Thackish

Proto-Majorcan Arabic is the variety of vernacular Arabic that eventually became Majorcan, specifically the stage before it started absorbing Irish influence.

(does van Putten's book have any leads on Neo-Arabic?)

Ultrashort yers? ŭ ĭ, with latter only allowed after nonemphatics

Phonological history of vowels

a i u ā ē ī ū

Final vowels

-ah, -ā, -ē, -ā2 > -ə

Phonological history of consonants

Segolate epenthesis

1V23 > 1V2ə3 (accounts for 3aSr vs 3aSR) if 3 = resonant

R vs. r

2i3rāb in an earlier stage gave rise to a phonemic split between emphatic and nonemphatic r. Final -r became nonemphatic in nouns and adjectives (from -ri) but emphatic in verbs (from -ru and -ra) and later in verbal nouns by analogy.

Emphasis spread

Sketch of emphasis spread rules:

1. T D S Z q x γ are emphasis sources
2. R is an emphasis source unless it's next to a [−emphasis] consonant where ir becomes r
3. In a cluster, emphasis sources spread emphasis onto consonants that are not marked for emphasis (nonemphatic consonants that are not š y j d s z)
4. The vowel before the cluster in the emphasis source is affected
5. Consonants with the [−emphasis] feature (š y j d s z), and the stressed vowel, absorb emphasis and emphasis (usually) can't spread past them
6. If the onset of the stressed syllable is emphatic, inflectional prefixes become emphatic
7. If the coda of the stressed syllable is emphatic, then the emphasis spreads all the way to the end of the word
8. Enclosing rule (rìġail et tedvìṟ): If the outer consonants of a consonantal root are emphatic then the middle consonant(s) become emphatic

   xalaqa > xaḻag̱ 'to create'

Some kind of metrical foot-based emphasis spread system?

dr DR yes, Dr no, dR not sure

rd RD yes, rD Rd no

Morphology

Hollow root ʔimālah depends on whether it's 2-y or 2-w:

• *zēd 'he increased', *māt 'he died'

Syntax

Proto-Majorcan Arabic had the emphatic "ADJ el-NOUN" construction (from earlier "STATIVE_VERB al-NOUN"), which got reinterpreted as 'an ADJ NOUN' under Irish influence: "In Classical Arabic syncopated forms do not usually occur, the only place where they occur is when the verbs naʕima ‘he is glad’ and baʔisa ‘he is miserable’ are employed as pseudo-verbs of emphatic qualification, such as niʕma r-raǧulu ‘what a wonderful man!’ and biʔsa n-nisāʔu ‘what evil women!’ (Fischer 2002, §259–263)." (van Putten)