Glottal stop

I like ʔ because, in addition to being identical to IPA, it looks like 2, commonly used in Arabizi for hamza. Yes, the big issue with ʕ is that it's hard to distinguish from ʔ and even more from the Arabic question mark ؟ when we romanize whole sentences in examples...

Fena will answer later (when back from holidays) but I want to add that it would be great to reach a consensus on Levantine romanization so that Levantine entries in the English Wiktionary, Levantine transcription used on the English Wikipedia (w:Levantine Arabic), and the Levantine course on Wikibooks all use the same Romanization. If one day the South Levantine Wikipedia is approved, we could even use it to automatically transliterate the Arabic script (similar to what is done on the Kazakh Wikipedia, which offers Cyrillic, Latin, and Arabic scripts, automatically converted).

On Wikibooks, original research is allowed so we can use any Romanization we want; on Wiktionary, I think there's also some freedom to set a different standard if there's a consensus. But on Wikipedia, I assume we should use an existing Romanization scheme from an external quality source.

So if we agree on this goal (a common Romanization scheme for Levantine on all Wikimedia projects), then we're limited to existing Romanizations of Arabic (because of Wikipedia's limitations). 10 are presented there: w:Romanization of Arabic and 6 for Levantine there: w:Levantine Arabic. Unfortunately, I'm not a big fan of any of them, but if had to choose I would go for Elihay, Aldrich, Wehr, or EALL. What do you think?