On the basis of the above negative criticism, we may perhaps formulate the following positive requirements for those reading selections which are to be the foundation for instruction in languages, namely that as far as possible they must

This order does not indicate the relative importance or value of the requirements, which might be difficult to determine. If there should be any disagreement between them, I suppose it is generally best to try to find some practical compromise. We must now pass on to examine some of these requirements more closely.

The use of connected texts in the elementary teaching of languages has already previously been tried, but it seems as if in the effort to avoid the Scylla of disconnected sentences it has been impossible to escape the Charybdis of such texts as Chateaubriand’s Atala, Dickens’ Christmas Carol (MÉthode Toussaint-Langenscheidt), the New Testament, or CÆsar’s Gallic War, etc. How often after such experiments, when the pupil was overwhelmed and did not learn anything because he was to learn everything at once, has not the teacher returned in despair to the disconnected sentences.

But between the two extremes there is no doubt room for the golden mean of beginning with quite short connected pieces, and then gradually, as each lesson may be lengthened, passing over to longer texts—of course this does not necessarily mean that a whole piece must always be taken for each lesson; the breaks in the lessons do not need to correspond to the breaks in the text-book.

Anecdotes meet the requirements in so far as they are short connected pieces, and therefore they play such an important part in many readers. But yet they are not quite the thing, especially when they are used in too great numbers. A pointed anecdote can only be really funny once; if it is to be repeated many times, it soon becomes stale and indeed more tiresome than most other things. And just the very quality which makes it amusing makes it less valuable for teaching purposes; that is, an anecdote must by its very nature contain as few words as possible; but it is better for beginners to get a little broader colouring, so that the most necessary words and phrases may recur frequently. If many anecdotes follow one upon the other, it is not easy to avoid frequent jumps between totally different spheres of thought and accordingly between totally different worlds of words; this increases the difficulty, and the result is apt to be that words and expressions once learned are soon forgotten. Anecdotes depending upon puns cannot be appreciated at all without full familiarity with the words resembling each other, and that can only in a minority of cases be assumed for our pupils. The best way to use anecdotes in teaching languages is to let them serve as spice in or in connection with other pieces, especially descriptive pieces, so that the words used in the anecdotes may there appear in their natural surroundings. This can best be done in short stories about animals; in my own books for beginners in English, I have taken several such pieces from purely scientific works by Sir John Lubbock, Romanes, Tylor, etc. I mention these as examples of a kind of texts which seem to me to be especially attractive (but which are neither so easy to get hold of nor to concoct), because they give entertaining and sensible information about things which are often neglected in the natural science instruction itself, and at the same time they give an opportunity of learning a good deal of useful language-material without being too difficult. The pieces which are merely descriptive of nature, and which Sweet lays so much stress upon, have the advantage that they in a still greater degree allow of the employment of the most indispensable material of language, and that a number of the sentences may be made self-explanatory (v. below). There are, however, but relatively few subjects that can be dealt with in this way—the most elementary natural phenomena—and when they are not written in such a masterly manner as in Sweet’s Elementarbuch des gesprochenen Englisch, there are apt to be so many well-known truths told in these pieces that the interest flags.

In deciding on what will be of interest as a selection for reading, differences in age must of course to a great extent be taken into consideration. But it is an experience which I myself have had, and in which many teachers bear me out, that beginners in a foreign language may very well be interested in certain reading matter even if they are beyond the age when corresponding things would interest them in their native language. So one must not be afraid of childish texts; but by this I do not mean to recommend a certain kind of juvenile literature which flourishes in all countries, and which aunts, especially the unmarried ones, often think that children appreciate, and so they themselves also proceed to produce it in large quantities, that is, milk-and-water stories and verses about the reward of good children and the frightful punishment of the naughty ones; both young and old find such “literature” nauseating, and it were best to avoid it in text-books in foreign languages. But there is another class of literature, that collected by folklorists, which is orally transmitted from generation to generation, and which shows its vigour by being continually amusing and by continually shooting new shoots. Much of it can successfully be used in teaching languages; and that which amuses a French child of five or six years may often amuse an English child of ten or eleven or even more, because in the foreign language it gets the charm that always is connected with the unknown.

Much of this material—and of other material, which, without belonging to popular tradition, is related to it—is in verse-form, which has the great advantage for our purpose, that rhythm and rhyme naturally rivet the words and expressions fast to each other, so that the memory gets hold of them like an unbreakable chain. It is only with great difficulty and with much repetition that prose sentences can be inculcated in a certain given form; but to learn verse is like play—it learns itself. If therefore the poetry of art, with its more or less unnatural language, is unsuitable for the beginner, the little witty natural verses of the genuine children’s literature are, on the other hand, excellent. But of course not even these are always pure pearls, and there are many of them to be rejected as containing impertinences, nonsense-words, fragments of antiquated language, or words which beginners have no use for; it seems to me, for instance, that ViËtor and DÖrr should not have transferred the nursery rhymes wholesale (even the old forms with —th in the third person, and much more) into their otherwise excellent English reader.

With respect to the requirement that the reading must be easy—or rather that there must be gradual progress from easy to difficult—it must be recognized that difficulty may depend upon several different things.

In the first place, the subject-matter may be too difficult; it ought never to be beyond the horizon of the pupils. As previously remarked, in the very beginning, one may even take something simpler than what would otherwise be suitable for persons of that age. But later, on the other hand, the subject-matter ought not to be too light; it is well, as soon as possible, to use matter which really has a permanent value of its own. A large part of the reading will no doubt always be taken from lighter literature, and most of it will not cause any real difficulty as far as the comprehension of the subject-matter is concerned. But in addition to that, there ought surely to be read to a far greater extent than has hitherto been the case in modern language instruction, matter which cannot be understood without some serious thinking, articles on natural science and on human relations in the widest sense of the word, political speeches, etc. Many teachers seem to be afraid to read anything else with their pupils than the most insignificant novel-literature whose contents furnish starvation food. A little friend of mine seven years old once said to his mother: “I like that best which I can scarcely understand.” He thereby expressed the same thought as Dante when he said that man is not happy unless he strains every nerve, or Stuart Mill in his remark: “A pupil who is never required to do what he cannot do never does what he can do.” All instruction must spur the pupil on with problems that are not too easy; in the first stage of instruction in languages, there are problems enough in the purely linguistic difficulties; later on the contents of the reading, too, ought to require some independent powers of assimilation. Sometimes it may even be best to choose selections where the language is very easy, but the matter rather weighty—especially in teaching according to the reform-method, where subject-matter is necessarily assigned a more important part than hitherto, and where even an easy text can in various ways be advantageously employed as a means of training in purely linguistic skill.

Even linguistic easiness or difficulty may depend upon different things. Difficulties in pronunciation ought not to be piled up, a caution applying especially to selections for the very first beginners. Some teachers try to begin with words which may be almost or wholly pronounced with sounds occurring in the native language of the pupils. Aside from the fact that in most cases it only leads to disappointment to exaggerate the resemblance between the foreign and native sounds, this principle may easily lead to slovenliness at a stage when it might involve the most dangerous consequences. The pupil ought from the very first lesson to have the clearest sensation of being on foreign ground, and he ought to realize that the foreign sounds cannot be learned without work. But the difficult sounds ought not to occur too many in succession or in too difficult combinations. It is perhaps best to begin with words of one syllable, but this need not be strictly carried through. I do not, however, attach so much importance to mere difficulties in pronunciation that I would advise an otherwise suitable opening selection in a French reader for beginners to be discarded because it contained such difficult words as manger and chien. It cannot be long, anyway, before the pupils must make acquaintance with, and, what is more, master all the sounds in the language they are about to learn. By difficulties in pronunciation here I mean the real ones, and not such apparent difficulties as are due to freaks of orthography; it is equally troublesome for a German to pronounce English pear and pair; such difficulties as are found in English scarce, fatigue, victuals, French eut, pupille, pitiÉ, balbutier, etc., may be overcome by a panacea which I shall come to later, namely, phonetical transcription.

Furthermore linguistic difficulty may be due to the use of too many new words, and in this respect the best principle at all stages is: as few new words as possible. Every one who has read such pages as often occur in Zola or Daudet, where technical expressions are abundantly piled up, will have had the experience that even with the most careful reading or study it did not take long before all the new words were just as unfamiliar as before the selection was read. Likewise, when one sets to work to learn systematic vocabularies like PlÖtz’s Vocabulaire SystÉmatique, it requires enormous exertion and a long time to learn them, and it takes an amazingly short time to unlearn them again. But if, in the course of one’s reading, the new words turn up occasionally at relatively large intervals, then the mind is able to absorb the one before the next appears; the intervening passages, which contain only familiar things, manure the soil, as it were, for the new things that are to be sown in it. Ten or twelve new words are more easily and more thoroughly learned when they are scattered over five pages than when they are crowded into ten lines, and then besides there is the benefit to be derived from the recurrence of a number of usual words, to say nothing of sentence-constructions, etc., so that he who has read those five pages has had more opportunity to familiarize himself with the idiosyncrasies of the foreign language than he would have had in ten lines; the apparent waste of time in reading the longer piece has really been profitable, for the capital which had already been acquired in the language has in that time borne interest and compound interest.

Now since it is also better, as we have said, to learn five absolutely necessary words than twenty-five of less importance, it is of course the duty of the editors of text-books in large part to revise the selections which they reprint, so that that which is of linguistic value for the pupils may be cultivated at the expense of everything that is unusual or odd. Texts whose subject-matter is good, but whose language makes them impossible for our purpose, may often be made pedagogically practicable by means of curtailing, paraphrasing, and adaptation in various ways; many popular fairy-tales in the collections of folklorists may be used if one only will take the trouble to translate them from the dialect in which they are written. Such a splendid little story as Mrs. Ewing’s Jackanapes, which is frequently read as it stands in German and Swedish schools, is, according to my judgment, too full of literary expressions and unnecessary words to be easily comprehended by our little pupils. In the passage which I have selected for my own primer, I have therefore in several places made considerable omissions, and the style has throughout been made more colloquial and direct, by means of corrections like these for instance: having ceased to entertain (given up) any hopes of his own recovery. " Tony tumbled off during the first revolution (before he had gone round once). " And what bright eyes peeped out of his dark forelock as it was blown by the wind! (he had!) " told him that he must be on his very best behaviour (behave properly) during the visit. If it had been feasible (possible) to leave off calling him Jackanapes and to get used to his baptismal (real Christian) name of Theodore before the day after to-morrow it would have been satisfactory (she would have done it) " said J., shaking his yellow mop (hair), and leaning back in his one of the two Chippendale armchairs in which they sat (the italicized words left out) " took their early promenade (went out for their walk) earlier than usual " His golden hair flew out, an aureole from which his cheeks shone red and distended with trumpeting (left out). It is very probable that on comparing the original with the revised text, it will be found that some of the colouring has been lost; I merely maintain that the pupils gain thereby. The more it is insisted upon (as according to the reform-method) that the selections are not only to be read but also to be mastered, so that their language becomes the mental property of the pupil, the more necessary is such revision. It is clear that as the pupil progresses, the texts may become more and more literary, and for various reasons the advisability of such curtailing and adaptation becomes more questionable.

As a sample of such revision, I shall reprint a part of an anecdote, (A) as it ought not to be given in a book for beginners (but as it stands in a certain English reader for foreigners) and (B) as it stands in Sweet’s excellent edition:—

Finally the difficulties may be grammatical. These are the difficulties that teachers have been most afraid of according to the old methods, so that they have even preferred to give up almost all sense and connection in the subject-matter rather than make a break in the systematical progress in grammar. Such a form as pu was not allowed to occur before the pupils had learned the whole conjugation of pouvoir pouvant pu je peux, etc.; these forms must be learned connectedly, it was said. But the irony of it all is that this “connectedly” means that they are learned out of all connection—and therefore to little profit. When the pupil is required to “understand” the forms which occur in his reader, it will be found on closer examination that this means merely that, for instance, il a is understood by the one who knows that it is 3 pers. sing. pres. of avoir, or who at least knows the formula j’ai, tu as, etc.; that yeux is “understood” by the one who has learned that it is an irregular plural belonging to the singular oeil, etc.; in short, to “understand” means here to know where the form in question belongs in the grammatical system; and the forms must be given in exactly the same order in which they are arranged in the grammar, the present before the past tense, etc. But what has the beginner got to do with all this system? The idea is not carried out consistently either, for when all the exercises on accidence have been gone through, it is generally the rule to pass over to connected (unrevised) texts, where such a form as puisse may occur, but the only thing that the pupils get to know about it is that it is subjunctive, for it may easily take a year or two before they learn why the subjunctive is used. Why is syntax less important than accidence? To be quite consistent, it ought no more to be permissible for a syntactical phenomenon than for a form in accidence to occur before the corresponding grammatical section has been learned. But since it seems to be inevitable that we must be inconsistent on some point or other, it is no use beating about the bush; in other words, we must not be afraid of using irregular forms in the very first selection.

Grammatical irregularities, viewed from a pedagogical point of view, fall into two entirely different classes, which are too apt to be treated as if they were co-ordinate. In the first place, all languages contain a number of irregularities which play a most insignificant part both in life and in literature, because they occur so seldom. When the users of the language produce them at long intervals, it is generally with the utmost caution, because they merely have a hazy conception of what the proper form of the expressions ought to be. But they are taken up in the grammars, and as soon as one grammarian has caught sight of one of them, it is carefully copied in all succeeding grammars for the sake of completeness. Foreign grammarians are even more inclined than the natives to pay attention to everything of that kind because they have no instinctive feeling of what is rare and what is common. In some English grammars which are used on the Continent, there may still be found I catched, I digged, I shined, I writ, as the preterite forms of I catch, I dig, I shine, I write; in one, I find given as two different verbs I weet, wit or wot, past tense wot, and I wis, past tense I wist. What a big mistake it is to include such musty and impracticable forms, we can best judge from our own language—but in those French and German grammars which we ourselves write there are things which are just as bad as the above offences in English. When I went to school, I learned the following rule about the plural of travail, “Travail has travails in the plural when it means a report from a minister to the king or from a subordinate official to the minister; likewise when it means a machine to hold unruly horses, while they are being shoed.” This rule is thus criticized by Storm: “Now I must say I have read many hundreds of French books in my day, but so far as I remember, I have never come across travails in modern literature! In the sense of report, it occurs in Mme. de SÉvignÉ. An educated Frenchman, when asked if the word was used with that meaning, answered me that he thought it was no longer used. So one would expect that the word had long ago ceased to have any show in modern grammars, but it seems to be continually creeping in again.”

However, it is easy enough to take a position with respect to this first kind of irregularities; they ought to be removed from the instruction as radically as possible; they ought to be weeded out root and all to a far greater extent than has yet been done in most text-books, even if it must be admitted that something has been done in this direction of late years. It is quite another matter when we come to the other kind of irregularities, which are found in the very commonest words, in words like German ist war, kann konnte, geht ging, ich mein, mann mÄnner. Those irregularities the pupil must learn, and learn thoroughly—there is no doubt about that. The only question is, at what stage? before or after the regular inflections? Most teachers will answer, after. That a systematic grammar first gives what is normal, that which can be expressed in general, comprehensive rules, and then afterwards mentions the exceptions, the isolated phenomena, that of course is all right. But it does not necessarily follow that the pupils ought to familiarize themselves with the forms in the same order. What is won thereby? Perhaps some advantage for the theoretical knowledge about the language. But the loss incurred by this method of procedure is undoubtedly far greater. For it will be found to be absolutely impossible to arrange texts which are the least bit suitable without using irregularly inflected words, so indispensable are they. The dread of being unsystematic by taking up exceptions immediately is one of the causes of the prevalence of the disheartening series of detached sentences without any sensible meaning. It is only by freeing ourselves from this principle which requires rules first and exceptions later that we shall be able to get good texts for the teaching of beginners. Furthermore, by beginning with the regular forms, we perhaps run the risk that the pupils will analogically apply the rule even to the exceptional words, whereas the irregular forms generally deviate so much that they preclude the possibility of such mistakes. Those who have learned that the plural in English is formed by adding s, may perhaps construct such improper forms as mans, childs, but the plural forms men and children are not apt to tempt the pupils to inflect other words after the same pattern. But the moral of this is not that we are to turn the customary method of procedure upside down, and systematically learn the exceptions first. Here, too, nature must be our guide; just as persons talking within a child’s hearing never stop to consider if the words they are using are regular or not, so we ought not to be too painfully careful in selecting or arranging the first reading-exercises in a foreign language; we ought to choose what is otherwise good and take the forms as they come, wasting no words at this stage to explain their place in the system. In other words, the deviating forms must be learned as if they were merely matters of vocabulary. If in one of the first pieces there stands Il y avait une fois un roi et une reine, it is enough for the time being if the pupil is told that il y avait = there was; the forms for “there is” and “there has been” he can learn another time when he has use for them, and then the teacher can refer back to this early piece and remind the pupil about the related form which he learned before. For beginners in French, peux—“can” is just as difficult (or easy) as peu—“little,” and faire—“make, do,” as fer—“iron,” and it makes no difference if the one is regular and the other irregular. Indeed, an irregular plural like geese is even easier for Danes than the regular bees (on account of the z-sound); likewise, it is easier for an Englishman to learn the German irregular forms of comparison besser best than regular forms like sÜsser sÜssest. Later when the time has come for a more systematic study of the grammar, it will be rather an advantage that a number of the “exceptions” already have occurred at so early a stage that they are not at all felt to be strange and unusual.[4]

On the other hand, the beginner ought to be spared such grammatical difficulties as are due to complicated sentence-structure. All sentences ought from the very beginning to be constructed as evenly, simply and clearly as possible; co-ordinate independent clauses ought to be, if not the only, at least the predominating type of sentence. Not even, for instance, in the second year of Latin instruction, although there are just as many hours devoted to Latin in a year as generally fall to the share of modern languages in the course of two or three years, is it justifiable to let the pupils read the long passages of indirect discourse in CÆsar; they ought not to occur until the pupils are so far advanced that they could easily understand the same matter when directly presented. This is also a point to be kept in mind for any one who undertakes to revise the selections for reading according to the suggestions given above.