22 BEACH-LA-MAR. 
standing of averb. Of coursethe paradeictic 7 is irreducible; atrophy 
of form is impossible. If we examine the sense we shall find no 
variation of meaning; tala means as before, “‘there-is-a-telling”’; 2 as 
before means ‘‘there-is-a-relation”’; therefore talai means ‘‘to tell to,”’ 
“to declare.’ It is accordingly a compacted word and not a com- 
posite, it is a mark of isolating and not of agglutinative speech. 
If it were not that for lexicographic ends it will be found convenient 
to arrange certain verbal uses of ta/a under talaz, I should treat the 
compacted form as no more than a record of fluent pronunciation. 
In our English we have words now in good usage, but wholly sub- 
versive of etymology, as a result of such fluent utterance as must 
always characterize the speech of any one in his native tongue. I 
note the familiar instances of adder and apron where a space dis- 
lodged westward has obliterated from our language the true words 
nadder and napron, and newt where an eastward dislodgement of the 
space leaves us puzzled over the simultaneous existence of newt and 
evet or eft as names for the same animal. 
The second of the parts of isolating speech is the demonstrative— 
a class far larger and far more detailed and specific than the para- 
deictic, yet still numerically small. Into this class fall those words 
which give vocal expression to cognition data which in daylight may 
be expressed almost, if not quite, as well by the pointing finger, which 
commonly are expressed doubly by word of mouth and digito mon- 
strari, a process now held inelegant, but which as recently as the 
brightest days of the Appian Way and the Via Sacra was welcomed 
as the best of good form. In this class we find what we know as pro- 
nouns, personal, demonstrative, the beginning of interrogatives, but 
no relatives have yet come into being. We find furthermore the basic 
adverbs of place and time, not yet discrete; we find a few of the 
adverbs of manner, the simplest ones. In general the demonstrative 
words in this group of isolating languages are the vocalization of the 
gesture language, the man’s provision against speechlessness soon as 
the evening shades prevail. 
Next we come to the attributive class, in which lies the great bulk 
of the vocabulary of each of these languages. In this part of isolating 
speech we include those vocables which in speech of higher develop- 
ment we have differentiated and have learned to designate nouns, 
verbs, adjectives, and most of the adverbs. 
Let us resume the consideration of tala. We have already seen the 
word in a usage which must, despite its violence, be designated a 
transitive noun, if we are to attempt to parse Samoan by means of 
our own grammatical apparatus. But that by no means exhausts 
the utilization of the word. Used absolutely, or with the definition 
of a weak demonstrative which we might denominate article if it 
were at all necessary to make a distinction where there is really no 
