THE TRANSITIOy I.V THE RACE, 385 



The epoch during which these sentence-words prevailed 

 tN'as probably immense ; and, as we have before seen, far 

 from having^ been inimical to gesticulation, must have greatly 

 encouraged it — raising, in fact, the indicative phase of 

 language to the level of elaborate pantomime. Out of the 

 complex of sentence-words and gesture-signs thus inaugurated, 

 grammatical forms became slowly evolved, as we know from 

 the independent witness of philology. But long before 

 grammatical forms of any sort began to be evolved, a kind 

 of uncertain differentiation must have taken place in this pro- 

 toplasmic material of speech, in such wise that some sentence- 

 words would have tended to become specially denotative of 

 particular objects, others of particular actions, states, qualities, 

 and relations. This "primitive streak," as it were, of what 

 was afterwards to constitute the vertebral column of articulated 

 language in the independent yet mutually related ** parts of 

 speech," must in large measure have owed its development to 

 gesture. Now, by this time, gesture itself must already have 

 acquired an elementary kind of syntax, such as belongs even 

 to semiotic movements of an infant who happens to be late in 

 beginning to speak.* This elementary kind of syntax would 

 necessarily be taken over by, or impressed upon, the growing 

 structure of speech, at all events so far as the principles and 

 the order of apposition were concerned. Moreover, this sign- 

 making value of apposition would at the same time have been 

 promoted within the sphere of articulate signs themselves. 

 For, as we have previously seen, as soon as words become in 

 any measure denotative, they immediately begin to undergo 

 a connotative extension ; f and with this progressive widening 

 of signification, words require to be more and more frequently 

 used in apposition. Quite independently of any as yet non- 

 existing powers of introspective thought, the external "logic 

 of events " must have constantly determined such apposition 

 of receptually connotative terms, as we have already so fully 

 seen in the case of the growing child. Thus the conditions 

 were laid for the tripartite division — the genitive case, the 

 • See pp. 220-222. t See pp. 179-181. 



