Economy and the Conflict with Uniformity in Syntactic Theory : The Case of Wh-Questions in Moroccan Arabic

| ABSTRACT This paper investigates the relevance of some syntactic principles in Generative Grammar. Specifically, the objective is to highlight the conflict between economy and uniformity in syntactic theory. This conflict is illustrated by some elements of wh-questions. Using the Minimalist Program as a current syntactic ‘approach’, we take MA wh-elements as a case study to test the relationship between the two principles. We, first, survey the importance of economy in the language itself. Second, we bring examples of the conflict in wh-interrogatives, specifically in wh-subjects, wh-objects, yes-no questions, and the [Q] feature. The significance of the research is that, in the absence of empirical evidence for uniformity, economy should be prioritized.


Introduction and General Background
Language has always been considered a social fact. Indeed, the social dimension of language has always been present in the literature. This can be summarized in the following lines: It is a treasure deposited by the practice of speaking in people belonging to the same community, a grammatical system that exists virtually in each brain, or more exactly in the brains of a collection of individuals; because the language is not complete in any individual, it exists perfectly only in the collective. (Saussure, 1916(Saussure, /1922. Starting from the 1950s, a refined language innateness theory was developed/suggested. To be more specific, Noam Chomsky wrote his The Logical Structure of Linguistic Theory in 1955, though he did not publish it until 1975. Syntactic theory was seen from a completely different angle: 1 Linguistic theory is concerned primarily with an ideal speaker-listener, in a completely homogeneous speech-community, who knows its language perfectly and is unaffected by such grammatically irrelevant conditions as memory limitations, distractions, shifts of attention and interest, and errors (random or characteristic) in applying his knowledge of the language in actual performance. (Chomsky, 1965, p. 3).
The debate between the social and cognitive perspectives can be compared to that between those who lean to either studying language as a form of output (i.e., performance) or input (i.e., competence) (Chomsky, 1965, p. 4).
For the latter, the central question that needs to be answered is the following: How do children, at a fast pace, with limited exposure to language, and in a pretty uniform way acquire their mother tongue and produce from a finite set of data an infinite number of utterances? This question is often called the Logical Problem of Language Acquisition or 'discrete infinity' (Chomsky, 2005, p. 11). By answering this question, syntacticians will minimize the language acquisition burden on the child and, consequently, maximize the learnability of grammar (Chomsky, 1995). This is summarized in the so-called Universal Grammar (henceforth UG) (Chomsky, 1965). UG is "the theory about the mental procedure implicit in [the I-language]" (Uriagereka, 1998, p. 36). Since the child acquires language quickly and produces an infinite number of utterances from limited data, we need a syntactic theory to replicate the same process as the child (i.e., our grammar or theory needs to be explanatorily adequate). In other words, according to Chomsky (2015), any 'natural' science aims to catch the simplicity of nature. Since language is a perfect system, then the theory that explains language should also reflect a perfect design in the form of the most uncomplicated grammar (Chomsky, 1995). This simplicity is embodied in UG, which needs to encourage the most economical computational operations. 2 For this reason, Chomsky (1998, p. 9) states that language is "an optimal solution to legibility conditions." Thus, economy 3 is one of the most central elements in the current syntactic theory. Our paper is divided as follows: the second section tries to search for the signs of economy on language itself. The third section investigates the conflict between economy and uniformity, taking wh-questions in Moroccan Arabic (henceforth, MA) as a case study. Our main objective is to question whether the concept of 'economy' plays a role in the syntactic theory, especially if it clashes with other principles such as uniformity.

Economy and its Signs on the Language Itself
Syntactic theory gives importance to economy because language itself displays this phenomenon. A first example is the three economy principles of Move that languages show in general. These are Last Resort, Greed, and Procrastinate. First, the Last Resort principle is a "condition of movement (which) requires that movement is permitted only to satisfy some condition" (Chomsky, 2015, p. 41). In other words, Last Resort happens when "Move F raises to target K only if F enters into a checking relation with a sublabel of K" (Chomsky, 1995, p. 280). This principle motivates the movement of elements. If the operation can occur without movement, then Move need not happen; hence, movement is a last resort or option. An example of this is wh-movement in MA, where wh-elements do not stay in-situ and must move (Nouhi, 1996;Announi, 2019Announi, , 2021. Second, Last Resort is later modified to become Greed (self-serving Last Resort), which means that "benefiting other elements is not allowed 4 " (Chomsky, 2015, p. 184). Let us take the following example (Chomsky, 2015, p. 239): (1) seems [(that) John is intelligent] (2) *John seems [(that) t is intelligent] John moves to the matrix clause to satisfy the Extended Projection Principle (EPP) (i.e., the subject position is unfilled) and Case. However, the principle of Greed does not allow movement since John already satisfied EPP and Case in the embedded clause. Finally, Procrastinate (Chomsky, 1991(Chomsky, , 1992 permits delaying movement whenever possible. Chomsky (2015, p. 181) states that "LF movement is 'cheaper' than overt movement." Delayed movements are covert movements where elements that did not move in overt syntax move at the covert syntax. An example of this is English, where main verbs do not move until after the Spell-Out (Pollock, 1989) and wh-questions in EA (Gad, 2011), where they stay in-situ. These three principles show how languages are given different options to convey elements of economy.
A second example of the sign of economy on language itself is feature stranding. To explain this phenomenon, let us bring these two examples: There are two options for the movement of elements concerning features (Chomsky, 2015, p. 242). The first option is for syntactic elements to move with the feature that needs checking; this process is called pied-piping. For example, in MA (see, e.g., 3), the [Q] feature moves with the wh-elements to be checked (Chomsky, 2015, p. 125). The second option is that syntactic elements stay insitu while the feature that needs checking moves in the covert syntax; this process is called stranding. For example, the [Q] feature Page | 144 moves in Egyptian Arabic while the wh-element stays in-situ (Chomsky, 1995). Feature stranding is an example of a procedure where it is more economical to leave a wh-element behind than to move them along with the feature in question.
A final example of economy in languages is the Attract Closest Principle, which entails that "a head which attracts a given kind of constituent attracts the closest constituent of the relevant kind" (Radford, 2004, p. 162 (5) and (6) are examples of multiple wh-construction. In MA, multiple wh-questions are possible but only as echo-questions (Announi, 2019). We observe that while it is correct to front one wh-element (see 5), it is ungrammatical to front both wh-elements to the left periphery. Let us show example (6) in the form of a tree representation to explain its ungrammaticality: According to Attract Closest Principle, the head C needs to attract the closest element to it. In our example, the wh-element /ʃkun/ who is closer than /ʃnu/ what. Thus, (6) is ungrammatical because it selects the wh-object further apart from the wh-subject. This principle is an example of economy wherein language prefers to front the closest element to the head in question.
This section highlighted examples of economy in the language itself. The following section will look closer at the relationship between economy and uniformity.

The Conflict between Economy and Uniformity: The Case of Wh-Questions
This section will sketch the relationship between economy and uniformity. It looks at the cases where economy and uniformity clash. These cases are wh-subjects, yes-no particles, wh-objects, and the [Q] feature. We will conclude the section by looking at a case of multiple wh-questions where neither economy nor uniformity work.

The Conflict between Economy and Uniformity: The Case of Wh-Subjects
Consider the following example: The economy proponents call this the Vacuous Movement Hypothesis (Chomsky, 1986b; see also Nouhi, 1996   As we can see, there is an apparent clash between the two principles of economy and uniformity. With all of these taken into account, we propose that when the wh-subject appears with the complementizer, then uniformity overrides economy. When the wh-subject does not occur with the complementizer, economy overrides uniformity.

The Conflict between Economy and Uniformity: Yes-No Particles
Consider examples from English, Standard Arabic, and MA: (12) Did you study today? English (13) hal qaraʔt-a l-jawma Standard Arabic Q-Prt read.PAST-2P the-today 'Did you study today?' (14) waʃ qrit-i l-juma Moroccan Arabic Q-Prt read.PAST-2P the-today 'Did you study today?' (12), (13), and (14) are yes-no interrogative sentences in English, Standard Arabic, and MA. In English, yes-no question formation results from T-to-C movement (Chomsky, 1995). Let us look at the figure below: (15) As shown in (15), yes-no question formation is the result of the verb moving from the T(ense) to C(omplementizer) position (Chomsky, 1995). Let us now turn to Standard Arabic: (16) In Standard Arabic, the question particle /hal/ is base generated in C; this is because yes-no questions in SA do not force any morphosyntactic operation (i.e., subject-verb inversion) and are, thus, only base generated in C ( does not force subject-verb inversion. The obvious conclusion is, then, that the yes-no particle /waʃ/ is base generated. Moreover, Announi (2019) questions whether the yes-no question is base generated in C or in [Spec, CP] as seen below: (17) Whether the yes-no particle is base generated in C or [Spec, CP] is a classic problem of the clash between economy and uniformity. If we would like to achieve uniformity, all interrogatives (i.e., yes-no particles and wh-questions) are positioned in [Spec, CP]. This provides rule unification for question formation. If we would like not to confuse the structure (i.e., achieve economy), then yes-no interrogatives should be situated in C while wh-questions should be in [Spec, CP]. A simple and economical structure provides a clear meaning to the structures. Semantically, yes-no questions and wh-questions differ; therefore, the structure should distinguish between the two. Empirically speaking, data suggest that uniformity overrides economy. /ila/ if occupies C. This means that /waʃ/ is in a position higher than C. This position is [Spec, CP]. There is empirical evidence that the yes-no particle precedes the complementizer /lli/ that (Announi, 2019, p. 56, e.g., 2). Therefore, yes-no question formation provides evidence for the clash between economy and uniformity.

The Conflict between Economy and Uniformity: The Case of Wh-Objects
Let us turn our attention to wh-objects:  (20) The first option is that the wh-object DP2 will be externally merged with the verb. The second option is the wh-object is in Spechead relation with v. In this situation, we have two vPs: one for the wh-subject position, the outer shell vP1, and one for the whobject position, the inner shell vP2. These two options are the result of a clash between uniformity and economy. If we prioritize uniformity, then the specifier proposal is the right one since both subjects and objects will be situated in the specifier of the verb phrase. If economy is prioritized, the wh-object can simply be externally merged with the verb, and there is no need for an extra projection. In the current syntactic theory, Uriagereka (1998) selects the external merge proposal because it is the cheapest, most economical option. Another reason for economy overriding uniformity in this situation is that the specifier proposal violates the Shortest Move principle; that is, the direct object moves across the subject in [Spec, vP2] to [Spec, vP1] but the subject in [Spec,vP2] is the closest element to [Spec, vP1] (Musabhien, 2008, p. 75). This situation is, however, solved by Chomsky (1995, p. 356), who proposes the Equidistance principle: the idea that if the object and subject are within the same minimal domain, then [Spec, vP1] is accessible by the object and the subject and the inner shell will not block movement. This also explains why subjects moving to [Spec, TP] is not blocked by the outer shell (Musabhien, 2008, p. 76). One can claim that even if the situation is solved, the economy option is better since, in the absence of empirical evidence for uniformity, economy should be prioritized. It should be noted that the current syntactic theory (the Minimalist Program; Chomsky, 2005) prioritizes uniformity. That is, if the wh-object stays in the complement position, the derivation will crash since as soon as the TP is introduced, VP will be shipped to PF and LF.

The Conflict between Economy and Uniformity: The Issue of the [Q] Feature
In the case of wh-elements. The [Q] feature interprets the interrogativity of the sentence. The following questions arise: If we assume that [Q] is an inherent feature integrated into wh-elements in the lexicon, how can we explain the appearance of whelements in affirmative sentences? In (21), the wh-object appears in affirmative sentences. If the [Q] is present in the lexicon, how do we explain wh-elements appearing in non-interrogative sentences? If the [Q] feature is introduced to (21), it will not be checked against C since the sentence does not introduce an interrogative force.  The second evidence is the case of multiple wh-questions. In the case of (22), does that mean each wh-element has its own [Q] feature to interpret the sentence as interrogative? Only one contributes to the interrogativity of the sentence. If we argue for uniformity, both wh-elements will come with uninterpretable [Q] features. While one of them will be checked against C (i.e., the wh-subject), what will happen to the [Q] feature on the wh-object? This will cause the derivation to crash. The natural conclusion is that the [Q] features arrive until the computational system.
It gets even worse if we consider our third evidence. What happens to Slavic languages such as Hungarian, which fronts all of its wh-questions in the left periphery. Does that mean all of those questions have the [Q] feature or just one of them? It seems that the economy proposal will solve this problem more than the uniformity one.
A final piece of evidence is when it comes to the nature of wh-elements. It is not always the case that wh-elements express interrogativity, whether direct or indirect. Wh-elements can also be relatives, indefinites, and quantifiers in other languages.
To sum up, it is more economical for the [Q] feature to appear until the computational system since not all wh-elements possess the [Q] feature in the lexicon.
We discussed cases where economy is in apparent conflict with uniformity. Before we end the paper, let us discuss the issue of wh-questions one final time. Economy dictates that we do not need to move wh-elements. This shows that economy is not always attested in languages. For example, in MA, it uses wh-movement as a default strategy while languages such as Chinese conform to economy by having wh-elements stay in-situ. Now, if all wh-questions stay in-situ, then we conform by both economy and uniformity. Economy works since all wh-questions do not have to move. Uniformity works since all wh-questions stay without others moving. Languages such as MA and English display a dilemma. Only one wh-question moves to the left periphery while the other remains in-situ. In this case, neither economy nor uniformity work in MA multiple wh-construction. The situation worsens with the so-called optional wh-movement languages (e.g., Egyptian Arabic; Gad, 2011), where wh-elements can stay in-situ and optionally front to the left periphery. By taking everything into account, we conclude that there is an evident clash between economy and uniformity. If there is empirical evidence for selecting uniformity, then it overrides economy; otherwise, economy should be prioritized.

Conclusion
This paper discussed the principle of economy and its relation to uniformity. We took MA wh-elements as a case study to test the relationship between the two principles regarding wh-subjects, wh-objects, yes-no questions, and the [Q] feature. We also noted a case where neither economy nor uniformity work in MA. Our final remark is that when there is a clash between economy and uniformity, and if there is empirical evidence for uniformity, we should pick it. In the absence of empirical evidence and we have the option for economy, syntactic theory should opt for economy since it should reflect language acquisition. One limitation for the study is that we did not explore further syntactic phenomena, apart from wh-questions, in order to further prove this conflict and the prioritizion of economy in the absence of empirical evidence. Implications for future research include the need to further investigate the relationship between the two principles in the light of other syntactic structrues.
Funding: This research was funded by the National Center for Scientific and Technical Research, grant number 2UIT2020.

Conflicts of Interest:
The authors declare no conflict of interest.