The Syntax of Yes/No Questions in Modern Standard Arabic

Interrogative structures have been investigated in wide range of languages including but not limited to English, Italian, French, and Mandarin Chinese. Thus, this paper presents an analysis of the syntactic structure of yes/no questions based on feature-checking analysis (i.e., [Q], phi-features, [T], [Polarity], and EPP). First, I briefly discuss the feature-checking analysis in the declarative clauses in Modern Standard Arabic. Then, I analyze the interrogative structure in main clauses (hal, ʔa-) and in embedded clauses (idhaa) in MSA. Finally, this paper displays and discusses the findings showing that there are three types of feature-checking in yes/no particles in Modern Standard Arabic.


Introduction
Modern Standard Arabic (MSA) (Note 1) is derived from Classical Arabic and is one of the most widely spoken languages in the world. Both MSA and Classical Arabic share morphological and syntactic structures, however, each exhibits some differences from the other in respect to vocabularies and stylistic features.
During the past two decades, numerous research papers discussing issues in Modern Standard Arabic or its many dialects have been written on topics such as word order, agreement, negation, and case theory, yet little attention has been paid to yes/no questions. In this paper I address the structure of yes/no questions in Modern Standard Arabic. I also analyze yes/no questions based on feature-checking analysis using Q-feature, phi-feature, EPP, tense, and mood (Note 2). I present my analysis of yes/no questions in Modern Standard Arabic in three parts. First, in Section 2, I show and discuss the structure of declarative clauses based on feature-checking which explains and helps to analyze the structure of yes/no questions in MSA. Second, in Section 3, I investigate and analyze the syntactic structure of yes/no questions in main clauses containing one the interrogative particles hal or ʔa-, and in embedded clauses containing the interrogative particle idhaa. Third, in Section 4, I conclude by summarizing the three types of feature-checking mechanisms in interrogative structures in Modern Standard Arabic.

Clause Structure in Modern Standard Arabic
In this section I present and discuss the ways in which verb movement, EPP, and phi-features impact feature-checking mechanisms and derivations in Modern Standard Arabic. Furthermore, I argue that the deep structure word order of interrogative clauses in Modern Standard Arabic is subject-verb-object (SVO) while the surface structure has a verb-subject-object (VSO) word order that is a result of strong features which derive the VSO word order in the surface structure.
Scholars such as Doner (2013), Aoun, Benmamoun, and Choueiri (2010), Rouveret (2010), Al-Horais (2009), Soltan (2007), McCloskey (2001) Carnie, Harley, and Pyatt (2000), Ouhalla (1994Ouhalla ( , 1996, AL-Shorafat (1998), and Aoun, Benmamoun, and Sportiche (1994) and Mohammed (1989) have investigated verb movement in a variety of languages having VSO word order. Their studies examine feature-checking systems such as EPP, phi-features, T, and case in order to determine the surface word order for each of the languages considered. Mohammed (1989), Ouhalla (1994), Benmamoun, and Sportiche (1994), Soltan (2007), and Aoun, Benmamoun, and Choueiri (2010) claim that Modern Standard Arabic has both a VSO word order and an SVO word order. This indicates some differences in the feature-checking system or the presence of a strong/weak feature which is responsible for the two-word orders in MSA. Additionally, Alsaeedi (2015) asserts that MSA exhibits both VSO which triggers a lexical item to Merge/Move which in turn satisfies the strong feature [+Q]. He also asserts that the Q-feature could be checked by the T-movement as in (9b.), or by the base-generated complementizer as in (9c.). For example, in yes/no questions in English, there should be a lexical item to check the strong feature [+Q] which is an auxillary (AUX) element in English as in the following example (9): (9) a. Sarah ate the apples.
b. Did Sarah eat the apples?
c. I wonder if/whether Sarah ate the apples?
Applying Chomsky's and Radford's analyses, I assert that MSA has in fact two approaches for checking the Q-feature. The first is the base-generated approach found in idhaa and ʔa-clauses similar to the complementizer whether/if in English. The second is the raising approach found in hal clauses which move from the head of the Polarity Phrase (PolP) to the head of the CP. In Section 3.1. I discuss the syntactic structure and the possible analysis based on Mohammed's analysis (1989) for the interrogative particles hal and ʔa-in main clauses. In Section 3.2. I present and discuss the syntactic structure of the interrogative particle idhaa in embedded clauses.

Interrogatives hal & ʔa-
The interrogative particles hal and ʔa-are defined in Almojam Alwaf as question particles which ask for agreement (yes) or disagreement (no). Interrogative hal is a free morpheme, while ʔa-is a bound morpheme, in this case a prefix. Both particles must be located at the beginning of a main clause. One difference between these two interrogatives is that interrogative hal must not appear in any sentence which has a negative phrase (NegP) or any complementizer, while the interrogative prefix ʔa-may appear in any sentence even if that sentence has a complementizer or NegP, Why? Another difference between hal and ʔa-is that hal must be followed by a verb-element with a full agreement on the verb (i.e. hal + VSO with full agreement), while ʔa-could be followed by the subject or the verb-element with a full agreement (i.e., ʔa-+ SVO/VSO with full agreement in SVO/VSO word order As mentioned previously, the first difference between interrogative hal and the interrogative prefix ʔa-is that interrogative hal must not appear in a sentence containing a NegP or a complementizer, while the interrogative prefix ʔa-may be used in any sentence. I propose that this difference in usage can be attributed to grammaticalization. Grammaticalization is defined by van Gelderen (2011, p. 5) as "a process whereby lexical item loses phonological weight and semantic specificity and gain grammatical functions". Example (11) from van Gelderen (2011, p. 7) schematizes grammaticalization from a historical linguistics point of view, (11): (11) a. phrase > word/head > clitic > affix > zero b. adjunct > argument > (argument) > agreement > zero Using this schematic, I traced the grammaticalization of the interrogative hal to determine why hal could not show up with NegP or complementizers such as ʔin. In fact, I found that hal can be divided into the ha-element and the l-element. ha could be traced to the demonstratives in MSA such as hatha (meaning this for singular masculine), hathih (meaning this for singular feminine), and hawlaii (meaning these for plural masculine or feminine). In Hebrew, Eid (1989) pointed out that hu is used as a copula as in (12): (12) David hu ha-more David he the-teacher-MS (Eid, 1989) "David is the teacher"  (24) and (25) show that the verb element which is [-perfective], and in this case the interrogative idhaa would go with the elsewhere condition kana-insertion, so the verb kana would check the [subjunctive mood] of idhaa; thus idhaa can't see/search because there is a barrier which is kana, and therefore the EPP feature would be strong and we will have spec-head relation in the lower CP which would result a full agreement on the verb.
In summary, the interrogative idhaa is generated in the head of its CP, and requires a perfective verb to follow it in order to check the subjunctive mood of idhaa. There are two ways to check the subjunctive mood: first, by moving the perfective verb to the head of TP, which would result a weak EPP. The other way is by inserting kana which would result a strong EPP because the interrogative idhaa can't see/search in the lower CP. Furthermore, kana and the perfective verb always show a partial set of phi-features because it is agreeing with an empty subject.

Conclusion
In this paper, I investigated the syntactic structure of the interrogative particles in Modern Standard Arabic in both the main and embedded clauses. First, I presented the four analyses which are used in studying the clauses in Modern Standard Arabic, and I adopted Mohammed's (1989) analysis in investigating the structure of clauses in Modern Standard Arabic. In the interrogative structure, Chomsky (1995, p. 291) claims that there is a strong feature [Q] which requires a lexical item to check this feature. In MSA, I have shown that there are three approaches for checking this strong feature [Q]. First, the interrogative affix ʔa-which is base-generated in the head of the CP to check the Q-feature. The interrogative ʔa-may appear in VSO or SVO word order with full agreement on both. The second approach is the negative polarity interrogative which starts its derivation in the head of the PolP and requires the verb to move from VP-to-TP-to-PolP to check hal's feature in the head of the PolP which creates VSO word order in hal clauses. As in ʔa-clauses, the verb shows the full set of phi-feature even if it is in VSO word order. The final approach is the interrogative idhaa which is base-generated in the head of its CP. This interrogative requires a perfective verb or kana to move/merge to check the subjunctive mood of idhaa.
Finally, more study is needed on the historical changes in the interrogative particles in order to further understanding of their function in MSA. Also, I am looking to do a study on wh-questions in MSA to see if they behave as the interrogative particles or if they have a different feature-checking system.