PREDICATIVE INFINITIVE WITH ACCUSATIVE SUBJECT. 
245 
stated with some fullness the grounds for this belief with respect to each, and 
in connection therewith have given a good deal of the history of the opinions 
concerning the construction. Here I need only add that the cumulative weight 
of the evidence should not be lost sight of. 
Perhaps I should add here that some scholars, among them Drs. Becker, 
Grimberg, Prime±, and Zeitlin, hold that the accusative with infinitive in the 
Germanic languages is in part a native development from the accusative with 
predicative participle, and that in the Germanic languages the latter idiom was 
prior to the former. In Chapters XIV and XV, however, I have tried to show 
that, while, in conformity with the Greek original, the accusative with predica¬ 
tive present participle is commoner in Gothic than is the accusative with predi¬ 
cative infinitive after verbs of sense perception, the reverse is the case in the 
Germanic languages as a whole, especially in Anglo-Saxon and in High German, 
and that in these languages the predicative infinitive was prior to the predica¬ 
tive present participle. 
AS SUBJECT. 
Despite Dr. StoffeFs contention that “ we are almost forced to the conclu¬ 
sion that the Ace. cum Inf. as the logical subject of a quasi-impersonal verb, 
must once have been as common in the Germanic tongues as we find it to have 
been in the classical languages/’ 1 1 must hold that, in the Germanic languages, 
as in Anglo-Saxon, the idiom is relatively rare, and occurs for the most part 
only in translations. 
In Gothic we occasionally have the infinitive phrase as subject to the verb 
to be plus an adjective {gup ist, cizetizo ist, gadob ist, etc.), but, as Apelt, 1 Z. c., 
pp. 290-291, shows, only because of the influence of the Greek original, the 
Goth usually translating otherwise the Greek accusative and infinitive in such 
expressions. Examples are: (1) of accusative and infinitive: L. 16.17: ip azetizo 
ist himin jah airpa hindarleipan J>au witodis ainana writ gadriusan = €VK077(j)T€p0V 
Se ecrn rov ovpavov koll rr/v yrjv 7rapeA.^etv etc.j — (2) of other translations: Mk . 
10.25: azetizo ist ulbandau J>airh pairko neplos galeipan, etc. = oVoWj-repoV ianv 
KdpLrjXov . . . SukOdv, etc. This view as to the foreign origin of the accusative 
and infinitive as the subject of impersonals, in Gothic, though once opposed by 
Albrecht 2 and by Miklosich, 3 is now generally accepted: see Apelt, 1 Z. c., p. 290; 
Bernhardt, 2 Z. c., p. 113; Streitberg, 2 Z. c., p. 212; Zeitlin, 1 Z. c., p. 28; and Wil- 
manns, Z. c., p. 119. Quite recently, however, Professor G. O. Curme/ 2 Z. c., pp. 
359-367, has attempted to overthrow this theory, but without success in my 
judgment. 
That the construction is rare, also, in Old Norse, I judge from the fact that 
I find no examples cited by Lund or by Falk and Torp. 
In Old High German, too, the construction is rare 4 with impersonals, and 
as a rule is found only in translation of the same idiom in Latin. Usually, 
however, the translator uses another idiom, generally a dative dependent on 
the chief verb, with a subjective infinitive, either uninflected or inflected. Ex¬ 
amples are: (1) of accusative and infinitive: Tatian 187.9: gilimphit mih gangen 
= oportet me ambulare ; 5 — (2) of dative and infinitive: Tatian 85.22: gilimphit 
1 Stoffel, 1 1. c., p. 54. 2 L. c., p. 18. 3 Miklosich,i 1. c., p. 483. 
4 No one of the five Latin examples of his original is retained by Isidor: see Rannow, l. c., pp. 87-88. 
6 From Denecke, l. c., p. 42. 
