OTOH, hieroglyphs and similar ideographic writing systems do have some figurative characteristics which can help in their interpretation. Not so with syllabary/alphabet based ones, in which you would prefer at least a cursory understanding of the spoken form before going for an interpretation.
Chinese is interesting in this connexion. As far as I understand it, the 'language' is made up of different spoken forms, all of which are covered by the same writing system, are mutually incomprehensible when spoken. I don't know if there are other examples of this in human languages. Computer languages do provide examples - treading on thin ice here in such company: one could instance a piece of C code that is compiled for several different systems. In this case, the machine code for each system would bear little relationship to that for another system. I'll not go on in this vein, for fear of a pratfall.
It is interesting to speculate about artificial grammars. Most human and computer languages (with interesting exceptions, such as the Hopi Indians' concept of time -- others can describe computer exceptions)
Such exceptions always arouse my curiosity. Any online sources?
Don't know about online sources. As far as I remember, the Hopi used two tenses, one to describe the present, the other to describe dream time, taking in the past (memory regarded as a sort of dream, which Fraud would have loved had he known about it), and the future, as well as any loose ends involving magical or fabricated events, legendary time, etc.
I don't know whether any work has been done on constructing a seriously structurally different artificial grammar. Jorges Luis Borges has an interesting riff on the idea. If anyone's interested, I'll dig out the details.
"Tln, Uqbar, Orbis Tertius -- A reference to an imaginary country leads the author deeper into a different linguistical reality." is how the story's described on http://www.themodernworld.com/borges As far as I remember, the author discovers a reference to the country in an old and eccentric printing of an encyclopedia. The language described in the story has no verbs or nouns, using a compound form of both. To stray into another dangerous analogy with computer languages, one *might* say that object-orientated languages, by encapsulating data and methods, do something similar, but I'm not standing by this, because it's not my area of expertise. Noam Chomsky argues that the deep structures of languages are fundamentally similar: that the difference between Chinese and Old Irish is a matter of vocabulary, rather than means of signification. All the best Tiarnan