「#1198. ic → I」 ([2012-08-07-1]) の記事で,古英語から中英語にかけて用いられた1人称単数代名詞の主格 ich が,語末の子音を消失させて近代英語の I へと発展した経緯について論じた.そこでは,純粋な音韻変化というよりは,機能語に見られる強形と弱形の競合が関わっているのではないかと提案した.
しかし,音韻的な要因が皆無というわけではなさそうだ.Schlüter によれば,後続する語頭の音に種類によって,従来の長形 ich か刷新的な短形 i かのいずれかが選ばれやすいという事実が,確かにある.
Schlüter は,Helsinki Corpus を用いて中英語期内で時代ごとに,そして後続音の種類別に,ich, everich, -lich それぞれの変異形の分布を調査した.以下に,Schlüter (224, 227, 226) に掲載されている,各々の分布表を示そう.
I | 1150--1250 (ME I) | 1250--1350 (ME II) | 1350--1420 (ME III) | 1420--1500 (ME IV) | |||||
---|---|---|---|---|---|---|---|---|---|
tokens | % | tokens | % | tokens | % | tokens | % | ||
before V | ich | 169 | 100 | 121 | 95 | 4 | 3 | 0 | 0 |
I | 0 | 0 | 6 | 5 | 135 | 97 | 253 | 100 | |
before <h> | ich | 171 | 100 | 105 | 97 | 3 | 2 | 0 | 0 |
I | 0 | 0 | 3 | 3 | 156 | 98 | 316 | 100 | |
before C | ich | 513 | 94 | 363 | 42 | 0 | 0 | 0 | 0 |
I | 33 | 6 | 494 | 58 | 1106 | 100 | 2043 | 100 | |
EVERY | 1150--1250 (ME I) | 1250--1350 (ME II) | 1350--1420 (ME III) | 1420--1500 (ME IV) | |||||
tokens | % | tokens | % | tokens | % | tokens | % | ||
before V | everich | - | 6 | 86 | 7 | 64 | 9 | 39 | |
everiche | - | 1 | 14 | 0 | 0 | 0 | 0 | ||
every | - | 0 | 0 | 4 | 36 | 14 | 61 | ||
before <h> | everich | - | 0 | 0 | 1 | 20 | - | ||
everiche | - | 1 | 100 | 1 | 20 | - | |||
every | - | 0 | 0 | 3 | 60 | - | |||
before C | everich | - | 6 | 29 | 2 | 2 | 0 | 0 | |
everiche | - | 10 | 48 | 2 | 2 | 0 | 0 | ||
every | - | 5 | 24 | 105 | 96 | 138 | 100 | ||
-LY | 1150--1250 (ME I) | 1250--1350 (ME II) | 1350--1420 (ME III) | 1420--1500 (ME IV) | |||||
tokens | % | tokens | % | tokens | % | tokens | % | ||
before V | -lich | 23 | 12 | 8 | 12 | 12 | 4 | 1 | 0 |
-liche | 162 | 87 | 51 | 77 | 23 | 8 | 21 | 5 | |
-ly | 1 | 1 | 7 | 11 | 251 | 88 | 421 | 95 | |
before <h> | -lich | 13 | 18 | 7 | 21 | 1 | 2 | 0 | 0 |
-liche | 59 | 82 | 24 | 73 | 8 | 14 | 0 | 0 | |
-ly | 0 | 0 | 2 | 6 | 49 | 84 | 76 | 100 | |
before C | -lich | 70 | 13 | 18 | 15 | 18 | 2 | 2 | 0 |
-liche | 468 | 85 | 93 | 77 | 39 | 5 | 23 | 2 | |
-ly | 11 | 2 | 10 | 8 | 788 | 93 | 947 | 97 |
. . . the affricate [ʧ] in final position has turned out to constitute another weak segment whose disappearance is codetermined by syllable structure constraints militating against the adjacency of two Cs or Vs across word boundaries. . . . [T]he three studies have shown that the demise of final [ʧ] proceeds at different speeds depending on the item concerned: it is given up fastest in the personal pronoun, not much later in the quantifier, and most hesitantly in the suffix. In other words, the phonetic erosion is overshadowed by lexical distinctions. Relics of the obsolescent long variants are typically found in high-frequency collocations like ich am or everichone, where the affricate is protected from erosion by the ideal phonotactic constellation it ensures.
関連して,「#40. 接尾辞 -ly は副詞語尾か?」 ([2009-06-07-1]) 及び「#832. every と each」 ([2011-08-07-1]) も参照.
・ Schlüter, Julia. "Weak Segments and Syllable Structure in ME." Phonological Weakness in English: From Old to Present-Day English. Ed. Donka Minkova. Basingstoke: Palgrave Macmillan, 2009. 199--236.
Powered by WinChalow1.0rc4 based on chalow