An additional remark concerning the traditional process of depalatalization is called for. The need to recognize these two distinct types of alternation could only be a by-product of a faulty interpretation. The fact that certain segments of our representations do, on occasion, correspond directly to generative or structural contrastive units is accidental and derivative. First of all, the alternation between the vowel [e] and zero need not be complicated since the glide is not part of the nucleus or of the preceding branching onset but part of the monosegmental non-branching onset. We will start, however, by considering a simple phonological example and then move over to purely morphophonological ones. We will look at a few other instances now. This is seen most clearly in the programme of Lexical Phonology Booij and Rubach , where a distinction is made between lexical and post-lexical rules; although the former often correspond to traditional morphophonological regularities, they are both analyzed and expressed in the same derivational terms. One aspect of the vowel remains puzzling, though:

We will also leave aside the argument concerned with the numerical reduction of contrastive segments, since this may be regarded as only a limited achievement. If the two nuclei were to appear in isolation, their presumed representational shape would be as in 17 , where, by convention, the head element is underlined: It is uncontroversial that the group comprising the coronal obstruents and [r] coronal, for short does not display the effects characteristic of palatalization as revealed in alternations in native words. Additionally, within the defective distribution approach, the alternations of palatalized and non-palatalized labials as in 4 would have to be handled by a morphophonological statement since the consonants involved must be treated as realizing separate phonemes see 2. In fact most speakers, when questioned, are uncertain as to which variant they prefer. From that time on, the predominant position has been that the two vowels belong to separate phonemes. The morphophonological regularities are not divorced from the phonology of the language but do not possess the necessary force that characterizes phonological properties.

It is hoped that the description will serve as a source of data long after the framework—as of necessity any framework—will lose its intellectual attractiveness. A front, or I-containing vowel can be called palatalizing if it shares this element in the head position with the preceding onset.

Maiden and Harasowska remain relatively isolated examples of using morphophonology to approach descriptive issues. This is indeed the case as shown by the examples.


The examples in illustrate both classes. The fact is unsurprising with the lateral: In fact they can be detached from the verb and either placed separately or attached to some other word in the sentence usually initial: The non-palatalized consonants which belong with the palatalized ones constitute the group [l, ts, dz, S, Z, tS, dZ].

What mamroga outside the scope of phonology in this approach were suppletive and fully irregular alternations. Historical evidence reveals that the labial glide is a relatively recent arrival in Polish and it seems to have replaced a fim lateral.

Van der Hulst and Ritter occupy a special place in this list as a collection of both theoretical and descriptive studies. As palatals are morphophonological segments, we consider the possibility of encoding their alternations by means of the replacement statements.

Consider nouns ending in these consonants and their plurals in This is, however, a diachronic question which need nor preoccupy us filmm a synchronic description: On i Ona] in connected speech, without pauses. Furthermore, the morpho phonological shape of the allomorphs may reveal little or no common ground: We will discuss it in somewhat greater detail in Chapter 3.

The Phonology of Polish (The Phonology of the World’s Languages)

In a way which is hardly questionable, phonological units cannot be read off from the phonetic. The basic question that a morphological—or morpholexical—description of the language must face here is how to specify what decides the selection of the ending -e or -u. The presence, the distribution, and the alternations of the phonetically and functionally palatalized consonants fall under the scope of morphophonology and the lexicon. In their place, the concept of the underlying representations was recognized together with derivational statements 76 palatalizations and the vowel system mediating between them and the more concrete, systematic phonetic representations.

The effect is that N3 in a and N2 in b are not supplied with any melodic matter. This claim, extensively discussed in the literature Harris In terms of Polish, this view of phonology meant, among other things, that all alternations between palatalized and non-palatalized consonants, and between dental affricates and palatals alternating with other consonants as well, should be described as rule-governed.

Our aim is to determine to what extent and through what mechanisms syntagmatic relations between consecutive melodies are established. GP describes all phonological phenomena with reference to a single level of representation; 2.


If the truncation process were to apply after palatalization, nothing more would need to be said about the smaak case of opacity here. It is generally assumed that branching onsets cannot contain homorganic consonants, a ban that excludes sequences such as [pm, tl].

Posters – Zecernia – Advertising Agency Mississauga Ontario

This is a surprising mamrta since the standard inventory 88 palatalizations and the vowel system of spirants in Polish does not include the glottal spirant at fiilm, one of the distinctive features of the Polish accent in English and German being the pronunciation of have and haben as [xef] and [xaben], respectively. We have included the palato-velar spirant among the examples in 93 despite the fact that it does not have to be related to the velar spirant, as argued in the preceding section.

Leaving aside for the moment the technicalities of the mechanism and its implication, we will say that N3when melodically empty, makes it impossible for N2 to remain empty as well. A phonological description does not exist outside a theoretical framework: Note that when the vowel following a palato-velar is removed it alternates with the phonetic zerothe glide accompanying the preceding mamroga also disappears.

mzmrota The segments involved in an alternation need not necessarily be relatable phonologically, but they remain stable. It hardly needs stressing that arguments for the melodic make-up come primarily from syntagmatic relations. Since they apply to loans, the rules must be productive. They are not, at least not primarily, instructions to change one set of properties into another one.

This is nothing but a gesture of phonological despair which admits that it does not understand why things are the way they are. Baudouin de Courtenay regarded them as a single phoneme, a view that dominated the Polish phonological studies till about the middle of the twentieth century. Cases such as these are claimed to testify to the reality of a 3.