Phonetic Decomposition is a method of obtaining rich languages through the use of only a source language and a basic idea of how the language shall work. The method is outlined in the steps below, along with alterations for adding to languages.

Step 1: Choosing Suitable Sample Sentences[]

Firstly, one needs to select a set of suitable sample sentences. There may be some constraints on these once one reaches specific grammatical details which one should like to obtain, but without these one can choose literally any sample sentence one should like. Make sure the sentence is transcribed phonetically, not using a system which may not have 1:1 correspondence between glyphs and phonemes.

Step 2: Fusion[]

Now things become messy. One removes all the spaces separating the words to create a single long phonetic string. In the case where one has a set of sentences in an ordered string (if one were using a story, for example) then parts may sometimes be moved between strings and added to the appropriate location (that to which it was adjacent).

Step 3: Pulling apart[]

Now one gets to exercise one's personal whim. The phonetic string we have created must be divided into words. Note that the divison is not necessarily connected in any way to the former words. The new words need not be pronouncable either. There may be oddities in the writing system which mean that there is not 1:1 correspondence between glyphs and phonemes.

Step 4: Grammatical Extraction[]

We are almost done. Now we simply need to divide the words up into their morphemes (if there are more morphemes in some words than just a root), and then examine the sentence for any syntactical or morphological information present (its meaning is not changed in the process undergone above). This need not be done straight after one sentence; sometimes it is better to wait for a pattern to appear in multiple sentences before generalising.

Notes By The Author[]

This method has been employed in the creation of Fast Fusion. Research consists of using this method to learn more about the grammar and vocabulary of a language. If you would like to help, simply submit sentences created by the above method with their meanings to Meuser2's talk page (not the article for Fast Fusion). They will be considered as material for use in the language.