Problems of language and modern writing systems

'To purify, fix and give splendour to the language and its spelling.'
Aims of the Tje Real Academia Española de la Lengua, founded in Spain, 1714.

Bolshevik soldier at a street barricade in Oslo: 'How is the revolution coming along in Oslo, comrades?'
Norwegian: 'We're still fighting over how to spell it.'
Danish satirical cartoon, 1919

There are differing problems in matching language and writing systems for:

Some languages by their very nature set problems for how they are represented in writing
Arabic, Finnish, Hungarian and Norwegian, Danish, Faroese, French, and Swedish are examples of writing systems with script-based difficulties.

Fortunate in language and spelling. Italian, Spanish, Malaysian and Indonesian are fortunate in the simple nature of both language and spelling systems, Other consistent alphabetic orthographies include German, Russian, Turkish, Czech, Malaysian, Indonesian, and most English pidgins, but not English itself.

Few alphabetic scripts can achieve or maintain perfect sound-symbol correspondence.

Languages that are use alphabetic writing differ greatly among themselves in their linguistic structures and also how they are represented. An alphabetic system's relation to the spoken word can range from the close and clear 'transparency' of Serbo-Croatian to the dubious 'opacity' of English and French. Some orthographies, such as Italian, Spanish, Indon-Malaysian and Finnish, have relatively 'regular' standardisations of how sounds relate to letters. Biases can be towards 'spell as you speak', as in Croatian and White Russian, or towards morphemic representations, like English and French, that conserve representations of meaning. Some are transitional systems like the Devanagari syllabo-alphabet and Ivrit and Arabic graphemics. Spoken languages can have different morphemic and inflectional structures, and these too can be represented in alphabetic writing systems in different ways.

Although English spelling is outstanding in its difficulties, it is not in a class completely on its own. Alphabetic orthographies of other languages can also have features which complicate them. Systems that develop 'naturally' over time without benefit of coda or committees are not to be taken on trust on those grounds as likely to be the best possible.

On the other hand, deliberate efforts to improve anything cannot expect to avoid some mistakes. ('It-Seemed-A-Good-Idea-at-the-Time'.) Other modern alphabetic orthographies besides English can still incorporate handicaps for readers, writers and learners, both accidentally, and with the best of intentions.

The language itself can make a writing system difficult.

FINNISH

Intrinsic difficulties in the language can cause problems for its spelling, as in Finnish.
Finnish orthography was hotly debated throughout the late 18th and early 19th centuries before it was made completely regular and the responsibility of the Finnish Academy of Science. Now Finland has perhaps the most regular of all European orthographies in its close sound-symbol correspondence and consistency. Even the names of the letters are user-friendly because they show their pronunciation.

But learners still have some reading and spelling problems - even though they are not in the same class as English. This fact has been claimed to show that a regular spelling is not a solution to learning difficulties, but the problems are due to the language and to the intrinsic difficulty of learning to read whatever the writing system. The regular spelling itself is a major reason why Finns have one of the highest literacy rates in the world.

Finns have tried to set the record straight (e.g. Kyöstiö, 1974). The special difficulties in learning to read result from two features of the spoken language - the remarkable length of words and some phonemes (speech sounds) that are barely distinguishable from each other.
Finnish words tend to be so long, often with seven or more syllables, that text in English is 25% shorter than the same text in Finnish. 'Long' words in English such as elephant and crocodile are, after all, only three syllables long. Long words are more difficult to learn to read and write than short ones, unless the length stands out so strikingly and distinctively among the short words that the length itself aids recognition - such as the words elephant and crocodile in a children's story.

The critical differences between some Finnish phonemes are very small and they are difficult for beginners to detect. There are also a large number of diphthongs and sliding vowels. tule, tulee, tulle, tullee, tuule, tuulle and tullee are words that have confusing tiny differences in pronunciation as well as in spelling. Sentence structure is also complex, which makes comprehension of text more difficult; writing lacks the intonation clues of spoken language, and the complexity can put a burden on the short-term memories of beginning readers who are still reading slowly.

Nevertheless, Finland claims with some justice to have one of the most literate and active reading populations in the world. Venezky (1973) concluded from a study of Finnish children that more rapid acquisition of literacy was possible when there was greater phoneme-grapheme correspondence, and Finns themselves give credit to the ease of learning the spelling system. When children begin school at the age of seven, fifteen per cent are already literate enough to start in second grade, since the spelling system requires only a few clues from adults for children to pick it up. Kyöstiö (1973) estimated that Finland has about half the proportion of students with reading difficulties as in the U. S., and their backwardness was attributed mostly to lower intelligence, general language difficulties and environmental problems. Consistent representation of speech sounds has not produced mere 'barking at print' in reading, as some English writers such as Gillooly (1976) would fear, since reading for meaning is well developed in the adult population.

NORWEGIAN

The Norwegian form of 'Keep-a-Problem' also includes the daunting length of words. Words of 15 letters long are common. However, language and spelling have been described as 'relatively regular', with a limited number of orthographic concepts, and few exceptions to them. Little rote-learning is needed. Many, if not most children are able to learn to read from untrained parents before starting school at seven.

Norwegian spelling was a reform of the Danish-origin 'book-language' and it was brought in to celebrate Norway's 19th century separation from Denmark, although the most 'patriotic' partisans wanted to replace it completely with Nynorsk (New Norse). The 1885 solution officially recognised both forms, but there is continuing contention that is influenced by regional and socioeconomic factors. Nynorsk has been preferred by rural areas, speakers of western dialects, and primary school teachers.

Gregersen(1988) commented that Norwegian reforms show that reform in stages is not impossible, but it is messy. Complications result from 'juggling nationalism with linguistic requirements'. According to Tarnapol (1976 ) and Douglass (1969), reading difficulties in Norway are relatively rare, 'no more than 5-6%' of Norwegian children have any difficulty at all in learning to read, and Norwegians usually say that the reason is phonetic regularity of the spelling.

HUNGARIAN

In Hungary, reading difficulties also result more from the language itself than from the spelling, which has regular sound/symbol correspondence. There are some difficult phoneme discriminations, and many long words are due to the way that suffixes are built up. A 'whole word' global method of learning to read would have difficulties in Hungarian, since nouns are suffixed in all cases except one, so that their visual image keeps changing (Illés & Meixner (1976).

GERMAN

The German language, with its long compound words and unwinding syntax, presents more difficulty to learners and readers than does its regular spelling. German spelling has been described as a 'phonologically deep' system, reflecting units of meaning (Scheerer,1987) but sound-symbol correspondence rules are relatively consistent. Middle-High-German spelling has been relatively regular since the early 18th century. It is generally considered to be no real problem for readers or writers, despite wide dialect differences and although a few claim that it is still not regular enough. Learning to read is not considered difficult; boys have even tended to be superior readers to girls - very different from the Anglo-American situation. There is concern about 'dyslexics' and there appears to be about the same incidence of possibly congenital serious disability as in English, but not the great mass of 'backward readers' that clog English-speaking classrooms.
Research comparisons with English are difficult to find, apart from Upward's classic experiment of 1992 that showed that English students in an English university could spell better in German than they could in their native English. However, more studies are now confirming the greater efficiency of German spelling compared with English.

The major twentieth century reform has been to accept modern Roman print instead of the eye-straining Gothic script. The change was slow and contentious, fuelled by intense patriotic concern for all things German, and Gothic is still around, usually for ceremonial use. The old stereotype cartoon German used to wear glasses, even while wearing an iron helmet - suggesting that modern Germans are better off without Gothic. (Now it is the stereotype Japanese with the spectacles.)

Reconstruction of shattered societies often include repair of writing systems. Germany's postwar reconstruction of its wrecked society naturally, therefore, included attention to spelling. Since little scope appeared for other improvement, there have been efforts to change <ph> to <f> , and whole conferences have been held on the stormy issue of capitalisation of common nouns. Just like Anglo-Americans, German scholars have tended to ignore linguistic evidence if it comes from outside their own language boundaries. They like to rely on argument rather than research about whether capitalised nouns make it easier to skim text quickly and parse sentences for rapid comprehension, regarding this as plausible enough in a language which has such cumulative word-structures and syntax, and such long words and sentences. Today there is rearguard action against some minor official reforms of spelling and language.

ARABIC LANGUAGE AND SCRIPT

The language problem for modern Standard Arabic is that it represents no language that is still spoken today. It is is 'nobody's mother tongue'. Arabic script is used for dialects of its language of origin, but also for many north African languages for which it is not at all suited. Arabic for Arabs omits vowels at an early stage in reading, but vowels are retained much longer in Arabic script for non-Arabs, especially in African countries.

Arabic also has script-based problems, through its complex conventions. The 28 symbols in the basic modern Arabic alphabet take different forms according to their position in words - similar to the way English vowel spellings can vary in medial and final positions in words, but rather worse. The numerous diacritics cause difficulties in printing, and resulting printers' errors can be a barrier to comprehension. Joined letters often represent syllable and must be learnt separately. At one time these numbered nearly 500, and there are still over 130. Short vowels are often omitted, so that readers must know the grammatical structure of Arabic well, in order to reconstruct the words. Arabic spelling is still not completely standardised, and there can be different versions of proper names - with further problems in romanising them for European consumption - as in the many media versions of 'Gadafey'.

Arabic script reform is controversial because conservatives and Islamics enthusiasts seek to revive classical Arabic, and they revere the script of the Qur'an, and pan-Arabism works against matching script to languages, while modernists seek to promote the vernacular. The problems for reform are worse than English, since the ambiguous orthography represents obsolete dialects and grammars, and Arab countries must agree on how to represent modern vocabulary. Turkey, Indonesia and Malaysia cut the Gordian knot by turning to the Roman alphabet.

 FRENCH LANGUAGE AND SPELLING

The major French problem is the number of silent letters. Some of these were never pronounced, but were introduced for supposedly scholarly reasons. French has also had many dropped syllables and fused vowels from language changes, comparable to the changes in English speech through the Great Vowel Shift and dropped inflections.

The French led the way for the modern world in both spelling reform and violent social change. The great French spelling reform of 1762 was part of the social upheaval that built up to the French Revolution. In this reform, the French Academy put out a radical third edition of their Dictionary that changed the spelling of around 5000 words - about a quarter of the current vocabulary. However, like Johnson's English dictionary seven years earlier, it tried to base spelling stability on the etymological principle, since it would have been too contentious to select out any one of the contemporary spoken versions of French to be the single standard. The result is a French spelling system with consistent principles that make it easy to deduce the spoken language from the written, but the reverse is difficult - it is hard to work out the written forms from the spoken language. English, however, is unpredictable both ways.

Since then, France has seen a long history of movements seeking to remove some of the superfluous diacritics and silent letters - that were often introduced by this early revision. Recently, pressure groups such as the the Centre National de Recherches Scientifiques (C.N.R.S.) and a reform journal, Néos, with the motto 'Not a single useless letter' have been opposed by self-proclaimed patriots and the Academie Française, which also leads nationalist resistance to the import of Franglais or 'International English. Government decrees, such as the 1975 Bas-Lauriol French Language Law have tried to ban Franglais and prohibited 1,105 foreign words with support from vigilantes such as the Association Generale des Usagers de la Langue Française. A commission on terminology attempts to find French replacements, which are usually longer, for words like data bank, software, hardware, batch, and processing. At the time of writing, some silent letters and accents are in process of removal by the Academie - racing ahead of any English reforms. However, the style of arguments and controversies over spelling reform in English and French continue to have much in common.

DANISH LANGUAGE AND SPELLING

Danish spelling is admitted to be difficult. It is the result of a history with some parallels to English. The English language was considered uncivilised by England's Norman-French conquerors until the fifteenth century; the Danish language was considered uncivilised by the literate classes right up to the 19th century. The aristocrats spoke and wrote German and French, and the Church and University used Latin and Greek. Written Danish was only kept alive by nationalists.

A standardisation based on common speech was attempted in 1826, but contained compromises in the hope that a Norse-derived orthography could be made common to all Scandinavia. Since then, capitalisation of nouns was dropped in 1948 after long controversies, and a principle to naturalise foreign spellings was set up in 1955. The Danish dictionary continues minor revisions, but there are practical barriers to change in the great variety of dialects, and in the perceived need for stability to preserve such a very small language area. So the Danes have tried alternative solutions for literacy, by improving teaching, by keeping good spelling synonymous with 'good manners', by retaining a spelling exam as the only examination left in primary school after 1974, and by continual public complaints about the deterioration of spelling.

HEBREW

The Hebrew writing system has been described as having 'all the orthographic problems that other languages have, plus some knotty ones of its own' (Weinberg 1975). The vowel-stripped Biblical spelling of written Hebrew was revived in the nineteenth century, with the expedient of inserting 'pointing' to indicate where vowels are omitted reintroduced in the 1860s to enable more people to read Hebrew. The problem of reconstituting vowels, however, is illustrated by the continuing disputes over the transcriptions of Jehovah/ Yahweh and the names of identities mentioned in the Dead Sea Scrolls.

Two different orthographic systems have since been developed, defective and plene, with four main styles and intermingled states. Only the Hebrew letters are held in common. Several roman-alphabet systems were proposed in 1898 and although they were strongly supported in the 1920s-1930s, they failed because 'even secular Jews can be emotionally aroused by the desire to preserve the 'holy script' and its 'twenty two sacred letters'. Continuing reform proposals still try to solve the problems of 'historical ballast' and missing vowels, in order 'to provide a vehicle for the vernacular, without meddling with historical and sacred treasures'. Official attempts to standardise a defective-plene orthography with grammatical principles borrowed from Arabic began in 1900. This was clumsy, and difficult to print before there was computerised printing. So to make text readable, printers would use more consonants as vowels, or add vowel points by hand. The Hebrew Language Council finally agreed in the 1930s to have two co-existing orthographies for different readerships - popular and scholarly , similar to the Greek parallel spelling systems of the demotic and the scholarly. This type of solution is a possible precedent for the English situation.

The close link of spelling systems with nationalism was demonstrated when the State of and reformed Hebrew spelling were proclaimed simultaneously in 1948. This system was unpointed. The official Academy of Hebrew Language set up in 1953 was followed by a militant citizen 'Movement for an Unambiguous Hebrew Spelling' in 1969. However, even by 1975, when Israel was reported to be producing an amount of print per capita unparalleled in the world, there was no effective spelling standardisation. Compromises, controversies and contradictions continued in a running battle between religious tradition and efficiency. Frost (1987) has made a multilingual study that suggests that Hebrew readers are comparatively disadvantaged by their need to rely on context and syntax to disambiguate vowelless consonant strings, but the efficiency for experienced readers of this extremely concise orthography is still to be fully evaluated.

THE ISRAELI EXPERIENCE OF TEACHING METHODS FOR HEBREW IN SCHOOLS

The Israeli experience of teaching Hebrew is a dramatic illustration of how teaching methods can be a mismatch to the

The Hebrew language is highly inflected, and words change their appearance when inflected, even in beginners' reading books - unlike English words, which tend to have stable forms inflected by adding prefixes and suffixes. Hebrew print for adults has no vowels, and different words can be composed of the same consonant string. The ambiguity can be realised by imagining another language written on this principle. In English it would be all very well in the advertising journal Dltn's Wkly ('Dalton's Weekly') where vocabulary is predictable, as in bdrm drssng tbl ('bedroom dressing table') but more ambiguous in general text, if words are written like LTR or SM. To reduce this difficulty for children's reading, Hebrew vowels are represented by pointing marks, but these can still give learners difficulty, because they are not full-sized letters. So although Hebrew spelling represents consonants consistently and phonemically, it does present other difficulties. Whole-word learning methods that proceed by rote memory through repetition of whole words in a controlled-vocabulary or Language-Experience approach are intrinsically unsuitable.

However, when the new Israel set up its schools, teachers naturally turned to America and Britain for the most progressive teaching methods, and so they taught reading by the current 'Look-and-Say', taking no advantage from Hebrew's reliable phonetic spelling. They were very successful in teaching reading to their young pupils. Then the next waves of immigrants were illiterate peasants from the Middle East. Their children failed to learn how to read in school, despite the same teaching methods. Investigation found that all through the time of apparent 'whole word' success, the literate and highly -motivated Jews from Europe had been giving their children at home the sort of phonics teaching that they had had themselves in their own native tongues. Following this discovery, phonics programs were set up in Israeli schools so that the children without literate parents to teach them could learn to read Hebrew too. (Feitelson, 1982)

FAROE ISLANDS

The writing system of the Faroe Islands was created relatively recently, but its particular problem is that it was deliberately given conventions of an 'etymological' character. These sought to keep the spelling closer to its geographically distant Scandinavian origins, but are often far from the actual pronunciation. We can observe today then, the drawbacks of Dr Samuel Johnson's etymological principles for spelling when they are put into practice.

SERBO-CROATIAN

Serbo-Croatian is unique. It is a dual writing system. For historical and political reasons, the Serbo-Croatian language may be written in either the Roman or Cyrillic alphabets. Both alphabets were taught in the first two grades in Yugoslavia, the roman one taught first in the Western Croat (Roman Catholic) regions, and Cyrillic first in the Serbian East (Orthodox). Literate adults can read both. The language is highly inflected, but both spellings have regular, consistent sound-symbol relationships which show pronunciation rather than morphemic structure. So far so good.

However, to make a fascinating field of study for research psychologists, the two scripts overlap in an odd and potentially confusing way. Each alphabet has some letters unique to itself, but shares eleven letters with the other alphabet. Of these, seven letters (A E O M K T J) share the same phonemic representation but four (H P C B) have different values in the two alphabets.

So researchers into reading processes can play with the spelling equivalent of twin studies, plus the additional element of ambiguity of symbol. They can study the way that Yugoslav readers adapt automatically to changes of spelling from one text to another - which is important practical knowledge for transitions when modern countries change their writing systems. Still to come are more socially-oriented studies of the practical consequences for learners and users from the combination of duplication and the partial overlapping. The visible segregation of the two spelling systems in print, with dominance and prominence varying according to locality, has certainly contributed to unfortunate political and social developments splitting the former Yugoslavia.