224 Mr. Sankey on the Analysis and Structure 



For example, it is usual to consider a large class of nouns as de- 

 rived from the middle perfect, cumbered though almost every 

 part of it must be with the different modifications of mood, tense, 

 number, and person ; whereas a just analysis will give us that 

 part of the noun which connects it with the middle perfect, as 

 a constituent part of both, and, consequently, as formed in 

 both, on the very same principles, directly from the same original 

 radix. 



Proceeding in this way upon any inflected language, it is 

 obvious that we shall approximate, at least, to a knowledge of 

 its roots, or most elementary parts, from which again, by syn- 

 thesis, according to the analogy of the language, the words 

 may be recombined into their present inflected forms. The 

 particles of the language, such as ptev, &c., it is obvious will, 

 for the most part, be found partaking of the simplest forms, as 

 not being subject to any modifying inflection ; hence, they will 

 at once be recognized as roots, and not abbreviations, accord- 

 ing to the erroneous hypothesis of the ingenious Home Tooke : 

 — iVbbreviation acted not so much in breaking down words as 

 in coalescing many into one, and so inflecting the language. 



If, now, we suppose the analysis carried into all languages, 

 it is clear we shall thus arrive at the roots or elementary parts 

 of each distinct language. Comparing, then, these elements 

 with one another, we shall be enabled to simplify them still 

 further, whilst very many of the elements of one language will 

 be found to be exactly the same as those of another. Olten, 

 indeed, we shall find that which is only an element in one 

 tongue, occupying a place as a distinct and significant word in 

 another ; thus, for example, analysis gives us er, as an elemen- 

 tary and constituent part in the English language, as in the 

 word great-er, while we have it a distinct word in the German 

 tongue. In this way, it is obvious that the aggregate sum of 

 the elements to be found throughout all languages will be 

 much diminished — the great families of languages as they 

 may ,be called, such as the Aramean, Greek, Teutonic, &c., 

 consisting, for the most part, of the same elements. Thus, 

 then, we might be enabled to construct a universal key to all 

 languages, by simply arranging these elements in alphabetic 

 order, from the recombining of which, according to the ana- 



