Page:Language and the Study of Language.djvu/101

III.] by the vowels of certain case-endings, which assimilated the vowels of the nouns to which these were attached. So little was the altered vowel in Anglo-Saxon a sign of plurality, that it was found also in one of the singular cases, while two of the plural cases exhibited the unchanged vowel of the theme. Man, for instance, was thus declined:

But the nominative and accusative singular exhibited one vowel, and the nominative and accusative plural another; and so this incidental difference of pronunciation between the forms of most frequent occurrence in the two numbers respectively came to appear before the popular apprehension as indicative of the distinction of number; its genesis was already long forgotten, as the case-endings which called it out had disappeared; and now it was fully invested with a new office—though only in a few rather arbitrarily selected cases: the word book, for example, has the same hereditary right to a plural beek, instead of books, as has foot to a plural feet, instead of foots. The case is quite the same as if, at present, because we pronounce nătional (with "short a") the adjective derived from nātion, we should come finally to neglect as unnecessary the suffix al, and should allow nātion and nătion to answer to one another as corresponding substantive and adjective.

A very similar case of substitution of distinctions originally accidental for others of formal and organic growth appears also in some of our verbs. From dælan, 'to deal,' the Anglo-Saxon formed, by the usual suffixes of conjugation, the imperfect dælde and the participle dæled. In our mouthing over of these forms to suit our ideas of convenient pronunciation, we have established a difference of vowel sound among them, saying I dēal, but he dĕalt and we have dĕalt. Here is an internal distinction, of euphonic