Thread by @matt_boot_

another consequential ⭐️sound change⭐️ in #EgyptianArabic is the loss of the short vowel /i/, when 2 conditions are met:

1️⃣ the vowel doesn't carry the word stress;

2️⃣ deleting it will not create a syllable with shape CVCC (i.e. with multiple coda consonants).

examples within a word:

ˈʕārif-a → عارفة ˈʕarfa "she knows"

θaˈmāniya → تمانية taˈmanya "eight"

naḍˈḍārit-i → نضارتي naḍˈḍarti "my glasses"

consider especially the effect this has on verb paradigms:

كتبت ˈkatab-it "she wrote"
but
ˈmišiy-it → مشيت ˈmišyit "she left"

ˈʕamal-u "they did"
but
ˈrikib-u → ركبوا ˈrikbu "they stole"

note the re-syllabification that can happen, even across multiple words:

ده كتير "that's too much"
ˈda kiˈtīr = dak | tīr

شقة كبيرة "a big apartment"
ˈšaʔʔa kiˈbīra = šaʔ | ʔak | bī | ra

عايز أعرف "i (m) want to know"
ˈʕāyiz ˈaʕraf = ʕay | zaʕ | raf

notable exceptions include:

1️⃣ the 3rd person fem. perfect with an object suffix;

you'd think شافت ˈšāfit + ـك -ak might yield *ˈšaftak .. but it's šaˈfitak "she saw you (m)", with the word stress moving forward.

...

... 2️⃣ in loans from (or code-switching into) Standard Arabic

ex. نادراً nāˈdiran "rarely" (not *ˈnadran)

this rule can apply to /u/ as well, in parts of the grammar where it varies with /i/ .. like the participle affix mi-, wh can be mu- in higher-register speech.

but i don't think native speakers would syncopate e.g. ˈkutub-i "my books" to *ˈkutbi (correct me if i'm wrong). 📚

this distinction — between unstressed /i/ that undergoes syncope & unstressed /a/ that does not — reminds me a LOT of #Coptic (at least Sahidic), which appears to contrast 2 unstressed vowel phonemes, one <ⲁ> /a/ that does not syncopate and one <ⲉ>~ø /ə/ that does.