Although spoken English is quite easy to learn, a fact that certainly in part helped to make it the most commonly learnt second language on the planet, English is also one of the toughest languages for any learner to master, native or not, as far as its spelling is concerned. English spelling is littered with irregularities: lurking mute letters, treacherous doubled consonants or slippery spellings with multiple pronunciations, which are, in most cases, far more confusing than those of French spelling, a close second when it comes to languages with a low graphy to sound correlation.

Before anything else is here said about the relationship of spoken and written English, it is important to understand that the problem of the ambiguity of the sound represented by any given letter of an alphabet like ours is an old one and indeed goes back to the very creation of the precursor of the Latin alphabet used to write English and most of the European languages today: the Greek alphabet. The Greek alphabet was invented around 800 BC by, most probably, a single gifted individual who achieved the feat of adapting a row of signs, used by Phoenician tradesmen to record their speech, to the very different requirements of his own native Greek. Phoenician is a semitic language and relies therefore, just like Arabic and Hebrew, two other members of the same family, very heavily on consonantal information in the production of meaningful units of sounds. Due to the heavy emphasis on consonantal information in their language the Phoenicians decided to dispense with any signs for vowels, as ambiguities could be kept at a tolerable minimum even without them. The adaptor thus had to adapt this purely consonantal alphabet to the requirements of Greek, which is a language where vocalic information is crucial, as there are words such as eaō (to allow) or aiei (always) which do not actually contain any consonants.

The adaptation was on the one hand a conscious effort of attributing new sounds to the letters the Greek adaptor learnt from his Phoenician contact, but the two seem, on the other hand, to have at times just not understood each other properly, which lead to certain adaptations being made by misunderstanding. One example is the letter A, which represents a vowel both in Greek and Latin, but a consonant in Phoenician. The consonant being a glottal stop, a sound produced by constriction of the throat, as in the cockney pronunciation bu’er for butter, and entirely alien to the soundscape of Greek, the adaptor just mistook the vowel a, with which the letters are pronounced by default in semitic languages, as the actual relevant information of the written symbol. This goes to show how arbitrary the attribution of any sound to any written sign is and how easily a sound can change without this being reflected in the way it is recorded; be it, as in the example above, when the symbol is handed from one language to another, or when the language employing it changes over time; in other words, when the symbol is handed from generation to generation of the same but constantly evolving language. This makes it apparent that the historically most significant feature of writing, the fact that it can fix the flux of speech and transmit it over vast spaces or even through time, is also its greatest flaw when it comes to the accurate phonetical representation of words and their sounds; the picture of which is handed down by tradition and bound by the same tradition to keep a recognisable, traditional shape, while the represented speech remains in constant flux.

As far as English is concerned, the reasons for the present state of its spelling are various, but all essentially derive from the paradox that writing fixes something in flux, which although graphically fixed will always remain in flux. Writing is like photography: it catches stills of moments, but does not therefore put and end to the flux of time; it allows us to see what we might not have been able to see, either because of separation in space or time, but at the same time never lets us forget that its reality is purely representational. Reading a language is, to a certain extent, always like looking at old pictures of familiar people. Depending on how much time is allowed to pass between the capture of the picture and moment at which it is beheld, some people will have changed more than others; some might have put on weight or lost weight, grown a beard or got bald, got some wrinkles or become tall and handsome after being pudgy and awkward. Words are not any different than people and their old pictures handed down by tradition never quite match what they have become in the present. The older the pictures, the more conservative the spelling and the more likely inconsistencies in the way they represent present speech. Conversely, if pictures are frequently updated to keep up with the flux of a language, the less likely inconsistencies are to arise between the way the language is written and the way it is spoken.

German, for example, is a language that updates its spelling-pictures quite frequently. This results in the representation of speech by writing with as little inconsistencies as possible. In German telephone is now spelled Telefon, with the native f having replaced the etymologically more correct, but phonetically identical, ph from Greek. The word style also has lost its y originating from Greek and is spelled Stil. The reason again being that the y-spelling is nothing more than a traditional feature and does not in the least affect the pronunciation of the word. But French and English are much more conservative in their spelling and have not updated some spelling-pictures for many centuries, which can, if the language’s flux is fast, result in a striking lack of correspondence between spelling and speech.  In French, the most obvious lack of correspondence between spelling and pronunciation is the loss of nearly all consonants that stand in absolute final position of a word, which is especially challenging for new learners of the language, who, for the most part, will learn the language through writing and reading. Any student of French will have struggled with the fact the the French singular avenue and its plural avenues have exactly the same pronunciation. While the language has moved on, the spelling has not and represents a state of the language which no longer exists as such. The fact that both forms are also English words having an English pronunciation entirely different from the French one again points to the limitations of an alphabet, which has, from initially being intended for a phonetically accurate representation of speech, become ever more reliant on conventions and traditions that vary from one language to another.

Another language that combines very conventional spelling with a pronunciation that has changed greatly over the centuries is Greek. The original spellings i, y, ei, oi and ē of ancient Greek have, to the inconvenience of millions of Greek school-kids, all merged with the sound represented by ee in English, while the spelling has been kept as a traditional feature handed down from the classical era to the present. A good example of how written language develops as a tradition separate from spoken language and how a cultural group can conceive of the act of changing age-old spellings as of the corruption of the own cultural heritage. Writing can, in the course of time and the development of a tradition, therefore become much more than just a visual representation of speech, but of an entire culture, the continued integrity of which is of greater significance than the precise rendering of spoken language. Thus traditional features are kept as part of the safeguard of cultural and linguistic identity, even though it is to the inconvenience of those born into the cultural group itself. In Greek, the retention of archaic features in the spelling means that any word containing the sound ee, can potentially be spelled in five different ways; and the frequency of the sound in modern Greek, due to large number of original sounds that have merged with it, makes it quite probable that there will be two rather than one of them in the same word, producing a dismaying array of interchangeable alternative spellings of which only one will be correct according to tradition. A similar state of affairs applies to the spelling ee itself, the sound of which can equally well be spelled ea, ie, or ei . Which of the identically sounding spellings appears in any given word only depends on tradition.

Considering the question of English spelling in the light of what has been said above, English can be characterised as a language that has a relatively conservative spelling. But while it is relatively conservative in its attitude to spelling English also has undergone extreme external influences in the course of its development from Old English, a Germanic language with three genders and four cases and a close relative of modern German, to the streamlined no-gender and no-case language spoken by millions today at home and at work, either as a native language or as a common medium of communication between non-native speakers.

Old English, and Middle English to a lesser extent, was a language that was written in a purely phonetic manner, and therefore reflecting the different dialectical variations in which the language was spoken across the country. For example both the consonants of the cluster kn were pronounced in the words know (cnawan), knee (cneo) or knight (cniht), which was not yet a homophone of night (niht). As can be seen from the previous examples, the origin of the many silent gh-s is also to be found in the original Old English pronunciation of the words. Both these spellings were still pronounced without change in the Middle English period when Chaucer wrote his Canterbury Tales around 1400. And it is his way of writing English that became the standard of spelling until the advent of printing. About which, more below. Later the sequence of the two consonants kn was, for some reason, no longer perceived to be aesthetically pleasing and went out of fashion, but not along with the spelling which was kept as a traditional feature, just as the several spellings for the sound ee were kept in modern Greek.

The first major external influence on the English language was the influx of marauding vikings, beginning with the raid on the abbey of Lindisfarne in 793, speaking Old Norse, a language different from but related to Old English. Many very common English words are of Old Norse origin, e.g. ill (illr), weak (veikr), ugly (uggligr), sky (ský, meaning cloud), window (vindauga, literally a wind-eye), leg (leggr) or even the so common verb to get (geta). As can be seen from the bracketed examples, the phonology of Old Norse exhibits some considerable deviation from that of English. The most obvious feature is the preservation in Old Norse of the marker for the nominative case, the -r at the end of the words, which had been dropped in Old English even before the earliest attested texts. The process of adopting such words and make them native would speed up the process by which the language as a whole changed and became more distinctively English to our modern ears, as by assimilation of foreign material a language is forced itself to assume new and initially foreign shapes.

The second event, and by far the more influential of the two, was the Norman conquest of 1066 that made England a part of France and French the language of the elite and the law courts for the couple centuries to come. During this time countless French words were borrowed into English as part of a process that fundamentally altered the fabric of the language and made an originally purely Germanic language into the mongrel language, a Germanic-Romance genetical hybrid, which English is today. Only very few languages ever underwent such deep-reaching influence from without for them to be reshaped from within.

English, as a language and therefore a cultural belief system, thus occupies a very special place in the history of the development of languages for having, in the course of the centuries, been asked so many questions – and all know with how much difficulty belief systems cope with too many questions. Generally speaking belief is only for those who accept it, not for those who question it. As soon as the values of a belief system are no longer self-referential but measured against the external reference of another system, the possibility of the establishment of meaning through conventional reference – the core principle of all languages, which are, without conventional reference, just complete and utter gibberish – becomes impossible. Seen from this angle it can be said that English is the belief system on earth the basic meaningfulness of which has been most frequently and vigorously challenged by other belief systems and therefore also was more likely to change at a faster pace than others. As has been already shortly pointed out above, difficulties in spelling mainly arise when the way a language is written does not keep up with the pace at which its spoken form changes. And this is exactly what happened in the case of English.

It is also in the period following the arrival of French that another significant change took place, which, although not related to spelling, shows the melting-pot that was English in the making: the loss of gender distinction. Since grammatical gender is a purely artificial feature and thus another kind of belief, it too does not cope with too much scrutiny from without. And such scrutiny came from the binocular gaze of the already simplified French gender system, which only distinguishes masculine from feminine gender. Probably through the revelation that grammatical gender is entirely arbitrary and ultimately only a morphological burden on the language, English gradually dropped all gender distinction.

We can now see that as a result of both these events, the Norse and French invasions, English not only adopted many Old Norse and French words still in use on today’s streets, but it was also subjected to much bigger forces that made its pronunciation change drastically until the early modern period, around 1600. A time at which, we will see later, English spelling was, for a number of reasons, in terrible disarray and many words had multiple spellings, even occurring in close proximity of each other. Is it head? Or hede, heede, hedde or heade? From these multiple spellings it also becomes obvious that the conservatism affecting some spellings does not extend to the entire language and the representation of one and the same sound, particularly of vowels, since English contains very few phonetically pure vowels, could vary widely.

Another source of difficult spellings in modern English is the heavy borrowing from the two classical languages Latin and Greek, the own peculiar spelling conventions of which are reflected in the frequently only very slightly adapted English spellings. Both languages are for example the source of many doubled consonants in English words, which can cause problems to the user since modern English does not distinguish length of either consonants or vowels. Examples from Latin are words containing the negative suffix in-, of which the final n is always assimilated to the following consonant: irrational, immortal or illiterate. The words occasion, arrogant or opposition too are of Latin origin. Many of these words will have entered the language not through Latin directly but rather through French, which, representing the borrowing of a borrowing, can produce even more unpredictable spellings. The most common example of a word of Greek origin containing a doubled consonant is parallel, the second element of which is derived from the Greek word allos, meaning other. The word parallel is derived from an adjective meaning next (para) to each other (allēlos).

Greek is also the source of all words where the letter y is used as a vowel, as opposed to words such as yoke or yes, where it is a semivowel. Since, by ear, the y in this position is not to be distinguished from i, a knowledge of ancient Greek, where the two letters also have two distinct pronunciations, is required to predict where which of the two letters will be employed in spelling a word. Among the many examples are analysis, physics or style, which is phonetically identical with the native stile in turnstile. Greek is also the source of all words such as chasm, character or chaos, where ch is pronounced like c. In the fiendish words haemoglobin and haemorrhage it is again the accurate preservation of Greek features, the representation of the original Greek diphthong ai as ae and the several other options for writing the sound it represents, the inaudible doubling of r and the fact that r is in Greek always followed by a breathing h, that poses difficulties to the modern user who does not know ancient Greek and makes the task of learning the words one of pure memorisation by rote.

Up to now the question of spelling has mainly been discussed in relation to the general limitations of all spelling systems and the very moved history of the English language which encouraged it to change more quickly than other languages. But, in the end, what makes the spelling is the writing and writing became more and more important as the art of printing books became socially and culturally more important, affordable and accessible.

The history of printing in English begins with William Caxton, when he came back from the Continent in the year 1476 and set up the first printing press in London. The problem with his publications was that, although Caxton was a native English speaker, most of his workers were not and thus made many spelling errors when setting texts. Some of their errors even slowly became the standard. The word busy, for example, was originally spelled bisy, something no one will have problems believing considering the pronunciation it still has. Being paid by the line, the same printers were also fond of adding letters to words wherever they could. Some such letters finally too made it into the standardised modern English spelling. The words friend and season, originally were spelled more phonetically as frend and seson. But although some of these blunders and business-minded inventions are now “proper” English, the activities of the foreign printers helped to make spelling more unpredictable and corrupt rather than anything else.

After Caxton matters became even worse when the Bible started to be translated into English. This had to be done outside the country because England was a catholic country and the translation of the holy scripture into the vernacular was banned under pain of being burned at the stake as a heretic. The first translation of the Bible into English was produced by William Tyndale in the middle of the 16th century and printed in protestant countries. Mainly Germany, Belgium and Holland. The frequency with which these copies were copied helped the spread of corrupt spellings. It was also the most commonly owned book at the time and the one from which many people would learn to read and write. Resulting in the fact that at the beginning of the 17th century no one really knew what the rules were for writing English.

To summarise, the reasons for the complicated nature of English spelling are: the inherent limitations of our alphabetical writing system, conservatism in the case of some spellings inherited from Old English, Greek and Latin, extreme external influence by foreign invasions, borrowing both from the languages which these invasions brought to England and from Latin and Greek, the lack of pure vowels in English and, finally, the conscious or unconscious interference by printers who did not know the language with which they were working. The interplay of these so different forces culminated in the great confusion of the English spelling around 1700. And only gradually could the great confusion of the English spelling be reduced to the organised confusion that can be found in a modern English dictionary.

Silvio Zinsstag,

teacher for ancient languages