Most Japanese place names are written in kanji ("Chinese characters"). There are a number of romanisation systems in use for rendering the Japanese readings of these characters into English - some more intutive than others.
(See http://en.wikipedia.org/wiki/Romaji for an introduction.)

Japanese place-names used on Diddlefinger must be easily type-able into search fields and not require special knowledge to pronounce.

Therefore...

• Place names on Diddlefinger use only the standard English (Latin) alphabet of 26 characters.
• Constructions that would likely be misread by a native English speaker are avoided. For example, we render 'Ootemon' ("owe-temon") as "Otemon" - which is far from ideal. But "Ootemon" would likely be pronounced 'u-temon' by a native English speaker, "Outemon" is just absurd and a Hepburn style macron would be tricky to type into a search box. So "Otemon" is the least bad option. However, "Iizuka" conveys no such confusion, so is left with its double 'i'.
• We have tried to apply a principle of "least surprise".

The outcome is a slightly modified and simplified version of the Hepburn method of romanisation, although eschewing its use of macrons.


The sokuon or small "tsu".

In general we have honoured the sokuon except where we feel it would confuse the reader.

E.g. "Boccha" would likely be pronounced "Boc-cha" by a native English speaker, so we drop the first 'c''. However, "Kappa" would cause no such confusion, so we allow the 'pp'.


Hyphenation

The basic purposes of hyphenation on Diddlefinger are to increase readability, suggest relationships between places and pick out semantically interesting elements - such as, for example, "ekimae" ("in front of the station").

But what to hyphenate? Higashi, Minami, Kita and Nishi are obvious candidates as they often appear in long compunds - but doing so produces such inanities as "Nishi-jin", rather than "Nishijin". So - perhaps - hyphenate them only when they do not start the word.... But that too produces a crop of anomolies.... Treat the anomolies on a case by case basis? No - better to have a guiding principle, as people must type things into search boxes.

The current hyphenation scheme is provisional and a work in progress.