If a transcription is shallower if closer to the phonetic end, then why are representations morphophoneMic, not morphophoneTIc?

Question

1. I read Differences between phonemic and phonetic transcriptions, but no avail. Please see the terms that I colored in gray below. The book merely put them in bold, not gray. Don't the two sentences, with terms colored in gray, conflict?

I'm baffled because the sentence with "phonetic" is hinting at phonetics. But then the next sentence with "morphophonemic" brings up "phonology". hy did the author switch so suddenly from phonetics to phonology?

The first time I read this, I thought this word was morphophoneTic! Because the previous sentence was discussing "phonetic end"!

2. Then I Googled morphophonetic, but got zero results. Does the term "morpho-phoneTic" truly not exist? Why not?

Linguistic Level

A linguistic representation may relate to different linguistic levels. It can be described by its position on a continuum between deep and shallow. A transcription is shallower if it is closer to the phonetic end; a representation is deeper if it gives more morphological information. A morphophonemic representation is one related to linguistic units between morphology and phonology. In the examples above of the past tense, the writing system of English regularly represents the past tense morpheme the same way: <-ed>, except for irregular verbs such as kept, sent, rode, sang (this simplifies the situation somewhat), even though the past tense morpheme has two different sounding allomorphs /t/ and <-d/. In this case, English is using a deep transcription.

Henry Rogers, Writing Systems (2004), p 284.

Answer 1

The term "morphophonemic" derives from an earlier theory of language, taxonomic structuralism, which focused on defining various levels of linguistic analysis with a strict separation between the levels. Each level of analysis had its own vocabulary (in the sense of "things that it operates on"). The first level of "analysis" is the phonetic, where the linguist writes down what he hears. In creating a corpus for a language, this gives you a bunch of phones, and these are the basic "data" of language.

Then you analyze the distribution of these transcriptional symbols, and discover "phonemes" where each phoneme is a bi-unique organization of phones (every phone is uniquely mapped to one phoneme, every phoneme maps to one phone in a given phonetic context, which a complication in wording having to do with free variation). Now you have a phonemic transcription. A phonemic transcription can't neutralize the difference between phonemes (example: "writing" and "riding" cannot be phonemically /raɪtɪŋ/ and /raɪdɪŋ/ since they are phonetically indistinguishable.

The phonemes of a morpheme in one morphological context (an "allomorph") are related to the phonemes of the same morpheme in a different context – for example in German, the phonemes of /hant/ and /hand/ which are two of the allomorphs of the word meaning "hand" have, at the morphemic level of analysis, a morphophoneme //D//, thus the morpheme for "hand" is //hanD// with the morphophoneme //D// which is realised as the phoneme /d/ or /t/ depending on the context, and then the phonemes /d/, /t/ are realised however the allophonic rules dictate. This gives you another kind of transcription, a morphophonemic transcription.

This notion of a "morphophoneme" was taken into modern linguistic practice, to refer to the situation where the difference between two distinct phonemes is eliminated by some rule. In post-structuralist theories we expanded the concept of a "phonemic" transcription to include cases where we add information to the transcription that can only be determined by comparing phonemic realizations of a particular morpheme, therefore we can morphophonemically transcribe German "hand" as //hand// even though we only know that the last consonant is /d/ by looking at related word forms (doing morphological analysis). A morphophonemic transcription is a phonemic transcription enhanced by restoring lost information, recovered by doing a morphological analysis.

You should also note that "phonology" used to be "the rules relating phonemes to allophones" and "morphophonemics" used to be "the rules morphophonemes to phonemes" (morphemes are made up of morphophonemes). However, we abandoned the theory that there are separate sets of rules for "allophonics" versus "morphophonemics", but kept some of the terminology. The notion of "morphophonetics" would have been a theoretical anomaly because of the order of levels of analysis: phones > phonemes > morphophonemes.