek...; also wilk, wilgoć, milknąć. . . , but pełny, zółty, tłusty, długi. . . .

The Historical Phonetic Changes Which

Took Place in the Period, or rather, in the

Various Periods of a Separate Polish Language

§38. In addition to the changes touched upon in §§34-36, we shall now reiterate or state for the first time, the following historical-phonetic processes:

1) the proto-Sl. fronted (palatalized or “softened”) consonants t’, d’ (from tj, jt, dj) became c, dz (§36);

2) the pre-Sl. groups or, er, ol, el (in proto-Sl., long syllabic sonants (§36)) underwent metathesis with the restoration of the pre-Sl. shortness of their vowels: ro (ró) (broda, gród), rze, rzo, rzó, re, ro, ró) (brzeg, trzewia, przodu, przód, trzoda, trzód, średni, źrebię, środa, śród), ło (łó) (głowa, głów, błoto, słoma, młot, głód. . .), le (lo) (mleko, pion. . .) (§37);

3) like all other Slavic languages, Polish underwent a general temporal shortening of words, and in particular the elimination in many cases of the short y (from pre-Sl. u) and i, and shortening of words through the loss of whole syllables, with all the far-reaching consequences which this development had for the general phonetic character of the language. Where the short vowels y, i were not eliminated, they changed historically to e: sen, mech, łeb. . . , pies, dzień, len. . . (§35);

4) in connection with the last change, Polish acquired a whole series of phonetic zeroes of vocalic origin, whereas the phonetic zeroes of proto-Sl. were mostly of IE consonantal origin (§28). In the course of the shift of the art.-aud. representations from proto-Sl. to Polish, and in the independent history of Polish, there also was, of course, a loss of some inherited consonants, so that there also appeared a phonetic zero of consonantal origin. As for the vowels which became phonetic zeroes in Polish, it should be noted that in addition to the articulatory-auditory weak and short y, i (which disappeared primarily in word final position, but also within the word), other vowels (which were either short in proto-Sl. such as o, e. . ., or long, such as a, y, i. . .) could likewise disappear if they were weakly morphologized and semasiologized. Thus, for example, the infinitive desinence —ć is historically derived from ci (with a long proto-Sl. i) and has a phonetic zero in place of the vowel i;

5) the secondarily lengthened vowels a, o, e (§§26, 35) were narrowed, creating a new historical category of narrow (pochylone) vowels: à (medial between a and o, with a tendency toward o or a), ó (medial between o and u, with a tendency toward u), and è (medial between e and i, with a tendency toward i or e). The distinction between á and a, ó and u, é and i was eventually lost;

6) a specifically Polish historical-phonetic phenomenon is the change in certain consonantal groups and combinations of ś ź, ć, dź, ść, źdź. . . into j: wiejski, wejrzy, ojca, ogrójca, Zamojski (cf. Zamość), Ujejski (cf. Ujazd). . . ;

7) a phenomenon which is also known in other languages, but which is, historically speaking, a particularly Polish phonetic development, is the change of the vowel i (from proto-Sl. long i and y) into e before a following r: sierota, ser, Serock, Zgierz, ubiera, umiera, rozpiera... but rozcina, zaczyna, zżyma. . . ; also ćwierć, śmierć, pierwej, wierzch, wierzba from older ćwirć, śmirć, pirwej, wirzch, wirzba. . . ;

8) in the transition from IE to proto-Sl., the syllabic nasal sonants n, m (which were distinguished as palatal and nonpalatal and as short and long) disappeared, yielding a nasal vowel ę or a pre-Sl. short u and a Slavic short y (§21). The syllabic “liquid” sonants r, l (which had the same distinctions as the sonants m, n) were replaced by various combinations of vowels with the consonants r (r, rz) and l (ł, l): ar, ru, ir (later er: er, erz); il, eł, oł, łu. . . (§37). The original syllabic “liquid” sonants r, l thus contributed to an increase both of corresponding nonsyllabic consonants, and of syllabic vowels. We have here to do with one of the cases illustrating the historical loss of intermediate, transitional categories in favor of opposite, clearly defined categories.

A similar breakdown of the proto-Sl. syllabic sonants r, l into combinations of vowels and consonants is found in the Northwest and East Slavic territories, as opposed to those of the Southwest (beginning with Czech and especially with Slovak), which have, to a minor or lesser extent and in one way or another, preserved the original syllabic sonants.

Table of Autonomous Independent

Phonemes of the Polish Linguistic System

§39. This table cannot be considered absolutely precise and exhaustive, for it is by its very nature in flux and variable, depending upon dialectal influences and the usage of individual speakers of Polish.

§40. If we compare the table of the Polish phonemes with that of the proto-Sl. phonemes, we arrive at the following conclusions:

1) the phoneme ƒ, which was lacking in IE and in proto-Sl., appeared in the Polish system mostly as an art.-aud. component of morphemes borrowed from other languages; in addition, it is characteristic of some onomatopoeic words. Only in rare cases is it the historical continuation of phonemes and phonemic combinations in morphemes or in morphemic groups which are of IE and proto-Sl. origin (e.g., ufać. . .);

NOTES: 1. The phonemes associated with the graphic combination rz are from an acoustic-auditory viewpoint ž or š; only a relatively small part of the Polish linguistic territory has a separate phoneme rz, which is a kind of consonantal diphthong, i.e., a combination of a weak r with ž or š(ř).

2. The digraph ch in this table stands for either ch or h of the conventional Polish orthography.

3. i is a symbol for the conventional Polish y (być, ty, syn) or i when not followed by a letter denoting another vowel (bić, ci, siny . . .).

4. I have used the letter a to denote both the representation of the vowel ą (nasal o) and the combination an.

2) the nasal phoneme ą occurs only in words and morphemes which are of foreign origin: ansa, awans, kwadrans, kontredans, rewansz, blamanże. Compare also Francja, bankiet, lampa, trampolina, gangrena. . . vs. ponsowy, bronz, bonza, koncha. . ., pensum, pensja, sensat. . . ;

3) in the phonemic table of proto-Sl. it was already possible, without further reservations, to group m, n with the consonantal phonemes, for, unlike IE, proto-Sl. lacked syllabic nasal sonants proper. The phonemes r, l, however, had to be listed under both the consonantal and the vocalic phonemes. In Polish the phonemes r and l (l, ł) have definitively switched to the series of nonsyllabic consonants. The once palatal (“soft”) r has likewise dropped out (in acoustic-auditory terms) of the class of semi-vowels capable of performing a syllabic function; it became either a purely nonsyllabic consonant rz, or it underwent further reduction to r when the elements ž or š were, in some cases, lost (środa, źrebak, źródło...). The hard ł, on the other hand, has lost its dental consonantal character in the speech of most Poles, who pronounce it as a bilabial, nonsyllabic vowel u (§22).

§41. 4) In the table of proto-Sl. phonemes (§32), the phonemes c, dz, sz, ż, cz are listed as palatal (“soft”), for according to their historical origin they really contained a palatal element; their Polish reflexes have lost this palatal element and have become nonpalatal (“hard”);

5) the proto-Sl. palatals t’, d’ have been replaced in the same positions in historically related morphemes by the Polish nonpalatal phonemes c, dz (§§36, 38); the palatal r’, which had greatly increased in frequency owing to the palatalization of dental and labial consonants before front vowels, gave way at a much later period to the phoneme rz (§37);

6) owing to this very palatalization of dental and labial consonants before front vowels and to the further development of these palatal consonants, Polish acquired a new category of palatals, ś, ź, ć, dź, which did not exist in proto-Sl. The series p’, b’, m’, f’, w’ belongs, in fact, to the same category and is partially matched (in a far smaller number of cases) by the proto-Sl. series pl’, bl’, ml’, vl’. The latter were actually groups of two consonants, but we treat them as indivisible units on the basis of comparison with other phonemes: in proto-Sl. they were variants of pi, bi, mi, vi, and their subsequent Slavic correspondences (e.g., the older pj, bj, mj, vj or younger (Pol.) p’, b’, m’, w’) are single phonemes. These indivisible units can, of course, in turn be broken down into psychologically no longer divisible elements such as kinemes, acousmemes, and kinakemes (§51);

§42. 7) The proto-Sl. phonemic system did not tolerate the palatal (“soft”) consonants k’, g’, ch’, inasmuch as the back phonemes k, g, ch, which became palatalized in their transition from pre-Sl. to proto-Sl. had changed into cz, ż, sz or c, dz, sz, whereas they did not undergo palatalization in the groups ky, gy, chy (where y was long or short). The new vocalic phonemes of a palatal character, which arose when long y merged psychologically with the historical reflex of long i (thus ceasing to be a separate phoneme), and when short y (which merged psychologically with the reflex of short i) became e (if it did not become a phonetic zero), failed to exert a palatalizing (“softening”) effect upon the labial and dental consonants. By compensation, the velar consonants, which are in general more susceptible to palatalization, in turn underwent the effect of palatalization. Thus there appeared in place of the original proto-Sl. combinations ky, gy (with long y) the groups k’i, g’i, and in place of the combinations ky, gy (in which the short y is preserved), the groups k’e, g’e.

The velar spirant ch, which involves acoustically not only backlingual but also glottal activity, did not undergo this new palatalization, and remained a hard consonant. Thus, although we have forms such as kisiel, kipi, boki, ręki. . . , ginie, gil, stogi, nogi. . . , kierz, kieł, bokiem.. . , giez, rogiem, we also have such forms as chytry, chybi, duchy, strzechy. . ., duchem, piachem. . . . Only iterative and durative verbs formed with the suffix -iv-a (-ywa) exhibit a soft ch;, thus we have not only forms such as przepłókiwać, odskakiwać, pobrzękiwać, usługiwać, obełgiwać. . . , but also rozdmuchiwać, rozmachiwać. This is obviously not the result of a historical-phonetic process, but rather the effect of a strong morphological assimilation, or “analogy,” inasmuch as the forms rozdmuchywać, rozmachywać are likewise possible. In this respect Polish differs, for example, from Great Russian which, along with the soft k’, g’, also has a “soft” ch. Before ę, which continues proto-Sl. ą, the nonpalatal (“hard”) k, g have been preserved: kępa, kędy, kęs. . . , gęś, gęgać. . . , rękę, nogę. . ., though dialectally we also have k’, g’: kiempa. . ., gięś. . . .

The number of cases containing “soft” k’, g’ has considerably increased since they appear in many words of foreign origin, or in adopted words: kierat, rakieta, kimono. . . , gienjusz, gieografja, gigant. . . . One should also keep in mind cases of onomatopoeia, such as chichotać. . . .

In Kashubian-speaking territory, this third phase of palatalization produced a new set of velar consonants which became palatal. In Kashubian we have not only examples of the oldest proto-Sl. layer to which we owe cz, ż, sz (piecze, ręczyć, rączka. . . , może, dłużyć, nóżka. . . , straszyć, dusza. . .) and the subsequent proto-Sl. layer to which we owe c, dz (ręce, wielcy, lice, owca. . . , nodze, drodzy. . . §34), but also examples of the youngest—no longer proto-Sl., but rather Kashubian stratum with its ć, dź (such as ćiwać, ciede, taći, dzinie, drodźi instead of kiwać, kiedy, taki, ginie, drogi. . .).

At any rate, velar consonants with a palatal element k’, g’, dz’ are a new phenomenon in Polish which was unknown in proto-Sl.

§43. As a result of the later historical-phonetic processes which marked the transition from proto-Sl. and Common Slavic to pre-Polish, and subsequently to an autonomous Polish language with all its historical varieties, the Polish-speaking community acquired the following phonemes and series of phonemes which were alien to proto-Sl.:

1) the consonants ć, dź, ś, ź;

2) the consonant rz;

3) the consonants p’, b’, m’, w’;

4) the consonants k’, g’, ch’;

5) the vowel i, continuing proto-Sl. i;

6) the vowel e, continuing short y (from pre-Sl. short u) and exerting no palatal effect (except when combined with the velar consonants ky, gy, §42);

7) new short vowels (later neutral, neither long nor short): i (i, y), u, a, e, ą, ę;

8) new long vowels formed through compensatory lengthening (later likewise neutral) o, e, i, u, a;

9) narrowed vowels: á ó é;

10) the vowels ą (from ę), ę (from ą ),and nasal a.

In addition, we have in contemporary Polish some sounds which are not psychologically independent phonemes, but speech-sounds which are due to the impossibility of implementing the phonetic (articulatory) intention:

1) weakened consonants at the end of words and of syllables: -p, -t, -k; -b, -d, -g; -p’, -b’, -m’, -f,’ -w’, -f, -,s, -sz, -ch, -ś, -w, -z, -ż, (-ż, -rz), -ż, -c, -cz, -ć; -dz, -dż, -dź;

2) voiceless -r, -ł, -l, -m, -n, -ń (§35).

The Strengthening of Consonantism
and the Weakening of Vocalism

§44. In the history of Polish, as in the history of many other languages, we can observe an ever greater strengthening of consonantism at the expense of vocalism. This phenomenon expresses itself, among other ways, in the ever greater importance of the palatal element, as well as in the ever growing influence of the consonantal distinctions upon the quality of vowels (e.g., the strong influence of dental consonants as opposed to (the minor influence of) other consonants, §37).

In comparison with the less remote pre-Sl. past, Polish displays a great increase and intensification of palatal activity, whereas in comparison with the more distant past, it displays an increase of dental (front-lingual) activities.

Although the “hard” labial and dental consonants in Polish underwent palatalization (or “softening”) under the influence of following palatal vowels (§37), this seeming increase of the influence of vowels upon consonants has gradually turned in favor of the consonants, which have acquired the distinction between non-palatals and palatals which was originally peculiar to the vowels. The two vocalic phonemes of proto-Sl., long y and i, merged into one i which was variable and dependent on the preceding consonant. The corresponding short vowels merged with e, if they did not disappear and become phonetic zeroes (§35). But their original psychological distinction is to this day reflected in the different preceding consonants, e.g., być but bić, chłopek but chłopiec, dech but dzień, ten but ciemno. ... A similar differentiation of consonants dating back to proto-Sl. and having its origin in a pre-Sl. and IE difference in the following vowels applies only to the original IE velar consonants; cf. on the one hand, ky, gy, chy (with proto-Sl. ch continuing IE s, and long or short y continuing u) and, on the other hand, czi, żi, szi (with short or long i); e.g., Polish kipi, ginie, chytry. . ., kieł, giez.. . ; czysty, żywy. . ., cześć. . . . Further, proto-Sl. ka, ga (with a going back to IE long a or o) vs. cza, ża (with a going back to IE long e); e.g., Polish każe, kara. . . , gani, gasi, gad. . . , but czas, czar. . ., żar, żal, żaden.

Proto-Sl. short e, the long e, ę, ą and the syllabic sonant r with a palatal element have split, either under the influence of the following dental consonant or from other causes; thus short e split into e and o, long e into e and a, ę into ę and ą, a into ą and ę. But whether these vowels were in proto-Sl. o or a, and short or long e, can be deduced only from the preceding consonants; e.g., wozi, nosi, płot, wyboru. . ., wada, maść, dar, tak, sadło, swat, łazi, raz. . ., but wiozę, niosę, plotę, biorę. . . , wiadomo, miasto, dziad, ciało, siadło, świat, lazł, rzadki. The same can be said about nasal vowels (§36).

Another phonemonon pointing to the strengthening of consonantism and weakening of vocalism is the stabilization of the Polish accent and the loss of the psychological distinction of quantity, i.e., the length and shortness of vowels (§27).

In line with the same historical tendency is the decrease of laryngeal activities in favor of activities and distinctions made within the oral cavity (§§28, 31, 27, 18).

Historical Changes in the
Structure of Syllables and Words

§45. The transition from IE and pre-Sl. to proto-Sl. and Common Slavic was marked by a tendency to eliminate narrowed or closed syllables. This general historical-phonetic process consisted of the following specific changes:

1) the elimination of consonants at the end of words and syllables. In other words, tautosyllabic syllables comprising vowels and consonants are continued by syllables ending in vowels; e.g., slovo from slovos, to from tod. . . (§28);

2) the change of vocalic diphthongs into homogeneous long vowels (§26);

3) the change of combinations of vowels with nasal consonants on, om, en, em into nasal vowels ą, ę (§21);

4) the change of the pre-Sl. combination of vowels with “liquid” sonants or, er, ol, el into presumed syllabic long r, l, which later yielded different combinations in various Slavic languages; in Polish ro, re, ło, le (§36).

In the transition from Common Slavic to Polish, there appeared a great variety of closed or narrowed syllables, due mostly to the loss of syllables containing a short y or i.

The Importance of Semasiologization and Morphologization for the Distinction of Phonemes and for the Preservation of their Distinctions

§46. If the distinctions between phonemes or, to be more precise, between the art.-aud. elements (kinemes, acousmemes, and kinakemes) which make up the phonemes did not play a part in distinguishing morphemes (the morphological-semasiological elements which are associated either with a particular meaning, that is, with a group of extralinguistic concepts, or with a particular form such as case, person, etc., in connection, of course, with some suffixes or desinences, including zero suffixes or desinences), or, in other words, if the distinctions between phonemes were not semasiologized and morphologized, they would not be so persistently preserved and would have gradually been lost in the transmission from one generation to another.

In contemporary Polish, as inescapably in all languages, all phonemic distinctions must be semanticized (semasiologized). In mama, as opposed to baba, the semanticized distinction is that of a lowered soft palate with the acousmeme of nasal resonance vs. the kineme of a raised soft palate (that prevents the passage of air to the nasal cavities) with the acousmeme of the absence of nasal resonance. In kosa as opposed to koza, the distinction is that of a kinakeme of the activity of the vocal cords. In tom and tam, kora and kura, bok and byk, there is a semanticized distinction of vocalic arrangements of the oral cavity with corresponding differences in their resonance.

The morphologization of certain art.-aud. distinctions is, on the other hand, only a transitory phenomenon in the history of languages. It is usually brought about by some historical-phonetic split of one phoneme into two or more phonemes. These new varieties of an original single phoneme acąuire the value of psychophonetic alternants which become associated with different forms. In Polish, for example, such psychophonetic alternants or correiatives consist of the distinction between non-palatal consonants and those which are, or were historically palatal; e.g., ryba, kopa, słoma, sowa, sofa, robota, woda, osa, koza, rana, siła, kora, ręka, noga, mucha. . . , but rybie, kopie, słomie, sowie, sofie, robocie, wodzie, osie, kozie, ranie, sile, korze, ręce, nodze, musze. . . ; ręka, suka, paka, noga, mucha. . . , but rączka, suczka, paczka, nóżka, muszka. Vocalic differences are similarly morphologized: głowa, broda, robota, noga, ręka, wstęga. . . but główka, bródka, robótka, nóżka . . . , rączka, wstążka. . . .

Morphologization supports phonemic distinctions more firmly than semanticization, but it operates only as long as there is a clear morphological division of words and a clear-cut distinction between various types of declension, conjugation, etc.

Alternants or Phonemic Alternations Going Back to Different Stages of the History of the Language

§47. The following alternants of Polish go back to IE:

1) alternants based on the distinction of tautosyllabism (membership in the same syllable) and heterosyllabism (membership in two syllables). The following specific categories are involved:

a) combinations of simple vowels with simple consonants; e.g., nieb-o beside niebi-os-a, nie-bi-es-ki. .., wymię beside wymienia . . . , cielę beside cielęci-a. . . ;

b) combinations of two vowels: the tautosyllabic or diphthongized ou with the heterosyllabic o-u or o-v; eg., ku-ć beside kow-al. . . ;

c) the tautosyllabic or, ol, er, el. . . with heterosyllabic o-r, o-l, e-r, e-l; e.g., pró-ć, kłóć, brze-mię, mle-ć. . . vs. porz-e, kol-e, bierz-e, bior-ę, miel-e;

2) alternations based on the split of the IE vowel e into e and o (short or long): of ei into ei and oi, of eu into eu and ou, of en into en and on, of er into er and or, and el into el and ol; e.g., wiezie, wiozę beside wozi; siedzi, siada beside sad, sadzi; widzieć beside wiedzieć wiadomo; lęk beside łęk; sięk beside sączyć; krzątać beside kręcié; wrzeciono beside wrota; wlecze beside zawłoka. . . ;

3) alternations due to the weakening and loss of vowels in unaccented syllables, i.e., alternations of a historical-phonetic zero with the vowels e and o plus sonant: zero vs. e and o; i vs. ei, oi; u vs. eu, ou, syllabic u; m vs. en, on, em, om; syllabic r vs. er, or; syllabic l vs. el, ol. Examples: jes-t beside s-ą, dech beside duch; sechł, schnie beside such-; wierci, wartki beside wrzeciono, wrota; dłubac beside dłóto. . . .

§48. To the continuation of alternations which were formed in proto-Sl. but had their beginning in IE, we owe:

1) ży, czy, szy beside ż, cz, sz; li, ni beside l, ń; rzy beside rz (e.g., trwoży, toczy, straszy, pali, rani, tworzy. . . beside trwożę, toczę, straszę, palę, ranię, tworzę. . .);

2) p’i, b’i, m’i, w’i, beside p’ b’ m’ w’ (e.g., topi, robi, tłumi, łowi... beside topię, robię, tłumię, łowię...);

3) śi, źi beside sz, ż; ći, dźi beside c, dz (e.g., nosi, gasi, wozi, grozi.. . noszę, gaszę, wożę, grożę. .. ; świeci, traci, chodzi, kadzi. . . świecę, tracę, chodzę, kadzę. . .).

Alternations of proto-Sl. origin, i.e., those brought about phonetically and art.-aud.-ly in the proto-Sl. period or in a transitional period from IE to proto-Sl.;

1) g, k beside ż, cz (e.g., naga, ręka, mogę, piekę. . . beside obnażyć, ręczyć, może, piecze. . .ż);

g, k beside dz, c (e.g., noga, ręka, druga, wielka. . . beside nodze, ręce, drudzy, wielcy. . .);

c, dz beside cz, ż (e.g., chłopiec, owca, lice, ksiądz, drudzy beside chłopcze, owieczka, liczko, księże, drużyna. . .);

2) z beside s (e.g., mazać beside masło, wiozę beside wiosło. . .);

3) w beside historical-phonetic zero (e.g., warzyć, wóz, włóczyć. . . vs. obarzanek, obóz, obłok. . .);

4) consonants alternating with phonetic zero at the end of words and syllables (e.g., niebiosa beside niebo, cielesny beside ciało. . . §45).

§49. The following alternations are due to tradition, but their phonetic causes go back to Old Polish:

1) ś, ź, ć, dź, ść, źdź beside j (§38);

2) reflexes of “hard” consonants beside “soft” consonants (e.g., chłopy, schody beside chłopi, schodzi. . . ; tęgi beside ciężki, wożę beside wiozę, wieś beside wioska, twór beside tworzy, dworek beside dworzec §44);

3) k’, g’ beside k, g (e.g., bok, boku, stóg, stogu, ręka, rąk, noga, nóg. . . beside boki, bokiem, stogi, stogiem, ręki, nogi. . . §42);

4) o beside u (ó), ę beside ą (e.g., rodu, mrozu, głogu; bobu, noża. . . beside ród, mróz, głóg, bób, nóz. . . ; woda, koza, ozdoba beside wód, kóz, ozdób. . ., wódka, kózka, ozdóbka. . . ; męża> dębu. . . beside mąż, dąb. . . ; część, gęś. . . beside cząstka, gąska. . . §38);

5) e beside phonetic zero (e.g., sen, mech, zamek, dech. . . beside snu, mchu, zamku, tchu. . . ; pies, chłopiec, goniec, wieś, szedł. . . beside psa, chłopca, gońca, wsi, szła... §55);

6) consonant beside phonetic zero (e.g., szedł beside szła, cztery vs. cz(t)woro, cześć vs. cz(ś)ci; ciskac vs. cis(k)nąć, pisk vs. pis(k)nąć... §55);

7) e beside o, e beside a, er beside ar: wiezie, bierze, plecie. . . beside wiozę, biorę, plotę. .. ; leci, w lesie, w mieście, strzeli, wierzy, ścienny. . . vs. lata, las, miasto, strzał wiara, ściana. . . ; twierdza, cierń, pierścień. . . vs. twardy, tarnina, naparstek. . . §37).

Traditional alternations, i.e., alternations due to traditional conditions but whose phonetic, art.-aud. conditions are still in effect today: d beside t, b beside p. . . ; e.g., dech beside tchu, łba beside łeb, łepek; z beside ź, s beside ś: zły beside źle. . . ; sumienie, sfora, smutek. . . beside śmierć, śmietana, śmiecie, ślub. .. ; cz beside ć: czworo, czwarty, cztery beside ćwierć.

§50. Finally, there is the most recent layer of alternations, which are alive and determined by present-day conditions:

1) the difference in energy and autonomy of the phonemes when the intended articulation is fully realized, on the one hand, and weakly or not at all realized on the other; e.g., słup, ryk. . . beside słupa, ryku... ; głód, rydz, wóz.. . beside głodu, rydza, wozu.. . §43); wiatr, myśl, pieśń, pasm, drachm. . . beside wiatru, myśli, pieśni, pasma, drachmy. . . §§22, 35); drop, paw, drób. . . vs. dropia, pawia, drobiu. . . ;

2) the difference in quantity when a phonetic zero or a facultative syllable (i.e., a possible but not obligatory one) alternates with a full syllable; as a unit of thought, as a phoneme, it retains its identity (?); e.g., niósł, padł, rzekł. . . beside niosła, padła, rzekła. .., jabłko beside jabłek. . . , §55.

These two kinds of alternations are of a quantitative character, since they imply a difference of intensity, or quantity. Other alternations are of a qualitative character; they are as follows:

3) alternations in the activity of the vocal cords with the corresponding acoustic effects: w gaju beside w krajuy z żyta beside z czego, pod bokiem beside pod pasem, nad bramą beside nad podłogą. . . ;

głodu, głodny, głodzić beside głód, męża beside mąż, prosić beside prośba. . . ;

wiatru beside wiatr, Piotrek beside Piotrka, mech beside mchu, pasmo beside pasm. . . ;

ryk beside ryk wód, jest beside jezdem, dwuch beside we dwucheśmy byli, rzekł beside rzegem, niech beside niech no przyjdzie... ;

4) alternations based on palatal activity: kostka beside kość, pasę beside paść, niosła beside nieśli. . . ;

pies beside psa, kupiec beside kupca, krawiec beside krawca. . . ;

dropia, drobiu, gołębia, pawia, karmi, Radomia beside drop, drób, gołąb, paw, karm, Radom. . . ;

5) alternations based on activities of the front part of the tongue: s, z || sz, ż || ś, ź;

z domu, z masła, z ogrodu beside z tańca, z kaszy beside z dżumy, z żyta beside z czego, z szafy beside zziąbł, z ziemi, z dziury beside z sił, z ciebie, z cicha. . . ;

6) alternations of the nasal element of the nasal “vowels” (§§21, 51).

The Reflection of Dialectal Differences in the Pronunciation of Common (Standard) Polish

§51. Nowadays it is assumed that the Common Polish language, which unifies the entire nation in a single community and is refleeted in its literary language, had its beginnings in Wielkopolska and was later influenced by other regions of the Polish linguistic territory. Of course, even today the dialectal variation which is observed in different parts of this territory, or more precisely community, leaves its imprint on the external form, i.e., on the art.-aud. perception of the Common Polish language and its literary language which, in turn, modifies and lends variety to its corresponding art.-aud. representations.

Three possibilities should be mentioned:

1) the same phoneme is “pronounced” in different ways;

2) dialectally colored speech presents differences and phonetic nuances which are not found in the larger part of the Polish linguistic territory;

3) dialectal pronunciation presents fewer differences and nuances than the Common language.

The different regional pronunciations affect the following phonemes and phonemic series:

a) the “hard” or “normal” pronunciation of the consonant ł (i.e., with dental occlusion and lowering of the ridges of the tongue) is found among only a minority of Poles; in rare cases the consonant is pronounced as l (like the German or Czech l). More often, and in the speech of most Poles, it is pronounced as a nonsyllabic vowel, either with a velar narrowing and a place of articulation close to the vowel a, or with a labial narrowing which is close to a bilabial w, that is a nonsyllabic u;

b) the “soft” l varies territorially in degree of “softness” (palatality), ranging from a nonmedial l to an l which resembles the French l “mouillé” with very strong palatality;

c) the series of palatal-dental consonants ś, ź, ć, dź is territorially graduated from an almost pure dental pronunciation with the addition of palatality (as, for example, in the “model” pronunciation of the Russian consonants s’, z’) through pure palatals ś, ź, ć, dź, to consonants which merge acoustically with sz, ż, cz, dż. Among the Kashubians the latter consonants have merged with the dental consonants s, z, c, dz;

d) as one moves northwards, the palatal element of the eonsonants p’, b’, m’, f’, w’ becomes intensified, developing into a separate phonetic art.-aud. unit which leads to a split of these phonemes into consonantal diphthongs p’j, b’j, m’j, f’j, w’j. The j acquires a definite degree of friction such as ź or ś, and like n’ after m’. This change of the labio-palatal phonemes into the groups pś, bź, wź, mń and into the simple phonemes ś, ź, ń is a characteristic dialectal feature of the northern part of the Polish speaking territory;

e) in the pronunciation of some Poles, a velar n appears when a nasal vowel is followed by a velar stop (męka, mąk, gęgać, drągal. . .) and in place of the dental consonant n followed by a velar stop (panienka, wanienka, wianka, piastunka. . .). Some Poles pronounce the velar n only in the first of the above cases, while others (e.g., in Lithuania) employ only the dental consonant. In the mixed Polish-Byelorussian territory, vowel plus consonant n is used instead of a nasal vowel, even in the case when the latter is followed by the spirants s, z, sz, ż, ś, ź, ch, w; thus mięso, wąs, kąsa, więzy, wiązać, ciężko, męża, zdąży, gęsi, gąsior, więzi, wącha, wąwóz. . . are here pronounced mienso, wons, konsa, wienzy, wionzać, cienszko, menza, zdonzy, gensi, gonsior, wienzi, woncha, wonwóz. . . . This phenomenon is of great importance for the ethnic history of these regions;

f) the pronunciation of the “soft״ k’, g’, ch’ varies regionally.

§52. Art.-aud. differences and nuances alien to most Polish-speaking areas and not admitted in the contemporary literary language:

a) The vast majority of Poles do not employ any vocalic narrowing intermediate between a narrow vowel u and a narrow o, pronouncing the vowel which is graphically rendered by the letters u and ó in the same way. In some parts of the Polish-speaking area, primarily in the Southwest, ó is pronounced as a narrow o, i.e., an intermediate vowel between u and o. Such words as mur and mór, buk (Bug) and bóg, gul and gól, stuł and stół, clearly differ here in their pronunciation. Similarly, there are some areas which distinguish three degrees of the front vowels i, é, e and other areas that have an á “pochylone” or a narrow labialized a, intermediary between o and a.

b) The phoneme which is graphically represented by rz is not, in contemporary pronunciation, distiguished from ż and sz (e.g., morze = może, rzeka—żeka, drzewo = dżewo, trzy = tszy, prze = psze, krzak = kszak. . .). However, in certain regions (e.g., in Silesia) this phoneme is pronounced with a prominent r-like vibration differing from both ż and sz. These regions thus have two additional consonantal diphthongs rż (rzeka, drzewo. . .) and rsz (trzy, prze, krzak. . .).

c) Although the great majority of Poles do nøt employ an autonomous voiced h as opposed to voiceless ch, in some regions (especially in those which are adjacent to Byelorussian, Ukrainian, and Czech or Slovak areas) there is such a phoneme, which is clearly distinct from ch.

d) In the pronunciation of some Poles speaking the common language, there is a greater number of nasal vowels. In addition to the nasal vowels o (ą) and e (ę), these Poles also use the nasal vowel a (ą) Such a nasal vowel a appears otherwise in the speech of all Poles who employ such words as ansa, pasjans, kwadrans. . . (§40; cf. however §51, ie).

3) Dialectal influence is responsible for the decrease of phonetic (art.-aud.) nuances in the case of the above-mentioned distinction (§34) of the two i’s, two e’s, and two ę’s when they are preceded by palatal (ni, pi, ci. . . nie, pie, cie. . . nię, pię, cię. . .) as opposed to nonpalatal consonants (ny, py, cy, ty..., ne, pe, ce, te..., nę, pę, cę, tę. . .). But in the Southeastern part of the Polish-speaking territory there is not the slightest difference between these vowels. Here the words syn, ryba, ty, być, łysy... (s’in, r’iba, t’i, b’ić, l’is’i...), have exactly the same i as the words siny, ci, bić, lisy. . . (śin’i, ći, b’ić, l’is’i. . .).