Bobdot-ph: A dictionary key spelling

Alan Beale

September 17, 2005


Bobdot is a spelling system devised by my friend and cohort Bob Boden which has several unusual virtues.  It is both extremely readable and extremely precise, a combination not found in many reformed English spelling systems.  It has a few weaknesses which you may or may not find important.  It is strictly for American English - spellings like fothur and sawftwer tend to make Englishmen wince.  It shows primary stress but not secondary stress.  It is very strictly phonemic, which means that it often obscures word relationships.  And it uses the same spelling, u, for the schwa and the unstressed short u.

Bobdot-ph is a slight variant of Bobdot that I devised while working with Bob on his Phondot project.  Bobdot-ph corrects the ambiguity of the letter u, and has a few other changes in order to be a bit easier for a computer program to analyze.  Bobdot-ph is as easy to read as Bobdot, and even more precise.  I believe Bobdot-ph to be a virtually ideal "dictionary key spelling", that is, a spelling which can be used to document pronunciation, and which could function as an auxiliary spelling system were one so inclined.

A 30,000 word Bobdot-ph dictionary is downloadable here.  The main reason you might want to download it is that it provides, in a most readable format, a sizable pronunciation dictionary for American English.  It may be useful to compare it to the FEWL database, which I've previously made available to aid reformers wishing to make a dictionary for their own systems.  The FEWL database contains significantly more information, including some guidance to British pronunciation as well as American, but its format is much more complex.  If you have a system which is strictly for American English, does not care about secondary stress, and does not make special provisions for compound words, prefixes or suffixes, it is likely that you can use the Bobdot-ph dictionary as a starting point for your own dictionary, with substantially less effort than use of FEWL would require.

Bobdot-ph should be blamed on me - Bob had no part in it.  I'm sure he would urge you to stick with the real thing.  I don't necessarily disagree; as I note above, the real thing is very good.  (I note that Bobdot-ph and Bobdot disagree on about 14 % of the words in the dictionary, ignoring the use by Bobdot of "citation pronunciation".  If one just drops the accents from ê and û, the differences become negligible.)

Differences between Bobdot and Bobdot-ph

The rules of Bobdot-ph are the same as those of Bobdot, with the following exceptions:

  1. The letter u in Bobdot-ph always designates the schwa, as in utëmpt, rivut and ambyuluns.  Where an unstressed short u which is not a schwa occurs, the symbol û is used in place of the Bobdot u, as in bañkrûpt or bûcunïr.  Additionally, ûr is used to indicate the unstressed version of the ür sound, as in hambûrgur or netwûrk.  Because this conflicts with the Bobdot use of û for the stressed short oo, that sound is represented in Bobdot-ph by ëu, as in òvurlëuk or sikyëur.  This is very consistent with Bobdot, which uses eu for the unstressed variant of this sound.

  2. Many words include a sound I call the "indistinct i", which is sometimes pronounced as a schwa, and sometimes a short i.  Bobdot-ph spells this sound as i when the traditional spelling uses i or a, and as u when the original spelling is o or u.  The original Bobdot spells this sound with an e when the original spelling is e, as in kíndnes or jaket.  This is like TS, and fits the Bobdot philosophy of "citation pronunciation", but is not completely accurate, as the sound is never a short e.  Bobdot-ph uses ê in this case, as in kíndnês and jakêt, which allows a program which translates Bobdot-ph to another form to make a different decision as to how to represent the sound.

  3. Bobdot-ph encourages the use of the underscore to separate misleading digraphs, as in mis_hap (avoiding a misleading sh) or alcu_hol (avoiding a misleading uh).  I call the underscore optional because, if anyone were to ever use Bobdot-ph as a practical orthography, I'm sure the underscores would be forgotten.  Still, they are very handy for a pronunciation guide.

  4. Bobdot uses the spelling ng for the soft ng sound in the last syllable of a word, but ñ elsewhere.  Bobdot-ph behaves slightly differently, in an  attempt to be easier to analyze yet no harder to read or less precise.  Bobdot-ph uses the ng spelling for this sound at the end of a word, before an inflection such as -ing, or before an underscore (in derived words).  Thus, Bobdot lengtþ becomes leñtþ, while Bobdot upbriñing becomes upbringing, and strawñman becomes strawng_man.

  5. Bobdot uses the spelling uh at the end of a word to indicate the schwa sound.  Bobdot-ph also uses this spelling in cases where a syllable whose only vowel is a schwa would otherwise be assumed to be stressed.  For instance, the unstressed word 'em would be spelled 'uhm.  This is primarily a dictionary notation.  Strictly speaking, this rule would spell the words "a", "of" and "the" as uh, uhv and thuh, but outside of the dictionary, it would be better to treat these very common words as sight words with a more familiar spelling (using ov for "of" to get the consonant part right).

Dictionary notations

The downloadable Bobdot-ph dictionary uses a few more notations that are not strictly part of Bobdot-ph, but which ease the programming task of establishing correspondences between the pronunciation and the traditional spelling of a word.  A dictionary entry looks like this:

Bdph_spelling  - TS (description)

where the description part is optional, and is used to distinguish multiple pronunciations for the same spelling.  A few entries contain special characters, as follows.

A colon (:) is used in place of a hyphen in a word where a letter is used as itself rather than phonetically, as with T:shurt or X:muhs.

The plus sign (+) is used to indicate divisions in unhyphenated compound words that might otherwise be misanalyzed.  For example:

cou+hand  - cow+hand

where the wh digraph in the traditional spelling might be misleading to a program.

The semicolon (;) is used to help the analysis of difficult words by dividing them up into sections which are more easily matched.  For instance,

Hu;wö;yè  - Ha;wa;ii

where both the aw and the ai in the traditional spelling might lead a program astray.

The broken bar character (¦) is used to represent an apostrophe, when the apostrophe does not serve as a separator.  For instance,

dón¦t  - don't

In a few words, the apostrophe in the TS is replaced by a backquote (`), to indicate that the quote should be regarded as a separator, for instance

wé;r  - we`re

where a program might become confused without a separator between the e and the r in the traditional spelling.

Finally, the tilde character (~) is used when the apostrophe is not a separator, and the sounds around the apostrophe are linked, as in

the~r  - they're

The apostrophe variants above are not particularly well-defined, and may be of use only for my specific application.  Probably, these notations can be ignored in most applications of the dictionary.

Notes on the dictionary pronunciations

As with the other dictionaries on this site, the pronunciations in the Bobdot-ph dictionary were chosen by a majority vote of three respected American dictionaries.  They do not always represent my own pronunciations.  In his Bobdot system, Bob Boden uses citation pronunciation, which sometimes calls for long rather than short vowels in words such as because and proceed, bècäwz and pròséd in Bobdot.  Bobdot-ph does not use citation pronunciation, so that these words are spelled bicäwz and pruséd in the Bobdot-ph dictionary.  The Bobdot-ph dictionary does show stressed pronunciations for normally unstressed function words, such as "to" (), "for" (for) and "and" (and).  Only "a", "of" and "the" are shown with an unstressed spelling.

A small number of words are listed with a different pronunciation from the consensus dictionary pronunciation (signature words).  Usually, a minority pronunciation was chosen to avoid a difference from a closely related word, so that, for instance, pairs like "what" and "whatever" or "washed" and "unwashed" are pronounced similarly.  The full list of signature words is:

   aburunt  - aberrant  [should be abërunt]
   arkàizum  - archaism  [should be arkèizum]
   archdíusès  - archdiocese  [should be archdíusês]
   uwösh  - awash  [should be uwäwsh]
   bulónyuh  - bologna  [should be bulónè]
   comptròlur  - comptroller  [should be contrólur]
   díusès  - diocese  [should be díusês]
   däwgäwn  - doggone  [should be dögön]
   Febrùerè  - February  [should be Febyùerè]
   Halòén  - Halloween  [should be Halòwén]
   Jezùit  - Jesuit  [should be Jezhùit]
   Júdàizum  - Judaism  [should be Júdèizum]
   last-minut  - last-minute  [should be last-minit]
   minut  - minute (n)  [should be minit]
   minutman  - minuteman  [should be minitman]
   minuts  - minutes  [should be minits]
   noév  - naive  [should be nìév]
   noévlè  - naively  [should be nìévlè]
   noèvtá  - naivete  [should be nìèvtá]
   nesuling  - nestling (v)  [should be nesling]
   recugnïshun  - recognition  [should be rekignïshun]

   recugnìzubul  - recognizable  [should be rekignìzubul]
   recugnìzublè  - recognizably  [should be rekignìzublè]
   recugnìz  - recognize  [should be rekignìz]
   ricönusuns  - reconnaissance  [should be ricönuzuns]
   sacurèn  - saccharine  [should be sacurin]
   serátêd  - serrated  [should be surátêd]
   stonch  - stanch  [should be stawnch]
   Tþurzdà  - Thursday  [should be Tþurzdè]
   trawmätik  - traumatic  [should be trumätik]
   Túzdà  - Tuesday  [should be Túzdè]
   ûnaksëptubul  - unacceptable  [should be ûniksëptubul]
   ûnaksëptublè  - unacceptably  [should be ûniksëptublè]
   ûncümfurtubul  - uncomfortable  [should be ûncümftubul]
   ûncümfurtublè  - uncomfortably  [should be ûncümftublè]
   ûnrecugnìzd  - unrecognized  [should be ûnrekignìzd]
   ûnwösht  - unwashed  [should be ûnwäwsht]
   up-tuh-thuh-minut  - up-to-the-minute  [should be up-tuh-thuh-minit]
   vurtêbrè  - vertebrae  [should be vurtêbrà]
   wirweulf  - werewolf  [should be werweulf]
   wirweulvz  - werewolves  [should be werweulvz]
   whûtëvur  - whatever  [should be whotëvur]
   Whïtsündà  - Whitsunday  [should be Whïtsündè]
   resuling  - wrestling  [should be resling]

In addition to these words, there are a few words in the dictionary with pronunciations using foreign sounds.  These words have been spelled in the Bobdot-ph dictionary using alternate pronunciations which avoid these sounds, as follows:

   cònsyërzh  - concierge
   contrutom  - contretemps
   dànùmön  - denouement
   àlön  - elan
   lok  - loch
   màlönzh  - melange
   rapròshmön  - rapprochement
   soñfrwäh  - sangfroid

To comment on this page, e-mail Alan at

Go to home page
Go to spelling system roster