INDONESIAN ANNOTATIONS ON THE VOCABULARY. 111 
material we seem to have followed a law of progressive devolution 
in accordance with which we have found ourselves passing from a 
higher, because more complex and more completely developed, form 
by successive stages of debilitation of the strong consonant to its 
weaker congener, eventually of reduction of consonant strength of 
whatever degree to semivowel weakness, and finally to extinction; 
and side by side with this process runs the attrition of the final vowel 
and at last of the newly become final consonant. ‘Thus, along strictly 
academic lines, wherein we postulate the validity of the rules which 
we create, we present the chain of devolution from kafuna to kau. 
In phonetics we have been trained so long in the study of the break- 
ing down of forms from the higher to the lower, progressive deformity, 
that we rest under the obsession of regarding our phonetic laws as 
invariable in their operation in that direction solely. In this group 
of languages it is becoming more and more clear that we have to do 
with evolution rather than devolution, that all growth is progress, that 
the motion is from below upward, in form from the simple to the com- 
plex, in sense from the broad and diffuse to the particular and precise. 
In this view of the situation there is no reason why we should not invert 
the foregoing series and present the evolution from kau to kafuna. 
It will be objected that we have no knowledge of laws governing the 
assumption of formative elements. Quite true; but that is not essen- 
tially objectionable. Weshould the rather rejoice in the opportunity 
to carry out in vivid and growing material the observations which 
afford us at some more advanced stage the opportunity of formu- 
lating the laws deducible from these observed phenomena. Further- 
more, we must never lose sight of the fact that in any living speech 
the law is secondary, it is but the average of fallible criticism of what 
is observed; supreme above the law lies the life of the speech and the 
will of the folk who use it for the communication of their thought. 
Without as yet attempting to assert the law of growth by assump- 
tion, let us examine the kau stem as the beginning rather than as the 
end product. For convenience we repeat the items which now come 
under review: 
1. kau. 3. kauwa. 5. kauuku. 28. auwou. 
2. kaua. 4. kauwek. 29. bwauwa. 
In 28 and 29 we find ourselves close to the ultimate of simplicity, 
an unprefaced stem au. Likewise we find ourselves close to the voice 
of the dog; we have to face the suggestion of onomatopoetic speech. 
It must be acknowledged that onomatopoeia has at times been over- 
worked in the theory of speech beginning; at other times it has been 
denied with somewhat too much of insistence, the truth being that 
each case in which it is sought to employ this principle must be judged 
independently on its own merits. 
