What is AJP?

AJP is the official ISO abbreviation for South Levantine Arabic. South Levantine Arabic is the name given by Ethnologue — a language classification organization — to the Spoken Arabic dialects of Palestine & Jordan. Since Wiktionary follows the ISO standard, the name given to this continuum of dialects on Wiktionary is South Levantine Arabic (AJP). AJP seems to stand for "the Arabic of Jordan & Palestine", so I have retained the abbreviation as a convenient shorthand term.

Having mentioned the origin of the term & the South Levantine Arabic dialect grouping, I must point out that these terms have little if any linguistic validity. Ethnologue — the ultimate arbiter in the matter, as far as the ISO is concerned — is not even an organization of linguists. No linguist that I'm aware of would argue that South Levantine Arabic is a "language" & I furthermore think that, even as a general grouping of dialects, the classification is more inaccurate, misleading and generalizing than not. I'm simply compelled to use the term, particularly on Wiktionary.

That brings us to the crucial question: What is AJP? Is it a dialect of Arabic? Does it have a meaningfully distinct identity — or, rather: What is it distinct from? The Arabic used in Jerusalem is different from that of Damascus. And it's different from that of Nazareth or even Amman as well, even if the differences between them get increasingly smaller. Where does one dialect end & another begin?

Mainstream Ideologies

Situating AJP is not an easy task, because it must be situated in relation to something else: It requires some theory of what Arabic itself is. Simplifying for the sake of brevity, among laypeople these theories generally fall into two camps, which flow from the same assumption:

1. Modern Standard Arabic (MSA) is canonical Arabic, whereas
1. the dialects are simply non-standard forms of MSA, not languages.
2. the dialects are too different from MSA; they are independent languages.

I believe language ideology surrounding Arabic is so polarized & politically charged precisely on account of how easy it is to subscribe to either of these theories. Indeed, they rely on the same assumptions: One is free to decide what degree of difference is sufficient to designate a dialect an independent language, something to which linguistics itself has never claimed to have the answer.

Notably, these mainstream accounts are strengthened by a synchronic analysis (i.e. at a single point in time) of the uses of MSA & Spoken Arabic. Linguistically, the Arab World is characterized by widespread diglossia: While everyone acquires a dialect of Spoken Arabic through socialization, anyone with a basic education — or, in the past century, any exposure to media — has some command of MSA as well. Given that most native speakers are proficient in both, they may overlook the fact that MSA would probably not be mutually intelligible with Spoken Arabic for someone who was not exposed to the former from childhood, like heritage speakers & the illiterate.

Linguistic Theories

Diachronic Analysis

In contrast, historical linguists focusing on Arabic have offered a diachronic analysis of the dialects (i.e. of their evolution over time) that entirely debunks the notion that they are directly descended from MSA. In fact, MSA is not equivalent to Quranic Arabic in the first place; MSA itself is a descendant of it. Meanwhile, today's dialects of Spoken Arabic are descended from several dialects that were siblings of the dialect that was canonized in the Quran. Although their ancestors were displaced officially by Quranic Arabic, they continued to enjoy popular use & undergo evolution to this day.

What this would imply is that MSA & Spoken Arabic are neither in a parent-child nor sibling-sibling relationship; they are cousins, meaning that the Arabic language family is far more complex than what mainstream theories imply. It means that the Arabic dialect of Jerusalem is not just a dialect of Arabic more generally, but a dialect of Spoken Arabic specifically — not a dialect of MSA, its distant cousin.

As we can see in the foregoing diagram, AJP would be a subcategory of Levantine Arabic, which in turn is a form of Spoken Arabic. Note that MSA remained a single unit through strict standardization, while Spoken Arabic was diversified into a group of dialects with no standard form. My conclusion from this analysis is thus quite distinct from that of the mainstream theories:

2. MSA & Spoken Arabic are two distinct varieties of one language (Arabic). While the former is a single unit, the latter is a highly diversified grouping of dialect continua (e.g. Levantine) with their own sub-groupings (e.g. Palestinian) & their various sociolects (e.g. Urban) & locolects (e.g. Jerusalemite).

Why distinct "varieties" & not distinct languages, or distinct dialects for that matter? (One should be cautious of using such terms altogether, as they have no definitive — well, definition.)

Semantically speaking, the term "dialect" usually refers to some form of vernacular speech; it does not usually account for fully-featured language varieties only used in particular settings (e.g. liturgical languages). It therefore would be a somewhat misleading designation for MSA. Moreover, MSA & Spoken Arabic are, in my view, too different to think of as merely dialects. Not only is mutual intelligibility low — it would likely fall outside the Levant, as Levantine dialects tend to have the largest percentage of shared vocabulary with MSA. Here are just some of the key differences between MSA & Spoken Arabic:

• Morphology — Spoken Arabic lost case marking, all passivity, dual agreement & often gender agreement in the plural. Essentially the same morphology & word derivation from roots.
• Lexicon — Both varieties have borrowed terms via language contact, but often not the same terms or to the same extent.
• Semantics — MSA & Spoken Arabic may prefer different lexemes even when alternatives are technically present in the lexicon. High-frequency lexemes are often present in both varieties with entirely different meanings.
• Syntax — Spoken Arabic has significantly less syntactic flexibility, but the formulations it does allow are generally valid in MSA.

However, MSA & Spoken Arabic are not distinct languages either — at least not yet. What this chart demonstrates is that, although MSA is significantly more complex than Spoken Arabic in essentially every regard, their underlying grammar is not significantly different. (Indeed, this is why I personally emphasize the benefit of learning Spoken Arabic first & then building toward MSA; it is — for what it's worth — the natural path of acquisition for native speakers too.)

In fact, several of the key differences between the two are deliberately enforced: It is generally accepted that languages have tended to lose some of their complexity over time; the trademark example of case marking happens to be one of the key features of MSA morphosyntax, while the dialects lost all case marking through organic change. Had the dialects that were canonized into Quranic Arabic evolved similarly (in fact, there is convincing evidence that some such dialects exist), they would likely not be as different to other Spoken Arabic dialects MSA is.

Synchronic Analysis

Now that we have offered a diachronic analysis of the origins of MSA & Spoken Arabic to argue that they are two different varieties of the same language, we need to return to the synchronic analysis of diglossia among Arabic speakers to ask: How can we know where one variety ends & the other begins? Consider the following two phrases:

Lexically, the items أراد & ذهب clearly mark the first phrase as MSA, represented by بدّ & راح — respectively — in the Spoken Arabic phrase, while the dative preposition with a directional meaning is replaced by a slightly different preposition in Spoken Arabic as well. Syntactically, MSA uses إنّ before the Subjunctive & uses word-final case markers to indicate grammatical function in the sentence. Phonologically, الـ is realized as [ɪl] in Spoken Arabic, while /q/ is replaced by /ʔ/.

Apparently, there are several indicators of the variety in use within any given utterance. But at the word level, things are not so clear. It is actually rare to mix the two varieties in any meaningful way beyond word choice. But should the use of terms "from MSA" be thought of as an instance of mixing varieties in the first place — or of mixing registers (i.e. formal vs. casual)? In the diglossic setting, the lexicons of MSA & Spoken Arabic coexist in the mind of the speaker & are accessible at any time. How, then, can we draw the line — if there is one — between the MSA lexicon & the Spoken Arabic lexicon?

Since MSA is highly normative, it is easier to say that using certain terms (e.g. بدّ) as they are used in Spoken Arabic is proscribed, if not invalid. However, whether an MSA term may be used in the dialect is more reliant on an individual judgment of whether or not its use seems "natural" in the dialect — a more subjective judgment that is influenced by the individual's own extent of diglossia. Due to sheer exposure, someone who produces MSA on a daily basis may find the use of a certain term in the dialect more "natural" than the next person. Using رَجُل instead of زَلَمة is arguably quite unnatural, yet using طِفل instead of وَلَد is certainly plausible, even if the former is certainly more marked.

How can we know, then, what terms to include in a dictionary of Palestinian Arabic? One place to start is by noticing that the Palestinian Arabic lexicon — like that of many other languages — has at least two etymological categories pertaining to diachronic origin: inherited terms & learned borrowings. In brief, inherited terms are those that have undergone all the evolutionary changes that distinguish the contemporary language from its ancestor at the time the term was coined or introduced. On the other hand, learned borrowings are introduced — often from an ancestor or prestige form of the language — with the morphology & phonological features of the donor language intact.

By taking inventory of the morphological & phonological features of the Spoken Arabic dialect in study, we can distinguish between terms that were inherited from earlier forms of the dialect versus borrowings from MSA. Urban Palestinian dialects, for instance, have no native /ð/ — it merged with /d/ over time. Consequently, inherited terms with /ð/ at the time of their inception now feature /d/, while learned borrowings maintain /ð/ (realized as [ð] or, more commonly, [z]).

However, terms that only feature phonemes that did not undergo change over time cannot be assessed in this way. Generally, "semantically precise" terms (e.g. اغتال "to assassinate") are more likely to be learned borrowings, yet since all the phonemes in اغتال are natively attested in the dialects, we cannot by means of phonology alone make such a judgment for this term.

Although phonology is the main way we can distinguish between inherited terms & learned borrowings in Spoken Arabic (the native phonology of Spoken Arabic is discussed in further detail in the section on Phonology), morphology plays a role as well. Form 4 (أفعل), for instance, is rare in the dialect, if not natively non-existent (with the possibly singular exception of أعطى). Aside from "semantically precise" terms (e.g. أعلن "to announce"), all terms that are attested in this form in MSA & have currency in the dialects have a native Form 1 (فعل) or Form 2 (فعّل) version. We can therefore argue that all Form 4 terms attested in the dialect are learned borrowings.

However, this approach has significant limitations as well. Given the very sparse attestation evidence of Spoken Arabic over time, rarely can we know when a term started being used in Spoken Arabic: Has أشتى been in the dialects since the medieval period, or has it enjoyed currency in the dialects more recently? (Let's not forget that before the global boom in literacy, MSA most likely had a far weaker influence on Spoken Arabic.) This is important because it means we don't have a way to properly determine whether certain terms are actually learned borrowings present in the Spoken Arabic lexicon versus lexical items from MSA that are seamlessly used on account of most native speakers having access to both lexicons. Just because someone uses the term رَجُل does not necessarily mean it's in the native Spoken Arabic lexicon.

One indicator of lexical attestation could be the whether the term may be subject to Spoken Arabic inflections. In the case of أشتى, the acceptability of بشتي suggests that the term is present in the Spoken Arabic lexicon. Conversely, consider that more educated speakers may summon MSA passive forms to convey a precise meaning due to the absence of native passive forms in Spoken Arabic (e.g. الكلمة تُعتبر مش منيحة); these passive inflections are not generalizable, but rather are attested purely on a case-by-case basis & usually without the imperfect marker بـ. However, this criterion is not very reliable: the imperfect marker بـ may be optionally attached regardless of the inflection's lack of general attestation in the dialect (الكلمة بتُعتبر مش منيحة).

In summary, داق (from ذاق) is definitely an inherited term, while ذوق is definitely a learned borrowing. But أشتى may be either a learned borrowing or — if inflectional flexibility is considered irrelevant — it may not be in the Spoken Arabic lexicon at all. Finally, اغتال could be an inherited term, a learned borrowing, or not in the lexicon. While it is probably fair to say that أشتى & اغتال are learned borrowings rather than unattested, we have no formal basis on which to make such a conclusion; there are many terms that will be far harder to make a call on.

Concluding Thoughts

In summary, I neither claim that MSA & Spoken Arabic are distinct languages nor that "the Arabic of Jordan & Palestine" itself has any sort of distinct identity. I use AJP as a shorthand. In reality, the notion is not well defined internally or externally. Questions remain: If the dialects of Palestine & Jordan have something in common that adjacent dialects lack, what exactly is that? Is there a more accurate way to categorize dialects according to shared features than via geography & country borders?

But defining the linguistic identities & phylogenetics of the dialects of the Levant is far beyond the scope of this project. Just as British schools teach British English — not American English — I can only teach what I know. If I say I'm teaching AJP, then, what exactly am I referring to? Often, the name of a country is used in the mainstream to refer to the dialect of its most influential region or city (i.e. the capital). As for this site, it's not really a full outline of "Palestinian Arabic" (does such a thing exist?), but a description of the Spoken Arabic dialect used in Jerusalem specifically. I have tried to note some of the salient features of surrounding dialects to furnish a fuller picture of the linguistic landscape of Palestine, but the starting point is the Urban dialects of Central Palestine.

I may not have all the answers you wanted, but I like to believe that the result is highly accurate.