symbols in CUBE
Sound searches must be input in ASCII characters. The following charts show the ASCII characters (green) in CUBE’s database and the IPA symbols which they correspond to. The IPA symbols in blue are CUBE’s default phonetic transcription. Where different, the IPA symbols in black are those of traditional ‘Gimsonian’ transcription, which CUBE will display if the ‘Gim’ box (under systems) is ticked. The phonetic and phonological inadequacies of the Gimsonian symbols are explained here.
Stressed vowels are specified in cube by entering them in UPPER CASE, not by means of any stress symbols. So, if you wish to specify a stressed vowel in a sound search, you should enter CUBE's ASCII vowel characters in capitals (except for x, which represents unstressed schwa). For example, the noun impact is Impakt in CUBE’s ASCII characters, while the verb impact is impAkt. The representation of stress in cube is explained in greater detail here.
Several natural classes of sounds can be searched for by inputting an exclamation mark followed by various letters:
|!H||any vless obstr|
|!L||any vd obstr|
|!v:||any long vowel|
|!V||any stressed vowel|
|!V:||any stressed long v|
|!y||any short vowel|
|!Y||any stressed short v|
|!D||any stressed diphth|
For example, oJ!c will find words with the vowel of CHOICE followed by any consonant (boing, paranoid, voice, etc). The sequence #s!p will find words beginning with s followed by any plosive. !c!c!c# will find all the words that end in three consonants. The string !d-!d finds any word that contains at least two diphthongs; the hyphen separating the abbreviations means that the two diphthongs do not have to be adjacent. (The symbol !b will find so-called branching onsets).
The consonantal natural classes above can also be negated by !~. For example, !V:!~P will find words with a stressed long vowel followed by a consonant, which is not labial. (Note that !~P matches any noncoronal consonant, but not vowels. The negation operator does not apply to the vocalic natural classes, and to !b.)
CUBE also allows the use of various regular expressions to enrich your searches. For example, s[ptk] will find words containing s followed by p or t or k, while s[^ptk] will find words containing s followed by any sound other than p or t or k. (In fact, the sound categories above are abbreviations for precompiled regular expressions.)
|.||any one character|
|a?||zero or one a|
|a*||zero or more a’s|
|a+||one or more a’s|
|a.*b||a + anything/nothing + b|
|[abc]||a or b or c|
|[abc]+||one or more chars of the set|
|[^abc]||anything but a or b or c|
|(foo|bar)||either foo or bar|
You can give several chunks of transcription separated by any of the following characters: = < > ~. CUBE will find words which contain these chunks in the given order. If you separate the chunks by a comma or a space, words in which the chunks match in any order will be returned. That is, a:~e: will find words that contain ɑː earlier and ɛː later, while a:,e: will find not only words of this type but also words in which ɛː occurs before ɑː.
|T~D||(no entries for this criterion)|
|a:, e:||hɑ́ːdwɛː, ɛ́ːkɹɑːft|
Since the permutation of sets increases rapidly as the sets grow larger, the maximum number of chunks that can be given is limited to six. Any further chunk is discarded. For example, the y (strut vowel) in the second line of the next table is ignored, because it is beyond the limit. Also note that single vowel symbols match diphthongs too.