some ideas (not all novel) about writing systems

Writing systems don't have to show phonological contrasts, they have to show morphemes. Alignments of form with meaning.

That's why logographic systems do just fine, from the familiar Mandarin one (though see the note to the preceding post) to the fascinating and still-woefully-underappreciated-for-its-significance to-the-linguistics-of-orthography Mi'kmaw komqwewi'kasikil system. They write morphemes, or subsets or combinations thereof.

That's also why even alphabetic writing systems dribble with morphological sensitivity.

English has the -s' vs. -'s singular and plural genitive contrast so beloved by pedants and so hated by regular folks (especially when adding to that the fun of adding to -s-final wordforms).

Arabic has its tāʾ marbūṭa, has its dropping of the alif with li- but not with bi- (justified, yes, but largely on graphic grounds rather than phonological ones), its floaty little alif after verbal 3plural -ū.

Irish writes "an" all the time for definite article a(n), even as it applies basically the same rules as Welsh y(r), whose system goes the purely phonological route.

Doubtless many more examples exist---certainly English heterography is: while often historically motivated by a original phonological difference, now it's being exploited purely morphologically. I have a three-way merry-Mary-marry distinction, but I wonder how many who phonologically contrast two or less of these then have this just filed under morphological graphic (morphographic?) contrast.

Graphic elements need to contrast morphology. They can do it by representing phonological contrasts---since those carry out said task pretty well in speech---but they are not limited to that. And tend to happily mix back and forth between this kind of phonologically mediated graphic-to-morphemic contrast, and a more direct graphic-to-morphemic contrast.

Perhaps if modern linguistics hadn't developed largely from a scholarly world steeped in strong phonology-to-orthography mappings---by this I mean Pāṇini, not English, for example---we would not tend to think, rather misguidedly, that writing is for marking down the sounds of languages.

It is, rather, for marking down words. In the loose sense of the term, i.e. morpholexemes of all shapes and sizes. And whatever works, works. As the evolution (convergent and divergent) and diversification of human-used orthographies tends to show.


And, as Rich Rhodes has helpfully pointed out, when it comes to representing phonology in writing, salience very often outweighs contrastiveness.

-Hence no prosody, either of the lexically contrastive types incompletely listed below, or especially of intonation. [Pretty much everybody, barring performance-detailing religious texts]
-Rarely if ever tones, and most often from their original segmental sources, and/or serious post hoc scholarly metalinguistic devisings. [Thai, Burmese, etc.]
-Often no vowels. [Tifinagh]
-And if vowels, only the salient ones, namely the long ones. [Arabic, etc.]
-No codas, even if there are at least some. [Sulawesi-area lontarak, and lots of other syllabaries]
-Weird treatments of obstruent phonation contrasts. [Old Irish, Mongolian (in both cases perhaps just because the donor orthography isn't well-matched, but still....)]