18 Owen—Meaning and Function of Thought-Connectives. 
attempt will now be made by the study of a single thought-con¬ 
nective to develop a more detailed theory. The whole body of 
thought-connectives will next be classified. The theory will 
then be tested by successive applications to all classes. 
For preliminary examination of thought-connectives let the 
word “ therefore ” serve as a type; e. g. “ He invited me. 
Therefore I came. ” The meaning of the word " therefore ” is es¬ 
sentially “on account of that ” or, more briefly, “from that.” Not 
only is this evident to ordinary observation, but etymology, his¬ 
torical and comparative, declares it. It is accordingly clear at 
the outset that the present case is one of multiple symboliza¬ 
tion. That is, the single word “ therefore ” stands for ideas 
which are often represented separately. 
The exact value of each of these ideas merits careful observa¬ 
tion. And plainly the usual demonstrative value of “ that ” is 
not sufficient. Its present value is reinstative. That is, it brings 
back to attention something which has preceded. In the present 
case, moreover, it brings back not a mere thought-element but 
a whole thought. The proper test of this opinion is to put the 
whole preceding thought in place of the word “ that. ” Accord¬ 
ingly “He invited me. On account of his inviting me I came.” 
Plainly this substitution neither adds to nor subtracts from the 
speaker’s meaning. “ His inviting me ” may be accepted, there¬ 
fore, as the true value of “ that. ” 1 This element of the connec¬ 
tive is, then, reinstative or anaphoric, not initial, not deiktic. 
1 The propriety of thus defining “that” may be further developed as 
follows. It is evident that the same mind-stuff can be regarded as one or 
as composed of parts, according to the speaker’s choice. To illustrate ob¬ 
jectively, a horse, buggy and connecting harness may be regarded as either 
one or three at convenience. Indeed the language of the people supplies 
in the word “rig” a name for the unanalyzed combination. But as 
rule such names do not exist for thoughts. Names for the most part are 
symbols merely of thought-elements or ideas. Even when conceived as a 
unit, the thought as a rule can be distinctly presented only by combining 
the names of its parts, one part appearing as nucleus and the others being 
clustered about it. The question rises then: which part shall be put as 
nucleus? In answering, we must avoid a common but erroneous valua¬ 
tion of thought-elements. 
The phraseology of grammar presents relations as of subordinate value. 
