50 



♦ KNOWLEDGE ♦ 



[January 1, 188i 



ill which the word whit-li is practically a root may be used 

 indifferently as one and another part of speech. This type 

 it is which, as 1 have said, approaches most nearly to what 

 we may imagine to have been the primitive form of root- 

 language. 



More than two thousand j'ears ago the Indian gramma- 

 rians had discovered that all the woi-ds of their language 

 could be traced back to certain jihonetic forms, which they 

 called " elements," and fierce controversy arose as to 

 whether the.se " elements," or, as we call them, roofs, were 

 verbal or nominal, whether the noun or the verb was the 

 older part of speech. But, as S.iyce observed : " We must 

 be careful not to fall into the mistake of the Indian 

 grammarians and their modern followers, and confound 

 these roots with verbs or any other of the constituents of 

 living speech. The roots of language are like the roots of 

 the tree with its stem and branches ; the one implies the 

 other, but all alike spring from the seed, which in language 

 is the undeveloped sentence of primitive man, the aboriginal 

 monad of speech." According to Sayce, roots " are the last 

 result of linguistic analysis, the elements out of which the 

 materi;il of speech is formed, like the elementary substances 

 of the chemist." * 



When, however, I am in favour of the theory that the 

 languages of various families may be reduced by analysis to 

 one primitive form, namely, the root-word, I do not wish 

 this to be understood as a defence of the hypothesis of the 

 common origin of all languages ; for, as I have said already, 

 the universal similarity of physiological structure, as well as 

 many of the phenomena of external nature, are sufficient to 

 account for any resemblance which may exist between the 

 tongues of widely-diverging races. 



In his rambling work on the " Origins of Religions and 

 Language," Canon Cook, supporting the theory of the 

 primeval unity of language, gives a list of two hundi-ed and 

 fifty words, which he maintains are identical in form and 

 meaning in Egyptian, Semitic, Aryan, and other families. 

 The resemblances are certainly in many cases very striking, 

 as the following instances will show ; but it will also be 

 observed that the majority of these words are imitative of 

 natural sounds, a fact which Canon Cook seems to have 

 overlooked. Thus in ancient ICgyptian ad is an ass, and 

 mau cat and lion; chefer (cJt like f(), sairabeus, omv chafer; 

 Cymric, cwyber and cirifio, onr quiver, czvnf meaning motion, 

 stir, representing the sound of a body shooting through the 

 air; Cornish ch^iyvyan, to escape; French esqu'iver; Old 

 High German chtvor, clieviro beetle ; in Middle High 

 ( Jerman kever ; Anglo-Saxon ceafor ; Egyptian refref, a 

 worm, reptile, Greek erjjeton, from erpo, to creejj ; Sanscrit 

 srp or sarp ; Latin serpo and repo, imitative of sound 

 produced by peculiar motion of snakes. Egyptian amani ; 

 Sanscrit i/am ; New Zealand amu, to eat, swallow, imitative 

 of the sound of the closure of the lips after a mouthful has 

 been taken. Egyptian aiii/ or ank, to squeeze, embrace ; 

 Sanscrit anhu-s ; Latin anr/o, to throttle, representative of 

 the guttural sound produced by the person who is being 

 throttled. Egyptian pad or pat, to flee ; j^et-pet, to cause to 

 llee ; Sanscrit pa/, to tiee, fall down ; palao, a fall or w-ing; 

 pattrin, a bird, imitative of sound. Compare our patter of 

 rain or footsteps. Egyptian «,<, swift ; Sanscrit arna, a 

 horse, indicative of the sound of rapid motion through air. 

 Compare our word swisli, also Egyptian 6--«, or s-in-s, horse ; 

 Hebrew s^is ; African so, S((mar, so-sa, and so-si ■, so-mnsa, a 

 mare. Egyptian \aru, to wage war ; Sanscrit JiCira, war ; 

 hr, to .seize violentlj^, probably imitative of the rapid 

 breathing of a man or animal when springing on an 

 enemy or prey. Egyptian purasha and |)«ras/«j, to hew ; 



* Loc. cit , p. ,'!. 



Sanscrit parnrti, an axe, sound of cutting through wood. 

 Simil.irly, Egyptian shad, to cut ; Sanscrit cid (representing 

 ancient kh sound) ; Old Dravidian, ciit ; Zend s/ienda. 

 Egyptian fper, wish, request; Sanscrit spr/i, spark, to 

 desire; Latin spero ; Old Norse spir, to a.sk, imitative 

 of sound in breathing. Compare Latin spiro, I breiAthe ; 

 our to aspire, literally " to breathe after." In Hebrew the 

 the ideas of wishing or desiring are expressed by several 

 words, oi'iginally meaning to breathe — as, for instance, ava, 

 and imitative of sounds produced in breathing. Egyptian 

 kerer, to burn ; Hebrew charar ()(), the .sound of which 

 woi-d is very suggestive of the roaring sound of flames ; 

 Sanscrit yJiar, to shine; gharmes, heat; Zend f/ar. 

 Egyptian ash, lis/ibu, to consume; Hebrew esh, tire; 

 Sanscrit ush, to burn ; us/tiia, hot ; usltra, flash. 

 Egyptian shu, sun or, with a different ideograph, light ; Rig 

 Veda rue, shine, blaze ; riiki, shining, flaming; rush, ritsliia, 

 the demon of drought: cui/i»ia, flame; cihhumi, blazing; 

 Basque sii, fire, from the hissing sound of flames. Egyptian 

 ka, (ja, bull or ox ; kaul, cows; Sanscrit and Zend go, and 

 Egyptian be/ies. Coptic bahsl, anciently Egyptian ah, calf. 

 Compai'e Greek bous, bullock ; ox, cow, are probably derived 

 from sounds similar to those made by the animals. The 

 word mer in Egyptian is used in three different senses, with 

 different ideographs : it means — 1, to die ; Sanscrit mf or 

 mar, maras, death ; 2, to love ; 3, the sea ; Coptic meire, 

 lake, inundation ; Cornish mor, and Old Turanian mor, 

 meaning the Caspian Sea. In the last sense it is imitative 

 of the murLuuring .sound of waters. 



More oljscure, however, is the ]5robable origin and rela- 

 tionship of the following words : — Egyptian dka or d(/a, to 

 consume by heat ; Hebrew ach {\), oven ; Sanscrit agni, 

 fire ; Egyptian bd, to shine brightly ; Sanscrit b/id ; Egyptian 

 rakJiu, to burn ; Sanscrit rue, Latin lucere ; Egyptian dam, 

 to subdue with violence ; Sanscrit dam., Latin domare, the 

 original meaning of which was probably to bind ; Egyptian 

 kan, to beat, conquer ; Sanscrit can or ksliait ; Egyptian 

 kuk, evening ; kak or kakiu, darkness, night; Sanscrit rdi/ci 

 for k/tdi/a, Latin ccpcus, blind ; Accadian gij and kug, dark- 

 ness; Finnish (/(', night; in the Veda used in the sense to 

 cover, tamas asit tamasd gudliam — " darkness was covered 

 with darkness " ; Egyptian »(«», ?»?«, to be steady, abide ; 

 Hebrew amen, Sanscrit man, puto ; Zend upaman, Latin 

 manere, Greek man, verily, in truth, like Hebrew amen, 

 also Greek men; Egyptian mdeh, measure, weigh, reflect; 

 Sanscrit ind/i, to measure; Egyptian mat, truth, justice; 

 Sanscrit, ma, measure. 



It is not so easy in the latter instances to see the 

 imitative origin which we may conjecture to lie at the 

 root of the similarity in these words, but in inquiries 

 such as the present we must not lose sight of the fact 

 that words or signs (as I have shown when discussing 

 the language of the deaf and dumb) at first distinctly 

 imitative, after they have been adopted into language, and 

 when there is no longer a risk of their being misunderstood, 

 become conventionalised on the principle of economy of 

 effort. And again, we must remember the powerful imagery 

 of primitive minds, which, on the grounds of .some resem- 

 blance not aiiparent to our more analytic brains, will transfer 

 a word from one meaning to another ; so that, to take a 

 familiar example, the word driim, imitative of the sound of 

 sticks or hands beating on parchment, is applied to the drum 

 of the ear, the similarity in this case being the tightly 

 stretched membrane, and having no reference to sound. It 

 is necessary to make these observations, as the facts men- 

 tioned are persistently ignored by those who deny what I 

 have called the " natural origin " of language, in order to 

 ascribe it to the supernatural. 



There can be little doubt that the earliest means by which 



