Speech Reform by Vivian Cook

'The perceived problem [with English spelling] is partly a matter of double standards: while people insist that words should have a single standardised form, with the partial exception of proper names, no one makes such claims for phonology or grammar: it would be just as advantageous to standardise dialect accents and words' Cook (2004, 173)
Aim	The reform of English speech for the benefit of learners and users everywhere
Objectives	To publicise the unnecessary difficulties of English speech and the benefits that its simplification would bring. To raise awareness of the phonemic principle, its corruption during the long history of spoken English, and its more rational application in other languages. To promote research and debate on ways of reforming English speech, and to prepare a graded set of proposals for simplifying speech-sounds. To help co-ordinate proposals for English speech reform across both English-speaking and non-English-speaking countries. To persuade the public, opinion-formers, policy-makers and relevant agencies of the need for and practical possibilities of reforming English speech.
Axioms on English Speech	The sounds of the language were designed to distinguish words in speech; that is to say the phonemic principle. The phonemic principle makes speaking easy, allowing the speaker to pronounce words more easily, and the listener to understand them. As pronunciation changes through the ages, the phonemic principle tends to be corrupted; the pronunciation of words then needs to be adapted to modern forms of English. English has not systematically modernised speech over the past 1,000 years, and today it only haphazardly observes the phonemic principle. Neglect of the phonemic principle now makes oracy unnecessarily difficult in English throughout the world, and learning, education and communication all suffer. Procedures are needed to manage improvements to English speech as a world communication system.
Eliminate unnecessary voicing contrasts / ~/, /~ / and /s~z/	The differences between voiced and unvoiced consonants were not needed for Old English and are not required in German and should be dispensed with. Most obviously this would apply to the pair // and //, which are used quite differently in that // goes with function words, // with content words. This would apply neatly also to /~ / and to the /s~z/ contrast, where voice is mostly superfluous. True this would make 'peace' /pi:z/ synonymous with 'piss' /pi:s/ but it would be a small price to pay if a few more homophones were created in the language. As with other languages, there would doubtless be allophonic variation between voiced and unvoiced sounds in context.
Eliminate unnecessary /h/	Since many English dialects manage without /h/, there is no reason to preserve it in the standard language. English would thus belatedly join the other European languages that lost /h/.
Eliminate unnecessary /l/	Many modern accents such as Estuary English already manage perfectly satisfactorily with a vocalised /l/. /w/ could then substitute for /l/ in all positions. Some homophones would be created but context would eliminate any ambiguity between say 'longl' and 'wrong'.
Eliminate unnecessary vowels and diphthongs	Phoneticians have often demonstrated that English is perfectly intelligible if all the vowels are reduced to schwa / /. The most logical solution is then to have only one vowel for English; any problems would be handled by context. A less radical solution would be to reduce English to the minimal 3-vowel system, one high front vowel /i/, one high back vowel /u/, and one central low vowel /a/ like Arabic, Greenland Inuit, and Dyribal, thus enabling 'bin' to be distinguished from 'boon' and 'bun'.
Eliminate unnecessary //	All // endings are also pronounced /n/ in some varieties of English; other //s are simply predicted by following /k/. So there is no further need for an // phoneme in English.
The optimum phoneme inventory for English is then: 11 Consonants: /p t k s t m n r w/ 3 Vowels: /i e a/ This represents a saving of 31 sounds, that is to say 71%. Clearly the spoken language would be far easier to learn and to use if these simple logical changes were implemented.