I have read the chapter by Norman, although I was not able to access it from the Jōdo Shinshū site. However it is available on books.google.com. The article is chapter five of this text.
Try this URL (paper V)-
http://www.shin-ibs.edu/academics/_forum/v5.php
Norman's contention is not that the language(s) that formed the basis for the written language of Pāḷi and the early oral recessions of the canon lacked these consonants, but that the early orthography used to write it down did.
Absolutely, I am only taking into account the facts about the original script (and not his theories about the language itself) from Norman's article.
I accept your claim that there was no change to the spoken language between the imperfectly written pre-Pali text, and the Pali texts that we currently possess i.e. the evolution was merely orthographic and not linguistic.
What he has shown is that the non-phonetic script originally used to write down the canon omitted double-consonants which are found in Pali (among other imperfections). Let us call this early written form (which lacks the geminates found in Pali) the
pre-Pali form.
Therefore we have the following known forms for the Pali word Dhamma (listed in chronological order):
Old-Indic form: Dharma (all pre-Buddhist literature)
Pre-Pali form: Dhama (furthest away from the Old-Indic form)
Pali form: Dhamma (close to Old-Indic)
Post-Pali (i.e. Buddhist Hybrid Sanskrit) form: Dharma (virtually identical to Old-Indic)
Therefore we see a trend between the Buddha's parinibbana and the start of the common era. Initially there was the Old-Indic form present in all pre-Buddhist literature where middle indic is completely absent. With the advent of writing in India, which happens to almost coincide with the the Buddha's parinibbana, we have the first written texts (i.e. the Buddhist canon), that too in the pre-pali form (which we now associate with the early Kharosthi texts, now called the Gandhari texts). However we have not found a copy of the entire canon from this written stage, all we have are a few suttas usually from some of the oldest parts of the canon (i.e. the sutta nipata), such as the Khaggavisana sutta etc.
The canon then either naturally evolves (or is artificially recast) into the pali form that we have today as the script evolves further and eventually becomes fully phonetic and capable of even expressing the Old-Indic conjunct consonants. This phonetic script is called the Brahmi script, which becomes the de-facto standard for both Pali and Old-Indic texts. The Kharosthi script also evolves to the extent of representing the conjunt consonants, but it still has limitations in displaying vowel length distinctions etc, and eventually is abandoned.
You said there was no change to the spoken language, and the evolution from pre-Pali to Pali was restricted to the script, which I accepted above. However I would like to extend the same logic both sides i.e. to before the pre-Pali stage and also to the post-Pali stage. I believe therefore that the language of the canon that was imperfectly written in the pre-Pali form, was probably spoken old-Indic. The spoken language never became middle-Indic, the existence of the intermediate forms i.e. pre-Pali, Pali etc. were due to introduction of writing and the evolution of orthography. This evidently completed a full circle by reaching the Buddhist Hybrid Sanskrit stage where the written form became virtually identical to Old-Indic.
Writing must have been first adopted in India by the Sangha as they evidently had no independent oral traditon to entrust their suttas to (considering also the sheer size of the canon). The canon therefore was probably always a written canon. Once the script had become fully phonetic, and Old-Indic could now be written phonetically, all new texts were read phonetically as they were written. The Pali texts (which earlier had not been read phonetically due to being in a non-phonetic script) now in the phonetic script were also read phonetically, giving rise to a new language, once called Paisachi, now called Pali.
However, the early orthographic versions were tied to a parallel oral tradition so even though the lack of indication of length for consonants and vowels allowed for the ambiguous readings of some items, the oral tradition was able to clarify these.
We don't know about that, the existence of such an independent oral tradition is entirely speculative. Norman assumes that they must have existed since he claims there is a word "bhanaka" in the commentarial literature which may have meant such oral reciters. However, these bhanakas, if they were indeed reciters of the suttas, cannot have formed an independent oral tradition so late in the canon's history i.e. nearly a 1000 years since the suttas were first recited, see below.
What he proposes is that there was a rupture at some point where the oral tradition ceased while the written tradition continued. When length was introduced to the orthography the ambiguous readings were more problematic for the scribes and some errors were introduced.
This means the rupture i.e. the demise of the independent oral tradition (if it was independent to begin with) happened when the written canon was at the the pre-Pali stage, since Pali has evidently inherited the ambiguities of that stage.
Therefore the word bhanaka found in the commentaries, if it referred to oral reciters, must not have meant an independent tradition, for such a tradition (assuming it did exist) evidently died out even before the pre-Pali text became our current Pali-canon.
I asked for proofs. All you present is just empty speculations.
That the canon was first reduced to writing in the 1st century BCE itself is an empty speculation. We don't have conclusive proofs for it.
As I said above, there exists a pre-Pali written form of suttas, which shows written forms that are older than Ashoka's rock edicts at Girnar. The pre-Pali written canon which must have once existed in its complete form must therefore pre-date Ashoka (3rd century BCE).
