Challenges of Writing in Arabic as a Foreign Language Context: Case of the Spanish School in Rabat

“Learning to write coherently... is something which many people never manage in their first language [ and ] the process is every bit as difficult in a second language” (Nunan, 1991, p.99). As such, interest in writing amongst applied linguists has increased quite importantly during the last two or three decades; soon, the discipline has rapidly and “clearly come of age thanks to the launch of several firsts in L2 writing” (Leki, Cumming & Silva, 2008, p.1). It is within this context and with the objective to uncover what challenges learners of Arabic as a foreign language (AFL) face while learning Arabic writing that this qualitative corpus study was conducted. Three prevalent challenges emerged: one was the students’ difficulty to spell Arabic words correctly; another emanated from their overall inability to learn the common Arabic order of syntactic constituents; and the other had to do with the challenge to write well-organized Arabic paragraphs.


Introduction
Writing as a skill poses significant challenges that are widely acknowledged by educators and learners alike.Writing is a productive process wherein learners demonstrate different levels of language understanding, awareness of the target structure and competences.Thus, within the scope of learning writing in foreign language classrooms, the process arises as complex and demanding for different levellearners.According to Machael Levy and Sarah Ransdell (1995, p.767) "both professional and nonprofessional writers often lament that the process of writing is arduous and complex [and that] writing is also one of the least understood language production tasks."Based on the researchers experience as AFL and EFL teachers, when foreign language learners embark on the journey of learning how to write or engage in writing tasks, they generally encounter a dichotomy of limited choices.The first entails learners' relinquishment of writing tasks after a number of unsuccessful attempts, while the second choice entails striving to join the minority club of good writers.Eventually, foreign language educators are invited to delve into the whereabouts of writing to ease the process of learning writing as a skill, and help learners to achieve their utmost potential.
The dilemma surrounding the skill of writing extends to the scope of AFL and gives rise to controversies about its learning.AFL learners not only face the common challenges that learners of other foreign languages meet, but they also confront issues related to Arabic language features such as script, lexico-grammar, and rhetoric.For instance, the fact that "the majority of letters have multiple shapes, are connected to adjacent letters…contribute to the difficulty in learning Arabic as a language" (Al Busaidi, 2015, p.708).Worth to mention is that most AFL learners' first languages have script systems that are different from that of Arabic.However, this difficulty is not inherent to the Arabic linguistic system or its rhetoric features as some foreign learners and first-language speakers of Arabic promulgate.The claim that Arabic is a difficult language, and that it is impossible to learn, is a myth which is not only "untrue" but also "counterproductive" (Bergman, 2009, p.3).
The present study is driven by two main objectives.The first is to investigate any potential challenges that students at the Spanish school in Rabat meet while carrying out writing as a skill.In this respect, the study used a corpus of 15 paragraphs composed by second and third-year middle school students and first and second-year high school students.The second objective is to examine whether these challenges persist across the aforementioned levels.The Spanish school in Rabat adopts the Common European Framework Reference for Learning Foreign Languages (CEFR) as a reference for teaching foreign languages, including Arabic; therefore, the analysis of the collected corpora is in concordance with the school's orientation.The current study advances the following research questions: 1.
Do students at the Spanish school in Rabat face any challenges when learning writing in their AFL classes?

2.
Is there a longitudinal progression of students' compositions over the academic year?

Literature Review
The relationship between learners of AFL and writing as a skill has long been documented in works of literature by both Arab and non-Arab scholars.The fact that Arabic writing system differs from Latin, its writing script and writing genre stand as major linguistic challenges for AFL learners.This section reviews antecedent research about writing in foreign language, and in AFL classrooms.

Major orientations to Approach the Learning of Writing
Research on writing can be undertaken from the perspectives of the text, the writer, or the reader (Hyland, 2009).Research oriented towards the exploration and analysis of writing from a textual perspective focuses on texts independently from the contexts in which they are produced.An application of this view may manifest in second language learning (L2) contexts wherein "students are asked to write simply to demonstrate their knowledge of facts [and] grammatical accuracy and clear exposition are often the main criteria for good writing" (Hyland, 2009).
Writer-oriented research, on the other hand, revolves around the writer as the main axis of research in the sense that it investigates "what it is that good writers do when confronted with a composing task, and seek to formulate the methods that will best help learners acquire these skills" (Hyland, 2009).Evanic (2004) declares that the workability of this approach made it highly attractive to teachers and policy-makers due to "its adaptability into a set of elements which have inherent sequence that can be taught explicitly" (p.231).
The core claim of reader-oriented writing is that writing is a social interaction and its effectiveness is complete only if the writer succeeds to satiate the rhetorical needs of readers.In other words, "an implicit text is what the writer expresses as his or her expectations for the text in terms of the reader's purposes (and expectations) as well as his or her own.A real text is what is actually produced.The ideal is what the reader expresses as his or her expectations for the author and the text" (Tierney and Lazansky, 1983, p. 21).

The Structure of Written Arabic
This section overviews the most important rules that govern the effective writing of Arabic sentence and paragraph.

The sentence
One definition of a sentence provides, it is "the largest unit of language that is the business of grammar to describe" (Leech, 2006, p. 104).The role of grammar in this sense is to provide the laws that govern the combination of words in the sentence in order to correctly and conventionally convey complete thought.Every language has a degree of governance that does not allow us to "arrange the units in any way that we like since every sentence has a structure, in that there are rules that decide the units that can co-occur in the sentence and the order in which they can occur" (Nelson & Greenbaum, 2016).It is also possible to discuss sentences from the perspectives of their simplicity and their complexity.A simple sentence is one that "consists of one clause", such as 'I am a teacher,' while a complex is "a sentence that contains one or more subordinate clauses," like 'I am a teacher who is committed to his job' (p.307).
The Arabic sentence, which is the concern of this paper, enjoys an important degree of suppleness, for it accepts SVO, VOS, and VSO word orders all by remaining free and away from being erroneous, which is not the case of English that is an SVO language (Bassam, et al., 2014, p. 612).For example:

SVO:
‫اﻟ‬ ‫اﻟدرس‬ ‫ﯾﻛﺗب‬ ‫ﺗﻠﻣﯾذ‬ (The student is writing the lesson.)VOS: ‫اﻟدرس‬ ‫اﻟﺗﻠﻣﯾذ‬ ‫ﯾﻛﺗب‬ (is writing the student the lesson.)VSO: ‫اﻟﺗﻠﻣﯾذ‬ ‫اﻟدرس‬ ‫ﯾﻛﺗب‬ (is writing the lesson the student.)This word order flexibility could be more helpful for learners of Arabic as a foreign language to produce a correct sentence because reversing its constituents would always entail acceptable structures.However, there are some other rules and conventions that learners of Arabic as a foreign language need to master in order to produce acceptable sentences.

The paragraph
The writing norms of the Arabic paragraph as we know of them today are not old.They have been adopted recently, especially after the appearance of the modern publishing technology.In fact, the exposure of Arab authors to the writings of Western writing has impacted the Arabic paragraph and made it look more or less like the western paragraph.As such, an Arabic paragraph is composed of a group of sentences that are strongly and clearly connected to convey one meaning or to explain one reality (Ahmad Chalabi, 2018, p. 80).This means that like English, French, and Spanish, a correct paragraph in Arabic has to include a topic sentence, developing sentences, and a concluding sentence.As such, the structure of the Arabic paragraph should look like what follows: This paragraph begins with a topic sentence: "Smoking is harmful for health."After, it provides a number of ideas that support the general idea: "smoking causes cancer, asthma, emphysema…).Ultimately, the paragraph finishes with a concluding sentence: "So, is smoking worth it?" The length of the Arabic paragraph depends on the idea it develops.If the idea is simple, the paragraph is supposed to be short.Contrarily, if the idea is complex, it becomes mandatory for the writer to develop a long paragraph.In any way, a paragraph does not have to be too short since the main idea might end up being unclear and/or insufficient, and not too long in order for the writer to avoid invoking more than one main idea in the same paragraph (Redwan, 1984, p. 62).
In addition, the inner build of a paragraph has to abide by a number of regulations.One is unity of thought.This means the paragraph must develop one main idea while all the other information serves to explain and support this main idea.Any information that is not relevant to the main idea that is presented in the topic sentence must be avoided for the criterion of unity of thought not to be infringed (Kharma, et al., 2005, p. 20).Another important criterion for the effective writing of Arabic paragraphs is cohesion, which means the ideas of the paragraph are supposed to be presented orderly, logically, and directly connected with the main idea; any aberration, major or minor, necessitates revision and reconstruction (p.20).The other important criterion for the inner structure of a paragraph to be effective is the smoothness of ideas that enables the reader to smoothly move from one sentence to another without interruptive gaps and mismatch between ideas (p.20).This last criterion would be worth more development if my paper were to approach writing from the reader's point of view, which is not the case.
Based on these criteria, it was possible for me to identify the extent to which the texts I analyzed were effectively structured in the sense that they meet the requirements necessary for AFL students to write correctly and effectively.

Common Features for the Analysis of FL Students' Written-Texts
An overview of the literature reveals researchers have targeted quite a large spectrum of features in analyzing FL students' written texts.Deciding on what features to maintain and what features to drain depends to a large extent on the research orientation the researcher takes as a departing point.Also, the selection of the features (linguistic categories) for research and analysis is naturally and academically predictable depending on the scope of research and the rational the researcher chooses and plans to abide by.For example, a researcher whose research is framed from within the textoriented tradition would be more concerned about the textual features of students' writing, such as the paragraph, sentence, grammar, spelling, and syntax, etc. Contrarily, a reader-oriented researcher would be more concerned about analyzing the samples' texts from within the reader's rhetoric satisfaction and understandability of the written material.Our research is not an exception and cannot at all be, for I am going to analyze the corpora based on a clear and subtle orientation: it is a text-oriented analysis that focuses on the learners' sentences, paragraphs, use of cohesive devices, and punctuation.However, I feel it is of special importance to this research to overview the other features as analyzed and discussed in a number of previous studies.

Temporality Analysis
Of relevance to our research topic is a study that was conducted with the objectives to explore the time L2 writers spend on formulating their texts and the nature of the problems they focus along this process.The study was cross-sectional and compared L2 students to L1 students writing about the same topic.To collect their data the authors asked the participants to write and at the same time say what they were writing.In other words, the participants were asked to say every word and every punctuation mark they were putting on their sheets while responding to the task (De Larios, et al., 2006, p. 104).Some of the major findings from this study are, "the writer invested in solving formulation problems in the L2 more than in the L1 task and that proficiency did not make any difference to the amount of time they spent solving problems in the L2" (p.110).Added to these two findings, the study provided there was a strong and clear relationship between the participants' mastery of the L2 and their overall writing competence.
Of equal importance, this study identified the main stages learners go through in their process to master the writing of an L2.These are summarized as

•
Noticing features in the input received; • Hypothesizing L2 rules; • Complexifying their L2 system; • Using rules with different levels of accuracy and speed; • To full automatization in the use of L2 resources (p.108).
A few important remarks can be made about these stages.One is that they are gradual as learners proceed from the simplest to the most complex writing.Another is that the learning of writing is a cognitive process that can be conscious and unconscious depending on the style of learning.The other is that time is an important feature to consider in collecting data for the analysis of students' difficulties in learning L2 writing.The three components are important in the analysis of written texts, especially if the objective is to explore the difficulties FL learners face in a level and/or across different levels.

Language Transfer Analysis
One other feature researchers have investigated while studying the difficulties that face FL students learning writing is language transfer.One definition of language transfer provides, it is "the full range of ways in which a person's knowledge of one language can affect that person's knowledge and use of another language" (Jarvis and Pavlenko, 2007, p. 3).It is not limited to and merely concerned about the skill of writing but about the other language skills as well.Linguists disagree as what exactly happens during language transfer or, even whether, 'transfer' is the term to describe what happens in the minds of L2 learners.For example, Cook (2002) indicates there are no boundaries that keep the L1 and L2 systems apart along the L2 learning process; instead, he hypothesizes they are "merged and overlapping systems" while nothing is clear about where one ends and where the other begins (p.18).Other research downgrades the meaningfulness of the term transfer and supports the claim that focus should be more on the learners' "reorganization, redirection, expansion, and transformation" of knowledge (Hall, Cheng & Carlson, 2006, p. 231).
To relate this debate more to the concern of my study, which is conducted to explore the writing difficulties facing AFL learners, I highlight the term "adaptive transfer" as it was introduced into the literature by DePalma and Ringer.The authors indicate, adaptive transfer is "a conscious or intuitive process in which composers apply and reshape their knowledge and experiences in order to negotiate new and potentially unfamiliar rhetorical tasks" (Depalma & Ringer, 2011, p. 137).For them, "writers have the agency to draw from among a variety of discourses and language varieties" (p.141).
The end product of this mental negotiation is either a correct or incorrect FL product, depending on whether the transfer is positive or negative.
An example of a study that approached the analysis of FL written-texts from the perspective of language transfer or adaptive transfer was conducted on a group of students learning Japanese.The objectives of this study were mainly to investigate "how the view of writer as agent explains the process of transfer across the two languages (Japanese and English)," and to identify how the Japanese learners employ transfer in their l2 writing process (Rinnert,et al.,p. 217).The findings of this study concluded that while writing FL essays, the participants "took an active role in assessing audience, selecting appropriate text features, and transforming selected features, influenced by contextual factors," and in this sense "a dynamic view of transfer provides insights into" how FL learners construct their written texts (p.213).
One other interesting finding was that the participants had "differing perceptions of L1 and L2 reader expectations and tried to tailor their English and Japanese to accommodate these perceptions" (p.228).

Cohesion and Coherence
The previous studies have targeted FL learning other than the learning of Arabic as a Foreign Language (AFL).And to relate the literature more to this current study, I will discuss hereafter studies that were conducted as part of AFL research with a focus on the features they identified for analysis.
One of these studies focused on students' development of cohesion and coherence when learning AFL writing.The authors focused on the following textual features: • Topic development; • Appropriateness of content organization as entailed by the rhetorical function of the text in question;

•
Relevance of the content to the global topic; • Audience awareness and contextualization; • Use of cohesive connectives appropriate to the type of text; • Use of verb tenses appropriate to the function of the text; • Naturalness of exposition, appropriate use of collocation expressions, L1 interference, etc. (Abdullah Shakir and Hussein Obeidat, 1992, p. 5).
This study concluded that coherence in the students' written texts was the result of interactivity between factors in the texts and factors beyond it.A few amongst these are, "cohesive connectives, verb tenses, collocations, and the proper employment of L2 syntactic structures" as well as a valid consideration of the reader's expectations from the writing task (p.14).The misuse of cohesive devices, however, led to the production of texts that are fragmentary in structure and blurry in meaning.The authors also identified a negative interference of the students' L1 in their writing of AFL.

Sentence Complexity
In another study, Shakir and Obeidat (1992) tested the maturity of their participants' FL written texts by investigating how they develop their AFL writing.The study was conducted from within the premise that writing maturity is not limited to the ability of the students to produce error-free texts; instead, it is about the development of the complexity of the content they produce and the organization through which they produce it ( Shakir and Obeidat, 1991, p. 70).The authors assumed that the participants' use of simple and short sentences meant they were not mature yet in the process of learning the writing of AFL.Contrarily, the ability of the participants to produce long and complex sentences meant they had matured as AFL writers.As for the parameters of writing maturity, the authors highlighted "the increase of the number of words… the use of embedding in writing… the employment of appropriate and varied cohesive devices" (p.78).
These are the features the study identified for analysis in order to investigate the development of their participants' writing from the stage preceding the writing course and the stage that followed it.

•
The number of words; • The number of simple sentences; • The number and percentage of missing words, and

•
The average scores obtained on each of the pre and post-course texts (p.69).

Challenges of Learning AFL Writing
The learning of Arabic as a FL, in general, and the learning of its writing in particular, might be difficult to many learners.However, this difficulty is not inherent to the Arabic linguistic system or its rhetoric features as some foreign learners and firstlanguage speakers of Arabic promulgate.The claim that Arabic is a difficult language and that it is impossible to learn is a myth, which is not only "untrue" but it is also "counterproductive" (Elizabeth Bergman, 2007, p. 3).It is untrue because, Arabic which is spoken by more than 200 million people, is a human language that all humans can learn provided there are the right condition for this learning to take place.It is counterproductive since foreigners who learn Arabic and believe it is a difficult language would have an affective barrier that would most probably slow the learning process or thoroughly halt it.Contrarily, if the learners are highly motivated and are convinced that the learning of Arabic is as possible as it is the case with all human languages, the learning would be attained by and far.The writing of Arabic is not an exception.
The claim that Arabic is not difficult to learn as a FL does not mean that it is easy to learn.In fact, there are many challenges that research has identified from studying the learning of AFL writing.Some of these are difficulties that are faced by FL learner of writing in other languages, while some are challenges that relate to the learning of Arabic writing because of the specifities of the language.The following is an overview of the most common difficulties learners of AFL face while learning writing.
The first challenge facing the mastery of AFL writing relates to the learning of Arabic alphabet, which is completely different from the alphabets of other languages.The letters of the Arabic alphabet are "completely unfamiliar" and "they must be connected together in a script rather than being written separately; they have up to four different shapes: initial, medial, final, and isolated; the short vowels and some grammatical endings are not indicated; and the writing goes from right to left" (Wahba, et al 2017, p. 42).For example, the letters ‫ت‬ /t/ and ‫ي‬ /y/ have the following shapes depending on where the fall in the word: So, for learners to master how to write the Arabic alphabet, each with its own shapes, might be quite challenging.However, Wahba et al (2018) indicates, Arabic has an alphabet that offers an "extraordinarily regular, one-to-one correspondence between letter and sound," and as a result makes its' writing easy to learn (p.42).
Another study -overviewing the learning of AFL writing from within the dimension of the Arabic script as being different from the mother tongues of students learning it-emphasized how difficult it is to learn the Arabic script.The study points out that FL learners of Arabic find it challenging to learn, for example, the Arabic 'hamza' ‫,)ء(‬ which has different spelling representations depending on where it occurs in the word, the vowel (Haraka in Arabic) on which it occurs, and/or the vowel that precedes it.For example, alhamza can be spelt as: The fact that the same sound has three spellings can cause some difficulty to students learning AFL.As such, the "learners face a lot of obstacles in this area, beginning with the joining and stopping hamza, and crossing the hamza that comes in the middle or the end of the word with its different vowel points" (Omari, 2017, p.83).
In addition to spelling, researchers have highlighted a learning difficulty at the level of the writing of the Arabic tense forms.For example, while Arabic "has only two tense forms, the English language has sixteen tense forms" (Alasmari et al, 2016, p. 3).In Spanish, which is the L1 of this study's samples, there are fourteen verb tenses.The switch from writing a language that has fourteen or seventeen tenses to writing a language whose tense system is limited to two tenses would certainly be challenging for the learners.The ability of Arabic to make up for the multi-tense system in other languages is achieved thanks to its ability to have developed rules of inflectional morphology.These are "the affixes that can be added to derived forms of the verb" in order to indicate "person, gender, number, aspect and mood" (Al-Saleemi, 1987, p. 38).Let's consider, for the sake of illustration, the following examples: • English: I will have left (Future perfect) • Spanish: habré salido • Arabic: ‫ﻏﺎدرت‬ ‫ﻗد‬ ‫ﺳﺄﻛون‬ (future morpheme -prefix for the first personal pronoun -to be in the present -emphatic word -verb to leave in the past -suffix for the first personal pronoun).
Obviously, an AFL learner has to learn all these structures in order to express the present perfect in their writing, which can but only be a challenging endeavor to undertake.
Last but not least, the difficulty to write Arabic might also emanate from the nature of the Arabic discourse.In fact, it is indispensable for L2 learners to know how to manage writing texts as objects, but it is also of equal importance for them to produce a writing discourse that abides by the discourse conventions produced by the native users of the L1 (Saif-Modhish, 2012).

Methodology
To answer the two research questions, the study used qualitative content analysis.This research method is "a subjective interpretation" used to investigate texts in their communicative contexts with the goal of yielding a valid description of recurrent language patterns and/or meanings in a text or a group of texts (Patton, 2002;Krippendorff, 2018).Despite being qualitative, this method of analysis can start "from the methodological basis of quantitative content analysis" in order to "conceptualize the process of assigning categories;" the latter can be assigned inductively (Bennett et al., 2019), and that is in the sense that researchers work on categories that emerge during data analysis.
Two AFL high school teachers, a university professor, a PhD graduate, and a PhD candidate participated in the collection, analysis, and rating of the corpus.The high school teachers were teaching the participants AFL at the time of study.First, the researchers prepared the rubrics, shared and discussed them with the students' teachers.Second, the teachers assigned the writing tasks to their students.Third, they excluded the compositions of non-AFL students.Fourth, a professor who did not take part in the study, coded the texts and prepared them for the analysis phase; the rationale of this measure was to avoid any bias in case raters knew about the level to which the paragraphs belonged.Finally, the coding professor proposed the following codes

Sampling
The collection of the sample texts was conducted using purposive sampling.This research technique is deliberate in the sense that researchers nonrandomly select their samples in order to reach their research objectives (Bernard, 2002, Lewis & Shepard 2006).Accordingly, the researchers initiated contact with the two AFL teachers in order to collect students' compositions for later analysis.The teachers were made aware that only students who spoke Arabic as a foreign language were eligible for the study in order to avoid any reliability issue with the data to be collected.Then, of the four classes the researchers approached, only 15 students met the AFL criterion.The figure below describes the percentages of students whose paragraphs were analyzed in this study.

Figure 2. Students' distribution by level of education
The variance amongst the percentages was unavoidable because the numbers of students who spoke Arabic as a foreign language differed from one category to another.As for determining inter-rater reliability, the researchers each analyzed and commented on the texts separately before they met to compare and conclude their findings.

Research instruments
The Prompts As summarized by Mohammad Adnan Latief (2009), the development of an effective prompt requires having clarity, challenge and task uniformity as crucial criteria.
Regarding the clarity of tasks, writing prompts ought to guide students through the writing stages in an explicit, unambiguous and recognizable manner.In other words, students are mandated to understand the writing prompt without the intervention of their teachers.As for the second criterion, prompts ought to be adequately challenging.
An effective prompt necessitates the use of texts, themes and language level that would go along with the students' level; worth to note that when a prompt is too or less challenging, then students' responses would be a failure.Concerning the third and last criterion, task uniformity entails that all students must respond to the same task since this would guarantee they are engaging in the same skills when writing.Considering these norms, the following prompt was designed in order to collect the data needed to respond to the questions we had asked previously: ‫ﺣﺪود‬ ‫ﻓﻲ‬ ‫أﺳﻄﺮ‬ ‫ﻋﺸﺮة‬ ‫اﻟﺘﺎﻟﻲ‬ ‫اﻟﺴﺆال‬ ‫ﻋﻦ‬ ‫ﻓﯿﮭﺎ‬ ‫ﺗﺠﯿﺐ)ﯾﻦ(‬ ‫ﻓﻘﺮة‬ ‫أﻛﺘﺐ)ي(‬ : Answer the following question in a paragraph form.Do not write more than 10 lines.A few rationales support the workability of this prompt for the valid and reliable collection of this paper's data.First, it can be responded to by the students of the two high school levels with the least difficulty because it constitutes equal challenge to all of them.Second, the topic is interesting enough to motivate students to willingly and excitedly write about it.Third, the prompt is very clear.Added to these criteria, the prompt was assigned as a mini-test, and this made the researchers ascertain that all students were responding seriously.Otherwise, nothing could guarantee the students would take the task as seriously as the purpose of the whole research procedure required.

‫ﺑﻌﺪ؟‬ ‫ﻋﻦ‬ ‫اﻟﺘﻌﻠﯿﻢ‬ ‫أم‬ ‫اﻟﺤﻀﻮري‬ ‫اﻟﺘﻌﻠﯿﻢ‬ ‫ﺗﻔﻀﻞ)ﯾﻦ(‬ ‫ھﻞ‬
The high school students were assigned the following rubric: Although the prompts differed at the level of the topic, they still had key common features which should justify their workability for our cross-sectional comparative analysis.One similarity is that they were both preference prompts.While the first asked students to express their preference with regards to online and onsite classes, the second requested they choose either to rent or to buy a house.Another common feature was that both prompts required students to provide arguments, reasons, and details to support and explain their preferences.The other similarity was that the two prompts required no previous knowledge for students to understand what they were supposed to write about and how they were supposed to write about it.

Students' Written-Texts
The texts the students wrote in response to the prompt constituted the second instrument this research used.Obviously, the use of students' written texts as a research instrument is a common tool for data collection in qualitative case studies, which is the case of this paper.In fact, "documents of all types can help the researcher uncover meaning, develop understanding, and discover insights relevant to the research problem" (Merriam, 1988, p. 118).Documents in qualitative research include amongst others "copies of participants' course syllabi, reading lists, assignment sheets…" and can help the researcher to "understand the demands being placed on his participants…" (Heigham & Croker, 2009, p. 76).As such, the choice of the texts as a research instrument is justifiable by its relevance to the objectives I set for this paper and by the qualitative research tradition, which accepts the use of documents as instruments for data collection, notably, in case studies.

Results and Discussion
The researchers (n=3) and two teachers of the students rated the paragraphs separately and then met to compare their findings.The raters did not agree on the themes to study; instead, they analyzed the texts and made categories based on the arising themes.These were spelling, sentence structure, paragraph organization.

Spelling
Spelling was the first theme to have emerged from the content analysis the researchers conducted to analyze the difficulties facing the AFL students of the Spanish Middle and High School in Rabat, Morocco.Clearly, Arabic spelling constituted a serious difficulty for students to overcome, especially in the second and the third year of middle school.Students spelling improved but slightly as of the first and second year of high school.The table below describes the teachers and the researchers rating of the students spelling.As the table shows, the students of the first two levels average was 1.1/4 while that of the other two levels increased slightly to reach 2.6/4.This average, according to the teachers, was lower than what the school expected from the students to achieve.
The difficulty the students encountered in learning the spelling of AFL and which this study has found is already documented in other research.For instance, Azzam (1993), Al-Mutawa (1995) and Hanson (2010), concluded the letter system of Arabic and the fact that they are connected sometimes and disconnected some other times constitute a difficulty for AFL learners to spell correctly.Nonetheless, it should be mentioned here that the Arabic writing system is "after all, an alphabet, with 28 letters" and "learning it is not an insurmountable task" (Stevens, 2006, p.60).

Order of Syntactic Constituents
Word order is the second significant theme the researchers identified after their analysis of the students' texts.The usual word order in Arabic is Verb-Subject-Object (VSO) (Parkinson, 1981;Al-Khuli 1982;Abdul-Raof 1998;).This was confirmed by Bakir (1980) and Fassi Fehri (1988), who argued the VSO sequence is the canonical word order in Arabic.Some other scholars, however, supported the SVO order as the canonical sequence in Arabic (Mohammad, 1990).The raters manually counted the number of each VSO and SVO structures in the texts written by the four groups of students and found the following: The table shows SVO word order was the common syntactic sequence in the students writing, which could mean the students were either taught to write in such an order or, probably, it was a transfer from their first languages -Spanish or French.Nevertheless, research shows that while Spanish is more of a free language in the sense that it accepts different word order, French is strict and, at least, does not accept VSO order, which is the usual order in Arabic (Lahousse and Lamiroy, 2012).In either case, the fact that the texts of the students were about 87% SVO despite Arabic being a VSO is worth identifying and researching.

Paragraph Organization
The organization of the students' paragraphs remained unchanged along the four levels.From level 1 until level 4, students did not manage to write paragraphs with a topic sentence, developing sentences, and a concluding sentence, which is the common paragraph layout in Arabic writing (Ahmad Chalabi, 2018).Of the 15 paragraphs this study analyzed, only 2 had topic sentences, and both these were written by students at the first level.The others were composed in a way that failed to meet the usual paragraph organization of Arabic.In fact, an interesting number of paragraphs were made of one sentence

Conclusions and recommendations
In response to research question one -whether students of the Spanish School in Rabat face challenges learning AFL writingthe researchers found these students encountered challenges to learn AFL writing at a variety of levels; on top of these are spelling, word order, and paragraph organization.Simultaneously, the analysis of the corpora revealed in response to research question two -whether there was any improvement of students AFL writing as they move from middle to high schoolthat the progress students made was very low.Comparing the compositions, the students wrote duing their middle and high school years reveals they still made almost the same mistakes despite composing lengthy and meaningful paragraphs.
Accordingly, the questions to respond to are why this is the case and how it can be overcome.Although responses to these questions require more research at the level of the Spanish school in Rabat in order to survey, for instance, the AFL qualifications of the teachers and the methods they use to teach writing, the researchers propose more focus on students' awareness of the AFL writing conventions.For instance, the fact that students did not write well-structured paragraphs is probably because they were not aware of the structure to follow in Arabic paragraph writing.
Imagine you needed to rent or buy an apartment.Describe how it should be.
, which was similar to what Montano-Harmon (1991) found in a study on Mexican students writing Spanish paragraphs.The author's finding provided, in many instances students' paragraphs were limited to one long sentence.Below are two examples of what two participants in this research wrote (the first is from second year middle school while the second is from second year high school).

Figure 2 .
Figure 2. Examples of students' writing Note: Each paragraph constitutes of one long sentence.As such, the topic sentences and the developing sentences are not presented as separate entities, and the concluding sentences are not provided in both paragraphs.