ANY comparison of the merits of different languages is likely to remind the reader of Rivarol's essay De l'Universalitd de la langue frangaise, which has given to the world the phrase, Ce qui n'est pas clair n'est pas franpais, and the consequent belief that French is especially distinguished by clearness. Those who have read Rivarol will have observed that the original special point of that saying was that French is 'clear' in that it arranges its words in the order spontaneous to man: subject, verb, object. In other words, Caesar was perverse and unnatural when in his first sentence he said unam incolunt Belgae, and Xenophon when he began with 'Of Darius and Parysatis are born two children,' and Moses when he started his Hebrew text with 'In the beginning created God.' That which a Frenchman automatically writes is the natural impulse of humanity; the ordinary narrative order of the Semitic languages is mere artificiality. If Rivarol had felt that it might be otherwise, that such impulses might depend on acquired habit, that Caesar had avoided the order dont les Belges habitent l'un for the very purpose of gaining clearness, then the doctrine that French is a peculiarly clear language would have lacked its original foundation. It will be generally found that a man's preference between his own and a foreign language depends not so much on their intrinsic merits as on whether he loves most that which is most familiar or that which he has studied most. Professor Aronstein's 'On Style and Styles in Lauguages' (American Speech, December 1934) raises one question which seems capable of a mathematical answer: What languages are most concise? And yet the answer cannot be purely mathematical. One must first settle the question of judgment: What constitutes conciseness? Is the conciseness of English af. fected by our decision whether to write barbershop with the new Webster's Dictionary or barber shop with the rest of the English-speaking world? Is a German who writes zu Gunsten equally concise with one who writes zugunsten? I maintain that the test of conciseness should be the number of syllables. If one says in Latin cucurrisset and in English he would have run, then the Latin cu means have and the curr means run and the iss means would and the et means he, and the two expressions are equally concise. Or take the longest word in Professor Aronstein's Hebrew sample, hosethikha: I brought thee out. Here ho corresponds to brought, se to out, thi to I, and kha to thee.