book will certainly act as an incentive to those who are interested in the study of Bantu, and will serve as a guide to those who study the Zulu language. It is too comprehensive to be examined further here. In the introduction the author gives general information about Zulu and the methods which he applied during his investigations, and in an appendix he discusses the orthographies of Lepsius, Steere, Meinhof, and the International Phonetic Association.

Mention must also be made of A Sechuana Reader in International Phonetic Orthography (with English translation) by Daniel Jones. (London. The University of London Press Ltd., 1916, xxxviii and 44 pages). If the book had been written for foreigners who did not know any Sechuana, it might be possible to accept the transcription; but the book was intended chiefly for Bechuana readers. The usual transliteration would therefore be preferable, because every Native will pronounce the sounds in his own language correctly even where the transcription is not adequate. But it must be simple. In my opinion the transcription that is nearest to the usual Latin one will be the most satisfactory for school purposes. The first part includes an exposé and explanation of Sechuana phonetics, and the corresponding transcription; the second part contains original texts with literal and fluent English translations. Rohan-Chabot Mission. Sous les auspices du Ministère de l'Instruction publique et de la Société de Géographie. Tome III, fasc. 1, Linguistique, Le Groupe sudouest des Langues Bantoues (Paul Geuthner, 1925, viii and 176 pages). Part of the investigations of the Rohan-Chabot Expedition, which worked in Angola from 1912 to 1914. The introduction contains a summary of African peoples and languages, which appears to me somewhat cursory and of little use. The first chapter describes in detail the tribes of that part of Angola in which the expedition worked and from which the languages were taken, for the most part they were obtained from two individuals who were brought to Europe. Other chapters contain shorter vocabularies, phonetic and grammatical texts. At the beginning the phonetic transcription is carefully explained. The Appendix gives various palatograms taken by Abbé Rousselot himself, but these are not explained. In any case the work is a valuable contribution to the knowledge of Bantu languages in the south-west district which, but for this, is little known.

PAUL SCHEBESTA.

THE PRACTICAL ORTHOGRAPHY OF AFRICAN LANGUAGES

BY A. LLOYD JAMES

IT is a very significant fact that the first memorandum issued by the

International Institute of African Languages and Cultures should be devoted to Practical Orthography,' for in no branch of African study is there such diversity of practice as in the orthography of the languages. Moreover it is probably no exaggeration to say that, inasmuch as the reading and writing of his language form the Native's first step in education, no question is of greater urgency than this matter of presenting the written language in a way that shall be logically satisfactory.

The conditions under which linguistic research has been carried on in the past have been responsible for the present diversity of practice. Each language was studied separately by different investigators, who differed among themselves in linguistic ability and in nationality; there was no central body to co-ordinate the work. The result is that each language has had its own system of orthography, reflecting in a remarkable manner the idiosyncrasies of its inventor. The same sound is represented in various ways, and it may happen that two separate sounds are represented by the same letter. More frequently it happens that essential features of a language are not indicated in any way, and it has never been quite clear whether a given orthography has been designed for the benefit of the Native, with a view to teaching him to read and write his language, or for the benefit of the European desirous of studying the language. The results are evident only to those who have tried to teach native children to read their language as written by the European orthographer, and to those who have had to learn languages from text books prepared on these lines, for it is often quite impossible to grasp any idea of the pronunciation from the orthography.

It must not, however, be imagined that this pioneer work has been in vain; it has served a very useful purpose, and it has enabled us to benefit from the experiences of our predecessors. Now that there exists some body whose function it is to co-ordinate African research, we may hope that isolated researchers will avail themselves of its advice and take advantage of its recommendations.

One fact would appear to be abundantly proved by experience; it is that no man should set out to write a grammar, a reader, or a dictionary of an African language merely because he has lived among the people and can

'Practical Orthography of African Languages, published by the International Institute of African Languages and Cultures, London, 1927. Price 6d.

speak their language. Such a man may, in the end, prove the ideal man for the task, but only if he undertakes the necessary training in linguistic research. On the other hand, a competent trained linguistic researcher may, although quite ignorant of the language, produce better work than the untrained speaker of the language.

This pamphlet on the Practical Orthography of African Languages should be the researcher's first step in his science, for it epitomizes, possibly in too concentrated a form, all the knowledge that recent research has brought to bear upon the very complicated problems that are involved in the establishment of an orthography. These problems are mainly of a three-fold nature; this pamphlet is the first public recognition of the triple aspect of the question, and it will, in after years, when time has revealed to a future generation the full stupidity of our rudimentary African orthographies, be regarded as the first attempt to handle the question with understanding. The three aspects of the question are:

1. Phonetic.

2. Psychological and Pedagogical.
3. Typographical.

I. PHONETIC. Much attention has been devoted to this question in recent. years but much remains to be done.

It must now be accepted that no system of orthography should be devised for any language unless and until a minutely accurate phonetic and tonal analysis of the language has been made. Existing systems of orthography have mostly been made without this preliminary work and the results are consequently unsatisfactory. Important distinctions are unrepresented and the written language is often nothing but a travesty of what it should be. A recent cursory investigation of Chikaranga, for instance, reveals that there are in this language two very distinct d sounds, as distinct shall we say, as English t and k. There is no indication of this distinction in the accepted orthography. What should we think of a Martian visiting these islands to report on our language, if he made no distinction between t and k, spelling king with a t, and making such a pair of words as tease and keys look alike? Any impartial investigation would at once condemn the Martian system of orthography, but it is no more unsound than our orthography of Chikaranga which in addition ignores other distinctions that our investigation is proving to be necessary, e.g. two separate t sounds, two separate ts sounds, and all the vital tonal distinctions. What is the African to make of a system that spells in the same way words which he and his parents know to be different? What impression does the European produce upon the native mind by speaking the language as represented in the current orthography? What impression does he, whether he be missionary, teacher or anthropologist, gather of the native mind and its content when he fails to hear these im

portant distinctions? How many misunderstandings, false impressions, false convictions, and false theories have arisen from this lack of preliminary analysis, without which there can be no scientifically sound system of orthography.

II. PSYCHOLOGICAL AND PEDAGOGICAL. This heading comprises a multitude of problems, or possibly aspects of one problem that we are only beginning to realize. It is beyond the scope of a short article to examine them in detail; we can but touch upon them. Armed with the data provided by the phonetic analysis, the orthographist is faced with the fact that generally he has not enough letters for his task. We have, in all, twenty-six, five of which must do duty for vowel sounds and twenty-one for consonants. How are we to represent, shall we say, thirty consonant sounds in any given language? How are we to show the two d's in Chikaranga? In the past, when an attempt has been made, it has been done by placing a diacritical mark over, under, or after one of the existing letters. Phonetically, this is excellent. Psychologically and pedagogically it is unsound, because recent investigation shows that in the process of reading the eye takes in a mass impression, and does not wander about letters, inspecting their every detail. A multiplicity of such diacritical marks necessitates an individual circumnavigation of letters by the eye, and naturally slows down considerably the speed of reading. In addition it lessens the legibility of the type by blurring its outline, and it is to be remembered that in the process of reading, it is the bold outline of the type, at any rate as far as Roman type is concerned, that is of importance; furthermore, it is the top outline of the word that matters most. Anything, therefore, in the nature of a dot, a comma, a stroke over or under letters, confuses the outline and impedes the recognition of the word.

This reacts pedagogically by increasing the difficulties attendant upon learning to read, and it is a matter of common experience that children, and indeed adults, omit most of these diacritics in writing. How many of us dot all our i's in writing? The crossing of the t is another frequent omission, not because it is a diacritic but because, in writing, it involves going back to the letter after the word is finished, as is the case in the writing of all diacritical marks.

III. TYPOGRAPHICAL. The letters that cover this page owe their present form to the fact that centuries of experience have convinced printers and type founders that these forms are most convenient, most durable, most legible, and most easily fitted into the body of the well-printed page. In the end it is the printer who settles what our printed language shall look like. On the whole, he has done his task well, but he would have saved us some anxiety if he had not been so eager in the past to rid our printed page of one or two very useful letters that we badly need now, e.g. the old English ð

(for the th in then), þ (for the y in thick), 3 (for the th sound in yes, and for the long lost gh sound in night); the long s would have been very useful too. As it is, we manage without them, but the printer does not hesitate to recommend new letter forms for the printing of African languages. To him, dots and dashes over and above letters are an abomination. Some people imagine that in the setting up of type the diacritical marks are put in separately; this is not so, for a letter with a diacritical mark must be separately cast, so that if we are faced with the alternative of deciding in favour of either a new letter or an old letter with a diacritic, as far as typecutting goes, there is no advantage in favour of the diacritic. Types with diacritics are less durable than others, because the dots get worn out, or knocked off in everyday wear and tear. The types are not easily distinguishable: and the printer makes frequent mistakes.1 Capital letters with dots and dashes over them require types so high that they spread into the line above and interfere with spacing. It is the printer, and only the printer, who decrees that French capital letters shall have no accents. Such accented capitals involve the process known as kerning, which reduces the strength of the type, for the kern, an unsupported projection of metal bearing the diacritic, is liable to be broken off.

We are left, then, with no alternative but to use new letters to do duty for sounds that we cannot represent with the twenty-six letters that are our inadequate inheritance from the printers of the middle ages. The psychologist, the educationist, and the typographer recommend this course; language specialists have been divided into two schools: the diacritic. school and the new letter school.

This pamphlet has ranged itself boldly on the side of the new letter. Indeed, in face of the evidence it quotes, it would be difficult for it to do otherwise. The new letters will probably disturb the European more than the African, but this is no cause for undue concern; we have advanced sufficiently far in our intercourse with Africa to realize that it is no longer the European's standpoint that is vital. Any step that facilitates the education of the Native needs no other recommendation; a sensible orthography of his language is the first step, and probably the most important step. To hand on to the African the hopeless chaos that prevails in English orthography, for instance, is a breach of trust. If it is complained that these new letters will be hard to learn, we can only suggest that to a native child all letters are new, and equally hard.

In Les Sons du Langage et leur représentation dans un Alphabet Linguistique général (P. W. Schmidt, translated by P. J. Hermes, Salzburg, 1907) there is a list of 27 errata, 21 of which are due to wrong diacritics.