CHAPTER V. 

 DISSECTION OF THE THEORY. 



The result of independent treatment of the data should be identical — ■ 

 As the other elements have long been known and carefully studied, 

 the Melanesian is the critical test — The computation of exactly 

 what material is now made available for study — In what pro- 

 portion the Efate vocabulary contributes to the solution of the 

 problem and the manner in which properly it may be employed. 



The force of Dr. Macdonald's argument, the proof of the theory 

 thus summarily outlined, must rest upon the data which he has 

 utilized for its development. Those data we find available in the 

 vocabulary which he has given us as a complete dictionary of Efate. 

 He has made use of this material in a certain fashion, such as most 

 commended itself to his thought ; he has prosecuted long and pain- 

 fully a certain method. All this was within his prerogative. If 

 the theory be valid, if the method be true, the same data should 

 yield the same result, and no other than the same result, when 

 studied in accordance with such other method, being valid, as may 

 commend itself to another philological investigator. 



This it is which is now to engage our attention. We are to take 

 his data, his vocabulary material assumed to be in itself accurate, 

 to argue it afresh and, I feel confident, without preconception or 

 other such prejudice, and let it lead us where it may. 



If really there be an Oceanic speech family tree, with its roots in 

 the Hadramaut and its distal twigs in Te Pito te Henua, then we 

 may expect to find in the material in the course of this examination 

 such a mass of words showing clearly a nexus of development — 

 enabling us thereby to establish a distinct and probable law of the 

 mutation of sounds — that we may finish the investigation with the 

 happy satisfaction that the Oceanic tongue has proved itself. 



We are not to ask that every Polynesian word shall reveal to 

 us through the operation of this law of mutation its primitive 

 triliteron in some Semitic household. We are by no means to 

 expect that we can take any Semitic stem and by the application 

 of the rule develop the succeeding forms in Indonesia, Melanesia, 

 and Polynesia. Not even Grimm's law will do that for us in the 

 Aryan family. But we do have the right to expect that, if there 

 prove to be a substantial base for this theory, there be a sufficiency of 

 examples in each direction and a consistency in their establishment. 



31 



