English Spelling Reform

From Andrewgdotcom

Version: 2beta2 (2005-08-27)\n\nEnglish Spelling Reform (transliterated). Caveat: This is a draft specification, and my spelling may not be totally consistent from one paragraph to the next... :-)\n\n=Overview=\n\nIt will never happen. Nobody wants to change. It will cause chaos. But nearly everyone agrees that English spelling is:\n\n* illogical\n* confusing\n* hard to learn\n* hard to teach\n\nMany languages have undergone spelling reform (most recently German), but these tend to be ones which are only spoken in a few countries and are thus amenable to political intervention. With the widespread and unregulated use of English and its blooming family of dialects, the effect of any imposed change would be negligible. With this in mind, the New English specification is designed as an alternative rather than a progression. It is intended that the new spelling be used concurrently with traditional orthography over an extended phasing period, to be determined not by edict but by voluntary choice. Those people who wish to learn to read and write the new system are encouraged to do so and to encourage others in turn.\n\nBecause the written and spoken languages have mutually influenced each other over the years, divorcing the orthography entirely from the original in favour of a minimal phonetic system would cause a lot of problems, most notably in the realm of derivative words, many of which would contain unpredictably altered vowel sounds. There have been many attempts to design such a phonetic system, but these tend to differ significantly from tradition and have never had much success. In addition, phonetic systems are too rigid to properly take account of the multitude of different accents and dialects that form the body of the spoken language.\n\nTherefore, New English is intentionally not phonetic. Rather, the basic units (graphemes) of the written system represent not the literal sounds but a common-denominator set of \'abstract phonemes\', each of which may be pronounced in several different ways depending on local practice. There is also a degree of redundancy in the grapheme set to allow them to indicate not only the sounds which make up the word but also the set of transformations which are applied when derived words are formed.\n\n=Notes on version 2=\n\nThis specification has proved remarkably fluid. Version 1.0 was published online in 2002 and was revised twice in the space of a year, the second revision partly undoing some of the first. I decided that I really needed to start a dictionary in order to winkle out all the bugs, and a couple of thousand words into the process I found enough annoyances to justify a major revision. A year of sleeping-on-it later, I finally got around to collecting my thoughts in one place (here), and after another year or so of thought I am pretty close to a specification that\'s solid enough for me to consider restarting the dictionary project.\n\nMajor changes include:\n* (re?)introduction of consonant doubling for stress, removing the need for many vowel accents\n* introduction of u-umlaut\n* explicit rules for handling declensions, common suffixes and contractions\n* formalisation of ASCII-limited method\n* loanword graphemes\n* some redundant forms removed (ö, ou, oi)\n* special cases of u and y\n* Icelandic characters replaced\n\nA full list can be found in the changelog at the end of the document.\n\nMy thanks to Clive McCaig for telling me when I\'ve thought up something ridiculous, and to my brother Colin for inspiring me to grapple with the problem in the first place.\n\n=Dictionary=\n\nI have started an old-to-new English dictionary. It is incomplete as yet, but I am starting with the short words so it should still be useful. It will be updated as I go on. The full dictionary will be approximately 63,000 words. It is currently 4.7% complete, and contains all words of 4 letters or less in the old spelling.\n\nUnfortunately, this dictionary is in v1.2 and will need updating again. Bah.\n\n=Graphemes=\n\nVersion 2 of the New English spelling proposal is a system of 44 graphemes, as follows: \n\n\n:\nA Á Ã… Ä B ÄŒ D Đ E É Ë F G H Ä° Í Ä°y J K L M N ÅŠ\n:\nO Ó Ow Oy P R S Å T Ŧ U Ù Ú Ãœ V W Å´ X Y Z Ž\n:\na á Ã¥ ä b č d Ä‘ e é ë f g h i í iy j k l m n Å‹\n:\no ó ow oy p r s Å¡ t ŧ u ù ú ü v w ŵ x y z ž\n:\n\n\n(44 may sound like a lot, but remember there are a lot of digraphs in English already, such as ch, th, sh, wh, rh, gh, ng; and that\'s just the consonants, not to mention trigraphs...)\n\nThese divide into six short vowels (one multipurpose), nine long vowels (two aliases), three digraph vowels and twenty-six standard consonants (one alias, one multipurpose).\n\nWe use seven diacritics, three of which have limited application:\n{|cellpadding=\"5\"\n|Accent \n|Name \n|Meaning \n|Used with\n|-\n|` \n|grave \n|\'hard\' short vowel \n|ù\n|-\n|° \n|ring \n|rounded long vowel \n|Ã¥\n|-\n|´ \n|acute \n|standard long vowel \n|á é í ó ú \n|-\n|¨ \n|umlaut \n|front/closed long vowel \n|ä ë Ã¼ \n|-\n| Ì‚ \n|circumflex (caret)\n|aspirated consonant \n|ẇ (also maybe lÌ‚ rÌ‚ in loanwords)\n|-\n|ˇ\n|caron (hachek)\n|soft (palatalised) consonant \n|č Å¡ ž\n|-\n|-\n|stroke\n|fricative consonant\n|Ä‘ ŧ (also Ç¥ ħ in loanwords)\n|}\nAlso included in the standard glyphs is a new letter (Å‹) taken from Sami, by way of the phonetic alphabet.\n\nWe also define a secondary set of eight glyphs for loanwords (I don\'t like the use of dots as diacritics, but I\'m not sure about using irreducible forms such as l- and r-caret, particularly since my copy of Safari renders them so badly):\n\n\n:\nC Ç Ǥ Ħ Ä¿/LÌ‚ Q Ṙ/RÌ‚ Åž\n:\nc ç Ç¥ ħ Å€/lÌ‚ q á¹™/rÌ‚ ÅŸ\n:\n\n\nThese make use of a further one/two diacritics:\n{|cellpadding=\"5\"\n|Accent \n|Name \n|Meaning \n|Used with\n|-\n|· \n|dot \n|aspirated consonant \n|Å€ á¹™\n|-\n|¸ \n|cedilla\n|\'whistled\' palatalised consonant\n|ç ÅŸ\n|}\n\n=Pronunciation Guide=\n\n==Short Vowels==\n{|cellpadding=\"5\"\n|Grapheme \n|Example \n|-\n|A a \n|pat \n|-\n|E e \n|pet \n|-\n|Ä° i \n|pit \n|-\n|O o \n|pot \n|-\n|U u \n|put, fur\n|-\n|Ù ù \n|bùt \n|}\nA short vowel carries the primary stress of the word if either the following consonant is doubled, or if it is in the first syllable of a word containing only short vowels. Unstressed short vowels can tend to a null-vowel sound, but this is highly accent-dependent. \n\nA special case is unaccented u when used before r, in which case it is \'\'always\'\' a null vowel, even when stressed. In many accents, the u and r blend to form a long null-like vowel, but the u is treated here as short in order to take advantage of short-vowel stress rules.\n\n==Long Vowels==\n{|cellpadding=\"5\"\n|Grapheme \n|Alias \n|Example \n|Source \n|-\n|Á á \n| \n|blah -> blá \n|-\n|Ã… Ã¥ \n|\n|ball -> bÃ¥l \n|Scandinavian languages\n|-\n|É é \n|Ä ä \n|rein -> rén, rain -> rän \n|-\n|Í í \n|Ë ë  \n|ski -> skí, been -> bën \n|-\n|Ó ó \n|\n|go -> gó \n|-\n|Ú ú \n|\n|blue -> blú\n|-\n|Ãœ ü\n|\n|few -> fü\n|~German\n|}\nLong vowels and digraphs are usually stressed, but not as much as stressed short vowels. The use of acute accents to lengthen vowels is standard in many European languages, e.g. French.\n\nThe aliased vowels are primarily for avoiding large changes to the spelling of root forms in derived words and compounds, but also to retain more traditional-looking spellings. The umlaut is stolen from German, in the sense that \'ä\' is an \'a\' that sounds like an \'e\'. By extension \'ë\' (not used in German) is an \'e\' that sounds like an \'i\'. These aliases actually represent the more common method of lengthening vowels in English (pal vs. pale, pet vs. Pete) but are treated here as exceptional in order to keep the acute accent systematic.\n\nThe umlaut is also used for \'ü\'. This is again cod-German, but for a different purpose. It is usually, but not always, a diphthong which is pronounced similarly to \'yú\'. Its primary purpose is to enable American and British pronunciations to be accommodated in a single grapheme (in American usage it is less often a diphthong, especially after t, d or n), but it also has the not-unrelated side effect of making spellings look more intuitive.\n\nNote that while é and ó are often pronounced as diphthongs this is not universal.\n\n==Digraph Vowels==\n{|cellpadding=\"5\"\n|Grapheme\n|Example \n|Source \n|Notes\n|-\n|Ä°y iy \n|die -> diy, fly -> fliy\n|~Dutch \n|The traditional \'long i\' is (usually) a diphthong. The grapheme is designed so it can take the place of either i or y, and also deliberately resembles the Dutch \'ij\'.\n|-\n|Ow ow \n|owl\n|-\n|Oy oy \n|boy \n|}\nDigraph vowels are (almost!) always diphthongs, but not vice versa.\n\n==Consonants==\n{|cellpadding=\"5\"\n|Grapheme\n|Example \n|Source\n|Notes\n|-\n|B b \n|bin \n|-\n|ÄŒ č \n|chin -> čin \n|Slavic languages \n|bare c is not normally used\n|-\n|D d \n|din \n| \n|always voiced, never like \'t\' as in, well, \'voiced\'\n|-\n|Đ Ä‘\n|this -> Ä‘is \n|~Icelandic\n|\n|-\n|F f \n|fin \n|-\n|G g \n|get \n| \n|always hard\n|-\n|H h \n|hot \n| \n|does not combine with preceding letter\n|-\n|J j \n|jet \n|-\n|K k \n|kin \n|-\n|L l \n|let \n|-\n|M m \n|met \n|-\n|N n \n|net \n| \n|does not combine with following g or k\n|-\n|ÅŠ Å‹ \n|sing -> siÅ‹ \n|phonetic alphabet / Sami \n|the letter \'eÅ‹\'\n|-\n|P p \n|pet \n|-\n|R r \n|rot \n|-\n|S s \n|set \n| \n|always aspirated, never voiced as in \'always\'\n|-\n|Å Å¡ \n|ship -> Å¡ip \n|Slavic languages\n|-\n|T t \n|tin \n|-\n|Ŧ ŧ \n|thin -> ŧin \n| \n|\n|-\n|V v \n|vet \n|-\n|W w \n|wet \n|-\n|Å´ ŵ\n|when -> ŵen \n| \n|often pronounced as \'w\'\n|-\n|X x \n|box\n|\n|alias for \'ks\'\n|-\n|Y y \n|yet, pity\n|\n|see notes \n|-\n|Z z \n|zip \n|-\n|Ž ž \n|measure -> mežur \n|Slavic languages\n|}\n\nNote: it\'s notable how many of these graphemes are used in Sami. I didn\'t realise this until I saw this page on Omniglot recently, but I suppose it\'s understandable since both this scheme and the Sami orthography were consciously designed by Germanic-speakers.\n\n==Loanword consonants==\n{|cellpadding=\"5\"\n|Grapheme\n|Example \n|Source\n|Notes\n|-\n|C c\n|\n|Chinese, slavic languages\n|alias for \'ts\'\n|-\n|Ç ç\n|\n|\n|Chinese \'q\' ~= \'ch\' (yes, really!)\n|-\n|Ǥ Ç¥\n|\n|-\n|Ħ ħ\n|loch -> loħ\n|Maltese\n|used mainly for Gaelic words\n|-\n|Ä¿ Å€ / LÌ‚ lÌ‚\n|llan -> Å€an / lÌ‚an\n|\n|used for Welsh\n|-\n|Q q\n|Iraq\n|Arabic\n|guttural k\n|-\n|Ṙ á¹™ / RÌ‚ rÌ‚\n|rhyl -> á¹™il / rÌ‚il\n|\n|Welsh again\n|-\n|Åž ÅŸ\n|\n|\n|Chinese \'x\', also German \'(i)ch\'\n|}\n\n=Spelling Notes=\n\n==Vowels==\n\n* There is no more \'helper e\'. Bliss.\n* Vowels are always short, even in the middle of words, unless indicated otherwise (cf. bit/biting).\n* Aliases should be used to match the historical spelling where necessary, and also when a and e are lengthened in compound words.\n* Long and digraph vowels are usually more stressed than short vowels, but if a short vowel is followed by a doubled consonant then it has most stress. Otherwise, the stress is on the first syllable: present(n) -> prezent, present(v) -> prezennt, desert(n) -> dezert, desert(v) -> dezurrt\n* Doubled consonants should be used at most once in any word, and should never be used in words with only one vowel. In addition, the stress falls by default on the first syllable, so doubling should not be used in that position unless absolutely necessary.\n* There are certain cases (particularly in compound or declined words) where a short vowel appears twice in succession. In these cases they should be pronounced separately. \n* \'y\' can be used as a trailing unstressed, non-null vowel with a similar pronunciation to \'i\', if accompanied by the twin rules\n# y before -iÅ‹ is always a vowel and\n# when adding a suffix other than -iÅ‹, y -> i (this is standard in modern English anyway, so shouldn\'t cause problems).\n* Accents should be regarded as a necessary evil and should be avoided if it can be got away with. :-)\n* \'r\'s are never dropped, even if you think they\'re not pronounced. They are pronounced in many dialects, and removing them increases ambiguity.\n* \'ór\' is preferable to \'Ã¥r\', except where the original word had no \'o\', e.g. boar -> bór, bored -> bórd, war -> wÃ¥r. This is due to pronunciation distinctions in some dialects.\n\n==Consonants==\n\n* These consonants have their usual meaning: b, f, j, k, l, m, p, r, t, v, w, x, z.\n* These consonants require special care: c, d, g, h, n, s, y.\n** c is not used, except with a caron (č, Czech etc.), when it is pronounced as in cello. Hard c (cat) -> k, soft c (centre) -> s.\n** d is always voiced. The d in past forms such as \'passed\' -> \'t\'.\n** g is always hard, as in get. The soft g in gin -> j\n** h is always pronounced as in hello. It never combines with a preceding letter, i.e. th as in pithead, sh as in messhall, etc.\n** n is always as in thin, n in think -> Å‹.\n** s is always as in sit. s in was -> z. sh in shop -> Å¡, s in measure -> ž (Czech, etc.).\n** y is never a vowel, unless part of a digraph, at the end of a word, or before -iÅ‹.\n* ŵ is an aspirated w. It is used for \'wh\' in words such as which and where.\n* The accented letters ŧ and Ä‘ are used for th as in thigh and thy, respectively. Try putting on a Dublin accent if you can\'t work out why. :-)\n* The Sami letter Å‹ is used for ng. Note that many cases of ng in modern English need to be spelled Å‹g, e.g. anger -> aÅ‹ger\n* q is not normally used, -> k\n\n==Misc==\n\n* \'I\' (\'Ä°y\') is only spelled with a capital at the beginning of a sentence, the same as other words.\n* Proper names will use their traditional spelling, at least initially. Exceptions to this rule are well-known place names and their derivatives, e.g. France -> Frans, French -> Frenč\n* The Americans and the British will probably want to spell their words differently because of the different pronunciations. This should be resisted as the differences are slight and the distinction likely to blur with the spread of communication technology. A common standard has advantages which should not be dismissed lightly.\n\n=Grammatical Notes=\n\n* Apostrophes are not used, apart from disambiguation in ASCII format. \n* Common contractions are treated by shortening the words appropriately, but not joining them together. Some common examples follow:\n{|cellpadding=\"5\"\n|Traditional contraction\n|Uncontracted form\n|New contraction\n|Not to be confused with\n|-\n|i\'m\n|iy am\n|iy m\n|-\n|it\'s\n|it iz\n|it z\n|its\n|-\n|he\'s\n|hë iz\n|hë z\n|-\n|she\'s\n|šë iz\n|šë z\n|-\n|we\'re\n|wë Ã¡r\n|wë r\n|wër (weir)\n|-\n|you\'re\n|yú ár\n|yú r\n|yúr (your)\n|-\n|they\'re\n|đé ár\n|đé r\n|đér (their, there)\n|-\n|that\'s\n|Ä‘at iz\n|Ä‘at z\n|-\n|who\'s\n|hú iz\n|hú z\n|húz (whose)\n|}\n* Genitives are similarly constructed using the word \'z\', which is not attached to the root, e.g. boys -> boyz, boy\'s -> boy z, boys\' -> boyz z.\n* Unfortunately many common contractions are easily mistaken for genitives. On the bright side, it\'s no worse than before.\n* Several common suffixes have particular forms which should be noted. These are:\n** -Å¡on (when after t, s or z; and a preceding t is silenced by it)\n** -äšon (not ätÅ¡on, despite the above)\n** -us (not -ous or -os), e.g. porous -> pórus\n** -ossity (despite the above), e.g. porosity -> pórossity\n** -ur (not -our or -or), e.g. colour/color -> kùlur\n** -orrity (despite the above), e.g. ???\n** (etc.?)\n\n=Encoding and Fonts=\n\nThis system cannot be represented in any of the current ISO8859 encodings, which were found to be overly restrictive. Version 2 uses glyphs from the Basic Multilingual Plane of Unicode 4.0 and the use of UTF-8 as an encoding mechanism is strongly encouraged.\n\nIt is suggested that a dot be used above capital I (i) (Ä°, Turkish) to help differentiate it from small l (L) in badly-designed fonts.\n\nNote that Đ is U0110 LATIN CAPITAL LETTER D WITH STROKE, not U00D0 LATIN CAPITAL LETTER ETH.\n\n==ASCII format==\n\nIn order to enable computer input, an ASCII format has been defined using the letters h and q as modifiers. Depending on the capability of the system, this can be transformed on the fly or as a batch process into the standard (typeset) format; or if no such facility is available, the ASCII format is sufficiently human-readable to be used directly.\n\nEach letter in the ASCII alphabet, apart from h and q, stands for itself. All accented consonants (and the letter eng) are denoted by appending h or q to an ASCII consonant, as follows:\n{|border=\"1\"\n|č\n|Ä‘\n|Å‹\n|Å¡\n|ŧ\n|ŵ\n|ž\n|.\n|ç\n|Ç¥\n|ħ\n|Å€\n|á¹™\n|ÅŸ\n|-\n|ch\n|dh\n|nq\n|sh\n|th\n|wh\n|zh\n|.\n|cq?\n|gh\n|kh\n|lh\n|rh\n|sq?\n|}\nIn cases where ANY ASCII consonant other than h or q (even w or y when part of a digraph) is followed by an h or q which is not being used as a modifier, an apostrophe should be inserted between the two in order to disambiguate (many combinations may currently have no special meaning, but that doesn\'t mean they won\'t in the future). When h or q follows a vowel, no disambiguation is necessary.\n\nWhen a consonant is doubled in ASCII form, the stem is repeated but any modifier is not, e.g. sh -> ssh, th -> tth, nq -> nnq.\n\nAccented vowels are denoted mainly by vowel pairs, as follows:\n{|border=\"1\"\n|á\n|ä\n|Ã¥\n|é\n|ë\n|í\n|ó\n|ù\n|ú\n|ü\n|-\n|aa\n|ae\n|au\n|ee\n|ea\n|ii\n|oo\n|u\\\n|uu\n|ue\n|}\nSince there are many cases in English where a long vowel is followed by a short one but few (none?) the other way around, if there are three vowels together then the first two represent a long vowel (or diphthong) and the third a short vowel. There are a few words (particularly compounds) where two short vowels follow each other, and in that case an intervening apostrophe should be used to disambiguate.\n\nIt is suggested that these mappings be generally adopted for computer applications. I have been using these mappings in a Unicode editor called yudit to create these pages.\n\nDownload the [newenglish.my] file for yudit. This should be placed in /usr/share/yudit/data/ (installation dependent). [sorry, not currently available as the spec is in a state of flux]\n\n=Rationale=\n\n* The development criteria were, in rough order of importance:\n# Unambiguous.\n# Easily learned by native English speakers.\n# Consistent with other European languages where possible.\n# Forgiving of dialect/accent (i.e. not strictly phonetic).\n# Minimal and logical use of diacritics and digraphs.\n* Although the Latin alphabet is totally unsuited for writing any modern languages other than Italian and Polynesian, its universal familiarity is unsurpassed and many other languages use similar diacritic/digraph systems to the one proposed.\n* c could have been used without the caron, but it is probably better to emphasise the unconventional usage. This also allows us to use bare c for Chinese and Slavic loanwords, where it is an alias for \'ts\'.\n* \'ow\' is pretty un-phonetic (unless you\'re from Norn Iron), but is in keeping with tradition. Of the more obvious alternatives, \'au\'/\'aw\' look like they should be pronounced as \'Ã¥\', and the Italian \'ao\' is a little too alien. Then again, \'iy\' is hardly phonetic either.\n* Some spelling-reform proposals (and the Government of China no less) reuse \'q\' for useful things. It is sad to see a good letter go to waste when English needs all the help it can get, but q has too much historical baggage to be a credible \'ch\' or \'ng\' (yes, I tried!). However, it has been retained in ASCII mode as a modifier, and for Arabic loanwords.\n* Around half or more instances of short vowels in text could be replaced by a null vowel symbol. This was rejected because a) it removes a lot of information from the text, making it harder to read , and b) many root words would change spelling significantly on adding suffixes - if the original vowel is kept in the root word then we need only double a consonant in its compounds.\n\n=Changelog=\n\n==Changes in v2==\n\n* Grave accents for stress have been abolished in favour of consonant doubling. This reduces the number of diacritics in typical usage considerably.\n* u-breve is now u-grave, mainly because I find u-breve difficult to write by hand. :-)\n* u-umlaut has been introduced because \'yu\' was unintuitive to American speakers when used after d, n and t. It also helps transition from traditional spelling.\n** h-circumflex has therefore been abolished, since it was only ever used before u.\n** words traditionally ending in -our and -ous can now be spelled -ur and -us consistently.\n** \'čú\' is now deprecated when the traditional spelling is \'-tu-\'.\n* trailing syllabic l and r no longer have a preceding short vowel when there was none in the traditional spelling, or when none is used on adding suffixes.\n* x has been restored as an alias for \'ks\'.\n* ħ has been introduced to take the place of x.\n** other loanword-enabling consonants have been defined (one of which can be pronounced the same as h-circumflex... fancy that) as extensions to the basic grapheme set.\n** by analogy, the Icelandic letters ð and þ have been replaced by Ä‘ and ŧ respectively. Þ in particular was too obscure for anyone unfamiliar with Old Norse or Beowulf.\n* y can now be a vowel again under special circumstances.\n* \'ö\' abolished in favour of special-circumstance \'u\'.\n* all instances of \'oi\' have been standardised as \'oy\', and \'ou\' as \'ow\'.\n* the common suffix -ion is spelled -Å¡on when used after t, s or z - and t is \'eaten\' by it.\n* ASCII format has been properly defined, based on yudit input method.\n** \'ng\' ASCII method has been changed to \'nq\', where q is a modifier\n** q and h are now classed as modifiers, and disambiguated using a preceding apostrophe.\n** modifiers are not doubled in ASCII format, only the stems.\n* floating z is now used for genitive, to disambiguate from plurals.\n* renamed \'compatibility forms\' to \'aliases\'.\n\n==Changes in v1.2==\n\n* Compatibility forms have been introduced to reconnect the new spelling with the historical form. I have retrospectively included \'ow\' and \'oy\' (was \'oj\', see below) in this category.\n* The umlaut has been introduced, allowing \'ä\' and \'ë\' to stand in for \'é\' and \'í\' respectively.\n* \'ö\' has been introduced so that \'Å•\' becomes a more natural \'ör\'.\n* \'ý\' becomes \'iy\', as a compromise between the two previous versions.\n* \'ĵ\' becomes \'Ä¥\', to allow for more natural spellings of the (few) words it appears in.\n* Following the rationalisation of \'y\' in 1.1, \'oj\' reverts to \'oy\'.\n* Further rationalising, and ending its long trek towards respectability, most usages of \'y\' become \'i\' so that \'j\' can become \'y\' and ǧ can thus become \'j\'. This eliminates the Germanic j, the major justification for which was the now defunct \'bizarre y\' (see below), and the soft g, which made such a horlicks of \'judged\' -> ǧŭǧd (eek!).\n\n==Changes in v1.1==\n\n* The stress rules have been modified to allow a \'default\' stress on the first syllable, reducing the need for explicit stress marking considerably.\n* The bizarre use of y has been abandoned in favour of an earlier idea, namely u-breve (Å­). This was initially considered and then rejected because thought I needed both stressed and unstressed versions. With the modified stress rules I now believe this is unnecessary, and I always hated that y.\n* Y is now used as an unstressed final vowel, allowing more natural spellings of words such as partying -> pártyiÅ‹. Ý is used to replace ij.\n* After dithering about whether to use ow,oj or ou,oi I have decided to allow both, depending on the historical spelling of the word.\n* The aspirated letters ĵ and ŵ have been introduced.