312 MEX7-AL EVOLUTIOX IN MAX. 



'by;' if it 'sits,' 'stands,' 'remains,' or 'appears,' \vc need 

 no ghost to tell us that it is, nor any grammarian or 

 metaphysician to proclaim that recondite fact in formal 

 terms." * 



Having thus briefly considered the philology of predicative 

 words, we must next proceed to the not less important matter 

 of the philology of predication itself And here we shall find 

 that the evidence is sufficiently definite. We have already 

 seen good reason for concluding that what Grimm has called 

 the "antediluvian" pronominal roots were the phonetic 

 equivalents of gesture-signs — or rather, that they implied 

 accompanying gesture-signs for the conveyance of their 

 meaning. Now, it is on all hands allowed that these 

 pronominal roots, or demonstrative elements, afterwards 

 became attached to nouns and verbs as affixes or suffixes, 

 and so in older languages constitute the machinery both of 

 declension and conjugation. Thus, we can trace back, stage 

 by stage, the form of predication as it occurs in the most 

 highly developed, or inflective, languages, to that earliest 

 stage of language in general, which I have called the 

 indicative. In order to show this somewhat more in detail, 

 I will begin by sketching these several stages, and then 

 illustrate the earliest of them that still happen to survive by 

 quoting the modes of predication which they actually present. 



As we thus trace language backwards, its structure is 

 found to undergo the following simplification. First of all, 

 auxiliary words, suffixes, affixes, prepositions, copulas, 

 particles, and, in short, all inflections, agglutinations, or 

 other parts of speech which are concerned in the indication 

 of relationship between the other component parts of a 

 sentence, progressively dwindle and disappear. When these, 

 which I will call relational words, are shed, language is left 

 with what may be termed object-words (including pronominal 

 words), attributive-words, action -words, and words expressive 

 of states of mind or body, which, therefore, may be designated 



* Garnett, On the Xatitrc and Analysis of the IWd, /'7'oc. PhUo. Soc, vol. iii. 



