1 Enter the word you want in lower case and when it appears in pink, click it to hear it spoken as often as you want. Please note that full stops cannot be included in an entry, e.g. "a.w.o.l." is entered as "awol". Each word is individually pre-recorded and no form of synthetic speech is used. You can compare multiple entries by entering them separated by a space, eg cat cut cart. Both American and British spellings are provided.
2 [Addendum Nov 17, 2011: Entries are researched using authoritative sources including dictionary.com, m-w.com, Merriam-Webster Unabridged, Oxford Learner's Dictionary, forvo.com, Macmillan Dictionary, GREdic.com and many others, including specialist fora. We generally look for evidence confirming a given pronunciation from at least two sources, in the form of sound files and/or IPA transcriptions. My own background as editor of howjsay is summarised here.]
3 Pronunciation is in Standard British English, with World English alternatives (see also para 6 below).
4 The dictionary shows the closest cluster of letters that it has to your request - for example "furious" used to return "curious". Accents and full stops are not recognised.
5 The lexical corpus includes all of the General Service List (Bauman and Culligan 1995 version). This is a collection of 2,284 commonly used English words. More words will be added progressively, selecting them on the basis that most non-native speakers of English would be unlikely to pronounce a selected word in a universally understandable way without first hearing it spoken by a native speaker. Another source of new words is you: unsuccessful searches are automatically considered for inclusion. [Addendum May 13, 2007: this last source has, it turns out, swamped all others.]
6 Where alternative pronunciations are given, they are either
a widely used World English alternative (Eg "cóntroversy" versus "contróversy"). In these cases the first alternative is the recommended one, on the basis of its having the widest acceptance globally, although in many cases the second alternative is equally common - choose the one with which you're most comfortable - or
the weak version of the pronunciation - Eg "some" pronounced [s + schwa + m]. In most of these cases a brief context is provided (Eg "some" as in "some bread"; "have" as in "They've gone") or
two different pronunciations corresponding to two different parts of speech (Eg "close", which is entered as "close, verb or close, preposition") or to two different meanings (Eg "bow", which is entered as "bow, as in bow tie, or bow as in ship's bow"). [Addendum: Sep 23, 2006: Some familiar American alternatives are also given. If they are not widely used outside the United States, they are prefixed with the words "Also American..." (Eg "Also American: [t + schwa + meidou]").]
7 Profane language and erotica are excluded from the dictionary in the interests of child learners. [Addendum: May 22, 2010: they are now to be included, by popular demand and also because it seems narrow-minded to arbitrarily censor the corpus in this way].
8 No phonetic transcriptions are provided in this dictionary, for two reasons: firstly, the public seldom uses them and when it does so is often misled by them; secondly, any transcription will always be an inferior imitation of the original sounds. Transcriptions were necessary in the age of the book, because books cannot speak. [Addendum May 30, 2007: we nevertheless developed our own phonetic alphabet for annotation purposes, designed for speed, conciseness and flexibility using a standard computer keyboard.][October 14, 2009: You can nevertheless see a phonetic transcription for the commoner entries by clicking on the "define your entry here" link].
9 This dictionary will be a work in progress for many years to come, so please do not condemn it yet for its limited size. Judge it rather on the entries to date of high-frequency words such as "do" or "has". And please feel free to criticise, report errata, and make suggestions.
Tim Bowyer, Amman, Jordan June 2006