Miksi astevaihtelu fonemaattistui?

• Kari Nahkola

Avainsanat: astevaihtelu, fonologia

Abstrakti

Why did consonant gradation become phonemic? (englanti)

1995 (99)

Kari Nahkola

Why did consonant gradation become phonemic?

Many of the differences between Baltic-Finnic languages can be explained in terms of the phonemicization of plosive consonant gradation: first, the phonemicization process itself gave rise to partly different alternations in the different languages (e.g. the various weak grades of t); and second, the additional changes made possible by this phonemicization were also different (e.g. the treatment of vowels made adjacent by the deletion of a plosive). Consonant gradation presumably arose in Proto-Finnic, but originally took the form of phonetic variation such that in positions requiring weak grade, geminate plosives were pronounced shorter than usual, with "short endings", and single plosives were pronounced more weakly. The article argues that the alternation began to become phonemicized when weak-grade geminates were shortened slightly further, with the result that they could no longer contain a syllable boundary. Geminates whose ends had been lopped too short thus came to function as syllable-closing single plosives; because of the Baltic-Finnic syllable construction rule these then shifted to become the initial consonants of the following syllable.

This change also caused the consonant gradation of single plosives to alter; the precise reason for this depends on the phonetic form of the weak-grade allophones of single plosives at that period. If they differed from strong-grade allophones so little that speakers heard the same phoneme in both positions, morphemes with both geminate and single plosives in strong-grade positions became structurally identical in weak grade: both geminate and single plosives were represented by single plosives. This gave rise not only to homonymy in the weak grade, but also to a complex and unstable morphophonological situation, for weak-grade positions now had single plosives behaving in two different ways.

If, on the other hand, the weak grades of single plosives were phonetically very different from their strong-grade allophones for instance voiced obstruents there was a risk that such sounds might become independent phonemes by virtue of word-pairs where the only distinctive feature separating meanings in weak-grade inflections was the quality of the syllable-initial consonant: (katto:) katon vs. (kato:) kadon (it has been traditionally assumed that weak-grade allophones of single plosives were voiced fricatives in this position).

Whatever the type of weak-grade single plosive, the problem caused by the shortening of geminates disappeared when other weak-grade alternates were adopted for single plosives: phonemes already belonging to the sound system (such as voiced consonants) or else deletion.

The article suggests that the reason why weak-grade geminates became thus shortened was a change in the word-stress contour, a centralization in which the stress peak shifted closer to the begining of the main stressed syllable. The result of this shift was that the stressed syllable became even more dominant than before while the unstressed syllables became even weaker. The assumption is that this change of stress also set other weakening changes in motion in Baltic-Finnic (e.g. deletion of unstressed vowels).

The writer argues further that the weak allophones of single plosives were voiceless stops at the Proto-Finnic period of phonetic alternation: either weakened tenuis-plosives or voiceless medial plosives. In the traditional view, these allophones would have been voiced fricatives, even voiced stops in post-nasal positions. In the writer's view, however, it is difficult to justify the notion that a weakening of articulation could have added an extra phonetic feature, voicedness, to voiceless plosives. In some areas of Baltic-Finnic, voiced fricatives have nevertheless arisen during the phonemicization process (e.g. in the southwestern dialects of Finnish).