Chiense characters are pictograms (ideographs) because "dialects" (some argue truly distinct languages) within the Chinese empire truly sound different from each other. The use of a non-phonetic alphabet or syllabery allowed the exercise of imperial control over the entire country and its widely dissonant (er, literally) population simply because at least people could communicate by drawing the same shapes in the sand. Even today, the unified Kanji set promises to extend a common DigitalCulture over China, Taiwan, Japan and elsewhere despite vastly different spoken languages. Yet if the DigitalNetwork facilitates the de-literacizing of culture, say with VoiceInVoiceOut interfaces or more practically with such oral phenomena as audioblogs and viral videos, are we looking forward to a re-babelization of society to go along with the re-tribalization of the world McLuhan predicted?
People have already complained loudly about the inability to index, search, and machine-read audio and video, saying such things as audioblogs will be lost to Google and other SearchEngines. Further, they prevent skimming, tmesis (WhatIsTmesis), glossing, and other short cuts so necessary to contemporary reading. Yet, of course, there are SpeechToText converters that will undoubtedly get better over time. Audio can be machine transcribed to text, moving text between media just as MachineTranslation can move text between languages. Surely the audio dark ages (muffled ages?) will be short, overcome by necessity breeding invention. But this will only be a temporary reprieve.
To consider seriously the longNow, we all know that spoken languages change, not just in the literary dimensions of vocabulary and grammar--whose stability incidentally are critical to most SpeechToText systems today, including our brains--but also in the oral dimensions of accents, pronunciation, and enunciation. After all, we all wonder if people really spoke like that way back when--the oddly nasal and punchy mid-20th century American baseball announcers and Communist witchhunters. Even today, English has spread widely throughout the globe, a lingua franca of business. Each corner of the globe speaks with its own accent rooted in an ethnicity sometimes millenia old, but also ever-shifting as the GlobalVillage condenses us into a fluid, common OralCulture. Who can predict the outcome of this common community, least of all on how the glue of our common language will morph through time under such intense pressure?
In a few hundred years, all these audioforms stored dilligently by the InternetArchive may be unintelligible to the average listener. Worse, SpeechToText software that once understood these recordings may be lost or broken. These voices recorded may become LostVoices.
An audio Rosetta Stone may be needed, but no Rosetta Stone will exist unless hopefully the undying traditions of Shakespeare (and similar) maintains an unbroken (and salient) digital fossil record. A digi-paleontological record through linguistic and oral sediment.
Text presents its challenges as well. The real Rosetta Stone argues eloquently that text forms from antiquity can be difficult or impossible to decipher. Even more recent specimens remain difficult as any high school student studying Shakespeare will attest. Yet, truly text has more longevity and universality than sound. Shakespeare, despite challenges, is still readable, while a contemporary theatrical rendition of Shakespeare by a Scotsman or a Welshman will be difficult to understand by North American audiences. The common written language provides opportunities for shared understanding, just like with the Chinese empire.
Thus, we may be faced with a perplexingly shorted timeline for digital audio recordings exacerbating the already ephemeral nature of the impermenant DigitalNetwork. The LongNow may be a paradoxically antiquated notion, a vestigial value held over from the archivist PrintCulture and one perhaps best left to St. Leibovitz and his Order until a new Age.