📄 Moroccan Arabic and Jewish‑Arabic in Hebrew as Spoken by Israelis of Moroccan Origin: The Case of the TV Show Zaguri Imperya

Summary

This article looks at how Moroccan Arabic (MA) shows up in the Hebrew spoken by characters of the Israeli TV series Zaguri Imperya, created by Maor Zaguri. The show follows a Moroccan-origin family in a Beersheba neighbourhood. The author calls this variety "Peripheral Hebrew," preferring it to "Mizrahi Hebrew" because it names a social and geographic margin rather than an ethnic one. Moroccans were probably the largest immigrant group in those peripheries, so MA is plausibly the main substrate of this variety. The data is limited to the pilot episode. The author treats it with caution: fiction concentrates and exaggerates features, so it says more about how the phenomenon is represented than about the phenomenon itself.

The analysis classifies the material into:

• mixed bilingual expressions,
• words whose origin could be Moroccan or Palestinian Arabic (PA),
• everyday MA expressions now absorbed into Israeli Hebrew (IH),
• whole stretches in MA.

The main finding is that many items have several plausible, converging origins. This matters for contact linguistics and for any work on mixed-language speech.

Key points

• The corpus is the first episode of Zaguri Imperya. The Zaguri family consists of two parents and eight children (five sons, three daughters). The father, Albert, is the only character who produces entire sequences in MA.
• In xaʁam al a yeled ("shame on the kid"), the MA formula ḥṛām + ˁla is kept, but the preposition is the Hebrew cognate al. Similarly, šuf ze pairs an MA imperative with a Hebrew demonstrative.
• Moroccan and Palestinian loans are hard to tell apart, and speakers themselves often cannot tell. Words used by all IH speakers (e.g. falafel, yalla) were left out of the analysis.
• madʁub (crazy/stupid) is discussed through participle patterns: PA has maCCūC, while MA has məCCōC/məCCūC. The word may be MA məḍṛōb with its vowels adapted to Hebrew, a direct PA borrowing, or a blend of both. xalas gets a similar three-way analysis.
• lla isteʁ, šufuni and b‑əẓ‑ẓăˁṭ have spread into general IH slang with no felt Moroccan link. šufuni is treated as a single word meaning "show-off," and bzzaˁṭ survives only in its figurative sense, "who cares."

Interesting ideas

• I1. The author keeps competing etymologies open (PA source, MA source with adaptation, or convergence) rather than forcing a single answer, and lays out the evidence for each. — For Mova (learn): Silver labels where witness parsers disagree could be stored as ranked alternative analyses with their evidence, instead of being collapsed to one tree.
• I2. Some mixed expressions take their structure from one language and their words from another, such as an MA idiom frame filled with a Hebrew cognate preposition. — For Mova (frames): Treat idiom frames as templates keyed on structure plus a slot for a preposition class, so that calqued or slightly altered idioms ("shame on X") still trigger the right frame.
• I3. A multi-morpheme phrase (imperative + object pronoun) can become a single lexical item with a new meaning, as šufuni did. — For Mova (tokens): Keep a lexicon of fused expressions that the tokenizer and lemmatizer handle as one unit with its own sense, so that "look at me" is not decomposed literally when it is used as a noun.
• I4. Items that belong to the general language are excluded before analysis, so that only variety-specific features are counted. — For Mova (learn): When evaluating a domain or dialect adaptation, filter out tokens that are frequent in the general corpus, so that measured gains reflect truly domain-specific phenomena (a kind of negative control).
• I5. The author flags that scripted fiction exaggerates features and therefore reflects representation more than real usage. — For Mova (answer): Tag the genre of a source text, such as fiction, dialogue or reportage, so that the reader does not take a character's stylized speech or claims as facts about the real world.