343 
lying apparently at the basis of speech ; they differ in different 
languages. Primary Aryan roots are, or, at all events, are 
generally, monosyllabic ; Semitic roots dissyllabic, or, if the 
vowels are sounded, trisyllabic. Old Egyptian roots may be 
either. Semitic roots show the principle of triconsonantism ; 
Aryan roots do not.* 
And here let me observe that nothing is more dangerous than 
to build a universal theory on the phenomena afforded by a 
single family of languages, e . <7., the Aryan dialects. If any 
one is inclined to be alarmed at the amount of knowledge which 
may be supposed to be requisite for linguistic inquiries, let me 
reassure him by the dictum of a master ; — 
44 I must protest, at the very outset of these lectures, against 
the supposition that the student of language must necessarily 
be a great linguist.” f 
But whilst this is a most consoling fact, yet be it remembered 
that the student of language should have a clear grasp of a 
subject upon which most people have but very shadowy notions 
— the principles of evidence. Suppose, e.g., that Aryan man 
started with the verb, in the abstract it is evidently possible 
that Semitic man may have started with the noun. Yet we 
find persons arguing or even asserting, with the utmost con- 
fidence, that what has occurred in some families of speech must 
be the rule in all. What is at fault, their knowledge, accord- 
ing to the saying 44 a little knowledge is a dangerous thing ? ” 
No, their knowledge is very valuable ; it is their imperfect 
logic, — their ignorance of the laws of evidence, which overthrows 
their efforts. 
We have, then, these mysterious roots, and arrived at this 
point, Prof. Muller observed : — 
44 The science of language, I felt, had done its work when it 
had reduced the vague problem of the origin of language to a 
more definite form, viz., What is the origin of roots ? Beyond 
that point, however, where the student of language is able to lay 
the primary elements of language at the feet of philosophers, 
the science of language alone , apart from the science of thought, 
will not carry us.”J Psychology, then, must' be summoned to 
assist. The problem, to use the words of Prof. Muller, is 44 How 
do mere cries become phonetic types ? ” This most difficult 
question Noire claims to have solved, and Prof. Muller appears 
* Vide List of Primary Roots of Proto- Aryan iB.M.A. Appendix B.). 
t Lects. Sci. Lang ., i. 25. 
+ Contemporary Beview, Feb., 1878, p. 466. 
2 c 2 
