We would suggest to our fellow-workers as general principles: 1) That we abandon our present conventions in the use of Roman, italic, and bold-face type. They entail a duplication of equipment in special symbols and serve only purposes that are essentially ornamental. Linguistic forms can (as far as necessary) be set off from each other and from the surrounding text by brackets, dashes, or what not. Variations in these can be used to distinguish (where needed) 'broad' and 'narrow' transcription, transliteration, etc., etc. The equipment thus released becomes automatically available for real work: e.g. italics for 'emphatic' consonants, bold-face for stressed vowels, etc. Foreign words in Roman type stand out as well as in italics; the latter can then be used for translations. For a specimen of such printing, see LANGUAGE 1. 130-56; 3. 9-11. 2) That we use to the utmost Latin letters, including capitals and small capitals together with italic and bold-face types. Only where these do not reach should we have recourse to other alphabets and to diacritic marks. This applies chiefly to new needs and to particular emergencies that may arise. Diacritic marks that are well-known and already in the printers' stock are not to be recklessly discarded. But even among these are many whose right to a continued existence should be scrutinized closely. In general the presumptions are against any symbol with two marks above or below the letter, and very strongly against any with more. In a system employing such symbols, the number of type required mounts rapidly, and their cost becomes prohibitive. When all that these marks indicate must be indicated, the solution will often be to put the diacritics after or before the letter, or to avoid one set at least of diacritics: higher and lower vowel-types, for instance can be distinguished better by different letters (small capitals, e, ) than by tails or dots under one (and sometimes both!) letters. In making such readjustments advantage should be taken of the opportunity to iron out inconsistencies like those that have arisen in IE grammar from the principle of adhering to various (and unstable) traditions. More frequently, however, the solution will be found by regarding the next two suggestions. 8 9 8 If only five vowel letters and only six diacritics be used, it can be calculated that the possible combinations of zero, one, two, or three of these six diacritics with the five vowels is 210, and will cost well over $1000.00 for the equipment. • To illustrate: symbols such as (oo) (high long oral) and (on) (low short nasalized), will express as much as the troublesome combinations mentioned in the preceding note, and their cost is practically negligible. 3) That we recognize frankly the will-o'-the-wisp nature of the effort to assign a separate symbol to each variety of sound. In part we do recognize this, as when we employ the same symbol p for the surd labial stops of both French and English in spite of their easily noticeable differences. But at other times we become too rigid. If (o) and (o) represent tense, rounded vowels in French; that is no reason why [0] may not represent a loose vowel in German or a loose, unrounded vowel in English. For those who do not know these facts, a once-for-all statement will suffice; neither for them nor for other readers is anything gained by diacritics, inverted v, or similar devices of the grimoire. 4) That normally we symbolize only phonemes (distinctive features) so far as we can determine them; and that always before we indicate more, we convince ourselves that more is demanded by the purpose in hand. The gain in elegance (in the mathematician's sense) will repay us for whatever nostalgia may result. We know today that no purpose was served by those who wrote the Irish symbol instead of g in OE, presumably because the sound was a spirant. Who would wish for different symbols for the l's in E little, or for the sibilants in Gr.mpéoßus ? If in German the long vowels are tense and the short vowels loose, the long-sign will suffice. If our diphthongs are much altered before [r] the (r) will symbolize this. To the present writers it seems that these suggestions should be the more welcome, because they lie in the direction not of crippling our science, but of greatly enhancing its power by giving it a suppler and more abstract symbolism. We are all working with Roman numerals and deceiving ourselves by attaching costly flourishes; let us stop discussion of the flourishes and adopt the Arabic digits. The history of our science, by blind accident, has trained us to great flexibility in responding to symbols; let us take advantage of this flexibility, now that we need to free ourselves from the magic of symbolism. Once we take advantage of the purely external character of our symbols and learn to make them do what we want, the door will be open for uniformityuniformity as between different languages and as between different scholars. 10 The existence of such persons is not to be assumed too lightly. Periodical articles are written for scholars, not for college students. BOOK REVIEWS Die Sprachfamilien und Sprachenkreise der Erde. Pp. xvi + 596 and Atlas of 14 maps. By P. W. SCHMIDT S.V.D. (Kulturgeschichtliche Bibliothek herausgegeben von W. Foy; 1. Reihe: Ethnologische Bibliothek mit Einschluss des altorientalischen Kulturgebietes 5.) Heidelberg: Carl Winter's Universitätsbuchhandlung, 1926. The Indo-Europeanist who opens this book will find himself in a strange world, colorful, adventurous, even heroic,—the world of Humboldt. One journeys to the ends of the earth (pp. 7, 9); danger and hardship are not considered. Into this world the reader is guided by the strong, kind, and warmly human personality of Father Schmidt. Then with a start one realizes that this strange and vast world of human language is only that larger land within which lies our own wellcultivated domain. Upon this realization follows a profound regret: why have not the methods of our field been carried to the ventures without? Thus, 'older' and less old languages play a part in the discussion (p. 5); one suspects the criterion of this to be one which would make modern English 'older than the language of King Alfred (cf. the note on Hottentot and Bushman, p. 11). The whole second part of the book, 'Die Sprachenkreise und ihr Verhältnis zu den Kulturkreisen', will fail to instruct or convince, for it ignores what Indo-European has taught us about the variety of linguistic structure (even within a single stock) and of its mutability in the course of time. No one would today set up the simple concept of a 'genitive case' for even all Indo-European (e.g., modern English, French, the German dialects); Father Schmidt does so for all languages, patiently observes whether the 'genitive' precedes or follows its headword, and compares his results with a similar ethnologic schedule of 'matriarchate', deciding that originally ('die älteste Stellung') the genitive preceded, and that postposition of the genitive goes hand in hand with the 'matriarchate'. Yet, along with Meillet and Cohen's Langues du monde, Father Schmidt's book will be indispensable for the general study of language, since the first part, 'Die Sprachenfamilien der Erde und die Geschichte ihrer Erforschung' contains a list of the languages of the earth (too a optimistic, to be sure, in the assumption of relationships) and a splendid bibliography. It seems incredible that one man can have read and can know so much. In a human and manly way Father Schmidt speaks (p.iv) of the devoted labor of his predecessors; for fear of seeming thankless, one scarcely dares express the wish that all these workers had used less of sentiment and philosophy and more of the simple methods of science. LEONARD BLOOMFIELD The English Language in America. Pp. xiï + 377 and 355. By GEORGE PHILIP KRAPP. New York: The Century Company, for the Modern Language Association of America, 1925. It is not easy to arrive at a clear conception of the merits of these two large volumes, the first to be published on the Rotary Fund of the Modern Language Association. There is a vast amount of material gathered here, much of it new; there are courageous, if somewhat loosely connected, attempts to interpret the material which in spite of its fulness gives us only a fragmentary account of past usage; and there is the serious attempt to correlate the history of the language with the ethnic, the social, and the cultural history of the country. The undertaking is so daring and accomplishes so much in the face of the deplorable lack of earlier systematic investigation, especially on the historical side, that we are under great obligation to the author for having so arduously and courageously performed this pioneer work. And if we do not agree with him in one point or another, or if we should disagree with him even on fundamentals, we are nevertheless deeply indebted to him. It is to be hoped that some of the problems formulated and tentatively treated by the author will engage the attention of other investigators. Much investigation is needed before we shall be able to outline the history of our language with a sure hand. The book is somewhat loosely organized. The first volume deals with the general historical problems, vocabulary, literary dialects, and style; the second with pronunciation and, rather summarily, with inflections and syntax. A good bibliography, an index of subjects and names, and a very helpful index of the words treated are appended to the second volume. The author records (II 8) the conviction that American English of the present day is more homogeneous than our English of a hundred years ago or of the time of the Revolution; that there has been in progress since colonial days a standardization of the language, although there has been no such definite ideal of good usage in America as in England. 'Good English in America has always been a matter of the opinion of those who know, or think they know, and opinion on this point has always been changing', and, of course, has never been the same in the East, the West, and the South. Since no one section of the country has ever for any length of time dominated the others either culturally or politically, no one standard, even of the shifting and elusive kind defined by the author, ever came into existence, with the result that we now have an Eastern, a Western, and a Southern type of polite speech, each of some degree of uniformity. The Western type, one should add, is current from the Hudson River to the Rocky Mountains. Krapp gives (I 37–40) convenient lists of the distinctive features of these three types. To the Eastern characteristics there should be added (a) the closer and more diphthongal articulation of the vowel sounds in bay and toe, and (b) the preference for unstressed [1] in houses, naked. For the Southern there should be recorded (a) the tendency to pronounce the vowels in bad, bed, bid, bud overlong and tense, (b) the still widely current close (o.] in four and afford (closer than in the Western type), and (c) the slighting of weak vowels and whole syllables when unstressed; moreover, (ju) is nearly universal after dentals, as in new, duty, tune. As for the Western type, the common distinction between mid-slack long [9] of four, hoarse and low-slack short [o] of forty, horse, which the author claims (II 137) to be 'rarely maintained in practice', should be added; see Kenyon, American Pronunciation 120. The author calls these three types the New England Local Type', the 'Southern Local Type', and the 'General or Western Type' which (I 35) covers the rest of the country, and also all speakers in New England and the South at the moments when their speech is not local in character'. Do the Bostonians, the New Yorkers, and the cultured Southerners really have ‘moments when their speech is not local in character', and do they really ever speak the General or Western type of American English? Hardly! There is no such generally recognized standard in America, although the Western type has made effective inroads on the Eastern (Grandgent, Old and New 27) and the Southern, as one can observe in the Blue Grass of Kentucky. Krapp's General Type is fictitious at least insofar as pronunciation is concerned, and leads him to assume much more uniformity than actually exists and to make indiscriminate assertions. Witness the following statements. (a) 'In this matter of cadence, it is quite obvious to one familiar with various types of British speech, that the |