hioxian phonemic alphabet can predict coarticulation

http://www.robotroom.com/AlphanumericDisplaySchematic.gif

http://www.infoverse.org/octomatics/octomatics.htm

D: why these three apparently unrelated images?
Cuz they're not.
At least, they won't be.

http://www.phon.ox.ac.uk/~jcoleman/MULTART.htm

D: this is ANOTHER beef of mine about Esperanto. Or, should I say, Espranto with an 'uh'.
My point is that colloquial speech does not use formal pronunciation.
We speed up the words. We don't leave discrete pauses between them.
They blur together, the syllables and words.
Some combinations of phonemes are more or less prone to this.

D: let's call this tendency a "mutation".
Here is an interesting piece of trivia. Sharks, being largely nigh-perfectly evolved to exist in a fairly changeless setting, would be made less fit by any new mutation. So they evolved more safeguards against mutations than other creatures.

Well, a well-designed language ought to be a shark. Each phoneme, each letter is lovingly crafted for meaning and significance. The loss (or even gain) of any one makes the language less fit. Just like a shark. The ocean, in this analogy, is the human physiology that supports speech and writing.
Unless we start talking about AIs or transhumanism, human nature is likely to remain precisely what it has been for quite some time.

Now back to Esperanto. The book I read on it advised me to carefully pronounce each word as instructed. That's right - an IAL that forbids speakers to use ... colloquial speech! <: Wow. Don't get me wrong. Esp-o is a century old! Great-grandpa's holding up pretty well for 100+. IT has an excuse. The latest crop of interlangs has NO such excuse.

I was looking over my Decimese prototype. PB TD SZ FV KG. LR WY. H. 5 vowels, long and short. My mechanistic approach to syllable construction has issues.
Vocabulary is made via CV(nasal consonant). CVCV...CV(nasal).
Limited options to make C-YV and C-WV... . Also, VC-L and VC-R(nasal).
I reread some articulation article recently. I realized that ending in an N would deform certain following words that begin with certain sounds.
Note: there are some language-specific trends. Do the vowels voice or devoice consonants? There are some near-universals though based in biomechanics.
I think French and English use opposing strategies.
I wanted ONLY voiced OR voiceless consonants to begin a word or to be in the middle of it.
Allowing for voiced/voiceless pairs, limited rules for acceptable word last-letters, syllable stress and recognizable vocabulary, I have an unprecedented FOUR ways to indicate word boundaries.
These are terrifically important, as a perusal of Eurekalert and Sciencedaily will indicate.

D: back now to my main point.
I need a simple way to predict how adjacent phonemes will interact.
Octomatics has a system where somebody innumerate can visually do arithmetic.
I can do the same with an alphabet (writing system, whatever. Pedant! <:). My Hioxian alphabet essentially maps the human articulation diagram onto a 16 segment alphanumeric character, plus diacritic.
By simply comparing any two phonemic letters, somebody unable to read at all could still predict what symbol would instead result due to coarticulation.
This is a labour-saving device for me.
I don't need to consider each pair of letters.
21 or 26, depending. That is many 100s of combinations!

Ever try to read the IPA definitions in a dictionary? It is not intuitive at all.
I still don't know half of those symbols myself.
The learning curve on Hioxian is steep!
Note: a steep learning curve would suggest rapid learning. I never understood why we use that turn of phrase! Leading to one of my favorite put-downs, "a learning curve like a horizontal line". <:
And there you have it.
The Future...