The affixation differentiating between nominal arithmetic and adjectival arithmetic

Question

Since a suprafix can be the change of stress somewhere in the word (or other suprasegmental elements), and since accentuation plays a role in differentiating the noun arithmetic from the adjective arithmetic, I suspect that a derivational suprafix is at play.

The noun: [uh-RITH-muh-ik]

The adjective: [ar-ith-MET-ik]

The pseudo-phoneticization is taken from Dictionary.com.

Now, the only difference isn't just the accentuation. There is a change in phonemes:

The noun: /əˈɹɪθmətɪk/

The adjective: /æɹɪθˈmɛtɪk/

The phonetic transcription taken from Wiktionary.com.

Now, assuming the noun to be the original lexeme, the adjective would be affixed with a simulfix located at the first and sixth phoneme, replacing the two əs in the noun with an æ and an ɛ respectively.

So, if in fact there is both a simfulfix and suprafix deriving the adjectival form of the noun arithmetic, how would one write the morphological structure then? I've tried to represent simulfixes before (in a comment responding to a comment by @Draconis), but I have no idea how the format is supposed to look. Less so when it comes to suprafixes.

Answer 1

I've never seen this kind of vowel alternation analyzed as a "simulfix" like that.

I would say that it would be preferable to use either of the following analyses:

• the different phonological forms of the words are simply stored separately in their whole forms: (/əˈɹɪθmətɪk/ and /æɹɪθˈmɛtɪk/). This is redundant (due to the common consonants between the forms) and doesn't fit too well with the whole idea of morphology where surface forms of related words are derived from abstract underlying representations; however, I'm not sure that this kind of redundancy doesn't actually exist. It obviously doesn't answer your question though.

• There is a derivational process where at at certain point in the derivation, the only difference between the words is the position of stress. E.g. you consider the underlying form to be something like /æɹɪθmɛtɪk/ with vowel reduction of unstressed vowels in the surface forms. Vowel reduction is a phonetic process, so there is no simulfix involved.

  I've seen a few accounts that treat -ic adjectives as having, in their underlying representation, extra material at the end that is not overtly expressed segmentally in the surface representation. Chomsky and Halle's The Sound Pattern of English (1968) proposes covert -al (I don't remember the details), which perhaps could be theorized to be absent in the noun, so the morphological structure would be something like æɹɪθmɛtɪk vs. æɹɪθmɛtɪk-æl where after stress assignment, some morphological process causes the æl to not be pronounced. "The architecture of the English lexicon" by JB Alcántara (1998) treats the stress of -ic words as derived from an underlying form with final schwa (that doesn't get pronounced on the surface), like "ɪkə"; in this case, we would have noun æɹɪθmɛtɪk vs. adjective æɹɪθmɛtɪk-ə I guess.