TRANSACTIONS OF SECTION H. 785 



follow that our first ancestors were cannibals. And here, too, Bunsen's words have 

 hecome so strikingly true that I may be allowed to quote them : ' The savage is 

 justly disclaimed as the prototype of natural, original man ; for linguistic inquiry 

 shows that the languages of savages are degraded and decaying fragments of nobler 

 formations.' 



I know well that in unreservedly adopting Bunsen's opinion on this point also 

 I run counter to the teaching of such well-known writers as Sir John Lubbock, 

 Reclus, and others. It might be supposed that Mr. Herbert Spencer also looked 

 upon savages as representing the primitive state of mankind. But if he ever did so, 

 he certainly does so no longer, and there is notliing I admire so much in Mr. 

 Herbert Spencer as this simple love of truth, which makes him confess openly 

 ■whenever he has seen occasion to change his views. ' What terms and what con- 

 ceptions are truly primitive,' he writes, ' would be easy if we had an account of 

 truly primitive men. But there are sundry reasons for suspecting that existing 

 men of the lowest type forming social groups of the simplest kind do not exemplify 

 men as they originally were. Probably most of them, if not all, had ancestors in 

 a higher state.' ^ 



Most important also is a hint which Bunsen gives that the students of language 

 should follow the same method that has been followed with so much success in 

 Geology; that they should begin by studying the modern strata of speech, and 

 then apply the principles, discovered there, to the lower or less accessible strata. 

 It is true that the same suggestion had been made by Leibniz, but many sugges- 

 tions are made and are forgotten again, and the merit of rediscovering an old truth 

 is often as great as the discovery of a new truth. This is what Bunsen said : ' In 

 order to arrive at the law which we are endeavouring to find (the law of the develop- 

 ment of language) let us first assume, as Geology does, that the same principles 

 which we see working in the (recent) development were also at work at the very 

 beginning, modified in degree and in form, but essentially the same in kind.' We 

 know how fruitful this suggestion has proved, and how much light an accurate 

 study of modern languages and of spoken dialects has thrown on some of the 

 darkest problems of the science of language. But fifty years ago it was Sanskrit 

 only, or Hebrew, or Chinese, that seemed to deserve the attention of the students 

 of Comparative Philology. Still more important is Bunsen's next remark, that 

 language hegins with the sentence, and that in the Ijeginning each word was a 

 sentence in itself. This view also has found strong supporters at a later time, for 

 instance, my friend Professor Sayce, though at the time we are speaking of it was 

 hardly thought of. I must here once more quote Bunsen's own words : ' The 

 supreme law of progress in all language shows itself to be the progress from the 

 substantial isolated word, as an undeveloped expression of a whole sentence, towards 

 such a construction of language as malves every single word subservient to the 

 general idea of a sentence, and shapes, modifies, and dissolves it accordingly.' 



And again : ' Every sound in language must originally have been significative 

 of something. The unity of sound (the syllable, pure or consonantised) must 

 therefore originally have corresponded to a unity of conscious plastic thought, and 

 every thought must have had a real or substantial object of perception. . . . Every 

 single word implies necessarily a complete proposition, consisting of subject, 

 predicate, and copula.' 



This is a most pregnant remark. It shows as clearly as daylight the enormous 

 difference there is between the mere utterance of the sound Pah and Mah, as a cry 

 of pleasure or distress, and the pronunciation of the same syllable as a sentence, 

 when Pah and Mah are meant for ' This is Pah,' ' This is Mah ' ; or, after a still more 

 characteristic advance of the human intellect, ' This is a Pah,' ' This is a Mah.' 

 which is not very far from saying, ' This man belongs to the class or genus of fathers.' 



Equally important is Bunsen's categorical statement that everything in lan- 

 guage must have been originally significant, that everything formal must originally 

 have been substantial. You know what a bone of contention this has baen of late 

 between what is called the old school and the new school of Comparative Philology. 



' Ojfen Court, No, 20P, p. 2896. 



isri. 3e 



